{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9993650793650795,
  "eval_steps": 500,
  "global_step": 3543,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008465608465608466,
      "grad_norm": 1.2471813174095268,
      "learning_rate": 9.999803440244217e-05,
      "loss": 1.6288,
      "step": 10
    },
    {
      "epoch": 0.016931216931216932,
      "grad_norm": 1.0863274390486097,
      "learning_rate": 9.999213776431165e-05,
      "loss": 0.9235,
      "step": 20
    },
    {
      "epoch": 0.025396825396825397,
      "grad_norm": 0.8727777966082182,
      "learning_rate": 9.998231054922511e-05,
      "loss": 0.76,
      "step": 30
    },
    {
      "epoch": 0.033862433862433865,
      "grad_norm": 0.9216466406898077,
      "learning_rate": 9.996855352983658e-05,
      "loss": 0.6727,
      "step": 40
    },
    {
      "epoch": 0.042328042328042326,
      "grad_norm": 0.8685947217114767,
      "learning_rate": 9.995086778777658e-05,
      "loss": 0.6129,
      "step": 50
    },
    {
      "epoch": 0.050793650793650794,
      "grad_norm": 1.2232366822709322,
      "learning_rate": 9.99292547135672e-05,
      "loss": 0.6334,
      "step": 60
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 1.0539472852477354,
      "learning_rate": 9.990371600651263e-05,
      "loss": 0.6105,
      "step": 70
    },
    {
      "epoch": 0.06772486772486773,
      "grad_norm": 0.9891412672344401,
      "learning_rate": 9.987425367456573e-05,
      "loss": 0.5657,
      "step": 80
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 1.2412242369215105,
      "learning_rate": 9.984087003416997e-05,
      "loss": 0.5737,
      "step": 90
    },
    {
      "epoch": 0.08465608465608465,
      "grad_norm": 1.040677823054172,
      "learning_rate": 9.980356771007745e-05,
      "loss": 0.59,
      "step": 100
    },
    {
      "epoch": 0.09312169312169312,
      "grad_norm": 1.2535421463288563,
      "learning_rate": 9.976234963514244e-05,
      "loss": 0.537,
      "step": 110
    },
    {
      "epoch": 0.10158730158730159,
      "grad_norm": 1.0340076526320245,
      "learning_rate": 9.971721905009086e-05,
      "loss": 0.5238,
      "step": 120
    },
    {
      "epoch": 0.11005291005291006,
      "grad_norm": 0.9971098275169218,
      "learning_rate": 9.966817950326541e-05,
      "loss": 0.5225,
      "step": 130
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 1.227890883155069,
      "learning_rate": 9.961523485034662e-05,
      "loss": 0.5357,
      "step": 140
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 1.0605213954707176,
      "learning_rate": 9.955838925404971e-05,
      "loss": 0.5363,
      "step": 150
    },
    {
      "epoch": 0.13544973544973546,
      "grad_norm": 0.9916561072816364,
      "learning_rate": 9.94976471837973e-05,
      "loss": 0.5107,
      "step": 160
    },
    {
      "epoch": 0.1439153439153439,
      "grad_norm": 0.9224406793350302,
      "learning_rate": 9.943301341536799e-05,
      "loss": 0.5291,
      "step": 170
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 0.9459998168670802,
      "learning_rate": 9.936449303052086e-05,
      "loss": 0.4909,
      "step": 180
    },
    {
      "epoch": 0.16084656084656085,
      "grad_norm": 1.1437589163268378,
      "learning_rate": 9.929209141659598e-05,
      "loss": 0.5365,
      "step": 190
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 1.0253073070364553,
      "learning_rate": 9.921581426609074e-05,
      "loss": 0.5505,
      "step": 200
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 1.2156932145478005,
      "learning_rate": 9.91356675762124e-05,
      "loss": 0.5099,
      "step": 210
    },
    {
      "epoch": 0.18624338624338624,
      "grad_norm": 1.0357139504839994,
      "learning_rate": 9.905165764840645e-05,
      "loss": 0.5016,
      "step": 220
    },
    {
      "epoch": 0.19470899470899472,
      "grad_norm": 1.0513703237178957,
      "learning_rate": 9.896379108786126e-05,
      "loss": 0.502,
      "step": 230
    },
    {
      "epoch": 0.20317460317460317,
      "grad_norm": 1.0533509834595176,
      "learning_rate": 9.88720748029887e-05,
      "loss": 0.4945,
      "step": 240
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 1.0842886505846092,
      "learning_rate": 9.877651600488099e-05,
      "loss": 0.4765,
      "step": 250
    },
    {
      "epoch": 0.2201058201058201,
      "grad_norm": 1.0079186405663672,
      "learning_rate": 9.867712220674374e-05,
      "loss": 0.4842,
      "step": 260
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.9843739521940763,
      "learning_rate": 9.857390122330522e-05,
      "loss": 0.461,
      "step": 270
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 1.0682125354651388,
      "learning_rate": 9.846686117020196e-05,
      "loss": 0.4867,
      "step": 280
    },
    {
      "epoch": 0.2455026455026455,
      "grad_norm": 0.9969252730804404,
      "learning_rate": 9.835601046334065e-05,
      "loss": 0.4765,
      "step": 290
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 1.145271346749039,
      "learning_rate": 9.82413578182364e-05,
      "loss": 0.5131,
      "step": 300
    },
    {
      "epoch": 0.2624338624338624,
      "grad_norm": 1.0296375563736684,
      "learning_rate": 9.81229122493276e-05,
      "loss": 0.4762,
      "step": 310
    },
    {
      "epoch": 0.2708994708994709,
      "grad_norm": 1.0014721034964151,
      "learning_rate": 9.800068306926708e-05,
      "loss": 0.4663,
      "step": 320
    },
    {
      "epoch": 0.27936507936507937,
      "grad_norm": 1.0076225606653653,
      "learning_rate": 9.787467988818997e-05,
      "loss": 0.4918,
      "step": 330
    },
    {
      "epoch": 0.2878306878306878,
      "grad_norm": 1.1916493242294524,
      "learning_rate": 9.774491261295807e-05,
      "loss": 0.4712,
      "step": 340
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 1.0123926586182874,
      "learning_rate": 9.761139144638094e-05,
      "loss": 0.4648,
      "step": 350
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 1.0273059230601556,
      "learning_rate": 9.747412688641372e-05,
      "loss": 0.4956,
      "step": 360
    },
    {
      "epoch": 0.31322751322751324,
      "grad_norm": 1.1785090929345263,
      "learning_rate": 9.733312972533182e-05,
      "loss": 0.4788,
      "step": 370
    },
    {
      "epoch": 0.3216931216931217,
      "grad_norm": 1.0144172117749461,
      "learning_rate": 9.71884110488822e-05,
      "loss": 0.4596,
      "step": 380
    },
    {
      "epoch": 0.33015873015873015,
      "grad_norm": 1.043899340203269,
      "learning_rate": 9.703998223541196e-05,
      "loss": 0.4579,
      "step": 390
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 1.1486794203887134,
      "learning_rate": 9.688785495497366e-05,
      "loss": 0.4699,
      "step": 400
    },
    {
      "epoch": 0.3470899470899471,
      "grad_norm": 1.0373759945758736,
      "learning_rate": 9.673204116840768e-05,
      "loss": 0.4794,
      "step": 410
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.9675582489645671,
      "learning_rate": 9.6572553126402e-05,
      "loss": 0.4506,
      "step": 420
    },
    {
      "epoch": 0.364021164021164,
      "grad_norm": 0.9254293000510532,
      "learning_rate": 9.640940336852884e-05,
      "loss": 0.4633,
      "step": 430
    },
    {
      "epoch": 0.3724867724867725,
      "grad_norm": 1.0325559808636224,
      "learning_rate": 9.624260472225882e-05,
      "loss": 0.45,
      "step": 440
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 1.012802385726723,
      "learning_rate": 9.607217030195242e-05,
      "loss": 0.4618,
      "step": 450
    },
    {
      "epoch": 0.38941798941798944,
      "grad_norm": 1.0486189015875456,
      "learning_rate": 9.589811350782884e-05,
      "loss": 0.4694,
      "step": 460
    },
    {
      "epoch": 0.3978835978835979,
      "grad_norm": 1.1012030940322182,
      "learning_rate": 9.572044802491247e-05,
      "loss": 0.4505,
      "step": 470
    },
    {
      "epoch": 0.40634920634920635,
      "grad_norm": 1.0060674054310905,
      "learning_rate": 9.553918782195688e-05,
      "loss": 0.452,
      "step": 480
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 1.022772188842258,
      "learning_rate": 9.535434715034656e-05,
      "loss": 0.4356,
      "step": 490
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 0.9036986631596517,
      "learning_rate": 9.516594054297642e-05,
      "loss": 0.454,
      "step": 500
    },
    {
      "epoch": 0.43174603174603177,
      "grad_norm": 0.8998291431871417,
      "learning_rate": 9.497398281310914e-05,
      "loss": 0.4162,
      "step": 510
    },
    {
      "epoch": 0.4402116402116402,
      "grad_norm": 0.9267883794464391,
      "learning_rate": 9.477848905321055e-05,
      "loss": 0.453,
      "step": 520
    },
    {
      "epoch": 0.4486772486772487,
      "grad_norm": 1.0322711884681597,
      "learning_rate": 9.45794746337629e-05,
      "loss": 0.4505,
      "step": 530
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 1.0203838607248878,
      "learning_rate": 9.43769552020565e-05,
      "loss": 0.4384,
      "step": 540
    },
    {
      "epoch": 0.4656084656084656,
      "grad_norm": 0.9881433507860394,
      "learning_rate": 9.417094668095932e-05,
      "loss": 0.4606,
      "step": 550
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 1.0302659809674581,
      "learning_rate": 9.396146526766522e-05,
      "loss": 0.4285,
      "step": 560
    },
    {
      "epoch": 0.48253968253968255,
      "grad_norm": 1.0511136894758872,
      "learning_rate": 9.374852743242037e-05,
      "loss": 0.4643,
      "step": 570
    },
    {
      "epoch": 0.491005291005291,
      "grad_norm": 0.9837494678922593,
      "learning_rate": 9.353214991722834e-05,
      "loss": 0.4461,
      "step": 580
    },
    {
      "epoch": 0.49947089947089945,
      "grad_norm": 0.9920985093408109,
      "learning_rate": 9.331234973453375e-05,
      "loss": 0.4655,
      "step": 590
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 0.980959216204584,
      "learning_rate": 9.308914416588468e-05,
      "loss": 0.4155,
      "step": 600
    },
    {
      "epoch": 0.5164021164021164,
      "grad_norm": 1.031774153957894,
      "learning_rate": 9.286255076057395e-05,
      "loss": 0.4213,
      "step": 610
    },
    {
      "epoch": 0.5248677248677248,
      "grad_norm": 0.8674398091480106,
      "learning_rate": 9.263258733425933e-05,
      "loss": 0.4076,
      "step": 620
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.0196562220963516,
      "learning_rate": 9.239927196756279e-05,
      "loss": 0.4496,
      "step": 630
    },
    {
      "epoch": 0.5417989417989418,
      "grad_norm": 1.015803047506094,
      "learning_rate": 9.216262300464893e-05,
      "loss": 0.4399,
      "step": 640
    },
    {
      "epoch": 0.5502645502645502,
      "grad_norm": 0.8444200783279879,
      "learning_rate": 9.192265905178269e-05,
      "loss": 0.4394,
      "step": 650
    },
    {
      "epoch": 0.5587301587301587,
      "grad_norm": 0.9377213246241002,
      "learning_rate": 9.167939897586647e-05,
      "loss": 0.4483,
      "step": 660
    },
    {
      "epoch": 0.5671957671957671,
      "grad_norm": 1.059524915742627,
      "learning_rate": 9.143286190295668e-05,
      "loss": 0.4307,
      "step": 670
    },
    {
      "epoch": 0.5756613756613757,
      "grad_norm": 0.923761212995823,
      "learning_rate": 9.118306721676008e-05,
      "loss": 0.4508,
      "step": 680
    },
    {
      "epoch": 0.5841269841269842,
      "grad_norm": 1.0801213897332445,
      "learning_rate": 9.09300345571097e-05,
      "loss": 0.4454,
      "step": 690
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 1.0572026584991792,
      "learning_rate": 9.067378381842063e-05,
      "loss": 0.4056,
      "step": 700
    },
    {
      "epoch": 0.6010582010582011,
      "grad_norm": 0.887247483674217,
      "learning_rate": 9.041433514812589e-05,
      "loss": 0.4565,
      "step": 710
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 0.9437825913047283,
      "learning_rate": 9.015170894509244e-05,
      "loss": 0.3808,
      "step": 720
    },
    {
      "epoch": 0.617989417989418,
      "grad_norm": 0.9193237876699505,
      "learning_rate": 8.988592585801719e-05,
      "loss": 0.4397,
      "step": 730
    },
    {
      "epoch": 0.6264550264550265,
      "grad_norm": 1.0095223700805926,
      "learning_rate": 8.961700678380359e-05,
      "loss": 0.4667,
      "step": 740
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 1.0449169193411385,
      "learning_rate": 8.93449728659187e-05,
      "loss": 0.4269,
      "step": 750
    },
    {
      "epoch": 0.6433862433862434,
      "grad_norm": 0.980016186986963,
      "learning_rate": 8.906984549273069e-05,
      "loss": 0.4324,
      "step": 760
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 0.9073046321616872,
      "learning_rate": 8.879164629582726e-05,
      "loss": 0.3978,
      "step": 770
    },
    {
      "epoch": 0.6603174603174603,
      "grad_norm": 1.0375673558788747,
      "learning_rate": 8.851039714831492e-05,
      "loss": 0.4076,
      "step": 780
    },
    {
      "epoch": 0.6687830687830688,
      "grad_norm": 0.9857310889370984,
      "learning_rate": 8.822612016309915e-05,
      "loss": 0.4167,
      "step": 790
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 0.8471996503361908,
      "learning_rate": 8.793883769114586e-05,
      "loss": 0.4043,
      "step": 800
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.8920445010354429,
      "learning_rate": 8.764857231972408e-05,
      "loss": 0.4277,
      "step": 810
    },
    {
      "epoch": 0.6941798941798942,
      "grad_norm": 1.0104461104380098,
      "learning_rate": 8.735534687062999e-05,
      "loss": 0.4141,
      "step": 820
    },
    {
      "epoch": 0.7026455026455026,
      "grad_norm": 0.9774326975200812,
      "learning_rate": 8.70591843983927e-05,
      "loss": 0.4257,
      "step": 830
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.8665516757336554,
      "learning_rate": 8.676010818846145e-05,
      "loss": 0.4399,
      "step": 840
    },
    {
      "epoch": 0.7195767195767195,
      "grad_norm": 1.0277004789770665,
      "learning_rate": 8.645814175537497e-05,
      "loss": 0.4514,
      "step": 850
    },
    {
      "epoch": 0.728042328042328,
      "grad_norm": 0.9204427078381576,
      "learning_rate": 8.615330884091261e-05,
      "loss": 0.4144,
      "step": 860
    },
    {
      "epoch": 0.7365079365079366,
      "grad_norm": 1.1301407306313882,
      "learning_rate": 8.584563341222765e-05,
      "loss": 0.3885,
      "step": 870
    },
    {
      "epoch": 0.744973544973545,
      "grad_norm": 0.9458002020904854,
      "learning_rate": 8.553513965996293e-05,
      "loss": 0.4365,
      "step": 880
    },
    {
      "epoch": 0.7534391534391535,
      "grad_norm": 0.9889479454065517,
      "learning_rate": 8.52218519963489e-05,
      "loss": 0.4229,
      "step": 890
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.850438656019796,
      "learning_rate": 8.490579505328424e-05,
      "loss": 0.4306,
      "step": 900
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 0.9370585479712853,
      "learning_rate": 8.458699368039913e-05,
      "loss": 0.4362,
      "step": 910
    },
    {
      "epoch": 0.7788359788359789,
      "grad_norm": 0.9677664011225651,
      "learning_rate": 8.426547294310159e-05,
      "loss": 0.4025,
      "step": 920
    },
    {
      "epoch": 0.7873015873015873,
      "grad_norm": 1.029715838177075,
      "learning_rate": 8.394125812060666e-05,
      "loss": 0.4093,
      "step": 930
    },
    {
      "epoch": 0.7957671957671958,
      "grad_norm": 0.9344248674021697,
      "learning_rate": 8.361437470394889e-05,
      "loss": 0.4039,
      "step": 940
    },
    {
      "epoch": 0.8042328042328042,
      "grad_norm": 0.9458006999291949,
      "learning_rate": 8.328484839397807e-05,
      "loss": 0.3985,
      "step": 950
    },
    {
      "epoch": 0.8126984126984127,
      "grad_norm": 0.9459246300643047,
      "learning_rate": 8.295270509933862e-05,
      "loss": 0.4092,
      "step": 960
    },
    {
      "epoch": 0.8211640211640212,
      "grad_norm": 1.0060543461216613,
      "learning_rate": 8.261797093443249e-05,
      "loss": 0.395,
      "step": 970
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 1.1118301896881613,
      "learning_rate": 8.228067221736595e-05,
      "loss": 0.3968,
      "step": 980
    },
    {
      "epoch": 0.8380952380952381,
      "grad_norm": 0.9258014283812922,
      "learning_rate": 8.19408354678804e-05,
      "loss": 0.4025,
      "step": 990
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 0.9611975674417222,
      "learning_rate": 8.159848740526723e-05,
      "loss": 0.4189,
      "step": 1000
    },
    {
      "epoch": 0.855026455026455,
      "grad_norm": 1.0106358106141422,
      "learning_rate": 8.125365494626706e-05,
      "loss": 0.4062,
      "step": 1010
    },
    {
      "epoch": 0.8634920634920635,
      "grad_norm": 1.0337925818103473,
      "learning_rate": 8.090636520295348e-05,
      "loss": 0.4064,
      "step": 1020
    },
    {
      "epoch": 0.8719576719576719,
      "grad_norm": 0.9347719089761312,
      "learning_rate": 8.055664548060131e-05,
      "loss": 0.4085,
      "step": 1030
    },
    {
      "epoch": 0.8804232804232804,
      "grad_norm": 0.9001085371362909,
      "learning_rate": 8.020452327553986e-05,
      "loss": 0.4206,
      "step": 1040
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.9005574040109541,
      "learning_rate": 7.9850026272991e-05,
      "loss": 0.4139,
      "step": 1050
    },
    {
      "epoch": 0.8973544973544973,
      "grad_norm": 0.9668702597620379,
      "learning_rate": 7.94931823448924e-05,
      "loss": 0.4152,
      "step": 1060
    },
    {
      "epoch": 0.9058201058201059,
      "grad_norm": 1.055087550179992,
      "learning_rate": 7.913401954770622e-05,
      "loss": 0.3971,
      "step": 1070
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.8974328370738465,
      "learning_rate": 7.877256612021312e-05,
      "loss": 0.4147,
      "step": 1080
    },
    {
      "epoch": 0.9227513227513228,
      "grad_norm": 1.0826599341716459,
      "learning_rate": 7.84088504812921e-05,
      "loss": 0.4012,
      "step": 1090
    },
    {
      "epoch": 0.9312169312169312,
      "grad_norm": 0.9939329938043986,
      "learning_rate": 7.804290122768601e-05,
      "loss": 0.3925,
      "step": 1100
    },
    {
      "epoch": 0.9396825396825397,
      "grad_norm": 0.8740614733530233,
      "learning_rate": 7.767474713175321e-05,
      "loss": 0.4102,
      "step": 1110
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 0.9603955711346013,
      "learning_rate": 7.73044171392054e-05,
      "loss": 0.4065,
      "step": 1120
    },
    {
      "epoch": 0.9566137566137566,
      "grad_norm": 1.127743100139354,
      "learning_rate": 7.69319403668317e-05,
      "loss": 0.3827,
      "step": 1130
    },
    {
      "epoch": 0.9650793650793651,
      "grad_norm": 0.860392518744168,
      "learning_rate": 7.655734610020952e-05,
      "loss": 0.4088,
      "step": 1140
    },
    {
      "epoch": 0.9735449735449735,
      "grad_norm": 0.9182302824948408,
      "learning_rate": 7.618066379140186e-05,
      "loss": 0.4154,
      "step": 1150
    },
    {
      "epoch": 0.982010582010582,
      "grad_norm": 0.9472264649293427,
      "learning_rate": 7.580192305664174e-05,
      "loss": 0.3986,
      "step": 1160
    },
    {
      "epoch": 0.9904761904761905,
      "grad_norm": 0.8736825227580656,
      "learning_rate": 7.542115367400376e-05,
      "loss": 0.3929,
      "step": 1170
    },
    {
      "epoch": 0.9989417989417989,
      "grad_norm": 0.9809536270766679,
      "learning_rate": 7.503838558106263e-05,
      "loss": 0.3868,
      "step": 1180
    },
    {
      "epoch": 1.0074074074074073,
      "grad_norm": 1.0636167972989774,
      "learning_rate": 7.465364887253949e-05,
      "loss": 0.3725,
      "step": 1190
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 1.107441272645184,
      "learning_rate": 7.426697379793572e-05,
      "loss": 0.3436,
      "step": 1200
    },
    {
      "epoch": 1.0243386243386243,
      "grad_norm": 0.9993424372015514,
      "learning_rate": 7.387839075915465e-05,
      "loss": 0.3434,
      "step": 1210
    },
    {
      "epoch": 1.0328042328042328,
      "grad_norm": 0.9466656285727851,
      "learning_rate": 7.348793030811111e-05,
      "loss": 0.3333,
      "step": 1220
    },
    {
      "epoch": 1.0412698412698413,
      "grad_norm": 0.9712036510252812,
      "learning_rate": 7.309562314432952e-05,
      "loss": 0.3205,
      "step": 1230
    },
    {
      "epoch": 1.0497354497354496,
      "grad_norm": 0.9567265246539378,
      "learning_rate": 7.270150011252994e-05,
      "loss": 0.3426,
      "step": 1240
    },
    {
      "epoch": 1.0582010582010581,
      "grad_norm": 1.1511281895276615,
      "learning_rate": 7.230559220020319e-05,
      "loss": 0.3197,
      "step": 1250
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 1.143315468276855,
      "learning_rate": 7.190793053517424e-05,
      "loss": 0.3317,
      "step": 1260
    },
    {
      "epoch": 1.0751322751322752,
      "grad_norm": 0.9958926341045374,
      "learning_rate": 7.1508546383155e-05,
      "loss": 0.3264,
      "step": 1270
    },
    {
      "epoch": 1.0835978835978837,
      "grad_norm": 1.028371010948524,
      "learning_rate": 7.110747114528604e-05,
      "loss": 0.3294,
      "step": 1280
    },
    {
      "epoch": 1.0920634920634922,
      "grad_norm": 1.0914872277776548,
      "learning_rate": 7.070473635566768e-05,
      "loss": 0.3272,
      "step": 1290
    },
    {
      "epoch": 1.1005291005291005,
      "grad_norm": 1.127102783013863,
      "learning_rate": 7.030037367888069e-05,
      "loss": 0.3382,
      "step": 1300
    },
    {
      "epoch": 1.108994708994709,
      "grad_norm": 0.9437414070048694,
      "learning_rate": 6.989441490749662e-05,
      "loss": 0.313,
      "step": 1310
    },
    {
      "epoch": 1.1174603174603175,
      "grad_norm": 1.018312148913929,
      "learning_rate": 6.948689195957829e-05,
      "loss": 0.3261,
      "step": 1320
    },
    {
      "epoch": 1.125925925925926,
      "grad_norm": 1.1453502643848026,
      "learning_rate": 6.907783687617017e-05,
      "loss": 0.3295,
      "step": 1330
    },
    {
      "epoch": 1.1343915343915345,
      "grad_norm": 1.0896304992556385,
      "learning_rate": 6.866728181877914e-05,
      "loss": 0.3208,
      "step": 1340
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 1.0673296804602648,
      "learning_rate": 6.825525906684593e-05,
      "loss": 0.2993,
      "step": 1350
    },
    {
      "epoch": 1.1513227513227513,
      "grad_norm": 1.155110295228161,
      "learning_rate": 6.784180101520717e-05,
      "loss": 0.3519,
      "step": 1360
    },
    {
      "epoch": 1.1597883597883598,
      "grad_norm": 1.0507064287253836,
      "learning_rate": 6.74269401715483e-05,
      "loss": 0.3468,
      "step": 1370
    },
    {
      "epoch": 1.1682539682539683,
      "grad_norm": 1.1767438102467995,
      "learning_rate": 6.701070915384776e-05,
      "loss": 0.3666,
      "step": 1380
    },
    {
      "epoch": 1.1767195767195768,
      "grad_norm": 1.1040760205989266,
      "learning_rate": 6.659314068781241e-05,
      "loss": 0.3516,
      "step": 1390
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 1.1334475282531191,
      "learning_rate": 6.617426760430457e-05,
      "loss": 0.3444,
      "step": 1400
    },
    {
      "epoch": 1.1936507936507936,
      "grad_norm": 1.020666210262446,
      "learning_rate": 6.575412283676063e-05,
      "loss": 0.3246,
      "step": 1410
    },
    {
      "epoch": 1.2021164021164021,
      "grad_norm": 1.088794196150716,
      "learning_rate": 6.533273941860174e-05,
      "loss": 0.3406,
      "step": 1420
    },
    {
      "epoch": 1.2105820105820106,
      "grad_norm": 1.0604023713287243,
      "learning_rate": 6.491015048063662e-05,
      "loss": 0.3383,
      "step": 1430
    },
    {
      "epoch": 1.2190476190476192,
      "grad_norm": 1.0596416270343272,
      "learning_rate": 6.448638924845662e-05,
      "loss": 0.3217,
      "step": 1440
    },
    {
      "epoch": 1.2275132275132274,
      "grad_norm": 1.1101059810936256,
      "learning_rate": 6.40614890398235e-05,
      "loss": 0.3481,
      "step": 1450
    },
    {
      "epoch": 1.235978835978836,
      "grad_norm": 1.0321989665703173,
      "learning_rate": 6.36354832620497e-05,
      "loss": 0.3313,
      "step": 1460
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 1.076200733585246,
      "learning_rate": 6.320840540937196e-05,
      "loss": 0.3286,
      "step": 1470
    },
    {
      "epoch": 1.252910052910053,
      "grad_norm": 1.2096711574512293,
      "learning_rate": 6.278028906031757e-05,
      "loss": 0.337,
      "step": 1480
    },
    {
      "epoch": 1.2613756613756615,
      "grad_norm": 1.10111033029835,
      "learning_rate": 6.235116787506458e-05,
      "loss": 0.3373,
      "step": 1490
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 1.0054036677530953,
      "learning_rate": 6.192107559279513e-05,
      "loss": 0.344,
      "step": 1500
    },
    {
      "epoch": 1.2783068783068783,
      "grad_norm": 1.1848267908504544,
      "learning_rate": 6.14900460290428e-05,
      "loss": 0.3273,
      "step": 1510
    },
    {
      "epoch": 1.2867724867724868,
      "grad_norm": 1.1024985637947646,
      "learning_rate": 6.105811307303391e-05,
      "loss": 0.3283,
      "step": 1520
    },
    {
      "epoch": 1.2952380952380953,
      "grad_norm": 0.9670540378850452,
      "learning_rate": 6.0625310685023006e-05,
      "loss": 0.3365,
      "step": 1530
    },
    {
      "epoch": 1.3037037037037038,
      "grad_norm": 1.1427343254939735,
      "learning_rate": 6.0191672893622766e-05,
      "loss": 0.3634,
      "step": 1540
    },
    {
      "epoch": 1.312169312169312,
      "grad_norm": 1.1927363160896671,
      "learning_rate": 5.975723379312851e-05,
      "loss": 0.3269,
      "step": 1550
    },
    {
      "epoch": 1.3206349206349206,
      "grad_norm": 1.0661933157446577,
      "learning_rate": 5.932202754083765e-05,
      "loss": 0.3635,
      "step": 1560
    },
    {
      "epoch": 1.3291005291005291,
      "grad_norm": 1.080671677917908,
      "learning_rate": 5.888608835436408e-05,
      "loss": 0.3288,
      "step": 1570
    },
    {
      "epoch": 1.3375661375661376,
      "grad_norm": 1.0117605430718117,
      "learning_rate": 5.844945050894779e-05,
      "loss": 0.3225,
      "step": 1580
    },
    {
      "epoch": 1.3460317460317461,
      "grad_norm": 0.9733421931133834,
      "learning_rate": 5.8012148334760077e-05,
      "loss": 0.3146,
      "step": 1590
    },
    {
      "epoch": 1.3544973544973544,
      "grad_norm": 1.0677601655228577,
      "learning_rate": 5.757421621420438e-05,
      "loss": 0.323,
      "step": 1600
    },
    {
      "epoch": 1.362962962962963,
      "grad_norm": 1.027062473997525,
      "learning_rate": 5.713568857921296e-05,
      "loss": 0.3071,
      "step": 1610
    },
    {
      "epoch": 1.3714285714285714,
      "grad_norm": 0.979546471375574,
      "learning_rate": 5.669659990853975e-05,
      "loss": 0.3317,
      "step": 1620
    },
    {
      "epoch": 1.37989417989418,
      "grad_norm": 1.065648765600161,
      "learning_rate": 5.62569847250495e-05,
      "loss": 0.3109,
      "step": 1630
    },
    {
      "epoch": 1.3883597883597885,
      "grad_norm": 0.9904468465024019,
      "learning_rate": 5.581687759300344e-05,
      "loss": 0.3295,
      "step": 1640
    },
    {
      "epoch": 1.3968253968253967,
      "grad_norm": 1.032660358149486,
      "learning_rate": 5.537631311534176e-05,
      "loss": 0.3112,
      "step": 1650
    },
    {
      "epoch": 1.4052910052910053,
      "grad_norm": 1.2430136826049512,
      "learning_rate": 5.493532593096289e-05,
      "loss": 0.3449,
      "step": 1660
    },
    {
      "epoch": 1.4137566137566138,
      "grad_norm": 1.1898277112027362,
      "learning_rate": 5.449395071200014e-05,
      "loss": 0.3341,
      "step": 1670
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 1.043179777209729,
      "learning_rate": 5.40522221610956e-05,
      "loss": 0.336,
      "step": 1680
    },
    {
      "epoch": 1.4306878306878308,
      "grad_norm": 1.0248865486416825,
      "learning_rate": 5.3610175008671736e-05,
      "loss": 0.3238,
      "step": 1690
    },
    {
      "epoch": 1.439153439153439,
      "grad_norm": 1.1689685076211649,
      "learning_rate": 5.316784401020064e-05,
      "loss": 0.346,
      "step": 1700
    },
    {
      "epoch": 1.4476190476190476,
      "grad_norm": 0.911540237077825,
      "learning_rate": 5.272526394347156e-05,
      "loss": 0.3282,
      "step": 1710
    },
    {
      "epoch": 1.456084656084656,
      "grad_norm": 1.0990655046574285,
      "learning_rate": 5.228246960585639e-05,
      "loss": 0.3362,
      "step": 1720
    },
    {
      "epoch": 1.4645502645502646,
      "grad_norm": 1.0636049790842568,
      "learning_rate": 5.18394958115739e-05,
      "loss": 0.3326,
      "step": 1730
    },
    {
      "epoch": 1.4730158730158731,
      "grad_norm": 0.9983006726426711,
      "learning_rate": 5.139637738895243e-05,
      "loss": 0.3277,
      "step": 1740
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 1.1542167670907033,
      "learning_rate": 5.0953149177691526e-05,
      "loss": 0.3094,
      "step": 1750
    },
    {
      "epoch": 1.48994708994709,
      "grad_norm": 1.1340593143811428,
      "learning_rate": 5.0509846026122796e-05,
      "loss": 0.3299,
      "step": 1760
    },
    {
      "epoch": 1.4984126984126984,
      "grad_norm": 1.058208878715113,
      "learning_rate": 5.006650278846991e-05,
      "loss": 0.3034,
      "step": 1770
    },
    {
      "epoch": 1.506878306878307,
      "grad_norm": 1.0744521042723432,
      "learning_rate": 4.962315432210827e-05,
      "loss": 0.3426,
      "step": 1780
    },
    {
      "epoch": 1.5153439153439154,
      "grad_norm": 1.1542452796166005,
      "learning_rate": 4.917983548482439e-05,
      "loss": 0.3218,
      "step": 1790
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 1.0168410109864015,
      "learning_rate": 4.873658113207526e-05,
      "loss": 0.3349,
      "step": 1800
    },
    {
      "epoch": 1.5322751322751322,
      "grad_norm": 0.9972037732726962,
      "learning_rate": 4.8293426114247755e-05,
      "loss": 0.3172,
      "step": 1810
    },
    {
      "epoch": 1.5407407407407407,
      "grad_norm": 1.0573820943940695,
      "learning_rate": 4.785040527391876e-05,
      "loss": 0.337,
      "step": 1820
    },
    {
      "epoch": 1.5492063492063493,
      "grad_norm": 0.9881874822329866,
      "learning_rate": 4.740755344311549e-05,
      "loss": 0.3067,
      "step": 1830
    },
    {
      "epoch": 1.5576719576719578,
      "grad_norm": 1.1012178266159593,
      "learning_rate": 4.6964905440577074e-05,
      "loss": 0.3361,
      "step": 1840
    },
    {
      "epoch": 1.566137566137566,
      "grad_norm": 1.1008138761037878,
      "learning_rate": 4.6522496069016823e-05,
      "loss": 0.3143,
      "step": 1850
    },
    {
      "epoch": 1.5746031746031746,
      "grad_norm": 1.0000855307005652,
      "learning_rate": 4.6080360112385914e-05,
      "loss": 0.3268,
      "step": 1860
    },
    {
      "epoch": 1.583068783068783,
      "grad_norm": 1.0946371145172427,
      "learning_rate": 4.5638532333138656e-05,
      "loss": 0.324,
      "step": 1870
    },
    {
      "epoch": 1.5915343915343916,
      "grad_norm": 1.1665682442710945,
      "learning_rate": 4.519704746949915e-05,
      "loss": 0.3056,
      "step": 1880
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.042537667391268,
      "learning_rate": 4.475594023273024e-05,
      "loss": 0.3228,
      "step": 1890
    },
    {
      "epoch": 1.6084656084656084,
      "grad_norm": 1.2077526566227794,
      "learning_rate": 4.4315245304404175e-05,
      "loss": 0.3219,
      "step": 1900
    },
    {
      "epoch": 1.6169312169312169,
      "grad_norm": 0.9726841884389219,
      "learning_rate": 4.3874997333676e-05,
      "loss": 0.3071,
      "step": 1910
    },
    {
      "epoch": 1.6253968253968254,
      "grad_norm": 1.0872803982697865,
      "learning_rate": 4.343523093455909e-05,
      "loss": 0.3169,
      "step": 1920
    },
    {
      "epoch": 1.633862433862434,
      "grad_norm": 1.094914483331633,
      "learning_rate": 4.2995980683203905e-05,
      "loss": 0.3152,
      "step": 1930
    },
    {
      "epoch": 1.6423280423280424,
      "grad_norm": 1.0181256444534053,
      "learning_rate": 4.255728111517929e-05,
      "loss": 0.3403,
      "step": 1940
    },
    {
      "epoch": 1.6507936507936507,
      "grad_norm": 1.126842585732799,
      "learning_rate": 4.211916672275722e-05,
      "loss": 0.3132,
      "step": 1950
    },
    {
      "epoch": 1.6592592592592592,
      "grad_norm": 0.9511685681942271,
      "learning_rate": 4.168167195220088e-05,
      "loss": 0.299,
      "step": 1960
    },
    {
      "epoch": 1.6677248677248677,
      "grad_norm": 0.9727568749103337,
      "learning_rate": 4.1244831201056375e-05,
      "loss": 0.3059,
      "step": 1970
    },
    {
      "epoch": 1.6761904761904762,
      "grad_norm": 1.1382165054479474,
      "learning_rate": 4.080867881544826e-05,
      "loss": 0.318,
      "step": 1980
    },
    {
      "epoch": 1.6846560846560847,
      "grad_norm": 1.0381544841427086,
      "learning_rate": 4.037324908737909e-05,
      "loss": 0.3076,
      "step": 1990
    },
    {
      "epoch": 1.693121693121693,
      "grad_norm": 1.1334587465684427,
      "learning_rate": 3.9938576252033256e-05,
      "loss": 0.3052,
      "step": 2000
    },
    {
      "epoch": 1.7015873015873015,
      "grad_norm": 0.9774012866079483,
      "learning_rate": 3.950469448508532e-05,
      "loss": 0.3095,
      "step": 2010
    },
    {
      "epoch": 1.71005291005291,
      "grad_norm": 1.0840882454796814,
      "learning_rate": 3.907163790001292e-05,
      "loss": 0.3246,
      "step": 2020
    },
    {
      "epoch": 1.7185185185185186,
      "grad_norm": 1.1578703078806882,
      "learning_rate": 3.863944054541474e-05,
      "loss": 0.3148,
      "step": 2030
    },
    {
      "epoch": 1.726984126984127,
      "grad_norm": 1.0210773839195724,
      "learning_rate": 3.8208136402333314e-05,
      "loss": 0.3256,
      "step": 2040
    },
    {
      "epoch": 1.7354497354497354,
      "grad_norm": 1.0794889187437258,
      "learning_rate": 3.7777759381583495e-05,
      "loss": 0.2897,
      "step": 2050
    },
    {
      "epoch": 1.7439153439153439,
      "grad_norm": 1.0452290154108226,
      "learning_rate": 3.73483433210861e-05,
      "loss": 0.3051,
      "step": 2060
    },
    {
      "epoch": 1.7523809523809524,
      "grad_norm": 1.0972215102884366,
      "learning_rate": 3.691992198320753e-05,
      "loss": 0.3165,
      "step": 2070
    },
    {
      "epoch": 1.7608465608465609,
      "grad_norm": 1.102108276028283,
      "learning_rate": 3.649252905210519e-05,
      "loss": 0.2911,
      "step": 2080
    },
    {
      "epoch": 1.7693121693121694,
      "grad_norm": 1.0290079519183777,
      "learning_rate": 3.6066198131079146e-05,
      "loss": 0.3129,
      "step": 2090
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 1.0920319523740303,
      "learning_rate": 3.564096273993012e-05,
      "loss": 0.313,
      "step": 2100
    },
    {
      "epoch": 1.7862433862433862,
      "grad_norm": 1.2412824592232703,
      "learning_rate": 3.521685631232392e-05,
      "loss": 0.3202,
      "step": 2110
    },
    {
      "epoch": 1.7947089947089947,
      "grad_norm": 0.9652001472992224,
      "learning_rate": 3.479391219316291e-05,
      "loss": 0.3058,
      "step": 2120
    },
    {
      "epoch": 1.8031746031746032,
      "grad_norm": 1.2084364290298202,
      "learning_rate": 3.437216363596418e-05,
      "loss": 0.3192,
      "step": 2130
    },
    {
      "epoch": 1.8116402116402117,
      "grad_norm": 1.0200010445985095,
      "learning_rate": 3.395164380024512e-05,
      "loss": 0.3239,
      "step": 2140
    },
    {
      "epoch": 1.82010582010582,
      "grad_norm": 1.1971574618602432,
      "learning_rate": 3.3532385748916164e-05,
      "loss": 0.3229,
      "step": 2150
    },
    {
      "epoch": 1.8285714285714287,
      "grad_norm": 1.1807655262100478,
      "learning_rate": 3.3114422445681425e-05,
      "loss": 0.3136,
      "step": 2160
    },
    {
      "epoch": 1.837037037037037,
      "grad_norm": 1.2349063591796927,
      "learning_rate": 3.269778675244684e-05,
      "loss": 0.2882,
      "step": 2170
    },
    {
      "epoch": 1.8455026455026455,
      "grad_norm": 1.0765110541279164,
      "learning_rate": 3.22825114267364e-05,
      "loss": 0.295,
      "step": 2180
    },
    {
      "epoch": 1.853968253968254,
      "grad_norm": 1.220972473914235,
      "learning_rate": 3.1868629119116794e-05,
      "loss": 0.3176,
      "step": 2190
    },
    {
      "epoch": 1.8624338624338623,
      "grad_norm": 1.1122856433998665,
      "learning_rate": 3.1456172370630124e-05,
      "loss": 0.3324,
      "step": 2200
    },
    {
      "epoch": 1.870899470899471,
      "grad_norm": 1.1724923812151078,
      "learning_rate": 3.10451736102355e-05,
      "loss": 0.3091,
      "step": 2210
    },
    {
      "epoch": 1.8793650793650793,
      "grad_norm": 1.1020637891223475,
      "learning_rate": 3.0635665152259295e-05,
      "loss": 0.2955,
      "step": 2220
    },
    {
      "epoch": 1.8878306878306879,
      "grad_norm": 1.0318878630934867,
      "learning_rate": 3.0227679193854513e-05,
      "loss": 0.3195,
      "step": 2230
    },
    {
      "epoch": 1.8962962962962964,
      "grad_norm": 1.1032921856519728,
      "learning_rate": 2.982124781246932e-05,
      "loss": 0.3039,
      "step": 2240
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 1.131389139262005,
      "learning_rate": 2.9416402963324896e-05,
      "loss": 0.2742,
      "step": 2250
    },
    {
      "epoch": 1.9132275132275134,
      "grad_norm": 1.1298513214593129,
      "learning_rate": 2.9013176476903147e-05,
      "loss": 0.3048,
      "step": 2260
    },
    {
      "epoch": 1.9216931216931217,
      "grad_norm": 1.081531730138643,
      "learning_rate": 2.8611600056443944e-05,
      "loss": 0.3206,
      "step": 2270
    },
    {
      "epoch": 1.9301587301587302,
      "grad_norm": 1.1056344145079804,
      "learning_rate": 2.821170527545254e-05,
      "loss": 0.2952,
      "step": 2280
    },
    {
      "epoch": 1.9386243386243387,
      "grad_norm": 1.144349889481454,
      "learning_rate": 2.7813523575217103e-05,
      "loss": 0.3037,
      "step": 2290
    },
    {
      "epoch": 1.947089947089947,
      "grad_norm": 1.187655971159083,
      "learning_rate": 2.7417086262336777e-05,
      "loss": 0.3054,
      "step": 2300
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 1.1111194236831317,
      "learning_rate": 2.7022424506260113e-05,
      "loss": 0.3161,
      "step": 2310
    },
    {
      "epoch": 1.964021164021164,
      "grad_norm": 1.0599427922691493,
      "learning_rate": 2.662956933683443e-05,
      "loss": 0.3171,
      "step": 2320
    },
    {
      "epoch": 1.9724867724867725,
      "grad_norm": 1.0305400657350645,
      "learning_rate": 2.623855164186625e-05,
      "loss": 0.2773,
      "step": 2330
    },
    {
      "epoch": 1.980952380952381,
      "grad_norm": 1.0485786323318997,
      "learning_rate": 2.5849402164692593e-05,
      "loss": 0.3122,
      "step": 2340
    },
    {
      "epoch": 1.9894179894179893,
      "grad_norm": 1.0781221238793617,
      "learning_rate": 2.546215150176394e-05,
      "loss": 0.3002,
      "step": 2350
    },
    {
      "epoch": 1.997883597883598,
      "grad_norm": 1.1041144624563863,
      "learning_rate": 2.507683010023857e-05,
      "loss": 0.3006,
      "step": 2360
    },
    {
      "epoch": 2.0063492063492063,
      "grad_norm": 1.0896577752739438,
      "learning_rate": 2.4693468255588752e-05,
      "loss": 0.2681,
      "step": 2370
    },
    {
      "epoch": 2.0148148148148146,
      "grad_norm": 1.0383913047696638,
      "learning_rate": 2.4312096109218684e-05,
      "loss": 0.2509,
      "step": 2380
    },
    {
      "epoch": 2.0232804232804233,
      "grad_norm": 1.1401361260342018,
      "learning_rate": 2.3932743646094736e-05,
      "loss": 0.2495,
      "step": 2390
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 1.327049946404094,
      "learning_rate": 2.3555440692387947e-05,
      "loss": 0.2394,
      "step": 2400
    },
    {
      "epoch": 2.0402116402116404,
      "grad_norm": 1.1859201205267982,
      "learning_rate": 2.3180216913128872e-05,
      "loss": 0.2344,
      "step": 2410
    },
    {
      "epoch": 2.0486772486772487,
      "grad_norm": 1.1188101435575566,
      "learning_rate": 2.2807101809875275e-05,
      "loss": 0.243,
      "step": 2420
    },
    {
      "epoch": 2.057142857142857,
      "grad_norm": 1.2283454742009063,
      "learning_rate": 2.2436124718392565e-05,
      "loss": 0.248,
      "step": 2430
    },
    {
      "epoch": 2.0656084656084657,
      "grad_norm": 1.2592754589750133,
      "learning_rate": 2.206731480634739e-05,
      "loss": 0.2434,
      "step": 2440
    },
    {
      "epoch": 2.074074074074074,
      "grad_norm": 1.237235140900468,
      "learning_rate": 2.1700701071014222e-05,
      "loss": 0.2417,
      "step": 2450
    },
    {
      "epoch": 2.0825396825396827,
      "grad_norm": 1.1749453984216076,
      "learning_rate": 2.1336312336995572e-05,
      "loss": 0.2402,
      "step": 2460
    },
    {
      "epoch": 2.091005291005291,
      "grad_norm": 1.1524857608371186,
      "learning_rate": 2.0974177253955657e-05,
      "loss": 0.2476,
      "step": 2470
    },
    {
      "epoch": 2.0994708994708993,
      "grad_norm": 1.264906725996053,
      "learning_rate": 2.0614324294367882e-05,
      "loss": 0.242,
      "step": 2480
    },
    {
      "epoch": 2.107936507936508,
      "grad_norm": 1.1340174663623572,
      "learning_rate": 2.025678175127623e-05,
      "loss": 0.247,
      "step": 2490
    },
    {
      "epoch": 2.1164021164021163,
      "grad_norm": 1.2832254402303773,
      "learning_rate": 1.9901577736070598e-05,
      "loss": 0.2418,
      "step": 2500
    },
    {
      "epoch": 2.124867724867725,
      "grad_norm": 1.2771941330999887,
      "learning_rate": 1.954874017627684e-05,
      "loss": 0.2504,
      "step": 2510
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 1.1685751238482767,
      "learning_rate": 1.9198296813360777e-05,
      "loss": 0.2442,
      "step": 2520
    },
    {
      "epoch": 2.1417989417989416,
      "grad_norm": 1.3155782809463616,
      "learning_rate": 1.8850275200547142e-05,
      "loss": 0.2453,
      "step": 2530
    },
    {
      "epoch": 2.1502645502645503,
      "grad_norm": 1.2290797778685365,
      "learning_rate": 1.8504702700653194e-05,
      "loss": 0.2488,
      "step": 2540
    },
    {
      "epoch": 2.1587301587301586,
      "grad_norm": 1.140239798893309,
      "learning_rate": 1.8161606483937433e-05,
      "loss": 0.2377,
      "step": 2550
    },
    {
      "epoch": 2.1671957671957673,
      "grad_norm": 1.2500773194969288,
      "learning_rate": 1.7821013525963316e-05,
      "loss": 0.2571,
      "step": 2560
    },
    {
      "epoch": 2.1756613756613756,
      "grad_norm": 1.2713658534451882,
      "learning_rate": 1.748295060547823e-05,
      "loss": 0.2449,
      "step": 2570
    },
    {
      "epoch": 2.1841269841269844,
      "grad_norm": 1.227362647774226,
      "learning_rate": 1.7147444302308273e-05,
      "loss": 0.2348,
      "step": 2580
    },
    {
      "epoch": 2.1925925925925926,
      "grad_norm": 1.3167422374415048,
      "learning_rate": 1.6814520995268207e-05,
      "loss": 0.2578,
      "step": 2590
    },
    {
      "epoch": 2.201058201058201,
      "grad_norm": 1.2496998676048467,
      "learning_rate": 1.6484206860087653e-05,
      "loss": 0.2358,
      "step": 2600
    },
    {
      "epoch": 2.2095238095238097,
      "grad_norm": 1.3111014362047553,
      "learning_rate": 1.6156527867352845e-05,
      "loss": 0.2288,
      "step": 2610
    },
    {
      "epoch": 2.217989417989418,
      "grad_norm": 1.2159616363632633,
      "learning_rate": 1.5831509780464947e-05,
      "loss": 0.2478,
      "step": 2620
    },
    {
      "epoch": 2.2264550264550262,
      "grad_norm": 1.2868454216177594,
      "learning_rate": 1.5509178153614233e-05,
      "loss": 0.2462,
      "step": 2630
    },
    {
      "epoch": 2.234920634920635,
      "grad_norm": 1.1236506253498473,
      "learning_rate": 1.5189558329771092e-05,
      "loss": 0.2475,
      "step": 2640
    },
    {
      "epoch": 2.2433862433862433,
      "grad_norm": 1.2767351170017573,
      "learning_rate": 1.48726754386933e-05,
      "loss": 0.2577,
      "step": 2650
    },
    {
      "epoch": 2.251851851851852,
      "grad_norm": 1.2742999056295827,
      "learning_rate": 1.4558554394950319e-05,
      "loss": 0.2257,
      "step": 2660
    },
    {
      "epoch": 2.2603174603174603,
      "grad_norm": 1.1388456844003385,
      "learning_rate": 1.4247219895964458e-05,
      "loss": 0.2366,
      "step": 2670
    },
    {
      "epoch": 2.268783068783069,
      "grad_norm": 1.1420407981573506,
      "learning_rate": 1.3938696420068887e-05,
      "loss": 0.246,
      "step": 2680
    },
    {
      "epoch": 2.2772486772486773,
      "grad_norm": 1.2447613234115074,
      "learning_rate": 1.3633008224583276e-05,
      "loss": 0.2362,
      "step": 2690
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 1.2110210828346344,
      "learning_rate": 1.3330179343906424e-05,
      "loss": 0.2271,
      "step": 2700
    },
    {
      "epoch": 2.2941798941798943,
      "grad_norm": 1.1743235820264755,
      "learning_rate": 1.303023358762669e-05,
      "loss": 0.2349,
      "step": 2710
    },
    {
      "epoch": 2.3026455026455026,
      "grad_norm": 1.2896552909609165,
      "learning_rate": 1.27331945386499e-05,
      "loss": 0.2586,
      "step": 2720
    },
    {
      "epoch": 2.311111111111111,
      "grad_norm": 1.192957421343539,
      "learning_rate": 1.2439085551345209e-05,
      "loss": 0.2417,
      "step": 2730
    },
    {
      "epoch": 2.3195767195767196,
      "grad_norm": 1.3450432316093512,
      "learning_rate": 1.2147929749708931e-05,
      "loss": 0.2412,
      "step": 2740
    },
    {
      "epoch": 2.328042328042328,
      "grad_norm": 1.3062883724180023,
      "learning_rate": 1.1859750025546362e-05,
      "loss": 0.235,
      "step": 2750
    },
    {
      "epoch": 2.3365079365079366,
      "grad_norm": 1.1575476428826368,
      "learning_rate": 1.1574569036671978e-05,
      "loss": 0.2429,
      "step": 2760
    },
    {
      "epoch": 2.344973544973545,
      "grad_norm": 1.2906490240628528,
      "learning_rate": 1.129240920512798e-05,
      "loss": 0.2359,
      "step": 2770
    },
    {
      "epoch": 2.3534391534391537,
      "grad_norm": 1.3845442578956821,
      "learning_rate": 1.101329271542143e-05,
      "loss": 0.2382,
      "step": 2780
    },
    {
      "epoch": 2.361904761904762,
      "grad_norm": 1.2463846728950012,
      "learning_rate": 1.0737241512779928e-05,
      "loss": 0.2458,
      "step": 2790
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 1.2537500156225916,
      "learning_rate": 1.046427730142625e-05,
      "loss": 0.2487,
      "step": 2800
    },
    {
      "epoch": 2.378835978835979,
      "grad_norm": 1.3091633501767908,
      "learning_rate": 1.0194421542871918e-05,
      "loss": 0.2369,
      "step": 2810
    },
    {
      "epoch": 2.3873015873015873,
      "grad_norm": 1.1820540371459332,
      "learning_rate": 9.927695454229713e-06,
      "loss": 0.2448,
      "step": 2820
    },
    {
      "epoch": 2.3957671957671955,
      "grad_norm": 1.2731529604104752,
      "learning_rate": 9.664120006545573e-06,
      "loss": 0.2419,
      "step": 2830
    },
    {
      "epoch": 2.4042328042328043,
      "grad_norm": 1.3027428739822569,
      "learning_rate": 9.403715923149742e-06,
      "loss": 0.2551,
      "step": 2840
    },
    {
      "epoch": 2.4126984126984126,
      "grad_norm": 1.3254079829534025,
      "learning_rate": 9.146503678027451e-06,
      "loss": 0.2178,
      "step": 2850
    },
    {
      "epoch": 2.4211640211640213,
      "grad_norm": 1.1613878892535392,
      "learning_rate": 8.892503494209132e-06,
      "loss": 0.227,
      "step": 2860
    },
    {
      "epoch": 2.4296296296296296,
      "grad_norm": 1.0705896579237715,
      "learning_rate": 8.641735342180402e-06,
      "loss": 0.2449,
      "step": 2870
    },
    {
      "epoch": 2.4380952380952383,
      "grad_norm": 1.2593614315530395,
      "learning_rate": 8.394218938311982e-06,
      "loss": 0.2287,
      "step": 2880
    },
    {
      "epoch": 2.4465608465608466,
      "grad_norm": 1.1359699627950683,
      "learning_rate": 8.149973743309414e-06,
      "loss": 0.2488,
      "step": 2890
    },
    {
      "epoch": 2.455026455026455,
      "grad_norm": 1.3447659451396483,
      "learning_rate": 7.909018960683052e-06,
      "loss": 0.2339,
      "step": 2900
    },
    {
      "epoch": 2.4634920634920636,
      "grad_norm": 1.056442260887814,
      "learning_rate": 7.671373535238168e-06,
      "loss": 0.25,
      "step": 2910
    },
    {
      "epoch": 2.471957671957672,
      "grad_norm": 1.240966774372524,
      "learning_rate": 7.437056151585514e-06,
      "loss": 0.2348,
      "step": 2920
    },
    {
      "epoch": 2.48042328042328,
      "grad_norm": 1.1279901312845086,
      "learning_rate": 7.206085232672155e-06,
      "loss": 0.2319,
      "step": 2930
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 1.3442581999609493,
      "learning_rate": 6.978478938333055e-06,
      "loss": 0.23,
      "step": 2940
    },
    {
      "epoch": 2.497354497354497,
      "grad_norm": 1.3087593444833816,
      "learning_rate": 6.75425516386326e-06,
      "loss": 0.2571,
      "step": 2950
    },
    {
      "epoch": 2.505820105820106,
      "grad_norm": 1.2133121259337931,
      "learning_rate": 6.533431538610923e-06,
      "loss": 0.2291,
      "step": 2960
    },
    {
      "epoch": 2.5142857142857142,
      "grad_norm": 1.1919153490283616,
      "learning_rate": 6.3160254245911825e-06,
      "loss": 0.2393,
      "step": 2970
    },
    {
      "epoch": 2.522751322751323,
      "grad_norm": 1.4097002557717668,
      "learning_rate": 6.102053915121092e-06,
      "loss": 0.2595,
      "step": 2980
    },
    {
      "epoch": 2.5312169312169313,
      "grad_norm": 1.3516367010865522,
      "learning_rate": 5.891533833475732e-06,
      "loss": 0.2505,
      "step": 2990
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 1.250969302934421,
      "learning_rate": 5.684481731565433e-06,
      "loss": 0.2415,
      "step": 3000
    },
    {
      "epoch": 2.5481481481481483,
      "grad_norm": 1.136756466811525,
      "learning_rate": 5.48091388863442e-06,
      "loss": 0.2325,
      "step": 3010
    },
    {
      "epoch": 2.5566137566137566,
      "grad_norm": 1.231370156024587,
      "learning_rate": 5.280846309980892e-06,
      "loss": 0.2462,
      "step": 3020
    },
    {
      "epoch": 2.565079365079365,
      "grad_norm": 1.3326225138678371,
      "learning_rate": 5.0842947256986014e-06,
      "loss": 0.2412,
      "step": 3030
    },
    {
      "epoch": 2.5735449735449736,
      "grad_norm": 1.1660313881807405,
      "learning_rate": 4.891274589440148e-06,
      "loss": 0.2448,
      "step": 3040
    },
    {
      "epoch": 2.582010582010582,
      "grad_norm": 1.3191473663837276,
      "learning_rate": 4.7018010772018175e-06,
      "loss": 0.2436,
      "step": 3050
    },
    {
      "epoch": 2.5904761904761906,
      "grad_norm": 1.2540952812205302,
      "learning_rate": 4.515889086130565e-06,
      "loss": 0.2479,
      "step": 3060
    },
    {
      "epoch": 2.598941798941799,
      "grad_norm": 1.2017713792968152,
      "learning_rate": 4.333553233352589e-06,
      "loss": 0.2566,
      "step": 3070
    },
    {
      "epoch": 2.6074074074074076,
      "grad_norm": 1.29908123750953,
      "learning_rate": 4.154807854824194e-06,
      "loss": 0.2602,
      "step": 3080
    },
    {
      "epoch": 2.615873015873016,
      "grad_norm": 1.4243808220545542,
      "learning_rate": 3.979667004204535e-06,
      "loss": 0.2398,
      "step": 3090
    },
    {
      "epoch": 2.624338624338624,
      "grad_norm": 1.3457308926738438,
      "learning_rate": 3.808144451750756e-06,
      "loss": 0.2308,
      "step": 3100
    },
    {
      "epoch": 2.632804232804233,
      "grad_norm": 1.3407484748421956,
      "learning_rate": 3.640253683235262e-06,
      "loss": 0.2511,
      "step": 3110
    },
    {
      "epoch": 2.641269841269841,
      "grad_norm": 1.2093265189235727,
      "learning_rate": 3.4760078988854305e-06,
      "loss": 0.2335,
      "step": 3120
    },
    {
      "epoch": 2.6497354497354495,
      "grad_norm": 1.287221034400132,
      "learning_rate": 3.315420012345777e-06,
      "loss": 0.2149,
      "step": 3130
    },
    {
      "epoch": 2.6582010582010582,
      "grad_norm": 1.191374675667794,
      "learning_rate": 3.1585026496625925e-06,
      "loss": 0.2462,
      "step": 3140
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 1.2788639981032432,
      "learning_rate": 3.005268148291296e-06,
      "loss": 0.2404,
      "step": 3150
    },
    {
      "epoch": 2.6751322751322753,
      "grad_norm": 1.4725222188972884,
      "learning_rate": 2.855728556126319e-06,
      "loss": 0.2345,
      "step": 3160
    },
    {
      "epoch": 2.6835978835978835,
      "grad_norm": 1.1514884763928235,
      "learning_rate": 2.709895630553977e-06,
      "loss": 0.2247,
      "step": 3170
    },
    {
      "epoch": 2.6920634920634923,
      "grad_norm": 1.2744441762954501,
      "learning_rate": 2.5677808375279443e-06,
      "loss": 0.2418,
      "step": 3180
    },
    {
      "epoch": 2.7005291005291006,
      "grad_norm": 1.3232083105521923,
      "learning_rate": 2.42939535066784e-06,
      "loss": 0.2304,
      "step": 3190
    },
    {
      "epoch": 2.708994708994709,
      "grad_norm": 1.270314937268037,
      "learning_rate": 2.294750050380662e-06,
      "loss": 0.2247,
      "step": 3200
    },
    {
      "epoch": 2.7174603174603176,
      "grad_norm": 1.256698656991292,
      "learning_rate": 2.1638555230053315e-06,
      "loss": 0.2406,
      "step": 3210
    },
    {
      "epoch": 2.725925925925926,
      "grad_norm": 1.3193829744509844,
      "learning_rate": 2.036722059980406e-06,
      "loss": 0.2396,
      "step": 3220
    },
    {
      "epoch": 2.734391534391534,
      "grad_norm": 1.4009607878677235,
      "learning_rate": 1.9133596570348468e-06,
      "loss": 0.2329,
      "step": 3230
    },
    {
      "epoch": 2.742857142857143,
      "grad_norm": 1.29045787788227,
      "learning_rate": 1.7937780134021808e-06,
      "loss": 0.2301,
      "step": 3240
    },
    {
      "epoch": 2.751322751322751,
      "grad_norm": 1.2476131982200767,
      "learning_rate": 1.6779865310578623e-06,
      "loss": 0.2284,
      "step": 3250
    },
    {
      "epoch": 2.75978835978836,
      "grad_norm": 1.0670988501341874,
      "learning_rate": 1.5659943139801126e-06,
      "loss": 0.2406,
      "step": 3260
    },
    {
      "epoch": 2.768253968253968,
      "grad_norm": 1.308449437483495,
      "learning_rate": 1.4578101674340372e-06,
      "loss": 0.2332,
      "step": 3270
    },
    {
      "epoch": 2.776719576719577,
      "grad_norm": 1.4177869984435298,
      "learning_rate": 1.3534425972794195e-06,
      "loss": 0.2233,
      "step": 3280
    },
    {
      "epoch": 2.785185185185185,
      "grad_norm": 1.3267608082684277,
      "learning_rate": 1.2528998093019051e-06,
      "loss": 0.2498,
      "step": 3290
    },
    {
      "epoch": 2.7936507936507935,
      "grad_norm": 1.2882825046480746,
      "learning_rate": 1.1561897085678341e-06,
      "loss": 0.2286,
      "step": 3300
    },
    {
      "epoch": 2.8021164021164022,
      "grad_norm": 1.2774964089508063,
      "learning_rate": 1.0633198988027115e-06,
      "loss": 0.2253,
      "step": 3310
    },
    {
      "epoch": 2.8105820105820105,
      "grad_norm": 1.3527960705502262,
      "learning_rate": 9.742976817933847e-07,
      "loss": 0.24,
      "step": 3320
    },
    {
      "epoch": 2.819047619047619,
      "grad_norm": 1.365122676946633,
      "learning_rate": 8.891300568139638e-07,
      "loss": 0.2592,
      "step": 3330
    },
    {
      "epoch": 2.8275132275132275,
      "grad_norm": 1.2427303633572429,
      "learning_rate": 8.078237200754624e-07,
      "loss": 0.2474,
      "step": 3340
    },
    {
      "epoch": 2.835978835978836,
      "grad_norm": 1.144641112857142,
      "learning_rate": 7.30385064199357e-07,
      "loss": 0.2332,
      "step": 3350
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 1.2745498061485745,
      "learning_rate": 6.568201777149729e-07,
      "loss": 0.251,
      "step": 3360
    },
    {
      "epoch": 2.852910052910053,
      "grad_norm": 1.422987862547148,
      "learning_rate": 5.871348445807501e-07,
      "loss": 0.2449,
      "step": 3370
    },
    {
      "epoch": 2.8613756613756616,
      "grad_norm": 1.0332383237986493,
      "learning_rate": 5.213345437295181e-07,
      "loss": 0.2467,
      "step": 3380
    },
    {
      "epoch": 2.86984126984127,
      "grad_norm": 1.4294499006460313,
      "learning_rate": 4.594244486376964e-07,
      "loss": 0.2396,
      "step": 3390
    },
    {
      "epoch": 2.878306878306878,
      "grad_norm": 1.3112777173225585,
      "learning_rate": 4.014094269185642e-07,
      "loss": 0.2386,
      "step": 3400
    },
    {
      "epoch": 2.886772486772487,
      "grad_norm": 1.2628517758813262,
      "learning_rate": 3.472940399395108e-07,
      "loss": 0.2229,
      "step": 3410
    },
    {
      "epoch": 2.895238095238095,
      "grad_norm": 1.235621137865247,
      "learning_rate": 2.9708254246344494e-07,
      "loss": 0.2386,
      "step": 3420
    },
    {
      "epoch": 2.9037037037037035,
      "grad_norm": 1.1595358547316623,
      "learning_rate": 2.5077888231422873e-07,
      "loss": 0.2407,
      "step": 3430
    },
    {
      "epoch": 2.912169312169312,
      "grad_norm": 1.1332523589105532,
      "learning_rate": 2.0838670006632643e-07,
      "loss": 0.224,
      "step": 3440
    },
    {
      "epoch": 2.9206349206349205,
      "grad_norm": 1.431443541686851,
      "learning_rate": 1.699093287585274e-07,
      "loss": 0.2334,
      "step": 3450
    },
    {
      "epoch": 2.929100529100529,
      "grad_norm": 1.2372896845957517,
      "learning_rate": 1.3534979363191701e-07,
      "loss": 0.2471,
      "step": 3460
    },
    {
      "epoch": 2.9375661375661375,
      "grad_norm": 1.296014823220347,
      "learning_rate": 1.0471081189200593e-07,
      "loss": 0.2205,
      "step": 3470
    },
    {
      "epoch": 2.9460317460317462,
      "grad_norm": 1.2004684728241466,
      "learning_rate": 7.799479249510633e-08,
      "loss": 0.2376,
      "step": 3480
    },
    {
      "epoch": 2.9544973544973545,
      "grad_norm": 1.2399487682175703,
      "learning_rate": 5.520383595891132e-08,
      "loss": 0.2289,
      "step": 3490
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 1.439024332184667,
      "learning_rate": 3.6339734197360365e-08,
      "loss": 0.2487,
      "step": 3500
    },
    {
      "epoch": 2.9714285714285715,
      "grad_norm": 1.1880176062141694,
      "learning_rate": 2.14039703797575e-08,
      "loss": 0.2373,
      "step": 3510
    },
    {
      "epoch": 2.97989417989418,
      "grad_norm": 1.2924381706571177,
      "learning_rate": 1.0397718814125768e-08,
      "loss": 0.2468,
      "step": 3520
    },
    {
      "epoch": 2.988359788359788,
      "grad_norm": 1.382013827064607,
      "learning_rate": 3.321844854925482e-09,
      "loss": 0.2325,
      "step": 3530
    },
    {
      "epoch": 2.996825396825397,
      "grad_norm": 1.2057884625346706,
      "learning_rate": 1.7690483496979326e-10,
      "loss": 0.2409,
      "step": 3540
    },
    {
      "epoch": 2.9993650793650795,
      "step": 3543,
      "total_flos": 652111388999680.0,
      "train_loss": 0.34479539529973763,
      "train_runtime": 32429.3956,
      "train_samples_per_second": 0.874,
      "train_steps_per_second": 0.109
    }
  ],
  "logging_steps": 10,
  "max_steps": 3543,
  "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": 652111388999680.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}