{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.1536892727204059,
  "eval_steps": 500,
  "global_step": 6104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007805320206966878,
      "grad_norm": 24.262310028076172,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 9.1957,
      "step": 31
    },
    {
      "epoch": 0.0015610640413933755,
      "grad_norm": 29.819591522216797,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.9757,
      "step": 62
    },
    {
      "epoch": 0.002341596062090063,
      "grad_norm": 14.961602210998535,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 6.5436,
      "step": 93
    },
    {
      "epoch": 0.003122128082786751,
      "grad_norm": 5.151533126831055,
      "learning_rate": 4.062909567496723e-06,
      "loss": 5.4812,
      "step": 124
    },
    {
      "epoch": 0.003902660103483439,
      "grad_norm": 7.370818138122559,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.8332,
      "step": 155
    },
    {
      "epoch": 0.004683192124180126,
      "grad_norm": 7.932437896728516,
      "learning_rate": 6.094364351245085e-06,
      "loss": 4.4411,
      "step": 186
    },
    {
      "epoch": 0.005463724144876814,
      "grad_norm": 7.834314823150635,
      "learning_rate": 7.110091743119267e-06,
      "loss": 4.1978,
      "step": 217
    },
    {
      "epoch": 0.006244256165573502,
      "grad_norm": 11.881230354309082,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.9897,
      "step": 248
    },
    {
      "epoch": 0.00702478818627019,
      "grad_norm": 6.348697185516357,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.8375,
      "step": 279
    },
    {
      "epoch": 0.007805320206966878,
      "grad_norm": 8.414978981018066,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.7681,
      "step": 310
    },
    {
      "epoch": 0.008585852227663565,
      "grad_norm": 6.531826972961426,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.6599,
      "step": 341
    },
    {
      "epoch": 0.009366384248360253,
      "grad_norm": 9.794255256652832,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.5324,
      "step": 372
    },
    {
      "epoch": 0.01014691626905694,
      "grad_norm": 8.084766387939453,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 3.4696,
      "step": 403
    },
    {
      "epoch": 0.010927448289753628,
      "grad_norm": 7.659717559814453,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 3.3832,
      "step": 434
    },
    {
      "epoch": 0.011707980310450316,
      "grad_norm": 5.131711959838867,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 3.3514,
      "step": 465
    },
    {
      "epoch": 0.012488512331147004,
      "grad_norm": 6.3673601150512695,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 3.2988,
      "step": 496
    },
    {
      "epoch": 0.013269044351843692,
      "grad_norm": 4.5014142990112305,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 3.239,
      "step": 527
    },
    {
      "epoch": 0.01404957637254038,
      "grad_norm": 5.1420769691467285,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 3.1892,
      "step": 558
    },
    {
      "epoch": 0.014830108393237068,
      "grad_norm": 5.227200508117676,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 3.1732,
      "step": 589
    },
    {
      "epoch": 0.015610640413933756,
      "grad_norm": 4.4621734619140625,
      "learning_rate": 2.031454783748362e-05,
      "loss": 3.0898,
      "step": 620
    },
    {
      "epoch": 0.016391172434630442,
      "grad_norm": 5.213263511657715,
      "learning_rate": 2.13302752293578e-05,
      "loss": 3.0832,
      "step": 651
    },
    {
      "epoch": 0.01717170445532713,
      "grad_norm": 4.23059606552124,
      "learning_rate": 2.234600262123198e-05,
      "loss": 3.0341,
      "step": 682
    },
    {
      "epoch": 0.017952236476023817,
      "grad_norm": 3.5276854038238525,
      "learning_rate": 2.336173001310616e-05,
      "loss": 3.0241,
      "step": 713
    },
    {
      "epoch": 0.018732768496720505,
      "grad_norm": 3.199700355529785,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.9659,
      "step": 744
    },
    {
      "epoch": 0.019513300517417193,
      "grad_norm": 9.667389869689941,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.9292,
      "step": 775
    },
    {
      "epoch": 0.02029383253811388,
      "grad_norm": 3.677541732788086,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.919,
      "step": 806
    },
    {
      "epoch": 0.02107436455881057,
      "grad_norm": 3.711683750152588,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.9141,
      "step": 837
    },
    {
      "epoch": 0.021854896579507257,
      "grad_norm": 3.1250174045562744,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.8585,
      "step": 868
    },
    {
      "epoch": 0.022635428600203945,
      "grad_norm": 3.4183220863342285,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.8206,
      "step": 899
    },
    {
      "epoch": 0.023415960620900633,
      "grad_norm": 2.8176679611206055,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.787,
      "step": 930
    },
    {
      "epoch": 0.02419649264159732,
      "grad_norm": 2.7993154525756836,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.7678,
      "step": 961
    },
    {
      "epoch": 0.02497702466229401,
      "grad_norm": 2.8017027378082275,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.7499,
      "step": 992
    },
    {
      "epoch": 0.025757556682990696,
      "grad_norm": 2.958606004714966,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.7211,
      "step": 1023
    },
    {
      "epoch": 0.026538088703687384,
      "grad_norm": 2.5942046642303467,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.6823,
      "step": 1054
    },
    {
      "epoch": 0.027318620724384072,
      "grad_norm": 3.1978206634521484,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.6943,
      "step": 1085
    },
    {
      "epoch": 0.02809915274508076,
      "grad_norm": 2.6260809898376465,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.6574,
      "step": 1116
    },
    {
      "epoch": 0.028879684765777448,
      "grad_norm": 2.6633715629577637,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.6394,
      "step": 1147
    },
    {
      "epoch": 0.029660216786474135,
      "grad_norm": 2.5701236724853516,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.6336,
      "step": 1178
    },
    {
      "epoch": 0.030440748807170823,
      "grad_norm": 2.4242734909057617,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.6285,
      "step": 1209
    },
    {
      "epoch": 0.03122128082786751,
      "grad_norm": 2.6410114765167236,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.5821,
      "step": 1240
    },
    {
      "epoch": 0.032001812848564196,
      "grad_norm": 2.419649839401245,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.5576,
      "step": 1271
    },
    {
      "epoch": 0.032782344869260883,
      "grad_norm": 2.314750909805298,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.5438,
      "step": 1302
    },
    {
      "epoch": 0.03356287688995757,
      "grad_norm": 2.288361072540283,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.5197,
      "step": 1333
    },
    {
      "epoch": 0.03434340891065426,
      "grad_norm": 2.328840494155884,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.5193,
      "step": 1364
    },
    {
      "epoch": 0.03512394093135095,
      "grad_norm": 2.2555599212646484,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.4907,
      "step": 1395
    },
    {
      "epoch": 0.035904472952047635,
      "grad_norm": 2.1358766555786133,
      "learning_rate": 4.672346002621232e-05,
      "loss": 2.4737,
      "step": 1426
    },
    {
      "epoch": 0.03668500497274432,
      "grad_norm": 2.1471140384674072,
      "learning_rate": 4.77391874180865e-05,
      "loss": 2.4733,
      "step": 1457
    },
    {
      "epoch": 0.03746553699344101,
      "grad_norm": 2.1371941566467285,
      "learning_rate": 4.875491480996068e-05,
      "loss": 2.432,
      "step": 1488
    },
    {
      "epoch": 0.0382460690141377,
      "grad_norm": 1.9789690971374512,
      "learning_rate": 4.977064220183487e-05,
      "loss": 2.4407,
      "step": 1519
    },
    {
      "epoch": 0.039026601034834386,
      "grad_norm": 2.0055742263793945,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 2.4178,
      "step": 1550
    },
    {
      "epoch": 0.039807133055531074,
      "grad_norm": 2.095475196838379,
      "learning_rate": 4.999955597496219e-05,
      "loss": 2.4198,
      "step": 1581
    },
    {
      "epoch": 0.04058766507622776,
      "grad_norm": 2.046337842941284,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 2.3936,
      "step": 1612
    },
    {
      "epoch": 0.04136819709692445,
      "grad_norm": 1.8528848886489868,
      "learning_rate": 4.999799067923527e-05,
      "loss": 2.3893,
      "step": 1643
    },
    {
      "epoch": 0.04214872911762114,
      "grad_norm": 1.883175253868103,
      "learning_rate": 4.999678487776908e-05,
      "loss": 2.3668,
      "step": 1674
    },
    {
      "epoch": 0.042929261138317826,
      "grad_norm": 1.8084770441055298,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 2.3702,
      "step": 1705
    },
    {
      "epoch": 0.043709793159014514,
      "grad_norm": 1.8336963653564453,
      "learning_rate": 4.999352703566763e-05,
      "loss": 2.357,
      "step": 1736
    },
    {
      "epoch": 0.0444903251797112,
      "grad_norm": 1.8369375467300415,
      "learning_rate": 4.999147503179668e-05,
      "loss": 2.3393,
      "step": 1767
    },
    {
      "epoch": 0.04527085720040789,
      "grad_norm": 1.8411383628845215,
      "learning_rate": 4.998914100252672e-05,
      "loss": 2.319,
      "step": 1798
    },
    {
      "epoch": 0.04605138922110458,
      "grad_norm": 1.785543441772461,
      "learning_rate": 4.998652497419696e-05,
      "loss": 2.3259,
      "step": 1829
    },
    {
      "epoch": 0.046831921241801265,
      "grad_norm": 1.7194079160690308,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 2.3046,
      "step": 1860
    },
    {
      "epoch": 0.04761245326249795,
      "grad_norm": 1.7909469604492188,
      "learning_rate": 4.998044704162613e-05,
      "loss": 2.307,
      "step": 1891
    },
    {
      "epoch": 0.04839298528319464,
      "grad_norm": 1.725806474685669,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 2.2992,
      "step": 1922
    },
    {
      "epoch": 0.04917351730389133,
      "grad_norm": 1.67300546169281,
      "learning_rate": 4.997324150843799e-05,
      "loss": 2.2763,
      "step": 1953
    },
    {
      "epoch": 0.04995404932458802,
      "grad_norm": 1.7416061162948608,
      "learning_rate": 4.99692159912661e-05,
      "loss": 2.2596,
      "step": 1984
    },
    {
      "epoch": 0.050734581345284704,
      "grad_norm": 1.843024492263794,
      "learning_rate": 4.996490869988546e-05,
      "loss": 2.2644,
      "step": 2015
    },
    {
      "epoch": 0.05151511336598139,
      "grad_norm": 1.6172534227371216,
      "learning_rate": 4.996031968290326e-05,
      "loss": 2.271,
      "step": 2046
    },
    {
      "epoch": 0.05229564538667808,
      "grad_norm": 1.6510816812515259,
      "learning_rate": 4.995544899210594e-05,
      "loss": 2.2445,
      "step": 2077
    },
    {
      "epoch": 0.05307617740737477,
      "grad_norm": 1.625775933265686,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 2.2432,
      "step": 2108
    },
    {
      "epoch": 0.053856709428071456,
      "grad_norm": 1.602299690246582,
      "learning_rate": 4.994486281210429e-05,
      "loss": 2.2266,
      "step": 2139
    },
    {
      "epoch": 0.054637241448768144,
      "grad_norm": 1.7036757469177246,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 2.2292,
      "step": 2170
    },
    {
      "epoch": 0.05541777346946483,
      "grad_norm": 1.5851162672042847,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 2.2286,
      "step": 2201
    },
    {
      "epoch": 0.05619830549016152,
      "grad_norm": 1.6328140497207642,
      "learning_rate": 4.992687246588743e-05,
      "loss": 2.2342,
      "step": 2232
    },
    {
      "epoch": 0.05697883751085821,
      "grad_norm": 1.4901413917541504,
      "learning_rate": 4.992031299767347e-05,
      "loss": 2.2094,
      "step": 2263
    },
    {
      "epoch": 0.057759369531554895,
      "grad_norm": 1.5897445678710938,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 2.2126,
      "step": 2294
    },
    {
      "epoch": 0.05853990155225158,
      "grad_norm": 1.5534899234771729,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 2.2125,
      "step": 2325
    },
    {
      "epoch": 0.05932043357294827,
      "grad_norm": 1.5713675022125244,
      "learning_rate": 4.989894757091861e-05,
      "loss": 2.2007,
      "step": 2356
    },
    {
      "epoch": 0.06010096559364496,
      "grad_norm": 1.473955750465393,
      "learning_rate": 4.989126368919158e-05,
      "loss": 2.1819,
      "step": 2387
    },
    {
      "epoch": 0.06088149761434165,
      "grad_norm": 1.5200265645980835,
      "learning_rate": 4.988329891293693e-05,
      "loss": 2.1944,
      "step": 2418
    },
    {
      "epoch": 0.061662029635038335,
      "grad_norm": 1.5236589908599854,
      "learning_rate": 4.987505333203608e-05,
      "loss": 2.1816,
      "step": 2449
    },
    {
      "epoch": 0.06244256165573502,
      "grad_norm": 1.5306485891342163,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 2.187,
      "step": 2480
    },
    {
      "epoch": 0.06322309367643171,
      "grad_norm": 1.5017462968826294,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 2.1706,
      "step": 2511
    },
    {
      "epoch": 0.06400362569712839,
      "grad_norm": 4.610604763031006,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 2.1528,
      "step": 2542
    },
    {
      "epoch": 0.06478415771782509,
      "grad_norm": 1.4580930471420288,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 2.1849,
      "step": 2573
    },
    {
      "epoch": 0.06556468973852177,
      "grad_norm": 1.5506151914596558,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 2.1504,
      "step": 2604
    },
    {
      "epoch": 0.06634522175921846,
      "grad_norm": 1.4275034666061401,
      "learning_rate": 4.981968838155888e-05,
      "loss": 2.1492,
      "step": 2635
    },
    {
      "epoch": 0.06712575377991514,
      "grad_norm": 1.4749577045440674,
      "learning_rate": 4.980947995086024e-05,
      "loss": 2.1579,
      "step": 2666
    },
    {
      "epoch": 0.06790628580061184,
      "grad_norm": 1.4220107793807983,
      "learning_rate": 4.979899154855234e-05,
      "loss": 2.1522,
      "step": 2697
    },
    {
      "epoch": 0.06868681782130852,
      "grad_norm": 1.4243831634521484,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 2.1258,
      "step": 2728
    },
    {
      "epoch": 0.06946734984200521,
      "grad_norm": 1.4925100803375244,
      "learning_rate": 4.977717530570768e-05,
      "loss": 2.1379,
      "step": 2759
    },
    {
      "epoch": 0.0702478818627019,
      "grad_norm": 1.4839624166488647,
      "learning_rate": 4.976584771136425e-05,
      "loss": 2.1588,
      "step": 2790
    },
    {
      "epoch": 0.07102841388339859,
      "grad_norm": 1.4130090475082397,
      "learning_rate": 4.975424063779547e-05,
      "loss": 2.1461,
      "step": 2821
    },
    {
      "epoch": 0.07180894590409527,
      "grad_norm": 1.3940472602844238,
      "learning_rate": 4.974235421598557e-05,
      "loss": 2.1416,
      "step": 2852
    },
    {
      "epoch": 0.07258947792479196,
      "grad_norm": 1.4502291679382324,
      "learning_rate": 4.973018858007122e-05,
      "loss": 2.1315,
      "step": 2883
    },
    {
      "epoch": 0.07337000994548865,
      "grad_norm": 1.4334006309509277,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 2.122,
      "step": 2914
    },
    {
      "epoch": 0.07415054196618534,
      "grad_norm": 1.4366185665130615,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 2.1179,
      "step": 2945
    },
    {
      "epoch": 0.07493107398688202,
      "grad_norm": 1.437264084815979,
      "learning_rate": 4.969201777632205e-05,
      "loss": 2.1171,
      "step": 2976
    },
    {
      "epoch": 0.07571160600757872,
      "grad_norm": 1.3713815212249756,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 2.1215,
      "step": 3007
    },
    {
      "epoch": 0.0764921380282754,
      "grad_norm": 1.5930180549621582,
      "learning_rate": 4.966517710419033e-05,
      "loss": 2.0961,
      "step": 3038
    },
    {
      "epoch": 0.07727267004897209,
      "grad_norm": 1.4208778142929077,
      "learning_rate": 4.965133917685858e-05,
      "loss": 2.1193,
      "step": 3069
    },
    {
      "epoch": 0.07805320206966877,
      "grad_norm": 1.3563286066055298,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 2.0957,
      "step": 3100
    },
    {
      "epoch": 0.07883373409036547,
      "grad_norm": 1.382861852645874,
      "learning_rate": 4.962282892045718e-05,
      "loss": 2.117,
      "step": 3131
    },
    {
      "epoch": 0.07961426611106215,
      "grad_norm": 1.326231837272644,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 2.0906,
      "step": 3162
    },
    {
      "epoch": 0.08039479813175884,
      "grad_norm": 1.3465954065322876,
      "learning_rate": 4.959320720608049e-05,
      "loss": 2.09,
      "step": 3193
    },
    {
      "epoch": 0.08117533015245552,
      "grad_norm": 1.3832777738571167,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 2.089,
      "step": 3224
    },
    {
      "epoch": 0.08195586217315222,
      "grad_norm": 1.4453243017196655,
      "learning_rate": 4.956247537083282e-05,
      "loss": 2.0867,
      "step": 3255
    },
    {
      "epoch": 0.0827363941938489,
      "grad_norm": 1.310168743133545,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 2.0915,
      "step": 3286
    },
    {
      "epoch": 0.0835169262145456,
      "grad_norm": 1.354078769683838,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 2.0888,
      "step": 3317
    },
    {
      "epoch": 0.08429745823524228,
      "grad_norm": 1.3768595457077026,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 2.0821,
      "step": 3348
    },
    {
      "epoch": 0.08507799025593897,
      "grad_norm": 1.3402544260025024,
      "learning_rate": 4.949768693662973e-05,
      "loss": 2.0776,
      "step": 3379
    },
    {
      "epoch": 0.08585852227663565,
      "grad_norm": 1.3178027868270874,
      "learning_rate": 4.948079823064559e-05,
      "loss": 2.0699,
      "step": 3410
    },
    {
      "epoch": 0.08663905429733235,
      "grad_norm": 1.7057585716247559,
      "learning_rate": 4.946363326218074e-05,
      "loss": 2.0615,
      "step": 3441
    },
    {
      "epoch": 0.08741958631802903,
      "grad_norm": 1.3453648090362549,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 2.0788,
      "step": 3472
    },
    {
      "epoch": 0.08820011833872572,
      "grad_norm": 1.3319638967514038,
      "learning_rate": 4.942847531574167e-05,
      "loss": 2.0619,
      "step": 3503
    },
    {
      "epoch": 0.0889806503594224,
      "grad_norm": 1.2845163345336914,
      "learning_rate": 4.941048273452008e-05,
      "loss": 2.0728,
      "step": 3534
    },
    {
      "epoch": 0.0897611823801191,
      "grad_norm": 1.3615431785583496,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 2.0538,
      "step": 3565
    },
    {
      "epoch": 0.09054171440081578,
      "grad_norm": 1.2946122884750366,
      "learning_rate": 4.93736713712897e-05,
      "loss": 2.0689,
      "step": 3596
    },
    {
      "epoch": 0.09132224642151247,
      "grad_norm": 1.294846534729004,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 2.0706,
      "step": 3627
    },
    {
      "epoch": 0.09210277844220915,
      "grad_norm": 1.298010230064392,
      "learning_rate": 4.93357597968886e-05,
      "loss": 2.0406,
      "step": 3658
    },
    {
      "epoch": 0.09288331046290585,
      "grad_norm": 1.7569706439971924,
      "learning_rate": 4.931639196334338e-05,
      "loss": 2.0736,
      "step": 3689
    },
    {
      "epoch": 0.09366384248360253,
      "grad_norm": 1.2595418691635132,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 2.0654,
      "step": 3720
    },
    {
      "epoch": 0.09444437450429922,
      "grad_norm": 1.3139164447784424,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 2.0467,
      "step": 3751
    },
    {
      "epoch": 0.0952249065249959,
      "grad_norm": 1.290247917175293,
      "learning_rate": 4.925664290937196e-05,
      "loss": 2.0418,
      "step": 3782
    },
    {
      "epoch": 0.0960054385456926,
      "grad_norm": 1.2755826711654663,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 2.0244,
      "step": 3813
    },
    {
      "epoch": 0.09678597056638928,
      "grad_norm": 1.2596077919006348,
      "learning_rate": 4.921544116753962e-05,
      "loss": 2.0609,
      "step": 3844
    },
    {
      "epoch": 0.09756650258708598,
      "grad_norm": 1.249387264251709,
      "learning_rate": 4.919443027766935e-05,
      "loss": 2.0455,
      "step": 3875
    },
    {
      "epoch": 0.09834703460778266,
      "grad_norm": 1.255379319190979,
      "learning_rate": 4.91731463569418e-05,
      "loss": 2.024,
      "step": 3906
    },
    {
      "epoch": 0.09912756662847935,
      "grad_norm": 1.2620748281478882,
      "learning_rate": 4.915158964554312e-05,
      "loss": 2.0355,
      "step": 3937
    },
    {
      "epoch": 0.09990809864917603,
      "grad_norm": 1.2736345529556274,
      "learning_rate": 4.912976038673786e-05,
      "loss": 2.0351,
      "step": 3968
    },
    {
      "epoch": 0.10068863066987273,
      "grad_norm": 1.18524169921875,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 2.029,
      "step": 3999
    },
    {
      "epoch": 0.10146916269056941,
      "grad_norm": 10.429583549499512,
      "learning_rate": 4.908528521534139e-05,
      "loss": 2.0226,
      "step": 4030
    },
    {
      "epoch": 0.1022496947112661,
      "grad_norm": 1.240295648574829,
      "learning_rate": 4.906263980464644e-05,
      "loss": 2.0401,
      "step": 4061
    },
    {
      "epoch": 0.10303022673196278,
      "grad_norm": 1.2191627025604248,
      "learning_rate": 4.903972285033178e-05,
      "loss": 2.0079,
      "step": 4092
    },
    {
      "epoch": 0.10381075875265948,
      "grad_norm": 1.2340337038040161,
      "learning_rate": 4.901653461101213e-05,
      "loss": 2.0305,
      "step": 4123
    },
    {
      "epoch": 0.10459129077335616,
      "grad_norm": 1.284839153289795,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 2.0212,
      "step": 4154
    },
    {
      "epoch": 0.10537182279405286,
      "grad_norm": 1.2955284118652344,
      "learning_rate": 4.896934532712084e-05,
      "loss": 2.0328,
      "step": 4185
    },
    {
      "epoch": 0.10615235481474954,
      "grad_norm": 1.165124535560608,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 2.0248,
      "step": 4216
    },
    {
      "epoch": 0.10693288683544623,
      "grad_norm": 1.1925835609436035,
      "learning_rate": 4.892107408306516e-05,
      "loss": 2.0086,
      "step": 4247
    },
    {
      "epoch": 0.10771341885614291,
      "grad_norm": 1.1841633319854736,
      "learning_rate": 4.889653340498669e-05,
      "loss": 2.0166,
      "step": 4278
    },
    {
      "epoch": 0.1084939508768396,
      "grad_norm": 1.2198415994644165,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 2.0164,
      "step": 4309
    },
    {
      "epoch": 0.10927448289753629,
      "grad_norm": 1.2284913063049316,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 2.015,
      "step": 4340
    },
    {
      "epoch": 0.11005501491823298,
      "grad_norm": 1.2090349197387695,
      "learning_rate": 4.882129447892753e-05,
      "loss": 2.0094,
      "step": 4371
    },
    {
      "epoch": 0.11083554693892966,
      "grad_norm": 1.215232491493225,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 2.0045,
      "step": 4402
    },
    {
      "epoch": 0.11161607895962636,
      "grad_norm": 1.2804045677185059,
      "learning_rate": 4.876979062282995e-05,
      "loss": 1.9899,
      "step": 4433
    },
    {
      "epoch": 0.11239661098032304,
      "grad_norm": 1.2405016422271729,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 2.0093,
      "step": 4464
    },
    {
      "epoch": 0.11317714300101973,
      "grad_norm": 1.1784151792526245,
      "learning_rate": 4.871721381433344e-05,
      "loss": 2.0036,
      "step": 4495
    },
    {
      "epoch": 0.11395767502171641,
      "grad_norm": 1.185738205909729,
      "learning_rate": 4.869052379269719e-05,
      "loss": 2.0033,
      "step": 4526
    },
    {
      "epoch": 0.11473820704241311,
      "grad_norm": 1.1984614133834839,
      "learning_rate": 4.866356642671985e-05,
      "loss": 2.025,
      "step": 4557
    },
    {
      "epoch": 0.11551873906310979,
      "grad_norm": 1.1734029054641724,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 2.0023,
      "step": 4588
    },
    {
      "epoch": 0.11629927108380649,
      "grad_norm": 1.198611855506897,
      "learning_rate": 4.860885088159626e-05,
      "loss": 2.0031,
      "step": 4619
    },
    {
      "epoch": 0.11707980310450317,
      "grad_norm": 1.1793649196624756,
      "learning_rate": 4.858109331990751e-05,
      "loss": 1.9806,
      "step": 4650
    },
    {
      "epoch": 0.11786033512519985,
      "grad_norm": 1.1945818662643433,
      "learning_rate": 4.855306964878567e-05,
      "loss": 1.9855,
      "step": 4681
    },
    {
      "epoch": 0.11864086714589654,
      "grad_norm": 1.1875253915786743,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 1.9883,
      "step": 4712
    },
    {
      "epoch": 0.11942139916659322,
      "grad_norm": 1.1512328386306763,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 1.9807,
      "step": 4743
    },
    {
      "epoch": 0.12020193118728992,
      "grad_norm": 1.2059484720230103,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 2.0103,
      "step": 4774
    },
    {
      "epoch": 0.1209824632079866,
      "grad_norm": 1.1759779453277588,
      "learning_rate": 4.843832023980392e-05,
      "loss": 1.9754,
      "step": 4805
    },
    {
      "epoch": 0.1217629952286833,
      "grad_norm": 1.2004414796829224,
      "learning_rate": 4.840897082510106e-05,
      "loss": 1.9893,
      "step": 4836
    },
    {
      "epoch": 0.12254352724937997,
      "grad_norm": 1.1744251251220703,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 1.9848,
      "step": 4867
    },
    {
      "epoch": 0.12332405927007667,
      "grad_norm": 1.1731071472167969,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 1.9945,
      "step": 4898
    },
    {
      "epoch": 0.12410459129077335,
      "grad_norm": 1.2002074718475342,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 1.9869,
      "step": 4929
    },
    {
      "epoch": 0.12488512331147004,
      "grad_norm": 1.147085189819336,
      "learning_rate": 4.828893485248369e-05,
      "loss": 1.9876,
      "step": 4960
    },
    {
      "epoch": 0.12566565533216673,
      "grad_norm": 1.1605637073516846,
      "learning_rate": 4.825826797411682e-05,
      "loss": 1.9858,
      "step": 4991
    },
    {
      "epoch": 0.12644618735286342,
      "grad_norm": 1.1823869943618774,
      "learning_rate": 4.822733862935702e-05,
      "loss": 1.9617,
      "step": 5022
    },
    {
      "epoch": 0.12722671937356012,
      "grad_norm": 1.1143817901611328,
      "learning_rate": 4.819614716723775e-05,
      "loss": 1.9756,
      "step": 5053
    },
    {
      "epoch": 0.12800725139425678,
      "grad_norm": 1.1646877527236938,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 2.0005,
      "step": 5084
    },
    {
      "epoch": 0.12878778341495348,
      "grad_norm": 1.1390529870986938,
      "learning_rate": 4.813297930184042e-05,
      "loss": 1.9943,
      "step": 5115
    },
    {
      "epoch": 0.12956831543565017,
      "grad_norm": 1.1759583950042725,
      "learning_rate": 4.810100361140314e-05,
      "loss": 1.9772,
      "step": 5146
    },
    {
      "epoch": 0.13034884745634687,
      "grad_norm": 1.1465727090835571,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 1.9563,
      "step": 5177
    },
    {
      "epoch": 0.13112937947704353,
      "grad_norm": 1.2573471069335938,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 1.9735,
      "step": 5208
    },
    {
      "epoch": 0.13190991149774023,
      "grad_norm": 1.1301871538162231,
      "learning_rate": 4.8003513848046e-05,
      "loss": 1.987,
      "step": 5239
    },
    {
      "epoch": 0.13269044351843692,
      "grad_norm": 1.1424096822738647,
      "learning_rate": 4.79704975853109e-05,
      "loss": 1.9661,
      "step": 5270
    },
    {
      "epoch": 0.13347097553913362,
      "grad_norm": 1.131800889968872,
      "learning_rate": 4.793722210363262e-05,
      "loss": 1.9828,
      "step": 5301
    },
    {
      "epoch": 0.13425150755983029,
      "grad_norm": 1.1227242946624756,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 1.975,
      "step": 5332
    },
    {
      "epoch": 0.13503203958052698,
      "grad_norm": 1.143832802772522,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 1.9619,
      "step": 5363
    },
    {
      "epoch": 0.13581257160122367,
      "grad_norm": 1.1281683444976807,
      "learning_rate": 4.783584411463221e-05,
      "loss": 1.9603,
      "step": 5394
    },
    {
      "epoch": 0.13659310362192037,
      "grad_norm": 1.134901762008667,
      "learning_rate": 4.780153554146274e-05,
      "loss": 1.9888,
      "step": 5425
    },
    {
      "epoch": 0.13737363564261704,
      "grad_norm": 1.1317532062530518,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 1.9793,
      "step": 5456
    },
    {
      "epoch": 0.13815416766331373,
      "grad_norm": 1.2516345977783203,
      "learning_rate": 4.773214684850662e-05,
      "loss": 1.9839,
      "step": 5487
    },
    {
      "epoch": 0.13893469968401043,
      "grad_norm": 1.107391119003296,
      "learning_rate": 4.769706751176193e-05,
      "loss": 1.9606,
      "step": 5518
    },
    {
      "epoch": 0.13971523170470712,
      "grad_norm": 1.1041492223739624,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 1.9688,
      "step": 5549
    },
    {
      "epoch": 0.1404957637254038,
      "grad_norm": 1.1099929809570312,
      "learning_rate": 4.762614083706258e-05,
      "loss": 1.9821,
      "step": 5580
    },
    {
      "epoch": 0.14127629574610048,
      "grad_norm": 1.167151927947998,
      "learning_rate": 4.759029429950581e-05,
      "loss": 1.9669,
      "step": 5611
    },
    {
      "epoch": 0.14205682776679718,
      "grad_norm": 1.0942577123641968,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 1.9565,
      "step": 5642
    },
    {
      "epoch": 0.14283735978749387,
      "grad_norm": 1.0859640836715698,
      "learning_rate": 4.751783684659e-05,
      "loss": 1.9634,
      "step": 5673
    },
    {
      "epoch": 0.14361789180819054,
      "grad_norm": 2.420607089996338,
      "learning_rate": 4.748122674890348e-05,
      "loss": 1.9365,
      "step": 5704
    },
    {
      "epoch": 0.14439842382888723,
      "grad_norm": 1.0996880531311035,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 1.9718,
      "step": 5735
    },
    {
      "epoch": 0.14517895584958393,
      "grad_norm": 1.1183338165283203,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 1.9564,
      "step": 5766
    },
    {
      "epoch": 0.14595948787028062,
      "grad_norm": 1.1093947887420654,
      "learning_rate": 4.736987593718397e-05,
      "loss": 1.9516,
      "step": 5797
    },
    {
      "epoch": 0.1467400198909773,
      "grad_norm": 1.141797423362732,
      "learning_rate": 4.733225355658999e-05,
      "loss": 1.929,
      "step": 5828
    },
    {
      "epoch": 0.14752055191167399,
      "grad_norm": 1.1214358806610107,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 1.9297,
      "step": 5859
    },
    {
      "epoch": 0.14830108393237068,
      "grad_norm": 1.1025431156158447,
      "learning_rate": 4.725625317347119e-05,
      "loss": 1.9226,
      "step": 5890
    },
    {
      "epoch": 0.14908161595306738,
      "grad_norm": 1.1094226837158203,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 1.946,
      "step": 5921
    },
    {
      "epoch": 0.14986214797376404,
      "grad_norm": 1.0909287929534912,
      "learning_rate": 4.717924815801832e-05,
      "loss": 1.9511,
      "step": 5952
    },
    {
      "epoch": 0.15064267999446074,
      "grad_norm": 1.0755189657211304,
      "learning_rate": 4.714036999763532e-05,
      "loss": 1.939,
      "step": 5983
    },
    {
      "epoch": 0.15142321201515743,
      "grad_norm": 1.1426901817321777,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 1.93,
      "step": 6014
    },
    {
      "epoch": 0.15220374403585413,
      "grad_norm": 1.1208311319351196,
      "learning_rate": 4.7061864565225e-05,
      "loss": 1.9516,
      "step": 6045
    },
    {
      "epoch": 0.1529842760565508,
      "grad_norm": 1.0931717157363892,
      "learning_rate": 4.702223817912081e-05,
      "loss": 1.9314,
      "step": 6076
    }
  ],
  "logging_steps": 31,
  "max_steps": 30517,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3052,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.263370839061168e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}