{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.21478588268411977,
  "eval_steps": 500,
  "global_step": 6104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010908195221506739,
      "grad_norm": 27.24474334716797,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 9.0396,
      "step": 31
    },
    {
      "epoch": 0.0021816390443013477,
      "grad_norm": 16.773351669311523,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.9471,
      "step": 62
    },
    {
      "epoch": 0.0032724585664520214,
      "grad_norm": 15.084428787231445,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 6.5868,
      "step": 93
    },
    {
      "epoch": 0.0043632780886026954,
      "grad_norm": 6.7851786613464355,
      "learning_rate": 4.062909567496723e-06,
      "loss": 5.3744,
      "step": 124
    },
    {
      "epoch": 0.0054540976107533695,
      "grad_norm": 7.081949234008789,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.6378,
      "step": 155
    },
    {
      "epoch": 0.006544917132904043,
      "grad_norm": 8.945405960083008,
      "learning_rate": 6.094364351245085e-06,
      "loss": 4.2292,
      "step": 186
    },
    {
      "epoch": 0.007635736655054717,
      "grad_norm": 15.078816413879395,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.9767,
      "step": 217
    },
    {
      "epoch": 0.008726556177205391,
      "grad_norm": 8.799291610717773,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.765,
      "step": 248
    },
    {
      "epoch": 0.009817375699356064,
      "grad_norm": 9.741459846496582,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.6496,
      "step": 279
    },
    {
      "epoch": 0.010908195221506739,
      "grad_norm": 9.403132438659668,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.4936,
      "step": 310
    },
    {
      "epoch": 0.011999014743657412,
      "grad_norm": 8.479109764099121,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.3879,
      "step": 341
    },
    {
      "epoch": 0.013089834265808085,
      "grad_norm": 4.988598823547363,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.3147,
      "step": 372
    },
    {
      "epoch": 0.01418065378795876,
      "grad_norm": 5.518200874328613,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 3.2022,
      "step": 403
    },
    {
      "epoch": 0.015271473310109434,
      "grad_norm": 5.685606002807617,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 3.1393,
      "step": 434
    },
    {
      "epoch": 0.01636229283226011,
      "grad_norm": 5.6373491287231445,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 3.05,
      "step": 465
    },
    {
      "epoch": 0.017453112354410782,
      "grad_norm": 5.174935340881348,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.992,
      "step": 496
    },
    {
      "epoch": 0.018543931876561455,
      "grad_norm": 5.266006946563721,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.9558,
      "step": 527
    },
    {
      "epoch": 0.019634751398712128,
      "grad_norm": 4.111412048339844,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.9089,
      "step": 558
    },
    {
      "epoch": 0.0207255709208628,
      "grad_norm": 4.149659156799316,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.879,
      "step": 589
    },
    {
      "epoch": 0.021816390443013478,
      "grad_norm": 4.711772441864014,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.8057,
      "step": 620
    },
    {
      "epoch": 0.02290720996516415,
      "grad_norm": 4.372698783874512,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.7954,
      "step": 651
    },
    {
      "epoch": 0.023998029487314824,
      "grad_norm": 3.9052813053131104,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.7286,
      "step": 682
    },
    {
      "epoch": 0.025088849009465498,
      "grad_norm": 3.544926404953003,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.6875,
      "step": 713
    },
    {
      "epoch": 0.02617966853161617,
      "grad_norm": 4.380484104156494,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.6703,
      "step": 744
    },
    {
      "epoch": 0.027270488053766848,
      "grad_norm": 4.843706130981445,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.6387,
      "step": 775
    },
    {
      "epoch": 0.02836130757591752,
      "grad_norm": 3.555110454559326,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.5996,
      "step": 806
    },
    {
      "epoch": 0.029452127098068194,
      "grad_norm": 3.3113982677459717,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.5823,
      "step": 837
    },
    {
      "epoch": 0.030542946620218867,
      "grad_norm": 3.4552953243255615,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.5643,
      "step": 868
    },
    {
      "epoch": 0.03163376614236954,
      "grad_norm": 2.9786770343780518,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.5426,
      "step": 899
    },
    {
      "epoch": 0.03272458566452022,
      "grad_norm": 3.227999210357666,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.4834,
      "step": 930
    },
    {
      "epoch": 0.03381540518667089,
      "grad_norm": 3.0225250720977783,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.4755,
      "step": 961
    },
    {
      "epoch": 0.034906224708821564,
      "grad_norm": 2.6292648315429688,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.4393,
      "step": 992
    },
    {
      "epoch": 0.03599704423097224,
      "grad_norm": 2.767667055130005,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.4124,
      "step": 1023
    },
    {
      "epoch": 0.03708786375312291,
      "grad_norm": 2.6588900089263916,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.3551,
      "step": 1054
    },
    {
      "epoch": 0.03817868327527359,
      "grad_norm": 2.7497403621673584,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.3576,
      "step": 1085
    },
    {
      "epoch": 0.039269502797424256,
      "grad_norm": 2.5148332118988037,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.3078,
      "step": 1116
    },
    {
      "epoch": 0.04036032231957493,
      "grad_norm": 3.0025594234466553,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.303,
      "step": 1147
    },
    {
      "epoch": 0.0414511418417256,
      "grad_norm": 2.6871118545532227,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.3062,
      "step": 1178
    },
    {
      "epoch": 0.04254196136387628,
      "grad_norm": 10.630929946899414,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.307,
      "step": 1209
    },
    {
      "epoch": 0.043632780886026956,
      "grad_norm": 2.571220874786377,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.2837,
      "step": 1240
    },
    {
      "epoch": 0.044723600408177626,
      "grad_norm": 3.5510289669036865,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.2797,
      "step": 1271
    },
    {
      "epoch": 0.0458144199303283,
      "grad_norm": 2.794853925704956,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.2797,
      "step": 1302
    },
    {
      "epoch": 0.04690523945247897,
      "grad_norm": 6.132640838623047,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.2504,
      "step": 1333
    },
    {
      "epoch": 0.04799605897462965,
      "grad_norm": 6.17495584487915,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.2414,
      "step": 1364
    },
    {
      "epoch": 0.049086878496780326,
      "grad_norm": 2.3518564701080322,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.2198,
      "step": 1395
    },
    {
      "epoch": 0.050177698018930995,
      "grad_norm": 2.2073466777801514,
      "learning_rate": 4.672346002621232e-05,
      "loss": 2.2057,
      "step": 1426
    },
    {
      "epoch": 0.05126851754108167,
      "grad_norm": 2.018749237060547,
      "learning_rate": 4.77391874180865e-05,
      "loss": 2.1965,
      "step": 1457
    },
    {
      "epoch": 0.05235933706323234,
      "grad_norm": 2.7125244140625,
      "learning_rate": 4.875491480996068e-05,
      "loss": 2.19,
      "step": 1488
    },
    {
      "epoch": 0.05345015658538302,
      "grad_norm": 1.9725641012191772,
      "learning_rate": 4.977064220183487e-05,
      "loss": 2.1669,
      "step": 1519
    },
    {
      "epoch": 0.054540976107533695,
      "grad_norm": 2.0198206901550293,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 2.1486,
      "step": 1550
    },
    {
      "epoch": 0.055631795629684365,
      "grad_norm": 2.068424940109253,
      "learning_rate": 4.999955597496219e-05,
      "loss": 2.1476,
      "step": 1581
    },
    {
      "epoch": 0.05672261515183504,
      "grad_norm": 1.8509161472320557,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 2.1129,
      "step": 1612
    },
    {
      "epoch": 0.05781343467398571,
      "grad_norm": 1.805915355682373,
      "learning_rate": 4.999799067923527e-05,
      "loss": 2.1014,
      "step": 1643
    },
    {
      "epoch": 0.05890425419613639,
      "grad_norm": 1.9447969198226929,
      "learning_rate": 4.999678487776908e-05,
      "loss": 2.1067,
      "step": 1674
    },
    {
      "epoch": 0.059995073718287065,
      "grad_norm": 2.131700277328491,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 2.0817,
      "step": 1705
    },
    {
      "epoch": 0.061085893240437734,
      "grad_norm": 1.9113484621047974,
      "learning_rate": 4.999352703566763e-05,
      "loss": 2.067,
      "step": 1736
    },
    {
      "epoch": 0.06217671276258841,
      "grad_norm": 1.8802495002746582,
      "learning_rate": 4.999147503179668e-05,
      "loss": 2.0617,
      "step": 1767
    },
    {
      "epoch": 0.06326753228473908,
      "grad_norm": 1.6704238653182983,
      "learning_rate": 4.998914100252672e-05,
      "loss": 2.0671,
      "step": 1798
    },
    {
      "epoch": 0.06435835180688976,
      "grad_norm": 4.303380966186523,
      "learning_rate": 4.998652497419696e-05,
      "loss": 2.0436,
      "step": 1829
    },
    {
      "epoch": 0.06544917132904043,
      "grad_norm": 1.8054529428482056,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 2.0499,
      "step": 1860
    },
    {
      "epoch": 0.06653999085119111,
      "grad_norm": 1.663872241973877,
      "learning_rate": 4.998044704162613e-05,
      "loss": 2.0227,
      "step": 1891
    },
    {
      "epoch": 0.06763081037334177,
      "grad_norm": 10.130581855773926,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 2.0022,
      "step": 1922
    },
    {
      "epoch": 0.06872162989549245,
      "grad_norm": 3.7727627754211426,
      "learning_rate": 4.997324150843799e-05,
      "loss": 2.0172,
      "step": 1953
    },
    {
      "epoch": 0.06981244941764313,
      "grad_norm": 1.8669508695602417,
      "learning_rate": 4.99692159912661e-05,
      "loss": 2.0214,
      "step": 1984
    },
    {
      "epoch": 0.0709032689397938,
      "grad_norm": 17.45224952697754,
      "learning_rate": 4.996490869988546e-05,
      "loss": 1.9959,
      "step": 2015
    },
    {
      "epoch": 0.07199408846194448,
      "grad_norm": 1.7007793188095093,
      "learning_rate": 4.996031968290326e-05,
      "loss": 2.0139,
      "step": 2046
    },
    {
      "epoch": 0.07308490798409514,
      "grad_norm": 1.6912076473236084,
      "learning_rate": 4.995544899210594e-05,
      "loss": 1.9963,
      "step": 2077
    },
    {
      "epoch": 0.07417572750624582,
      "grad_norm": 1.6827666759490967,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 1.973,
      "step": 2108
    },
    {
      "epoch": 0.0752665470283965,
      "grad_norm": 1.767269253730774,
      "learning_rate": 4.994486281210429e-05,
      "loss": 1.9949,
      "step": 2139
    },
    {
      "epoch": 0.07635736655054717,
      "grad_norm": 7.507472038269043,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 1.9837,
      "step": 2170
    },
    {
      "epoch": 0.07744818607269784,
      "grad_norm": 1.6840234994888306,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 1.9789,
      "step": 2201
    },
    {
      "epoch": 0.07853900559484851,
      "grad_norm": 1.6442159414291382,
      "learning_rate": 4.992687246588743e-05,
      "loss": 1.9436,
      "step": 2232
    },
    {
      "epoch": 0.07962982511699919,
      "grad_norm": 1.546859622001648,
      "learning_rate": 4.992031299767347e-05,
      "loss": 1.9341,
      "step": 2263
    },
    {
      "epoch": 0.08072064463914987,
      "grad_norm": 1.4932371377944946,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 1.9583,
      "step": 2294
    },
    {
      "epoch": 0.08181146416130054,
      "grad_norm": 1.4531840085983276,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 1.9414,
      "step": 2325
    },
    {
      "epoch": 0.0829022836834512,
      "grad_norm": 1.5285487174987793,
      "learning_rate": 4.989894757091861e-05,
      "loss": 1.9502,
      "step": 2356
    },
    {
      "epoch": 0.08399310320560188,
      "grad_norm": 1.4159945249557495,
      "learning_rate": 4.989126368919158e-05,
      "loss": 1.9299,
      "step": 2387
    },
    {
      "epoch": 0.08508392272775256,
      "grad_norm": 1.4361426830291748,
      "learning_rate": 4.988329891293693e-05,
      "loss": 1.9266,
      "step": 2418
    },
    {
      "epoch": 0.08617474224990324,
      "grad_norm": 1.4310742616653442,
      "learning_rate": 4.987505333203608e-05,
      "loss": 1.9116,
      "step": 2449
    },
    {
      "epoch": 0.08726556177205391,
      "grad_norm": 1.3780311346054077,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.9189,
      "step": 2480
    },
    {
      "epoch": 0.08835638129420458,
      "grad_norm": 1.3917787075042725,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 1.9107,
      "step": 2511
    },
    {
      "epoch": 0.08944720081635525,
      "grad_norm": 1.3963998556137085,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 1.899,
      "step": 2542
    },
    {
      "epoch": 0.09053802033850593,
      "grad_norm": 1.3573299646377563,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.9084,
      "step": 2573
    },
    {
      "epoch": 0.0916288398606566,
      "grad_norm": 1.362552523612976,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.8954,
      "step": 2604
    },
    {
      "epoch": 0.09271965938280728,
      "grad_norm": 1.3538340330123901,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.9035,
      "step": 2635
    },
    {
      "epoch": 0.09381047890495794,
      "grad_norm": 1.407291054725647,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.8778,
      "step": 2666
    },
    {
      "epoch": 0.09490129842710862,
      "grad_norm": 1.3554202318191528,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.8783,
      "step": 2697
    },
    {
      "epoch": 0.0959921179492593,
      "grad_norm": 1.3513182401657104,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.8857,
      "step": 2728
    },
    {
      "epoch": 0.09708293747140997,
      "grad_norm": 1.345706582069397,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.8882,
      "step": 2759
    },
    {
      "epoch": 0.09817375699356065,
      "grad_norm": 1.3450254201889038,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.8746,
      "step": 2790
    },
    {
      "epoch": 0.09926457651571131,
      "grad_norm": 1.7820961475372314,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.8557,
      "step": 2821
    },
    {
      "epoch": 0.10035539603786199,
      "grad_norm": 1.3493785858154297,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.8683,
      "step": 2852
    },
    {
      "epoch": 0.10144621556001267,
      "grad_norm": 1.3066306114196777,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.8777,
      "step": 2883
    },
    {
      "epoch": 0.10253703508216334,
      "grad_norm": 1.2991000413894653,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.8605,
      "step": 2914
    },
    {
      "epoch": 0.10362785460431402,
      "grad_norm": 1.3748496770858765,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.8683,
      "step": 2945
    },
    {
      "epoch": 0.10471867412646468,
      "grad_norm": 1.3229252099990845,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.858,
      "step": 2976
    },
    {
      "epoch": 0.10580949364861536,
      "grad_norm": 1.2874318361282349,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.8435,
      "step": 3007
    },
    {
      "epoch": 0.10690031317076604,
      "grad_norm": 1.3341971635818481,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.8663,
      "step": 3038
    },
    {
      "epoch": 0.10799113269291671,
      "grad_norm": 1.276282548904419,
      "learning_rate": 4.965133917685858e-05,
      "loss": 1.854,
      "step": 3069
    },
    {
      "epoch": 0.10908195221506739,
      "grad_norm": 1.2999404668807983,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 1.8498,
      "step": 3100
    },
    {
      "epoch": 0.11017277173721805,
      "grad_norm": 1.2905164957046509,
      "learning_rate": 4.962282892045718e-05,
      "loss": 1.8436,
      "step": 3131
    },
    {
      "epoch": 0.11126359125936873,
      "grad_norm": 1.9952430725097656,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 1.8316,
      "step": 3162
    },
    {
      "epoch": 0.1123544107815194,
      "grad_norm": 1.2637739181518555,
      "learning_rate": 4.959320720608049e-05,
      "loss": 1.8391,
      "step": 3193
    },
    {
      "epoch": 0.11344523030367008,
      "grad_norm": 1.3711034059524536,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 1.814,
      "step": 3224
    },
    {
      "epoch": 0.11453604982582076,
      "grad_norm": 1.3050326108932495,
      "learning_rate": 4.956247537083282e-05,
      "loss": 1.8635,
      "step": 3255
    },
    {
      "epoch": 0.11562686934797142,
      "grad_norm": 1.230220079421997,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 1.8203,
      "step": 3286
    },
    {
      "epoch": 0.1167176888701221,
      "grad_norm": 1.2535375356674194,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 1.8319,
      "step": 3317
    },
    {
      "epoch": 0.11780850839227278,
      "grad_norm": 1.227552890777588,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 1.8157,
      "step": 3348
    },
    {
      "epoch": 0.11889932791442345,
      "grad_norm": 1.2049622535705566,
      "learning_rate": 4.949768693662973e-05,
      "loss": 1.8135,
      "step": 3379
    },
    {
      "epoch": 0.11999014743657413,
      "grad_norm": 1.1608529090881348,
      "learning_rate": 4.948079823064559e-05,
      "loss": 1.8191,
      "step": 3410
    },
    {
      "epoch": 0.12108096695872479,
      "grad_norm": 1.2294845581054688,
      "learning_rate": 4.946363326218074e-05,
      "loss": 1.7977,
      "step": 3441
    },
    {
      "epoch": 0.12217178648087547,
      "grad_norm": 1.1883920431137085,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 1.8153,
      "step": 3472
    },
    {
      "epoch": 0.12326260600302615,
      "grad_norm": 1.2770321369171143,
      "learning_rate": 4.942847531574167e-05,
      "loss": 1.8064,
      "step": 3503
    },
    {
      "epoch": 0.12435342552517682,
      "grad_norm": 1.5442144870758057,
      "learning_rate": 4.941048273452008e-05,
      "loss": 1.8134,
      "step": 3534
    },
    {
      "epoch": 0.12544424504732749,
      "grad_norm": 1.1753895282745361,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 1.7909,
      "step": 3565
    },
    {
      "epoch": 0.12653506456947816,
      "grad_norm": 1.2292894124984741,
      "learning_rate": 4.93736713712897e-05,
      "loss": 1.7894,
      "step": 3596
    },
    {
      "epoch": 0.12762588409162884,
      "grad_norm": 1.1696261167526245,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 1.7913,
      "step": 3627
    },
    {
      "epoch": 0.12871670361377952,
      "grad_norm": 1.1935744285583496,
      "learning_rate": 4.93357597968886e-05,
      "loss": 1.8103,
      "step": 3658
    },
    {
      "epoch": 0.1298075231359302,
      "grad_norm": 1.1842865943908691,
      "learning_rate": 4.931639196334338e-05,
      "loss": 1.7778,
      "step": 3689
    },
    {
      "epoch": 0.13089834265808087,
      "grad_norm": 1.187715768814087,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 1.7955,
      "step": 3720
    },
    {
      "epoch": 0.13198916218023155,
      "grad_norm": 1.1973644495010376,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 1.7786,
      "step": 3751
    },
    {
      "epoch": 0.13307998170238222,
      "grad_norm": 1.129618525505066,
      "learning_rate": 4.925664290937196e-05,
      "loss": 1.7819,
      "step": 3782
    },
    {
      "epoch": 0.13417080122453287,
      "grad_norm": 1.1466953754425049,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 1.7732,
      "step": 3813
    },
    {
      "epoch": 0.13526162074668355,
      "grad_norm": 1.1432143449783325,
      "learning_rate": 4.921544116753962e-05,
      "loss": 1.7872,
      "step": 3844
    },
    {
      "epoch": 0.13635244026883422,
      "grad_norm": 1.1669626235961914,
      "learning_rate": 4.919443027766935e-05,
      "loss": 1.7708,
      "step": 3875
    },
    {
      "epoch": 0.1374432597909849,
      "grad_norm": 1.1273548603057861,
      "learning_rate": 4.91731463569418e-05,
      "loss": 1.7767,
      "step": 3906
    },
    {
      "epoch": 0.13853407931313558,
      "grad_norm": 1.146063208580017,
      "learning_rate": 4.915158964554312e-05,
      "loss": 1.7695,
      "step": 3937
    },
    {
      "epoch": 0.13962489883528625,
      "grad_norm": 1.1724731922149658,
      "learning_rate": 4.912976038673786e-05,
      "loss": 1.7709,
      "step": 3968
    },
    {
      "epoch": 0.14071571835743693,
      "grad_norm": 1.122382402420044,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 1.7871,
      "step": 3999
    },
    {
      "epoch": 0.1418065378795876,
      "grad_norm": 1.2005928754806519,
      "learning_rate": 4.908528521534139e-05,
      "loss": 1.7896,
      "step": 4030
    },
    {
      "epoch": 0.14289735740173828,
      "grad_norm": 1.1000254154205322,
      "learning_rate": 4.906263980464644e-05,
      "loss": 1.7645,
      "step": 4061
    },
    {
      "epoch": 0.14398817692388896,
      "grad_norm": 1.1797596216201782,
      "learning_rate": 4.903972285033178e-05,
      "loss": 1.7702,
      "step": 4092
    },
    {
      "epoch": 0.1450789964460396,
      "grad_norm": 1.145271897315979,
      "learning_rate": 4.901653461101213e-05,
      "loss": 1.7541,
      "step": 4123
    },
    {
      "epoch": 0.1461698159681903,
      "grad_norm": 1.1336746215820312,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 1.7604,
      "step": 4154
    },
    {
      "epoch": 0.14726063549034096,
      "grad_norm": 1.1435718536376953,
      "learning_rate": 4.896934532712084e-05,
      "loss": 1.7461,
      "step": 4185
    },
    {
      "epoch": 0.14835145501249164,
      "grad_norm": 1.1105029582977295,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 1.7691,
      "step": 4216
    },
    {
      "epoch": 0.14944227453464232,
      "grad_norm": 1.1397055387496948,
      "learning_rate": 4.892107408306516e-05,
      "loss": 1.7599,
      "step": 4247
    },
    {
      "epoch": 0.150533094056793,
      "grad_norm": 1.1158910989761353,
      "learning_rate": 4.889653340498669e-05,
      "loss": 1.7537,
      "step": 4278
    },
    {
      "epoch": 0.15162391357894367,
      "grad_norm": 1.1238468885421753,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 1.7672,
      "step": 4309
    },
    {
      "epoch": 0.15271473310109435,
      "grad_norm": 1.1303473711013794,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 1.7542,
      "step": 4340
    },
    {
      "epoch": 0.15380555262324502,
      "grad_norm": 1.117262840270996,
      "learning_rate": 4.882129447892753e-05,
      "loss": 1.7378,
      "step": 4371
    },
    {
      "epoch": 0.15489637214539567,
      "grad_norm": 1.1086496114730835,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 1.7439,
      "step": 4402
    },
    {
      "epoch": 0.15598719166754635,
      "grad_norm": 1.276014804840088,
      "learning_rate": 4.876979062282995e-05,
      "loss": 1.7699,
      "step": 4433
    },
    {
      "epoch": 0.15707801118969703,
      "grad_norm": 1.1294457912445068,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 1.7292,
      "step": 4464
    },
    {
      "epoch": 0.1581688307118477,
      "grad_norm": 1.1111900806427002,
      "learning_rate": 4.871721381433344e-05,
      "loss": 1.7487,
      "step": 4495
    },
    {
      "epoch": 0.15925965023399838,
      "grad_norm": 1.0868422985076904,
      "learning_rate": 4.869052379269719e-05,
      "loss": 1.7397,
      "step": 4526
    },
    {
      "epoch": 0.16035046975614906,
      "grad_norm": 1.0772205591201782,
      "learning_rate": 4.866356642671985e-05,
      "loss": 1.7377,
      "step": 4557
    },
    {
      "epoch": 0.16144128927829973,
      "grad_norm": 1.0649960041046143,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 1.7253,
      "step": 4588
    },
    {
      "epoch": 0.1625321088004504,
      "grad_norm": 1.0653575658798218,
      "learning_rate": 4.860885088159626e-05,
      "loss": 1.7319,
      "step": 4619
    },
    {
      "epoch": 0.16362292832260109,
      "grad_norm": 1.0669118165969849,
      "learning_rate": 4.858109331990751e-05,
      "loss": 1.7425,
      "step": 4650
    },
    {
      "epoch": 0.16471374784475176,
      "grad_norm": 1.0559724569320679,
      "learning_rate": 4.855306964878567e-05,
      "loss": 1.7364,
      "step": 4681
    },
    {
      "epoch": 0.1658045673669024,
      "grad_norm": 1.038191318511963,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 1.7299,
      "step": 4712
    },
    {
      "epoch": 0.1668953868890531,
      "grad_norm": 1.0602952241897583,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 1.7256,
      "step": 4743
    },
    {
      "epoch": 0.16798620641120376,
      "grad_norm": 1.0466082096099854,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 1.7148,
      "step": 4774
    },
    {
      "epoch": 0.16907702593335444,
      "grad_norm": 1.051708698272705,
      "learning_rate": 4.843832023980392e-05,
      "loss": 1.722,
      "step": 4805
    },
    {
      "epoch": 0.17016784545550512,
      "grad_norm": 1.065427541732788,
      "learning_rate": 4.840897082510106e-05,
      "loss": 1.7088,
      "step": 4836
    },
    {
      "epoch": 0.1712586649776558,
      "grad_norm": 1.0983659029006958,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 1.7422,
      "step": 4867
    },
    {
      "epoch": 0.17234948449980647,
      "grad_norm": 1.0919371843338013,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 1.7253,
      "step": 4898
    },
    {
      "epoch": 0.17344030402195715,
      "grad_norm": 1.0881919860839844,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 1.7185,
      "step": 4929
    },
    {
      "epoch": 0.17453112354410782,
      "grad_norm": 1.0374606847763062,
      "learning_rate": 4.828893485248369e-05,
      "loss": 1.7255,
      "step": 4960
    },
    {
      "epoch": 0.1756219430662585,
      "grad_norm": 1.0947892665863037,
      "learning_rate": 4.825826797411682e-05,
      "loss": 1.7095,
      "step": 4991
    },
    {
      "epoch": 0.17671276258840915,
      "grad_norm": 1.1086746454238892,
      "learning_rate": 4.822733862935702e-05,
      "loss": 1.7233,
      "step": 5022
    },
    {
      "epoch": 0.17780358211055983,
      "grad_norm": 1.0662850141525269,
      "learning_rate": 4.819614716723775e-05,
      "loss": 1.7426,
      "step": 5053
    },
    {
      "epoch": 0.1788944016327105,
      "grad_norm": 1.0137659311294556,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 1.7358,
      "step": 5084
    },
    {
      "epoch": 0.17998522115486118,
      "grad_norm": 1.049338698387146,
      "learning_rate": 4.813297930184042e-05,
      "loss": 1.6981,
      "step": 5115
    },
    {
      "epoch": 0.18107604067701186,
      "grad_norm": 1.08591628074646,
      "learning_rate": 4.810100361140314e-05,
      "loss": 1.7178,
      "step": 5146
    },
    {
      "epoch": 0.18216686019916253,
      "grad_norm": 1.0983690023422241,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 1.7179,
      "step": 5177
    },
    {
      "epoch": 0.1832576797213132,
      "grad_norm": 1.0783159732818604,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 1.7198,
      "step": 5208
    },
    {
      "epoch": 0.1843484992434639,
      "grad_norm": 1.0519956350326538,
      "learning_rate": 4.8003513848046e-05,
      "loss": 1.7154,
      "step": 5239
    },
    {
      "epoch": 0.18543931876561456,
      "grad_norm": 1.0484226942062378,
      "learning_rate": 4.79704975853109e-05,
      "loss": 1.7081,
      "step": 5270
    },
    {
      "epoch": 0.18653013828776524,
      "grad_norm": 1.0263608694076538,
      "learning_rate": 4.793722210363262e-05,
      "loss": 1.6961,
      "step": 5301
    },
    {
      "epoch": 0.1876209578099159,
      "grad_norm": 0.9863409399986267,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 1.7198,
      "step": 5332
    },
    {
      "epoch": 0.18871177733206657,
      "grad_norm": 1.0293922424316406,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 1.6958,
      "step": 5363
    },
    {
      "epoch": 0.18980259685421724,
      "grad_norm": 0.9902015924453735,
      "learning_rate": 4.783584411463221e-05,
      "loss": 1.7162,
      "step": 5394
    },
    {
      "epoch": 0.19089341637636792,
      "grad_norm": 1.0472259521484375,
      "learning_rate": 4.780153554146274e-05,
      "loss": 1.7189,
      "step": 5425
    },
    {
      "epoch": 0.1919842358985186,
      "grad_norm": 1.0958000421524048,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 1.7288,
      "step": 5456
    },
    {
      "epoch": 0.19307505542066927,
      "grad_norm": 1.0132781267166138,
      "learning_rate": 4.773214684850662e-05,
      "loss": 1.6915,
      "step": 5487
    },
    {
      "epoch": 0.19416587494281995,
      "grad_norm": 1.0274685621261597,
      "learning_rate": 4.769706751176193e-05,
      "loss": 1.7063,
      "step": 5518
    },
    {
      "epoch": 0.19525669446497063,
      "grad_norm": 1.0145214796066284,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 1.7072,
      "step": 5549
    },
    {
      "epoch": 0.1963475139871213,
      "grad_norm": 1.053489327430725,
      "learning_rate": 4.762614083706258e-05,
      "loss": 1.6897,
      "step": 5580
    },
    {
      "epoch": 0.19743833350927198,
      "grad_norm": 1.0426928997039795,
      "learning_rate": 4.759029429950581e-05,
      "loss": 1.6868,
      "step": 5611
    },
    {
      "epoch": 0.19852915303142263,
      "grad_norm": 1.0594595670700073,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 1.6978,
      "step": 5642
    },
    {
      "epoch": 0.1996199725535733,
      "grad_norm": 1.0251655578613281,
      "learning_rate": 4.751783684659e-05,
      "loss": 1.6834,
      "step": 5673
    },
    {
      "epoch": 0.20071079207572398,
      "grad_norm": 1.0762938261032104,
      "learning_rate": 4.748122674890348e-05,
      "loss": 1.6825,
      "step": 5704
    },
    {
      "epoch": 0.20180161159787466,
      "grad_norm": 1.0672742128372192,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 1.6811,
      "step": 5735
    },
    {
      "epoch": 0.20289243112002533,
      "grad_norm": 1.0267951488494873,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 1.7024,
      "step": 5766
    },
    {
      "epoch": 0.203983250642176,
      "grad_norm": 1.0520440340042114,
      "learning_rate": 4.736987593718397e-05,
      "loss": 1.7008,
      "step": 5797
    },
    {
      "epoch": 0.2050740701643267,
      "grad_norm": 1.0106167793273926,
      "learning_rate": 4.733225355658999e-05,
      "loss": 1.6721,
      "step": 5828
    },
    {
      "epoch": 0.20616488968647737,
      "grad_norm": 0.9883444905281067,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 1.6884,
      "step": 5859
    },
    {
      "epoch": 0.20725570920862804,
      "grad_norm": 0.9873078465461731,
      "learning_rate": 4.725625317347119e-05,
      "loss": 1.7011,
      "step": 5890
    },
    {
      "epoch": 0.2083465287307787,
      "grad_norm": 0.9925514459609985,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 1.6965,
      "step": 5921
    },
    {
      "epoch": 0.20943734825292937,
      "grad_norm": 1.0054575204849243,
      "learning_rate": 4.717924815801832e-05,
      "loss": 1.6923,
      "step": 5952
    },
    {
      "epoch": 0.21052816777508004,
      "grad_norm": 1.0051021575927734,
      "learning_rate": 4.714036999763532e-05,
      "loss": 1.6864,
      "step": 5983
    },
    {
      "epoch": 0.21161898729723072,
      "grad_norm": 1.0312145948410034,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 1.6776,
      "step": 6014
    },
    {
      "epoch": 0.2127098068193814,
      "grad_norm": 0.9964466094970703,
      "learning_rate": 4.7061864565225e-05,
      "loss": 1.6811,
      "step": 6045
    },
    {
      "epoch": 0.21380062634153207,
      "grad_norm": 0.9997463226318359,
      "learning_rate": 4.702223817912081e-05,
      "loss": 1.6805,
      "step": 6076
    }
  ],
  "logging_steps": 31,
  "max_steps": 30517,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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
}