{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.25752014512930854,
  "eval_steps": 500,
  "global_step": 6104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0013078513268362654,
      "grad_norm": 31.18772315979004,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 9.3832,
      "step": 31
    },
    {
      "epoch": 0.0026157026536725307,
      "grad_norm": 24.045747756958008,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.9011,
      "step": 62
    },
    {
      "epoch": 0.003923553980508796,
      "grad_norm": 16.459999084472656,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 6.3655,
      "step": 93
    },
    {
      "epoch": 0.0052314053073450615,
      "grad_norm": 5.892439365386963,
      "learning_rate": 4.062909567496723e-06,
      "loss": 5.1577,
      "step": 124
    },
    {
      "epoch": 0.0065392566341813275,
      "grad_norm": 7.445889949798584,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.4767,
      "step": 155
    },
    {
      "epoch": 0.007847107961017593,
      "grad_norm": 8.36017894744873,
      "learning_rate": 6.094364351245085e-06,
      "loss": 4.0756,
      "step": 186
    },
    {
      "epoch": 0.009154959287853858,
      "grad_norm": 13.23757266998291,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.7893,
      "step": 217
    },
    {
      "epoch": 0.010462810614690123,
      "grad_norm": 10.12263011932373,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.6154,
      "step": 248
    },
    {
      "epoch": 0.01177066194152639,
      "grad_norm": 6.559150218963623,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.4391,
      "step": 279
    },
    {
      "epoch": 0.013078513268362655,
      "grad_norm": 6.898993015289307,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.3229,
      "step": 310
    },
    {
      "epoch": 0.01438636459519892,
      "grad_norm": 8.434200286865234,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.225,
      "step": 341
    },
    {
      "epoch": 0.015694215922035185,
      "grad_norm": 7.735746383666992,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.1381,
      "step": 372
    },
    {
      "epoch": 0.01700206724887145,
      "grad_norm": 7.964824676513672,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 3.06,
      "step": 403
    },
    {
      "epoch": 0.018309918575707716,
      "grad_norm": 6.081613063812256,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 3.0052,
      "step": 434
    },
    {
      "epoch": 0.01961776990254398,
      "grad_norm": 6.740341663360596,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 2.9035,
      "step": 465
    },
    {
      "epoch": 0.020925621229380246,
      "grad_norm": 4.908828258514404,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.8602,
      "step": 496
    },
    {
      "epoch": 0.02223347255621651,
      "grad_norm": 5.6150360107421875,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.8223,
      "step": 527
    },
    {
      "epoch": 0.02354132388305278,
      "grad_norm": 3.5981647968292236,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.7488,
      "step": 558
    },
    {
      "epoch": 0.024849175209889045,
      "grad_norm": 4.921584606170654,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.741,
      "step": 589
    },
    {
      "epoch": 0.02615702653672531,
      "grad_norm": 3.8303065299987793,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.6923,
      "step": 620
    },
    {
      "epoch": 0.027464877863561575,
      "grad_norm": 3.3674638271331787,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.6402,
      "step": 651
    },
    {
      "epoch": 0.02877272919039784,
      "grad_norm": 4.096519947052002,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.6372,
      "step": 682
    },
    {
      "epoch": 0.030080580517234105,
      "grad_norm": 3.10866641998291,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.5979,
      "step": 713
    },
    {
      "epoch": 0.03138843184407037,
      "grad_norm": 3.3147943019866943,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.549,
      "step": 744
    },
    {
      "epoch": 0.032696283170906636,
      "grad_norm": 3.045280933380127,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.5487,
      "step": 775
    },
    {
      "epoch": 0.0340041344977429,
      "grad_norm": 3.2536299228668213,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.5138,
      "step": 806
    },
    {
      "epoch": 0.035311985824579166,
      "grad_norm": 3.108548641204834,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.5006,
      "step": 837
    },
    {
      "epoch": 0.03661983715141543,
      "grad_norm": 3.048689603805542,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.4434,
      "step": 868
    },
    {
      "epoch": 0.037927688478251696,
      "grad_norm": 2.8378429412841797,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.4403,
      "step": 899
    },
    {
      "epoch": 0.03923553980508796,
      "grad_norm": 2.731419324874878,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.4028,
      "step": 930
    },
    {
      "epoch": 0.04054339113192423,
      "grad_norm": 2.4825587272644043,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.3695,
      "step": 961
    },
    {
      "epoch": 0.04185124245876049,
      "grad_norm": 2.5020129680633545,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.3513,
      "step": 992
    },
    {
      "epoch": 0.04315909378559676,
      "grad_norm": 2.374310255050659,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.3221,
      "step": 1023
    },
    {
      "epoch": 0.04446694511243302,
      "grad_norm": 2.4097862243652344,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.3366,
      "step": 1054
    },
    {
      "epoch": 0.045774796439269294,
      "grad_norm": 2.573655128479004,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.2997,
      "step": 1085
    },
    {
      "epoch": 0.04708264776610556,
      "grad_norm": 2.5408434867858887,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.2729,
      "step": 1116
    },
    {
      "epoch": 0.048390499092941824,
      "grad_norm": 2.222527265548706,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.2716,
      "step": 1147
    },
    {
      "epoch": 0.04969835041977809,
      "grad_norm": 2.342715263366699,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.2661,
      "step": 1178
    },
    {
      "epoch": 0.051006201746614355,
      "grad_norm": 2.134208917617798,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.2206,
      "step": 1209
    },
    {
      "epoch": 0.05231405307345062,
      "grad_norm": 2.2482688426971436,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.2207,
      "step": 1240
    },
    {
      "epoch": 0.053621904400286885,
      "grad_norm": 2.2316832542419434,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.1847,
      "step": 1271
    },
    {
      "epoch": 0.05492975572712315,
      "grad_norm": 2.2321085929870605,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.2079,
      "step": 1302
    },
    {
      "epoch": 0.056237607053959415,
      "grad_norm": 2.115349054336548,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.1458,
      "step": 1333
    },
    {
      "epoch": 0.05754545838079568,
      "grad_norm": 2.210914373397827,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.1435,
      "step": 1364
    },
    {
      "epoch": 0.058853309707631946,
      "grad_norm": 2.1882164478302,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.1389,
      "step": 1395
    },
    {
      "epoch": 0.06016116103446821,
      "grad_norm": 2.01705265045166,
      "learning_rate": 4.672346002621232e-05,
      "loss": 2.1599,
      "step": 1426
    },
    {
      "epoch": 0.061469012361304476,
      "grad_norm": 1.9447391033172607,
      "learning_rate": 4.77391874180865e-05,
      "loss": 2.1254,
      "step": 1457
    },
    {
      "epoch": 0.06277686368814074,
      "grad_norm": 1.9278348684310913,
      "learning_rate": 4.875491480996068e-05,
      "loss": 2.1109,
      "step": 1488
    },
    {
      "epoch": 0.064084715014977,
      "grad_norm": 167.2142791748047,
      "learning_rate": 4.977064220183487e-05,
      "loss": 2.12,
      "step": 1519
    },
    {
      "epoch": 0.06539256634181327,
      "grad_norm": 15.058513641357422,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 3.2174,
      "step": 1550
    },
    {
      "epoch": 0.06670041766864954,
      "grad_norm": 2.8855061531066895,
      "learning_rate": 4.999955597496219e-05,
      "loss": 2.191,
      "step": 1581
    },
    {
      "epoch": 0.0680082689954858,
      "grad_norm": 1.9484020471572876,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 2.1053,
      "step": 1612
    },
    {
      "epoch": 0.06931612032232207,
      "grad_norm": 5.792853832244873,
      "learning_rate": 4.999799067923527e-05,
      "loss": 2.1058,
      "step": 1643
    },
    {
      "epoch": 0.07062397164915833,
      "grad_norm": 5.8536553382873535,
      "learning_rate": 4.999678487776908e-05,
      "loss": 2.0906,
      "step": 1674
    },
    {
      "epoch": 0.0719318229759946,
      "grad_norm": 2.8606152534484863,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 2.1356,
      "step": 1705
    },
    {
      "epoch": 0.07323967430283086,
      "grad_norm": 2.423034191131592,
      "learning_rate": 4.999352703566763e-05,
      "loss": 2.0468,
      "step": 1736
    },
    {
      "epoch": 0.07454752562966713,
      "grad_norm": 1.7297399044036865,
      "learning_rate": 4.999147503179668e-05,
      "loss": 2.0687,
      "step": 1767
    },
    {
      "epoch": 0.07585537695650339,
      "grad_norm": 4.29679536819458,
      "learning_rate": 4.998914100252672e-05,
      "loss": 2.026,
      "step": 1798
    },
    {
      "epoch": 0.07716322828333966,
      "grad_norm": 1.942807674407959,
      "learning_rate": 4.998652497419696e-05,
      "loss": 2.034,
      "step": 1829
    },
    {
      "epoch": 0.07847107961017592,
      "grad_norm": 1.8823152780532837,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 2.0186,
      "step": 1860
    },
    {
      "epoch": 0.07977893093701219,
      "grad_norm": 2.135873556137085,
      "learning_rate": 4.998044704162613e-05,
      "loss": 2.0047,
      "step": 1891
    },
    {
      "epoch": 0.08108678226384845,
      "grad_norm": 3.113102436065674,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 1.9954,
      "step": 1922
    },
    {
      "epoch": 0.08239463359068472,
      "grad_norm": 2.253596305847168,
      "learning_rate": 4.997324150843799e-05,
      "loss": 1.9744,
      "step": 1953
    },
    {
      "epoch": 0.08370248491752098,
      "grad_norm": 5.765998363494873,
      "learning_rate": 4.99692159912661e-05,
      "loss": 1.9908,
      "step": 1984
    },
    {
      "epoch": 0.08501033624435725,
      "grad_norm": 2.8442440032958984,
      "learning_rate": 4.996490869988546e-05,
      "loss": 1.9803,
      "step": 2015
    },
    {
      "epoch": 0.08631818757119351,
      "grad_norm": 1.7528963088989258,
      "learning_rate": 4.996031968290326e-05,
      "loss": 1.9419,
      "step": 2046
    },
    {
      "epoch": 0.08762603889802978,
      "grad_norm": 1.7754414081573486,
      "learning_rate": 4.995544899210594e-05,
      "loss": 1.9485,
      "step": 2077
    },
    {
      "epoch": 0.08893389022486604,
      "grad_norm": 1.6553430557250977,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 1.9469,
      "step": 2108
    },
    {
      "epoch": 0.09024174155170232,
      "grad_norm": 1.7296561002731323,
      "learning_rate": 4.994486281210429e-05,
      "loss": 1.9195,
      "step": 2139
    },
    {
      "epoch": 0.09154959287853859,
      "grad_norm": 1.7001283168792725,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 1.9324,
      "step": 2170
    },
    {
      "epoch": 0.09285744420537485,
      "grad_norm": 7.332984447479248,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 1.9171,
      "step": 2201
    },
    {
      "epoch": 0.09416529553221112,
      "grad_norm": 1.6546969413757324,
      "learning_rate": 4.992687246588743e-05,
      "loss": 1.9152,
      "step": 2232
    },
    {
      "epoch": 0.09547314685904738,
      "grad_norm": 1.843203067779541,
      "learning_rate": 4.992031299767347e-05,
      "loss": 1.9015,
      "step": 2263
    },
    {
      "epoch": 0.09678099818588365,
      "grad_norm": 4.51094388961792,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 1.917,
      "step": 2294
    },
    {
      "epoch": 0.09808884951271991,
      "grad_norm": 1.5267099142074585,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 1.8985,
      "step": 2325
    },
    {
      "epoch": 0.09939670083955618,
      "grad_norm": 2.0826663970947266,
      "learning_rate": 4.989894757091861e-05,
      "loss": 1.8968,
      "step": 2356
    },
    {
      "epoch": 0.10070455216639244,
      "grad_norm": 1.4929964542388916,
      "learning_rate": 4.989126368919158e-05,
      "loss": 1.8711,
      "step": 2387
    },
    {
      "epoch": 0.10201240349322871,
      "grad_norm": 1.478498935699463,
      "learning_rate": 4.988329891293693e-05,
      "loss": 1.8672,
      "step": 2418
    },
    {
      "epoch": 0.10332025482006497,
      "grad_norm": 1.4785969257354736,
      "learning_rate": 4.987505333203608e-05,
      "loss": 1.8435,
      "step": 2449
    },
    {
      "epoch": 0.10462810614690124,
      "grad_norm": 1.524807333946228,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.8543,
      "step": 2480
    },
    {
      "epoch": 0.1059359574737375,
      "grad_norm": 1.4232748746871948,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 1.8463,
      "step": 2511
    },
    {
      "epoch": 0.10724380880057377,
      "grad_norm": 1.7178603410720825,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 1.8512,
      "step": 2542
    },
    {
      "epoch": 0.10855166012741004,
      "grad_norm": 1.4624053239822388,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.8421,
      "step": 2573
    },
    {
      "epoch": 0.1098595114542463,
      "grad_norm": 2.0191211700439453,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.8393,
      "step": 2604
    },
    {
      "epoch": 0.11116736278108257,
      "grad_norm": 1.4346847534179688,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.8149,
      "step": 2635
    },
    {
      "epoch": 0.11247521410791883,
      "grad_norm": 1.4674941301345825,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.8215,
      "step": 2666
    },
    {
      "epoch": 0.1137830654347551,
      "grad_norm": 1.4122453927993774,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.8265,
      "step": 2697
    },
    {
      "epoch": 0.11509091676159136,
      "grad_norm": 1.4748708009719849,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.8149,
      "step": 2728
    },
    {
      "epoch": 0.11639876808842763,
      "grad_norm": 1.40476655960083,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.8316,
      "step": 2759
    },
    {
      "epoch": 0.11770661941526389,
      "grad_norm": 1.4909706115722656,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.8208,
      "step": 2790
    },
    {
      "epoch": 0.11901447074210016,
      "grad_norm": 1.3661363124847412,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.8038,
      "step": 2821
    },
    {
      "epoch": 0.12032232206893642,
      "grad_norm": 1.422453761100769,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.7926,
      "step": 2852
    },
    {
      "epoch": 0.12163017339577269,
      "grad_norm": 1.3766013383865356,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.8192,
      "step": 2883
    },
    {
      "epoch": 0.12293802472260895,
      "grad_norm": 1.3904545307159424,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.7653,
      "step": 2914
    },
    {
      "epoch": 0.12424587604944522,
      "grad_norm": 1.369018316268921,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.7879,
      "step": 2945
    },
    {
      "epoch": 0.12555372737628148,
      "grad_norm": 1.5040982961654663,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.7777,
      "step": 2976
    },
    {
      "epoch": 0.12686157870311776,
      "grad_norm": 1.3161810636520386,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.7654,
      "step": 3007
    },
    {
      "epoch": 0.128169430029954,
      "grad_norm": 1.3379234075546265,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.7699,
      "step": 3038
    },
    {
      "epoch": 0.1294772813567903,
      "grad_norm": 1.3384974002838135,
      "learning_rate": 4.965133917685858e-05,
      "loss": 1.7915,
      "step": 3069
    },
    {
      "epoch": 0.13078513268362654,
      "grad_norm": 1.2991316318511963,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 1.7636,
      "step": 3100
    },
    {
      "epoch": 0.13209298401046282,
      "grad_norm": 2.500105857849121,
      "learning_rate": 4.962282892045718e-05,
      "loss": 1.7573,
      "step": 3131
    },
    {
      "epoch": 0.13340083533729907,
      "grad_norm": 1.2635610103607178,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 1.7537,
      "step": 3162
    },
    {
      "epoch": 0.13470868666413535,
      "grad_norm": 1.2823940515518188,
      "learning_rate": 4.959320720608049e-05,
      "loss": 1.7709,
      "step": 3193
    },
    {
      "epoch": 0.1360165379909716,
      "grad_norm": 1.2606983184814453,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 1.7611,
      "step": 3224
    },
    {
      "epoch": 0.13732438931780788,
      "grad_norm": 1.305782437324524,
      "learning_rate": 4.956247537083282e-05,
      "loss": 1.7547,
      "step": 3255
    },
    {
      "epoch": 0.13863224064464413,
      "grad_norm": 1.2986756563186646,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 1.7241,
      "step": 3286
    },
    {
      "epoch": 0.1399400919714804,
      "grad_norm": 1.3163083791732788,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 1.7352,
      "step": 3317
    },
    {
      "epoch": 0.14124794329831666,
      "grad_norm": 1.2844524383544922,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 1.7316,
      "step": 3348
    },
    {
      "epoch": 0.14255579462515294,
      "grad_norm": 1.2861442565917969,
      "learning_rate": 4.949768693662973e-05,
      "loss": 1.7233,
      "step": 3379
    },
    {
      "epoch": 0.1438636459519892,
      "grad_norm": 1.2419766187667847,
      "learning_rate": 4.948079823064559e-05,
      "loss": 1.7278,
      "step": 3410
    },
    {
      "epoch": 0.14517149727882547,
      "grad_norm": 1.2518367767333984,
      "learning_rate": 4.946363326218074e-05,
      "loss": 1.7399,
      "step": 3441
    },
    {
      "epoch": 0.14647934860566172,
      "grad_norm": 1.217373251914978,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 1.7386,
      "step": 3472
    },
    {
      "epoch": 0.147787199932498,
      "grad_norm": 1.2229036092758179,
      "learning_rate": 4.942847531574167e-05,
      "loss": 1.7096,
      "step": 3503
    },
    {
      "epoch": 0.14909505125933425,
      "grad_norm": 1.2012135982513428,
      "learning_rate": 4.941048273452008e-05,
      "loss": 1.7195,
      "step": 3534
    },
    {
      "epoch": 0.15040290258617053,
      "grad_norm": 1.2632161378860474,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 1.7377,
      "step": 3565
    },
    {
      "epoch": 0.15171075391300679,
      "grad_norm": 1.2119094133377075,
      "learning_rate": 4.93736713712897e-05,
      "loss": 1.7276,
      "step": 3596
    },
    {
      "epoch": 0.15301860523984306,
      "grad_norm": 1.2516822814941406,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 1.7224,
      "step": 3627
    },
    {
      "epoch": 0.15432645656667932,
      "grad_norm": 1.2437561750411987,
      "learning_rate": 4.93357597968886e-05,
      "loss": 1.703,
      "step": 3658
    },
    {
      "epoch": 0.1556343078935156,
      "grad_norm": 1.4856292009353638,
      "learning_rate": 4.931639196334338e-05,
      "loss": 1.7139,
      "step": 3689
    },
    {
      "epoch": 0.15694215922035185,
      "grad_norm": 1.202919840812683,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 1.7081,
      "step": 3720
    },
    {
      "epoch": 0.15825001054718812,
      "grad_norm": 1.1720086336135864,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 1.7086,
      "step": 3751
    },
    {
      "epoch": 0.15955786187402438,
      "grad_norm": 1.2187249660491943,
      "learning_rate": 4.925664290937196e-05,
      "loss": 1.6871,
      "step": 3782
    },
    {
      "epoch": 0.16086571320086065,
      "grad_norm": 1.1918401718139648,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 1.7058,
      "step": 3813
    },
    {
      "epoch": 0.1621735645276969,
      "grad_norm": 1.1956909894943237,
      "learning_rate": 4.921544116753962e-05,
      "loss": 1.6986,
      "step": 3844
    },
    {
      "epoch": 0.16348141585453319,
      "grad_norm": 1.1646822690963745,
      "learning_rate": 4.919443027766935e-05,
      "loss": 1.6797,
      "step": 3875
    },
    {
      "epoch": 0.16478926718136944,
      "grad_norm": 1.2145137786865234,
      "learning_rate": 4.91731463569418e-05,
      "loss": 1.678,
      "step": 3906
    },
    {
      "epoch": 0.16609711850820572,
      "grad_norm": 1.2347453832626343,
      "learning_rate": 4.915158964554312e-05,
      "loss": 1.672,
      "step": 3937
    },
    {
      "epoch": 0.16740496983504197,
      "grad_norm": 1.1764047145843506,
      "learning_rate": 4.912976038673786e-05,
      "loss": 1.6864,
      "step": 3968
    },
    {
      "epoch": 0.16871282116187825,
      "grad_norm": 1.1730256080627441,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 1.7021,
      "step": 3999
    },
    {
      "epoch": 0.1700206724887145,
      "grad_norm": 1.1880205869674683,
      "learning_rate": 4.908528521534139e-05,
      "loss": 1.6832,
      "step": 4030
    },
    {
      "epoch": 0.17132852381555078,
      "grad_norm": 1.1810024976730347,
      "learning_rate": 4.906263980464644e-05,
      "loss": 1.6674,
      "step": 4061
    },
    {
      "epoch": 0.17263637514238703,
      "grad_norm": 1.2133933305740356,
      "learning_rate": 4.903972285033178e-05,
      "loss": 1.6887,
      "step": 4092
    },
    {
      "epoch": 0.1739442264692233,
      "grad_norm": 1.1253775358200073,
      "learning_rate": 4.901653461101213e-05,
      "loss": 1.6884,
      "step": 4123
    },
    {
      "epoch": 0.17525207779605956,
      "grad_norm": 1.215759038925171,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 1.7013,
      "step": 4154
    },
    {
      "epoch": 0.17655992912289584,
      "grad_norm": 1.1305285692214966,
      "learning_rate": 4.896934532712084e-05,
      "loss": 1.6573,
      "step": 4185
    },
    {
      "epoch": 0.1778677804497321,
      "grad_norm": 1.127212405204773,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 1.6769,
      "step": 4216
    },
    {
      "epoch": 0.17917563177656837,
      "grad_norm": 1.1379367113113403,
      "learning_rate": 4.892107408306516e-05,
      "loss": 1.6583,
      "step": 4247
    },
    {
      "epoch": 0.18048348310340465,
      "grad_norm": 1.0926482677459717,
      "learning_rate": 4.889653340498669e-05,
      "loss": 1.6674,
      "step": 4278
    },
    {
      "epoch": 0.1817913344302409,
      "grad_norm": 1.1098601818084717,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 1.6598,
      "step": 4309
    },
    {
      "epoch": 0.18309918575707718,
      "grad_norm": 1.129784345626831,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 1.6786,
      "step": 4340
    },
    {
      "epoch": 0.18440703708391343,
      "grad_norm": 1.1000723838806152,
      "learning_rate": 4.882129447892753e-05,
      "loss": 1.6623,
      "step": 4371
    },
    {
      "epoch": 0.1857148884107497,
      "grad_norm": 1.2229663133621216,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 1.6675,
      "step": 4402
    },
    {
      "epoch": 0.18702273973758596,
      "grad_norm": 1.1608648300170898,
      "learning_rate": 4.876979062282995e-05,
      "loss": 1.6499,
      "step": 4433
    },
    {
      "epoch": 0.18833059106442224,
      "grad_norm": 1.1060220003128052,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 1.6453,
      "step": 4464
    },
    {
      "epoch": 0.1896384423912585,
      "grad_norm": 1.1028501987457275,
      "learning_rate": 4.871721381433344e-05,
      "loss": 1.6393,
      "step": 4495
    },
    {
      "epoch": 0.19094629371809477,
      "grad_norm": 1.1736187934875488,
      "learning_rate": 4.869052379269719e-05,
      "loss": 1.6563,
      "step": 4526
    },
    {
      "epoch": 0.19225414504493102,
      "grad_norm": 1.1284164190292358,
      "learning_rate": 4.866356642671985e-05,
      "loss": 1.6535,
      "step": 4557
    },
    {
      "epoch": 0.1935619963717673,
      "grad_norm": 1.1155307292938232,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 1.6417,
      "step": 4588
    },
    {
      "epoch": 0.19486984769860355,
      "grad_norm": 1.1379483938217163,
      "learning_rate": 4.860885088159626e-05,
      "loss": 1.6432,
      "step": 4619
    },
    {
      "epoch": 0.19617769902543983,
      "grad_norm": 1.1166200637817383,
      "learning_rate": 4.858109331990751e-05,
      "loss": 1.6545,
      "step": 4650
    },
    {
      "epoch": 0.19748555035227608,
      "grad_norm": 1.081626296043396,
      "learning_rate": 4.855306964878567e-05,
      "loss": 1.6402,
      "step": 4681
    },
    {
      "epoch": 0.19879340167911236,
      "grad_norm": 1.0875532627105713,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 1.6498,
      "step": 4712
    },
    {
      "epoch": 0.2001012530059486,
      "grad_norm": 1.0928492546081543,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 1.6341,
      "step": 4743
    },
    {
      "epoch": 0.2014091043327849,
      "grad_norm": 1.1116279363632202,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 1.6389,
      "step": 4774
    },
    {
      "epoch": 0.20271695565962114,
      "grad_norm": 1.0984513759613037,
      "learning_rate": 4.843832023980392e-05,
      "loss": 1.6277,
      "step": 4805
    },
    {
      "epoch": 0.20402480698645742,
      "grad_norm": 1.09713876247406,
      "learning_rate": 4.840897082510106e-05,
      "loss": 1.6244,
      "step": 4836
    },
    {
      "epoch": 0.20533265831329367,
      "grad_norm": 1.089237928390503,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 1.6136,
      "step": 4867
    },
    {
      "epoch": 0.20664050964012995,
      "grad_norm": 1.1337491273880005,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 1.641,
      "step": 4898
    },
    {
      "epoch": 0.2079483609669662,
      "grad_norm": 1.1337246894836426,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 1.6163,
      "step": 4929
    },
    {
      "epoch": 0.20925621229380248,
      "grad_norm": 1.112273097038269,
      "learning_rate": 4.828893485248369e-05,
      "loss": 1.6384,
      "step": 4960
    },
    {
      "epoch": 0.21056406362063873,
      "grad_norm": 1.049856424331665,
      "learning_rate": 4.825826797411682e-05,
      "loss": 1.6245,
      "step": 4991
    },
    {
      "epoch": 0.211871914947475,
      "grad_norm": 2.468637704849243,
      "learning_rate": 4.822733862935702e-05,
      "loss": 1.6165,
      "step": 5022
    },
    {
      "epoch": 0.21317976627431126,
      "grad_norm": 1.0854454040527344,
      "learning_rate": 4.819614716723775e-05,
      "loss": 1.6252,
      "step": 5053
    },
    {
      "epoch": 0.21448761760114754,
      "grad_norm": 1.0815082788467407,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 1.6096,
      "step": 5084
    },
    {
      "epoch": 0.2157954689279838,
      "grad_norm": 1.053945541381836,
      "learning_rate": 4.813297930184042e-05,
      "loss": 1.6296,
      "step": 5115
    },
    {
      "epoch": 0.21710332025482007,
      "grad_norm": 1.06475031375885,
      "learning_rate": 4.810100361140314e-05,
      "loss": 1.629,
      "step": 5146
    },
    {
      "epoch": 0.21841117158165632,
      "grad_norm": 1.1072969436645508,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 1.6244,
      "step": 5177
    },
    {
      "epoch": 0.2197190229084926,
      "grad_norm": 1.0536466836929321,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 1.6474,
      "step": 5208
    },
    {
      "epoch": 0.22102687423532885,
      "grad_norm": 1.1039215326309204,
      "learning_rate": 4.8003513848046e-05,
      "loss": 1.6354,
      "step": 5239
    },
    {
      "epoch": 0.22233472556216513,
      "grad_norm": 1.0463244915008545,
      "learning_rate": 4.79704975853109e-05,
      "loss": 1.6154,
      "step": 5270
    },
    {
      "epoch": 0.22364257688900138,
      "grad_norm": 1.0931519269943237,
      "learning_rate": 4.793722210363262e-05,
      "loss": 1.6037,
      "step": 5301
    },
    {
      "epoch": 0.22495042821583766,
      "grad_norm": 1.0838814973831177,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 1.6352,
      "step": 5332
    },
    {
      "epoch": 0.2262582795426739,
      "grad_norm": 1.0400842428207397,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 1.5924,
      "step": 5363
    },
    {
      "epoch": 0.2275661308695102,
      "grad_norm": 1.0707345008850098,
      "learning_rate": 4.783584411463221e-05,
      "loss": 1.6211,
      "step": 5394
    },
    {
      "epoch": 0.22887398219634644,
      "grad_norm": 1.0422611236572266,
      "learning_rate": 4.780153554146274e-05,
      "loss": 1.5969,
      "step": 5425
    },
    {
      "epoch": 0.23018183352318272,
      "grad_norm": 1.0401543378829956,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 1.5974,
      "step": 5456
    },
    {
      "epoch": 0.23148968485001897,
      "grad_norm": 1.0429288148880005,
      "learning_rate": 4.773214684850662e-05,
      "loss": 1.6043,
      "step": 5487
    },
    {
      "epoch": 0.23279753617685525,
      "grad_norm": 1.098827600479126,
      "learning_rate": 4.769706751176193e-05,
      "loss": 1.5984,
      "step": 5518
    },
    {
      "epoch": 0.2341053875036915,
      "grad_norm": 1.0090724229812622,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 1.6012,
      "step": 5549
    },
    {
      "epoch": 0.23541323883052778,
      "grad_norm": 1.0732818841934204,
      "learning_rate": 4.762614083706258e-05,
      "loss": 1.5996,
      "step": 5580
    },
    {
      "epoch": 0.23672109015736403,
      "grad_norm": 1.082144856452942,
      "learning_rate": 4.759029429950581e-05,
      "loss": 1.6127,
      "step": 5611
    },
    {
      "epoch": 0.2380289414842003,
      "grad_norm": 1.0642670392990112,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 1.5962,
      "step": 5642
    },
    {
      "epoch": 0.23933679281103656,
      "grad_norm": 1.1351134777069092,
      "learning_rate": 4.751783684659e-05,
      "loss": 1.5983,
      "step": 5673
    },
    {
      "epoch": 0.24064464413787284,
      "grad_norm": 1.09076726436615,
      "learning_rate": 4.748122674890348e-05,
      "loss": 1.6162,
      "step": 5704
    },
    {
      "epoch": 0.24195249546470912,
      "grad_norm": 1.0234689712524414,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 1.596,
      "step": 5735
    },
    {
      "epoch": 0.24326034679154537,
      "grad_norm": 1.1209464073181152,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 1.602,
      "step": 5766
    },
    {
      "epoch": 0.24456819811838165,
      "grad_norm": 1.0940710306167603,
      "learning_rate": 4.736987593718397e-05,
      "loss": 1.5992,
      "step": 5797
    },
    {
      "epoch": 0.2458760494452179,
      "grad_norm": 1.0460273027420044,
      "learning_rate": 4.733225355658999e-05,
      "loss": 1.5881,
      "step": 5828
    },
    {
      "epoch": 0.24718390077205418,
      "grad_norm": 1.0842856168746948,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 1.594,
      "step": 5859
    },
    {
      "epoch": 0.24849175209889043,
      "grad_norm": 1.0244876146316528,
      "learning_rate": 4.725625317347119e-05,
      "loss": 1.583,
      "step": 5890
    },
    {
      "epoch": 0.2497996034257267,
      "grad_norm": 1.0182656049728394,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 1.5839,
      "step": 5921
    },
    {
      "epoch": 0.25110745475256296,
      "grad_norm": 1.0451878309249878,
      "learning_rate": 4.717924815801832e-05,
      "loss": 1.6051,
      "step": 5952
    },
    {
      "epoch": 0.2524153060793992,
      "grad_norm": 1.0340505838394165,
      "learning_rate": 4.714036999763532e-05,
      "loss": 1.5969,
      "step": 5983
    },
    {
      "epoch": 0.2537231574062355,
      "grad_norm": 1.0457735061645508,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 1.5971,
      "step": 6014
    },
    {
      "epoch": 0.2550310087330718,
      "grad_norm": 1.034166693687439,
      "learning_rate": 4.7061864565225e-05,
      "loss": 1.5982,
      "step": 6045
    },
    {
      "epoch": 0.256338860059908,
      "grad_norm": 1.0326176881790161,
      "learning_rate": 4.702223817912081e-05,
      "loss": 1.5691,
      "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
}