{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.2336434831708933,
  "eval_steps": 500,
  "global_step": 100500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00613752976701937,
      "grad_norm": 0.7345873117446899,
      "learning_rate": 4.9897707837216344e-05,
      "loss": 1.0473,
      "step": 500
    },
    {
      "epoch": 0.01227505953403874,
      "grad_norm": 1.5779420137405396,
      "learning_rate": 4.9795415674432687e-05,
      "loss": 1.0447,
      "step": 1000
    },
    {
      "epoch": 0.01841258930105811,
      "grad_norm": 1.1132322549819946,
      "learning_rate": 4.969312351164903e-05,
      "loss": 1.0383,
      "step": 1500
    },
    {
      "epoch": 0.02455011906807748,
      "grad_norm": 1.4636728763580322,
      "learning_rate": 4.959083134886538e-05,
      "loss": 1.0395,
      "step": 2000
    },
    {
      "epoch": 0.03068764883509685,
      "grad_norm": 2.259823799133301,
      "learning_rate": 4.948853918608172e-05,
      "loss": 1.0436,
      "step": 2500
    },
    {
      "epoch": 0.03682517860211622,
      "grad_norm": 3.276669979095459,
      "learning_rate": 4.938624702329807e-05,
      "loss": 1.0375,
      "step": 3000
    },
    {
      "epoch": 0.04296270836913559,
      "grad_norm": 2.432560443878174,
      "learning_rate": 4.928395486051441e-05,
      "loss": 1.0189,
      "step": 3500
    },
    {
      "epoch": 0.04910023813615496,
      "grad_norm": 2.106665849685669,
      "learning_rate": 4.918166269773075e-05,
      "loss": 1.0282,
      "step": 4000
    },
    {
      "epoch": 0.05523776790317433,
      "grad_norm": 10.379436492919922,
      "learning_rate": 4.9079370534947094e-05,
      "loss": 1.0355,
      "step": 4500
    },
    {
      "epoch": 0.0613752976701937,
      "grad_norm": 1.7805677652359009,
      "learning_rate": 4.897707837216344e-05,
      "loss": 1.0286,
      "step": 5000
    },
    {
      "epoch": 0.06751282743721307,
      "grad_norm": 4.699483394622803,
      "learning_rate": 4.8874786209379785e-05,
      "loss": 1.0264,
      "step": 5500
    },
    {
      "epoch": 0.07365035720423244,
      "grad_norm": 3.2066714763641357,
      "learning_rate": 4.877249404659613e-05,
      "loss": 1.0174,
      "step": 6000
    },
    {
      "epoch": 0.07978788697125182,
      "grad_norm": 4.304272651672363,
      "learning_rate": 4.867020188381247e-05,
      "loss": 1.0238,
      "step": 6500
    },
    {
      "epoch": 0.08592541673827118,
      "grad_norm": 1.9387307167053223,
      "learning_rate": 4.856790972102881e-05,
      "loss": 1.0217,
      "step": 7000
    },
    {
      "epoch": 0.09206294650529055,
      "grad_norm": 7.0195488929748535,
      "learning_rate": 4.846561755824516e-05,
      "loss": 1.0203,
      "step": 7500
    },
    {
      "epoch": 0.09820047627230992,
      "grad_norm": 5.841785907745361,
      "learning_rate": 4.836332539546151e-05,
      "loss": 1.0201,
      "step": 8000
    },
    {
      "epoch": 0.10433800603932929,
      "grad_norm": 4.427718162536621,
      "learning_rate": 4.826103323267785e-05,
      "loss": 1.0266,
      "step": 8500
    },
    {
      "epoch": 0.11047553580634865,
      "grad_norm": 2.067218780517578,
      "learning_rate": 4.815874106989419e-05,
      "loss": 1.0379,
      "step": 9000
    },
    {
      "epoch": 0.11661306557336804,
      "grad_norm": 4.303891181945801,
      "learning_rate": 4.8056448907110535e-05,
      "loss": 1.0367,
      "step": 9500
    },
    {
      "epoch": 0.1227505953403874,
      "grad_norm": 5.363577842712402,
      "learning_rate": 4.795415674432688e-05,
      "loss": 1.0384,
      "step": 10000
    },
    {
      "epoch": 0.12888812510740677,
      "grad_norm": 7.8037028312683105,
      "learning_rate": 4.785186458154322e-05,
      "loss": 1.0291,
      "step": 10500
    },
    {
      "epoch": 0.13502565487442614,
      "grad_norm": 2.8838307857513428,
      "learning_rate": 4.774957241875956e-05,
      "loss": 1.0362,
      "step": 11000
    },
    {
      "epoch": 0.1411631846414455,
      "grad_norm": 18.838226318359375,
      "learning_rate": 4.764728025597591e-05,
      "loss": 1.0354,
      "step": 11500
    },
    {
      "epoch": 0.14730071440846487,
      "grad_norm": 1.8900924921035767,
      "learning_rate": 4.754498809319225e-05,
      "loss": 1.0204,
      "step": 12000
    },
    {
      "epoch": 0.15343824417548424,
      "grad_norm": 8.636752128601074,
      "learning_rate": 4.74426959304086e-05,
      "loss": 1.0401,
      "step": 12500
    },
    {
      "epoch": 0.15957577394250363,
      "grad_norm": 3.1361780166625977,
      "learning_rate": 4.734040376762494e-05,
      "loss": 1.0237,
      "step": 13000
    },
    {
      "epoch": 0.165713303709523,
      "grad_norm": 6.990635871887207,
      "learning_rate": 4.7238111604841285e-05,
      "loss": 1.0271,
      "step": 13500
    },
    {
      "epoch": 0.17185083347654237,
      "grad_norm": 2.6320559978485107,
      "learning_rate": 4.713581944205763e-05,
      "loss": 1.0235,
      "step": 14000
    },
    {
      "epoch": 0.17798836324356174,
      "grad_norm": 5.751638889312744,
      "learning_rate": 4.7033527279273976e-05,
      "loss": 1.0329,
      "step": 14500
    },
    {
      "epoch": 0.1841258930105811,
      "grad_norm": 10.333562850952148,
      "learning_rate": 4.693123511649032e-05,
      "loss": 1.0474,
      "step": 15000
    },
    {
      "epoch": 0.19026342277760047,
      "grad_norm": 9.57418155670166,
      "learning_rate": 4.682894295370666e-05,
      "loss": 1.0547,
      "step": 15500
    },
    {
      "epoch": 0.19640095254461984,
      "grad_norm": 6.002443313598633,
      "learning_rate": 4.6726650790923e-05,
      "loss": 1.0557,
      "step": 16000
    },
    {
      "epoch": 0.2025384823116392,
      "grad_norm": 17.77951431274414,
      "learning_rate": 4.6624358628139344e-05,
      "loss": 1.0626,
      "step": 16500
    },
    {
      "epoch": 0.20867601207865857,
      "grad_norm": 10.126110076904297,
      "learning_rate": 4.652206646535569e-05,
      "loss": 1.0551,
      "step": 17000
    },
    {
      "epoch": 0.21481354184567794,
      "grad_norm": 3.7376019954681396,
      "learning_rate": 4.641977430257204e-05,
      "loss": 1.0526,
      "step": 17500
    },
    {
      "epoch": 0.2209510716126973,
      "grad_norm": 19.0529842376709,
      "learning_rate": 4.6317482139788384e-05,
      "loss": 1.125,
      "step": 18000
    },
    {
      "epoch": 0.2270886013797167,
      "grad_norm": 1.9759446382522583,
      "learning_rate": 4.6215189977004726e-05,
      "loss": 1.2996,
      "step": 18500
    },
    {
      "epoch": 0.23322613114673607,
      "grad_norm": 2.7964680194854736,
      "learning_rate": 4.611289781422107e-05,
      "loss": 1.0703,
      "step": 19000
    },
    {
      "epoch": 0.23936366091375544,
      "grad_norm": 18.77906036376953,
      "learning_rate": 4.601060565143741e-05,
      "loss": 1.0601,
      "step": 19500
    },
    {
      "epoch": 0.2455011906807748,
      "grad_norm": 3.129147529602051,
      "learning_rate": 4.590831348865375e-05,
      "loss": 1.0835,
      "step": 20000
    },
    {
      "epoch": 0.2516387204477942,
      "grad_norm": 3.018688917160034,
      "learning_rate": 4.58060213258701e-05,
      "loss": 1.0583,
      "step": 20500
    },
    {
      "epoch": 0.25777625021481354,
      "grad_norm": 32.16483688354492,
      "learning_rate": 4.570372916308644e-05,
      "loss": 1.141,
      "step": 21000
    },
    {
      "epoch": 0.2639137799818329,
      "grad_norm": 8.431964874267578,
      "learning_rate": 4.5601437000302785e-05,
      "loss": 1.0802,
      "step": 21500
    },
    {
      "epoch": 0.2700513097488523,
      "grad_norm": 14.990534782409668,
      "learning_rate": 4.5499144837519134e-05,
      "loss": 1.0589,
      "step": 22000
    },
    {
      "epoch": 0.27618883951587164,
      "grad_norm": 6.134013652801514,
      "learning_rate": 4.5396852674735476e-05,
      "loss": 1.086,
      "step": 22500
    },
    {
      "epoch": 0.282326369282891,
      "grad_norm": 15.587114334106445,
      "learning_rate": 4.529456051195182e-05,
      "loss": 1.0743,
      "step": 23000
    },
    {
      "epoch": 0.2884638990499104,
      "grad_norm": 20.36590003967285,
      "learning_rate": 4.519226834916816e-05,
      "loss": 1.0605,
      "step": 23500
    },
    {
      "epoch": 0.29460142881692974,
      "grad_norm": 6.840169906616211,
      "learning_rate": 4.508997618638451e-05,
      "loss": 1.0602,
      "step": 24000
    },
    {
      "epoch": 0.3007389585839491,
      "grad_norm": 22.10498046875,
      "learning_rate": 4.498768402360085e-05,
      "loss": 1.0732,
      "step": 24500
    },
    {
      "epoch": 0.3068764883509685,
      "grad_norm": 2.3044190406799316,
      "learning_rate": 4.488539186081719e-05,
      "loss": 1.0614,
      "step": 25000
    },
    {
      "epoch": 0.31301401811798785,
      "grad_norm": 19.358158111572266,
      "learning_rate": 4.4783099698033535e-05,
      "loss": 1.0666,
      "step": 25500
    },
    {
      "epoch": 0.31915154788500727,
      "grad_norm": 47.77769088745117,
      "learning_rate": 4.468080753524988e-05,
      "loss": 1.0669,
      "step": 26000
    },
    {
      "epoch": 0.32528907765202664,
      "grad_norm": 31.26322364807129,
      "learning_rate": 4.4578515372466226e-05,
      "loss": 1.0627,
      "step": 26500
    },
    {
      "epoch": 0.331426607419046,
      "grad_norm": 31.237651824951172,
      "learning_rate": 4.4476223209682575e-05,
      "loss": 1.0659,
      "step": 27000
    },
    {
      "epoch": 0.33756413718606537,
      "grad_norm": 10.651726722717285,
      "learning_rate": 4.437393104689892e-05,
      "loss": 1.07,
      "step": 27500
    },
    {
      "epoch": 0.34370166695308474,
      "grad_norm": 10.767593383789062,
      "learning_rate": 4.427163888411526e-05,
      "loss": 1.0783,
      "step": 28000
    },
    {
      "epoch": 0.3498391967201041,
      "grad_norm": 13.177048683166504,
      "learning_rate": 4.41693467213316e-05,
      "loss": 1.078,
      "step": 28500
    },
    {
      "epoch": 0.3559767264871235,
      "grad_norm": 41.34538269042969,
      "learning_rate": 4.406705455854794e-05,
      "loss": 1.0832,
      "step": 29000
    },
    {
      "epoch": 0.36211425625414284,
      "grad_norm": 2.813678026199341,
      "learning_rate": 4.3964762395764285e-05,
      "loss": 1.0695,
      "step": 29500
    },
    {
      "epoch": 0.3682517860211622,
      "grad_norm": 32.63833236694336,
      "learning_rate": 4.3862470232980634e-05,
      "loss": 1.0714,
      "step": 30000
    },
    {
      "epoch": 0.3743893157881816,
      "grad_norm": 6.144928932189941,
      "learning_rate": 4.3760178070196976e-05,
      "loss": 1.3794,
      "step": 30500
    },
    {
      "epoch": 0.38052684555520094,
      "grad_norm": 21.691205978393555,
      "learning_rate": 4.365788590741332e-05,
      "loss": 1.0838,
      "step": 31000
    },
    {
      "epoch": 0.3866643753222203,
      "grad_norm": 5.830935955047607,
      "learning_rate": 4.355559374462967e-05,
      "loss": 1.0541,
      "step": 31500
    },
    {
      "epoch": 0.3928019050892397,
      "grad_norm": 5.59116792678833,
      "learning_rate": 4.345330158184601e-05,
      "loss": 1.0799,
      "step": 32000
    },
    {
      "epoch": 0.39893943485625905,
      "grad_norm": 5.144099712371826,
      "learning_rate": 4.335100941906235e-05,
      "loss": 1.0622,
      "step": 32500
    },
    {
      "epoch": 0.4050769646232784,
      "grad_norm": 30.603145599365234,
      "learning_rate": 4.324871725627869e-05,
      "loss": 1.0597,
      "step": 33000
    },
    {
      "epoch": 0.4112144943902978,
      "grad_norm": 14.602336883544922,
      "learning_rate": 4.314642509349504e-05,
      "loss": 1.0731,
      "step": 33500
    },
    {
      "epoch": 0.41735202415731715,
      "grad_norm": 11.165698051452637,
      "learning_rate": 4.3044132930711384e-05,
      "loss": 1.0871,
      "step": 34000
    },
    {
      "epoch": 0.4234895539243365,
      "grad_norm": 7.897899627685547,
      "learning_rate": 4.2941840767927726e-05,
      "loss": 1.0535,
      "step": 34500
    },
    {
      "epoch": 0.4296270836913559,
      "grad_norm": 34.061134338378906,
      "learning_rate": 4.283954860514407e-05,
      "loss": 1.3215,
      "step": 35000
    },
    {
      "epoch": 0.43576461345837525,
      "grad_norm": 5.050889015197754,
      "learning_rate": 4.273725644236041e-05,
      "loss": 1.1501,
      "step": 35500
    },
    {
      "epoch": 0.4419021432253946,
      "grad_norm": 11.23651123046875,
      "learning_rate": 4.263496427957676e-05,
      "loss": 1.1088,
      "step": 36000
    },
    {
      "epoch": 0.44803967299241404,
      "grad_norm": 11.963942527770996,
      "learning_rate": 4.25326721167931e-05,
      "loss": 1.0844,
      "step": 36500
    },
    {
      "epoch": 0.4541772027594334,
      "grad_norm": 19.694747924804688,
      "learning_rate": 4.243037995400945e-05,
      "loss": 1.0981,
      "step": 37000
    },
    {
      "epoch": 0.4603147325264528,
      "grad_norm": 32.396446228027344,
      "learning_rate": 4.232808779122579e-05,
      "loss": 1.0856,
      "step": 37500
    },
    {
      "epoch": 0.46645226229347214,
      "grad_norm": 30.794092178344727,
      "learning_rate": 4.2225795628442134e-05,
      "loss": 1.0956,
      "step": 38000
    },
    {
      "epoch": 0.4725897920604915,
      "grad_norm": 25.20745086669922,
      "learning_rate": 4.2123503465658476e-05,
      "loss": 1.0813,
      "step": 38500
    },
    {
      "epoch": 0.4787273218275109,
      "grad_norm": 13.46246337890625,
      "learning_rate": 4.202121130287482e-05,
      "loss": 1.0843,
      "step": 39000
    },
    {
      "epoch": 0.48486485159453024,
      "grad_norm": 9.27540397644043,
      "learning_rate": 4.191891914009117e-05,
      "loss": 1.1053,
      "step": 39500
    },
    {
      "epoch": 0.4910023813615496,
      "grad_norm": 3.5075111389160156,
      "learning_rate": 4.181662697730751e-05,
      "loss": 1.092,
      "step": 40000
    },
    {
      "epoch": 0.497139911128569,
      "grad_norm": 6.219689846038818,
      "learning_rate": 4.171433481452385e-05,
      "loss": 1.081,
      "step": 40500
    },
    {
      "epoch": 0.5032774408955883,
      "grad_norm": 16.255443572998047,
      "learning_rate": 4.161204265174019e-05,
      "loss": 1.0919,
      "step": 41000
    },
    {
      "epoch": 0.5094149706626078,
      "grad_norm": 32.27680206298828,
      "learning_rate": 4.150975048895654e-05,
      "loss": 1.0681,
      "step": 41500
    },
    {
      "epoch": 0.5155525004296271,
      "grad_norm": 107.78765869140625,
      "learning_rate": 4.1407458326172884e-05,
      "loss": 1.0826,
      "step": 42000
    },
    {
      "epoch": 0.5216900301966465,
      "grad_norm": 8.029512405395508,
      "learning_rate": 4.1305166163389226e-05,
      "loss": 1.0809,
      "step": 42500
    },
    {
      "epoch": 0.5278275599636658,
      "grad_norm": 44.91604232788086,
      "learning_rate": 4.1202874000605575e-05,
      "loss": 1.093,
      "step": 43000
    },
    {
      "epoch": 0.5339650897306852,
      "grad_norm": 11.860628128051758,
      "learning_rate": 4.110058183782192e-05,
      "loss": 1.0709,
      "step": 43500
    },
    {
      "epoch": 0.5401026194977046,
      "grad_norm": 30.72344398498535,
      "learning_rate": 4.099828967503826e-05,
      "loss": 1.0623,
      "step": 44000
    },
    {
      "epoch": 0.546240149264724,
      "grad_norm": 19.160911560058594,
      "learning_rate": 4.08959975122546e-05,
      "loss": 1.0801,
      "step": 44500
    },
    {
      "epoch": 0.5523776790317433,
      "grad_norm": 12.648110389709473,
      "learning_rate": 4.079370534947094e-05,
      "loss": 1.0833,
      "step": 45000
    },
    {
      "epoch": 0.5585152087987627,
      "grad_norm": 25.488170623779297,
      "learning_rate": 4.0691413186687285e-05,
      "loss": 1.0816,
      "step": 45500
    },
    {
      "epoch": 0.564652738565782,
      "grad_norm": 8.982975006103516,
      "learning_rate": 4.0589121023903634e-05,
      "loss": 1.0697,
      "step": 46000
    },
    {
      "epoch": 0.5707902683328014,
      "grad_norm": 12.776642799377441,
      "learning_rate": 4.048682886111998e-05,
      "loss": 1.0864,
      "step": 46500
    },
    {
      "epoch": 0.5769277980998208,
      "grad_norm": 70.14471435546875,
      "learning_rate": 4.0384536698336325e-05,
      "loss": 1.0744,
      "step": 47000
    },
    {
      "epoch": 0.5830653278668402,
      "grad_norm": 26.72732162475586,
      "learning_rate": 4.028224453555267e-05,
      "loss": 1.0897,
      "step": 47500
    },
    {
      "epoch": 0.5892028576338595,
      "grad_norm": 22.69961166381836,
      "learning_rate": 4.017995237276901e-05,
      "loss": 1.0727,
      "step": 48000
    },
    {
      "epoch": 0.5953403874008789,
      "grad_norm": 6.76914119720459,
      "learning_rate": 4.007766020998535e-05,
      "loss": 1.0621,
      "step": 48500
    },
    {
      "epoch": 0.6014779171678982,
      "grad_norm": 41.32302474975586,
      "learning_rate": 3.99753680472017e-05,
      "loss": 1.0582,
      "step": 49000
    },
    {
      "epoch": 0.6076154469349176,
      "grad_norm": 47.626644134521484,
      "learning_rate": 3.987307588441804e-05,
      "loss": 1.0681,
      "step": 49500
    },
    {
      "epoch": 0.613752976701937,
      "grad_norm": 27.24641227722168,
      "learning_rate": 3.9770783721634384e-05,
      "loss": 1.083,
      "step": 50000
    },
    {
      "epoch": 0.6198905064689564,
      "grad_norm": 22.27688217163086,
      "learning_rate": 3.9668491558850726e-05,
      "loss": 1.0632,
      "step": 50500
    },
    {
      "epoch": 0.6260280362359757,
      "grad_norm": 43.488407135009766,
      "learning_rate": 3.9566199396067075e-05,
      "loss": 1.0809,
      "step": 51000
    },
    {
      "epoch": 0.6321655660029951,
      "grad_norm": 8.742218017578125,
      "learning_rate": 3.946390723328342e-05,
      "loss": 1.0525,
      "step": 51500
    },
    {
      "epoch": 0.6383030957700145,
      "grad_norm": 35.785621643066406,
      "learning_rate": 3.936161507049976e-05,
      "loss": 1.6649,
      "step": 52000
    },
    {
      "epoch": 0.6444406255370339,
      "grad_norm": 17.245643615722656,
      "learning_rate": 3.925932290771611e-05,
      "loss": 1.0873,
      "step": 52500
    },
    {
      "epoch": 0.6505781553040533,
      "grad_norm": 11.474211692810059,
      "learning_rate": 3.915703074493245e-05,
      "loss": 1.0649,
      "step": 53000
    },
    {
      "epoch": 0.6567156850710726,
      "grad_norm": 57.05070114135742,
      "learning_rate": 3.905473858214879e-05,
      "loss": 1.0591,
      "step": 53500
    },
    {
      "epoch": 0.662853214838092,
      "grad_norm": 28.446794509887695,
      "learning_rate": 3.8952446419365134e-05,
      "loss": 1.061,
      "step": 54000
    },
    {
      "epoch": 0.6689907446051113,
      "grad_norm": 34.050048828125,
      "learning_rate": 3.8850154256581476e-05,
      "loss": 1.0682,
      "step": 54500
    },
    {
      "epoch": 0.6751282743721307,
      "grad_norm": 40.37173843383789,
      "learning_rate": 3.874786209379782e-05,
      "loss": 1.0616,
      "step": 55000
    },
    {
      "epoch": 0.68126580413915,
      "grad_norm": 17.34925079345703,
      "learning_rate": 3.864556993101417e-05,
      "loss": 1.0827,
      "step": 55500
    },
    {
      "epoch": 0.6874033339061695,
      "grad_norm": 57.942806243896484,
      "learning_rate": 3.8543277768230516e-05,
      "loss": 1.0799,
      "step": 56000
    },
    {
      "epoch": 0.6935408636731888,
      "grad_norm": 27.28282928466797,
      "learning_rate": 3.844098560544686e-05,
      "loss": 1.0705,
      "step": 56500
    },
    {
      "epoch": 0.6996783934402082,
      "grad_norm": 34.14970016479492,
      "learning_rate": 3.83386934426632e-05,
      "loss": 1.0821,
      "step": 57000
    },
    {
      "epoch": 0.7058159232072275,
      "grad_norm": 24.639619827270508,
      "learning_rate": 3.823640127987954e-05,
      "loss": 1.0763,
      "step": 57500
    },
    {
      "epoch": 0.711953452974247,
      "grad_norm": 45.23350524902344,
      "learning_rate": 3.8134109117095884e-05,
      "loss": 1.0842,
      "step": 58000
    },
    {
      "epoch": 0.7180909827412663,
      "grad_norm": 10.940398216247559,
      "learning_rate": 3.803181695431223e-05,
      "loss": 1.0813,
      "step": 58500
    },
    {
      "epoch": 0.7242285125082857,
      "grad_norm": 95.62271881103516,
      "learning_rate": 3.7929524791528575e-05,
      "loss": 1.0678,
      "step": 59000
    },
    {
      "epoch": 0.730366042275305,
      "grad_norm": 39.37797164916992,
      "learning_rate": 3.782723262874492e-05,
      "loss": 1.0894,
      "step": 59500
    },
    {
      "epoch": 0.7365035720423244,
      "grad_norm": 14.900856971740723,
      "learning_rate": 3.772494046596126e-05,
      "loss": 1.0733,
      "step": 60000
    },
    {
      "epoch": 0.7426411018093437,
      "grad_norm": 10.382057189941406,
      "learning_rate": 3.762264830317761e-05,
      "loss": 1.0688,
      "step": 60500
    },
    {
      "epoch": 0.7487786315763632,
      "grad_norm": 7.631415367126465,
      "learning_rate": 3.752035614039395e-05,
      "loss": 1.0665,
      "step": 61000
    },
    {
      "epoch": 0.7549161613433825,
      "grad_norm": 38.29065704345703,
      "learning_rate": 3.74180639776103e-05,
      "loss": 1.0681,
      "step": 61500
    },
    {
      "epoch": 0.7610536911104019,
      "grad_norm": 44.250144958496094,
      "learning_rate": 3.731577181482664e-05,
      "loss": 1.301,
      "step": 62000
    },
    {
      "epoch": 0.7671912208774213,
      "grad_norm": 3.7524375915527344,
      "learning_rate": 3.721347965204298e-05,
      "loss": 1.0907,
      "step": 62500
    },
    {
      "epoch": 0.7733287506444406,
      "grad_norm": 41.10980987548828,
      "learning_rate": 3.7111187489259325e-05,
      "loss": 1.3941,
      "step": 63000
    },
    {
      "epoch": 0.77946628041146,
      "grad_norm": 70.5405044555664,
      "learning_rate": 3.700889532647567e-05,
      "loss": 1.1355,
      "step": 63500
    },
    {
      "epoch": 0.7856038101784794,
      "grad_norm": 84.906005859375,
      "learning_rate": 3.690660316369201e-05,
      "loss": 1.2196,
      "step": 64000
    },
    {
      "epoch": 0.7917413399454988,
      "grad_norm": 86.1518325805664,
      "learning_rate": 3.680431100090835e-05,
      "loss": 1.225,
      "step": 64500
    },
    {
      "epoch": 0.7978788697125181,
      "grad_norm": 15.497425079345703,
      "learning_rate": 3.67020188381247e-05,
      "loss": 1.1437,
      "step": 65000
    },
    {
      "epoch": 0.8040163994795375,
      "grad_norm": 77.67127227783203,
      "learning_rate": 3.659972667534105e-05,
      "loss": 1.0787,
      "step": 65500
    },
    {
      "epoch": 0.8101539292465568,
      "grad_norm": 54.82986068725586,
      "learning_rate": 3.649743451255739e-05,
      "loss": 1.0686,
      "step": 66000
    },
    {
      "epoch": 0.8162914590135762,
      "grad_norm": 16.840795516967773,
      "learning_rate": 3.639514234977373e-05,
      "loss": 1.1017,
      "step": 66500
    },
    {
      "epoch": 0.8224289887805956,
      "grad_norm": 9.104973793029785,
      "learning_rate": 3.6292850186990075e-05,
      "loss": 1.0999,
      "step": 67000
    },
    {
      "epoch": 0.828566518547615,
      "grad_norm": 21.70631980895996,
      "learning_rate": 3.6190558024206417e-05,
      "loss": 1.0771,
      "step": 67500
    },
    {
      "epoch": 0.8347040483146343,
      "grad_norm": 26.37040901184082,
      "learning_rate": 3.6088265861422765e-05,
      "loss": 1.1015,
      "step": 68000
    },
    {
      "epoch": 0.8408415780816537,
      "grad_norm": 17.264562606811523,
      "learning_rate": 3.598597369863911e-05,
      "loss": 1.0826,
      "step": 68500
    },
    {
      "epoch": 0.846979107848673,
      "grad_norm": 46.9713249206543,
      "learning_rate": 3.588368153585545e-05,
      "loss": 1.1046,
      "step": 69000
    },
    {
      "epoch": 0.8531166376156925,
      "grad_norm": 21.621822357177734,
      "learning_rate": 3.578138937307179e-05,
      "loss": 1.0735,
      "step": 69500
    },
    {
      "epoch": 0.8592541673827118,
      "grad_norm": 23.59171485900879,
      "learning_rate": 3.567909721028814e-05,
      "loss": 1.1056,
      "step": 70000
    },
    {
      "epoch": 0.8653916971497312,
      "grad_norm": 52.57392883300781,
      "learning_rate": 3.557680504750448e-05,
      "loss": 1.1077,
      "step": 70500
    },
    {
      "epoch": 0.8715292269167505,
      "grad_norm": 4.959003448486328,
      "learning_rate": 3.547451288472083e-05,
      "loss": 1.0897,
      "step": 71000
    },
    {
      "epoch": 0.8776667566837699,
      "grad_norm": 10.618963241577148,
      "learning_rate": 3.537222072193717e-05,
      "loss": 1.0775,
      "step": 71500
    },
    {
      "epoch": 0.8838042864507892,
      "grad_norm": 34.424346923828125,
      "learning_rate": 3.5269928559153515e-05,
      "loss": 1.0907,
      "step": 72000
    },
    {
      "epoch": 0.8899418162178087,
      "grad_norm": 16.29790496826172,
      "learning_rate": 3.516763639636986e-05,
      "loss": 1.0705,
      "step": 72500
    },
    {
      "epoch": 0.8960793459848281,
      "grad_norm": 39.85658645629883,
      "learning_rate": 3.50653442335862e-05,
      "loss": 1.0776,
      "step": 73000
    },
    {
      "epoch": 0.9022168757518474,
      "grad_norm": 42.42219161987305,
      "learning_rate": 3.496305207080254e-05,
      "loss": 1.0896,
      "step": 73500
    },
    {
      "epoch": 0.9083544055188668,
      "grad_norm": 67.0084228515625,
      "learning_rate": 3.4860759908018884e-05,
      "loss": 1.0889,
      "step": 74000
    },
    {
      "epoch": 0.9144919352858861,
      "grad_norm": 39.18657684326172,
      "learning_rate": 3.475846774523523e-05,
      "loss": 1.0804,
      "step": 74500
    },
    {
      "epoch": 0.9206294650529055,
      "grad_norm": 76.6010513305664,
      "learning_rate": 3.4656175582451574e-05,
      "loss": 1.1087,
      "step": 75000
    },
    {
      "epoch": 0.9267669948199249,
      "grad_norm": 7.220675945281982,
      "learning_rate": 3.455388341966792e-05,
      "loss": 1.0738,
      "step": 75500
    },
    {
      "epoch": 0.9329045245869443,
      "grad_norm": 32.431453704833984,
      "learning_rate": 3.4451591256884265e-05,
      "loss": 1.0806,
      "step": 76000
    },
    {
      "epoch": 0.9390420543539636,
      "grad_norm": 89.43970489501953,
      "learning_rate": 3.434929909410061e-05,
      "loss": 1.0899,
      "step": 76500
    },
    {
      "epoch": 0.945179584120983,
      "grad_norm": 25.720674514770508,
      "learning_rate": 3.424700693131695e-05,
      "loss": 1.0814,
      "step": 77000
    },
    {
      "epoch": 0.9513171138880023,
      "grad_norm": 44.011478424072266,
      "learning_rate": 3.41447147685333e-05,
      "loss": 1.0723,
      "step": 77500
    },
    {
      "epoch": 0.9574546436550218,
      "grad_norm": 67.16317749023438,
      "learning_rate": 3.404242260574964e-05,
      "loss": 1.0923,
      "step": 78000
    },
    {
      "epoch": 0.9635921734220411,
      "grad_norm": 116.56925201416016,
      "learning_rate": 3.394013044296598e-05,
      "loss": 1.0748,
      "step": 78500
    },
    {
      "epoch": 0.9697297031890605,
      "grad_norm": 99.30931091308594,
      "learning_rate": 3.3837838280182324e-05,
      "loss": 1.0745,
      "step": 79000
    },
    {
      "epoch": 0.9758672329560798,
      "grad_norm": 60.88938522338867,
      "learning_rate": 3.3735546117398667e-05,
      "loss": 1.081,
      "step": 79500
    },
    {
      "epoch": 0.9820047627230992,
      "grad_norm": 38.99333572387695,
      "learning_rate": 3.3633253954615015e-05,
      "loss": 1.08,
      "step": 80000
    },
    {
      "epoch": 0.9881422924901185,
      "grad_norm": 125.93304443359375,
      "learning_rate": 3.3530961791831364e-05,
      "loss": 1.0905,
      "step": 80500
    },
    {
      "epoch": 0.994279822257138,
      "grad_norm": 39.0662841796875,
      "learning_rate": 3.3428669629047706e-05,
      "loss": 1.1064,
      "step": 81000
    },
    {
      "epoch": 1.0004173520241573,
      "grad_norm": 105.9404296875,
      "learning_rate": 3.332637746626405e-05,
      "loss": 1.0929,
      "step": 81500
    },
    {
      "epoch": 1.0065548817911767,
      "grad_norm": 74.30028533935547,
      "learning_rate": 3.322408530348039e-05,
      "loss": 1.0815,
      "step": 82000
    },
    {
      "epoch": 1.0126924115581961,
      "grad_norm": 14.825672149658203,
      "learning_rate": 3.312179314069673e-05,
      "loss": 1.0768,
      "step": 82500
    },
    {
      "epoch": 1.0188299413252153,
      "grad_norm": 58.09488296508789,
      "learning_rate": 3.3019500977913074e-05,
      "loss": 1.1004,
      "step": 83000
    },
    {
      "epoch": 1.0249674710922347,
      "grad_norm": 33.13391876220703,
      "learning_rate": 3.2917208815129416e-05,
      "loss": 1.0835,
      "step": 83500
    },
    {
      "epoch": 1.0311050008592542,
      "grad_norm": 23.691511154174805,
      "learning_rate": 3.2814916652345765e-05,
      "loss": 1.0765,
      "step": 84000
    },
    {
      "epoch": 1.0372425306262736,
      "grad_norm": 57.42161178588867,
      "learning_rate": 3.271262448956211e-05,
      "loss": 1.0906,
      "step": 84500
    },
    {
      "epoch": 1.043380060393293,
      "grad_norm": 52.550350189208984,
      "learning_rate": 3.2610332326778456e-05,
      "loss": 1.096,
      "step": 85000
    },
    {
      "epoch": 1.0495175901603122,
      "grad_norm": 89.49881744384766,
      "learning_rate": 3.25080401639948e-05,
      "loss": 1.1085,
      "step": 85500
    },
    {
      "epoch": 1.0556551199273316,
      "grad_norm": 11.250481605529785,
      "learning_rate": 3.240574800121114e-05,
      "loss": 1.0913,
      "step": 86000
    },
    {
      "epoch": 1.061792649694351,
      "grad_norm": 31.0516357421875,
      "learning_rate": 3.230345583842748e-05,
      "loss": 1.0974,
      "step": 86500
    },
    {
      "epoch": 1.0679301794613705,
      "grad_norm": 94.23751068115234,
      "learning_rate": 3.220116367564383e-05,
      "loss": 1.0763,
      "step": 87000
    },
    {
      "epoch": 1.0740677092283897,
      "grad_norm": 26.741689682006836,
      "learning_rate": 3.209887151286017e-05,
      "loss": 1.0894,
      "step": 87500
    },
    {
      "epoch": 1.080205238995409,
      "grad_norm": 39.82439422607422,
      "learning_rate": 3.1996579350076515e-05,
      "loss": 1.1065,
      "step": 88000
    },
    {
      "epoch": 1.0863427687624285,
      "grad_norm": 7.871860027313232,
      "learning_rate": 3.189428718729286e-05,
      "loss": 1.0898,
      "step": 88500
    },
    {
      "epoch": 1.092480298529448,
      "grad_norm": 35.93474578857422,
      "learning_rate": 3.17919950245092e-05,
      "loss": 1.1036,
      "step": 89000
    },
    {
      "epoch": 1.0986178282964671,
      "grad_norm": 146.91851806640625,
      "learning_rate": 3.168970286172555e-05,
      "loss": 1.0825,
      "step": 89500
    },
    {
      "epoch": 1.1047553580634866,
      "grad_norm": 187.8535614013672,
      "learning_rate": 3.15874106989419e-05,
      "loss": 1.1164,
      "step": 90000
    },
    {
      "epoch": 1.110892887830506,
      "grad_norm": 77.92440795898438,
      "learning_rate": 3.148511853615824e-05,
      "loss": 1.0811,
      "step": 90500
    },
    {
      "epoch": 1.1170304175975254,
      "grad_norm": 62.233360290527344,
      "learning_rate": 3.138282637337458e-05,
      "loss": 1.115,
      "step": 91000
    },
    {
      "epoch": 1.1231679473645446,
      "grad_norm": 54.17692947387695,
      "learning_rate": 3.128053421059092e-05,
      "loss": 1.1123,
      "step": 91500
    },
    {
      "epoch": 1.129305477131564,
      "grad_norm": 49.207454681396484,
      "learning_rate": 3.1178242047807265e-05,
      "loss": 1.0846,
      "step": 92000
    },
    {
      "epoch": 1.1354430068985835,
      "grad_norm": 49.67575454711914,
      "learning_rate": 3.107594988502361e-05,
      "loss": 1.0917,
      "step": 92500
    },
    {
      "epoch": 1.1415805366656029,
      "grad_norm": 154.799072265625,
      "learning_rate": 3.0973657722239956e-05,
      "loss": 1.0879,
      "step": 93000
    },
    {
      "epoch": 1.1477180664326223,
      "grad_norm": 146.50338745117188,
      "learning_rate": 3.08713655594563e-05,
      "loss": 1.0969,
      "step": 93500
    },
    {
      "epoch": 1.1538555961996415,
      "grad_norm": 155.8907928466797,
      "learning_rate": 3.076907339667264e-05,
      "loss": 1.0849,
      "step": 94000
    },
    {
      "epoch": 1.159993125966661,
      "grad_norm": 26.285480499267578,
      "learning_rate": 3.066678123388899e-05,
      "loss": 1.0852,
      "step": 94500
    },
    {
      "epoch": 1.1661306557336804,
      "grad_norm": 31.642507553100586,
      "learning_rate": 3.056448907110533e-05,
      "loss": 1.0774,
      "step": 95000
    },
    {
      "epoch": 1.1722681855006996,
      "grad_norm": 19.160383224487305,
      "learning_rate": 3.0462196908321673e-05,
      "loss": 1.0984,
      "step": 95500
    },
    {
      "epoch": 1.178405715267719,
      "grad_norm": 38.4587516784668,
      "learning_rate": 3.0359904745538015e-05,
      "loss": 1.0936,
      "step": 96000
    },
    {
      "epoch": 1.1845432450347384,
      "grad_norm": 54.26088333129883,
      "learning_rate": 3.0257612582754364e-05,
      "loss": 1.0842,
      "step": 96500
    },
    {
      "epoch": 1.1906807748017578,
      "grad_norm": 22.72319221496582,
      "learning_rate": 3.0155320419970706e-05,
      "loss": 1.0895,
      "step": 97000
    },
    {
      "epoch": 1.1968183045687772,
      "grad_norm": 88.04389190673828,
      "learning_rate": 3.0053028257187048e-05,
      "loss": 1.09,
      "step": 97500
    },
    {
      "epoch": 1.2029558343357964,
      "grad_norm": 49.89582061767578,
      "learning_rate": 2.9950736094403394e-05,
      "loss": 1.0813,
      "step": 98000
    },
    {
      "epoch": 1.2090933641028159,
      "grad_norm": 107.11421966552734,
      "learning_rate": 2.9848443931619736e-05,
      "loss": 1.0965,
      "step": 98500
    },
    {
      "epoch": 1.2152308938698353,
      "grad_norm": 9.076912879943848,
      "learning_rate": 2.9746151768836078e-05,
      "loss": 1.1017,
      "step": 99000
    },
    {
      "epoch": 1.2213684236368547,
      "grad_norm": 64.24601745605469,
      "learning_rate": 2.9643859606052427e-05,
      "loss": 1.1045,
      "step": 99500
    },
    {
      "epoch": 1.227505953403874,
      "grad_norm": 38.064544677734375,
      "learning_rate": 2.954156744326877e-05,
      "loss": 1.093,
      "step": 100000
    },
    {
      "epoch": 1.2336434831708933,
      "grad_norm": 98.80583190917969,
      "learning_rate": 2.9439275280485114e-05,
      "loss": 1.0903,
      "step": 100500
    }
  ],
  "logging_steps": 500,
  "max_steps": 244398,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}