{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0526225042242836,
  "eval_steps": 500,
  "global_step": 6104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002672505948480982,
      "grad_norm": 3.2076423168182373,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 1.4511,
      "step": 31
    },
    {
      "epoch": 0.0005345011896961964,
      "grad_norm": 2.221529006958008,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 1.3342,
      "step": 62
    },
    {
      "epoch": 0.0008017517845442946,
      "grad_norm": 2.961315393447876,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 1.2649,
      "step": 93
    },
    {
      "epoch": 0.0010690023793923928,
      "grad_norm": 2.311617612838745,
      "learning_rate": 4.062909567496723e-06,
      "loss": 1.2278,
      "step": 124
    },
    {
      "epoch": 0.001336252974240491,
      "grad_norm": 3.4343724250793457,
      "learning_rate": 5.078636959370905e-06,
      "loss": 1.2061,
      "step": 155
    },
    {
      "epoch": 0.0016035035690885892,
      "grad_norm": 2.421396017074585,
      "learning_rate": 6.094364351245085e-06,
      "loss": 1.1848,
      "step": 186
    },
    {
      "epoch": 0.0018707541639366874,
      "grad_norm": 2.8230600357055664,
      "learning_rate": 7.110091743119267e-06,
      "loss": 1.1586,
      "step": 217
    },
    {
      "epoch": 0.0021380047587847856,
      "grad_norm": 2.021296501159668,
      "learning_rate": 8.125819134993446e-06,
      "loss": 1.1552,
      "step": 248
    },
    {
      "epoch": 0.002405255353632884,
      "grad_norm": 2.2329111099243164,
      "learning_rate": 9.141546526867629e-06,
      "loss": 1.1342,
      "step": 279
    },
    {
      "epoch": 0.002672505948480982,
      "grad_norm": 2.8933441638946533,
      "learning_rate": 1.015727391874181e-05,
      "loss": 1.1092,
      "step": 310
    },
    {
      "epoch": 0.0029397565433290805,
      "grad_norm": 2.510772705078125,
      "learning_rate": 1.117300131061599e-05,
      "loss": 1.108,
      "step": 341
    },
    {
      "epoch": 0.0032070071381771785,
      "grad_norm": 2.170926094055176,
      "learning_rate": 1.218872870249017e-05,
      "loss": 1.0982,
      "step": 372
    },
    {
      "epoch": 0.003474257733025277,
      "grad_norm": 2.1116831302642822,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 1.0977,
      "step": 403
    },
    {
      "epoch": 0.003741508327873375,
      "grad_norm": 2.5462026596069336,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 1.0797,
      "step": 434
    },
    {
      "epoch": 0.004008758922721473,
      "grad_norm": 2.5239830017089844,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 1.0691,
      "step": 465
    },
    {
      "epoch": 0.004276009517569571,
      "grad_norm": 2.640591859817505,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 1.0547,
      "step": 496
    },
    {
      "epoch": 0.004543260112417669,
      "grad_norm": 1.9566192626953125,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 1.0526,
      "step": 527
    },
    {
      "epoch": 0.004810510707265768,
      "grad_norm": 2.515118360519409,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 1.0614,
      "step": 558
    },
    {
      "epoch": 0.005077761302113866,
      "grad_norm": 2.1588094234466553,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 1.0461,
      "step": 589
    },
    {
      "epoch": 0.005345011896961964,
      "grad_norm": 2.448280096054077,
      "learning_rate": 2.031454783748362e-05,
      "loss": 1.0333,
      "step": 620
    },
    {
      "epoch": 0.005612262491810062,
      "grad_norm": 2.9539694786071777,
      "learning_rate": 2.13302752293578e-05,
      "loss": 1.0224,
      "step": 651
    },
    {
      "epoch": 0.005879513086658161,
      "grad_norm": 2.6804676055908203,
      "learning_rate": 2.234600262123198e-05,
      "loss": 1.0145,
      "step": 682
    },
    {
      "epoch": 0.006146763681506259,
      "grad_norm": 2.176100730895996,
      "learning_rate": 2.336173001310616e-05,
      "loss": 1.0164,
      "step": 713
    },
    {
      "epoch": 0.006414014276354357,
      "grad_norm": 1.9107022285461426,
      "learning_rate": 2.437745740498034e-05,
      "loss": 1.004,
      "step": 744
    },
    {
      "epoch": 0.006681264871202455,
      "grad_norm": 3.175410270690918,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 0.9986,
      "step": 775
    },
    {
      "epoch": 0.006948515466050554,
      "grad_norm": 1.7941040992736816,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 0.9979,
      "step": 806
    },
    {
      "epoch": 0.007215766060898652,
      "grad_norm": 2.428041458129883,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 1.0056,
      "step": 837
    },
    {
      "epoch": 0.00748301665574675,
      "grad_norm": 2.1440467834472656,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 0.9908,
      "step": 868
    },
    {
      "epoch": 0.007750267250594848,
      "grad_norm": 2.138157844543457,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 0.986,
      "step": 899
    },
    {
      "epoch": 0.008017517845442947,
      "grad_norm": 2.144076108932495,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 0.9768,
      "step": 930
    },
    {
      "epoch": 0.008284768440291045,
      "grad_norm": 2.116927146911621,
      "learning_rate": 3.148754914809961e-05,
      "loss": 0.975,
      "step": 961
    },
    {
      "epoch": 0.008552019035139143,
      "grad_norm": 2.0087051391601562,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 0.9746,
      "step": 992
    },
    {
      "epoch": 0.00881926962998724,
      "grad_norm": 2.008014440536499,
      "learning_rate": 3.351900393184797e-05,
      "loss": 0.9612,
      "step": 1023
    },
    {
      "epoch": 0.009086520224835339,
      "grad_norm": 2.5932323932647705,
      "learning_rate": 3.453473132372215e-05,
      "loss": 0.9611,
      "step": 1054
    },
    {
      "epoch": 0.009353770819683437,
      "grad_norm": 1.7766073942184448,
      "learning_rate": 3.555045871559633e-05,
      "loss": 0.9529,
      "step": 1085
    },
    {
      "epoch": 0.009621021414531536,
      "grad_norm": 1.699992299079895,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 0.9638,
      "step": 1116
    },
    {
      "epoch": 0.009888272009379634,
      "grad_norm": 2.139831304550171,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 0.9518,
      "step": 1147
    },
    {
      "epoch": 0.010155522604227732,
      "grad_norm": 2.510404109954834,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 0.9557,
      "step": 1178
    },
    {
      "epoch": 0.01042277319907583,
      "grad_norm": 1.7539429664611816,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 0.949,
      "step": 1209
    },
    {
      "epoch": 0.010690023793923928,
      "grad_norm": 1.7679015398025513,
      "learning_rate": 4.062909567496724e-05,
      "loss": 0.9285,
      "step": 1240
    },
    {
      "epoch": 0.010957274388772026,
      "grad_norm": 1.6701771020889282,
      "learning_rate": 4.164482306684142e-05,
      "loss": 0.9391,
      "step": 1271
    },
    {
      "epoch": 0.011224524983620124,
      "grad_norm": 1.7283812761306763,
      "learning_rate": 4.26605504587156e-05,
      "loss": 0.9328,
      "step": 1302
    },
    {
      "epoch": 0.011491775578468222,
      "grad_norm": 1.363319754600525,
      "learning_rate": 4.367627785058978e-05,
      "loss": 0.9353,
      "step": 1333
    },
    {
      "epoch": 0.011759026173316322,
      "grad_norm": 1.847582459449768,
      "learning_rate": 4.469200524246396e-05,
      "loss": 0.9272,
      "step": 1364
    },
    {
      "epoch": 0.01202627676816442,
      "grad_norm": 1.6437714099884033,
      "learning_rate": 4.570773263433814e-05,
      "loss": 0.9219,
      "step": 1395
    },
    {
      "epoch": 0.012293527363012518,
      "grad_norm": 1.979443073272705,
      "learning_rate": 4.672346002621232e-05,
      "loss": 0.91,
      "step": 1426
    },
    {
      "epoch": 0.012560777957860616,
      "grad_norm": 1.601218342781067,
      "learning_rate": 4.77391874180865e-05,
      "loss": 0.9229,
      "step": 1457
    },
    {
      "epoch": 0.012828028552708714,
      "grad_norm": 2.36148738861084,
      "learning_rate": 4.875491480996068e-05,
      "loss": 0.9199,
      "step": 1488
    },
    {
      "epoch": 0.013095279147556812,
      "grad_norm": 4.211123466491699,
      "learning_rate": 4.977064220183487e-05,
      "loss": 1.0419,
      "step": 1519
    },
    {
      "epoch": 0.01336252974240491,
      "grad_norm": 1.620786428451538,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 0.954,
      "step": 1550
    },
    {
      "epoch": 0.01362978033725301,
      "grad_norm": 3.2490246295928955,
      "learning_rate": 4.999955597496219e-05,
      "loss": 0.9464,
      "step": 1581
    },
    {
      "epoch": 0.013897030932101108,
      "grad_norm": 1.9302624464035034,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 0.9294,
      "step": 1612
    },
    {
      "epoch": 0.014164281526949206,
      "grad_norm": 1.5750601291656494,
      "learning_rate": 4.999799067923527e-05,
      "loss": 0.922,
      "step": 1643
    },
    {
      "epoch": 0.014431532121797304,
      "grad_norm": 53.0409049987793,
      "learning_rate": 4.999678487776908e-05,
      "loss": 1.0611,
      "step": 1674
    },
    {
      "epoch": 0.014698782716645402,
      "grad_norm": 1.841700553894043,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 1.0363,
      "step": 1705
    },
    {
      "epoch": 0.0149660333114935,
      "grad_norm": 1.2549737691879272,
      "learning_rate": 4.999352703566763e-05,
      "loss": 0.938,
      "step": 1736
    },
    {
      "epoch": 0.015233283906341597,
      "grad_norm": 1.3289740085601807,
      "learning_rate": 4.999147503179668e-05,
      "loss": 0.9113,
      "step": 1767
    },
    {
      "epoch": 0.015500534501189695,
      "grad_norm": 1.5228811502456665,
      "learning_rate": 4.998914100252672e-05,
      "loss": 0.921,
      "step": 1798
    },
    {
      "epoch": 0.015767785096037795,
      "grad_norm": 1.5058027505874634,
      "learning_rate": 4.998652497419696e-05,
      "loss": 0.9214,
      "step": 1829
    },
    {
      "epoch": 0.016035035690885893,
      "grad_norm": 9.535687446594238,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 1.3068,
      "step": 1860
    },
    {
      "epoch": 0.01630228628573399,
      "grad_norm": 2.5820651054382324,
      "learning_rate": 4.998044704162613e-05,
      "loss": 1.7033,
      "step": 1891
    },
    {
      "epoch": 0.01656953688058209,
      "grad_norm": 148.47412109375,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 1.6935,
      "step": 1922
    },
    {
      "epoch": 0.016836787475430187,
      "grad_norm": 8.416805267333984,
      "learning_rate": 4.997324150843799e-05,
      "loss": 1.6607,
      "step": 1953
    },
    {
      "epoch": 0.017104038070278285,
      "grad_norm": 5.2713470458984375,
      "learning_rate": 4.99692159912661e-05,
      "loss": 1.1301,
      "step": 1984
    },
    {
      "epoch": 0.017371288665126383,
      "grad_norm": 2.096306800842285,
      "learning_rate": 4.996490869988546e-05,
      "loss": 0.9931,
      "step": 2015
    },
    {
      "epoch": 0.01763853925997448,
      "grad_norm": 1.5476640462875366,
      "learning_rate": 4.996031968290326e-05,
      "loss": 0.9514,
      "step": 2046
    },
    {
      "epoch": 0.01790578985482258,
      "grad_norm": 1.727232813835144,
      "learning_rate": 4.995544899210594e-05,
      "loss": 0.9409,
      "step": 2077
    },
    {
      "epoch": 0.018173040449670677,
      "grad_norm": 26.08816909790039,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 0.9368,
      "step": 2108
    },
    {
      "epoch": 0.018440291044518775,
      "grad_norm": 1.7082202434539795,
      "learning_rate": 4.994486281210429e-05,
      "loss": 0.9155,
      "step": 2139
    },
    {
      "epoch": 0.018707541639366873,
      "grad_norm": 1.5962992906570435,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 0.9233,
      "step": 2170
    },
    {
      "epoch": 0.018974792234214975,
      "grad_norm": 2.024268627166748,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 0.9128,
      "step": 2201
    },
    {
      "epoch": 0.019242042829063073,
      "grad_norm": 1.4828873872756958,
      "learning_rate": 4.992687246588743e-05,
      "loss": 0.8957,
      "step": 2232
    },
    {
      "epoch": 0.01950929342391117,
      "grad_norm": 1.745612382888794,
      "learning_rate": 4.992031299767347e-05,
      "loss": 0.9005,
      "step": 2263
    },
    {
      "epoch": 0.01977654401875927,
      "grad_norm": 1.354095458984375,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 0.9002,
      "step": 2294
    },
    {
      "epoch": 0.020043794613607367,
      "grad_norm": 2.6227681636810303,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 0.9004,
      "step": 2325
    },
    {
      "epoch": 0.020311045208455464,
      "grad_norm": 1.3870893716812134,
      "learning_rate": 4.989894757091861e-05,
      "loss": 0.9018,
      "step": 2356
    },
    {
      "epoch": 0.020578295803303562,
      "grad_norm": 1.2884349822998047,
      "learning_rate": 4.989126368919158e-05,
      "loss": 0.8993,
      "step": 2387
    },
    {
      "epoch": 0.02084554639815166,
      "grad_norm": 1.215851068496704,
      "learning_rate": 4.988329891293693e-05,
      "loss": 0.8925,
      "step": 2418
    },
    {
      "epoch": 0.02111279699299976,
      "grad_norm": 1.4246628284454346,
      "learning_rate": 4.987505333203608e-05,
      "loss": 0.8893,
      "step": 2449
    },
    {
      "epoch": 0.021380047587847856,
      "grad_norm": 1.5495259761810303,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 0.8749,
      "step": 2480
    },
    {
      "epoch": 0.021647298182695954,
      "grad_norm": 1.0121793746948242,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 0.8803,
      "step": 2511
    },
    {
      "epoch": 0.021914548777544052,
      "grad_norm": 1.1507561206817627,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 0.8843,
      "step": 2542
    },
    {
      "epoch": 0.02218179937239215,
      "grad_norm": 1.612584114074707,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 0.8733,
      "step": 2573
    },
    {
      "epoch": 0.02244904996724025,
      "grad_norm": 1.4475393295288086,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 0.8669,
      "step": 2604
    },
    {
      "epoch": 0.022716300562088346,
      "grad_norm": 1.9612644910812378,
      "learning_rate": 4.981968838155888e-05,
      "loss": 0.8791,
      "step": 2635
    },
    {
      "epoch": 0.022983551156936444,
      "grad_norm": 1.6032893657684326,
      "learning_rate": 4.980947995086024e-05,
      "loss": 0.8646,
      "step": 2666
    },
    {
      "epoch": 0.023250801751784546,
      "grad_norm": 1.6368708610534668,
      "learning_rate": 4.979899154855234e-05,
      "loss": 0.871,
      "step": 2697
    },
    {
      "epoch": 0.023518052346632644,
      "grad_norm": 1.0245656967163086,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 0.8534,
      "step": 2728
    },
    {
      "epoch": 0.023785302941480742,
      "grad_norm": 0.9378401041030884,
      "learning_rate": 4.977717530570768e-05,
      "loss": 0.8478,
      "step": 2759
    },
    {
      "epoch": 0.02405255353632884,
      "grad_norm": 1.4262951612472534,
      "learning_rate": 4.976584771136425e-05,
      "loss": 0.8634,
      "step": 2790
    },
    {
      "epoch": 0.024319804131176938,
      "grad_norm": 1.3320890665054321,
      "learning_rate": 4.975424063779547e-05,
      "loss": 0.8505,
      "step": 2821
    },
    {
      "epoch": 0.024587054726025036,
      "grad_norm": 1.043971061706543,
      "learning_rate": 4.974235421598557e-05,
      "loss": 0.863,
      "step": 2852
    },
    {
      "epoch": 0.024854305320873134,
      "grad_norm": 1.387088418006897,
      "learning_rate": 4.973018858007122e-05,
      "loss": 0.8505,
      "step": 2883
    },
    {
      "epoch": 0.025121555915721232,
      "grad_norm": 1.1267746686935425,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 0.8528,
      "step": 2914
    },
    {
      "epoch": 0.02538880651056933,
      "grad_norm": 0.9636305570602417,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 0.8453,
      "step": 2945
    },
    {
      "epoch": 0.025656057105417428,
      "grad_norm": 1.1140860319137573,
      "learning_rate": 4.969201777632205e-05,
      "loss": 0.8664,
      "step": 2976
    },
    {
      "epoch": 0.025923307700265526,
      "grad_norm": 1.1886147260665894,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 0.8569,
      "step": 3007
    },
    {
      "epoch": 0.026190558295113624,
      "grad_norm": 1.0308467149734497,
      "learning_rate": 4.966517710419033e-05,
      "loss": 0.85,
      "step": 3038
    },
    {
      "epoch": 0.02645780888996172,
      "grad_norm": 1.6353480815887451,
      "learning_rate": 4.965133917685858e-05,
      "loss": 0.8414,
      "step": 3069
    },
    {
      "epoch": 0.02672505948480982,
      "grad_norm": 1.2031575441360474,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 0.8435,
      "step": 3100
    },
    {
      "epoch": 0.026992310079657918,
      "grad_norm": 1.2575979232788086,
      "learning_rate": 4.962282892045718e-05,
      "loss": 0.8415,
      "step": 3131
    },
    {
      "epoch": 0.02725956067450602,
      "grad_norm": 1.338584303855896,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 0.8414,
      "step": 3162
    },
    {
      "epoch": 0.027526811269354117,
      "grad_norm": 1.3699461221694946,
      "learning_rate": 4.959320720608049e-05,
      "loss": 0.8397,
      "step": 3193
    },
    {
      "epoch": 0.027794061864202215,
      "grad_norm": 1.1471112966537476,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 0.8304,
      "step": 3224
    },
    {
      "epoch": 0.028061312459050313,
      "grad_norm": 1.1752680540084839,
      "learning_rate": 4.956247537083282e-05,
      "loss": 0.8408,
      "step": 3255
    },
    {
      "epoch": 0.02832856305389841,
      "grad_norm": 1.089552879333496,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 0.8286,
      "step": 3286
    },
    {
      "epoch": 0.02859581364874651,
      "grad_norm": 1.0688341856002808,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 0.8341,
      "step": 3317
    },
    {
      "epoch": 0.028863064243594607,
      "grad_norm": 0.9290914535522461,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 0.8307,
      "step": 3348
    },
    {
      "epoch": 0.029130314838442705,
      "grad_norm": 1.3211214542388916,
      "learning_rate": 4.949768693662973e-05,
      "loss": 0.8349,
      "step": 3379
    },
    {
      "epoch": 0.029397565433290803,
      "grad_norm": 1.8893773555755615,
      "learning_rate": 4.948079823064559e-05,
      "loss": 0.8289,
      "step": 3410
    },
    {
      "epoch": 0.0296648160281389,
      "grad_norm": 1.3086936473846436,
      "learning_rate": 4.946363326218074e-05,
      "loss": 0.8231,
      "step": 3441
    },
    {
      "epoch": 0.029932066622987,
      "grad_norm": 1.356311559677124,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 0.8343,
      "step": 3472
    },
    {
      "epoch": 0.030199317217835097,
      "grad_norm": 1.13872492313385,
      "learning_rate": 4.942847531574167e-05,
      "loss": 0.828,
      "step": 3503
    },
    {
      "epoch": 0.030466567812683195,
      "grad_norm": 1.1912773847579956,
      "learning_rate": 4.941048273452008e-05,
      "loss": 0.8236,
      "step": 3534
    },
    {
      "epoch": 0.030733818407531293,
      "grad_norm": 0.9231215715408325,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 0.8226,
      "step": 3565
    },
    {
      "epoch": 0.03100106900237939,
      "grad_norm": 1.0268129110336304,
      "learning_rate": 4.93736713712897e-05,
      "loss": 0.8218,
      "step": 3596
    },
    {
      "epoch": 0.03126831959722749,
      "grad_norm": 0.974876344203949,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 0.8149,
      "step": 3627
    },
    {
      "epoch": 0.03153557019207559,
      "grad_norm": 1.2175925970077515,
      "learning_rate": 4.93357597968886e-05,
      "loss": 0.8152,
      "step": 3658
    },
    {
      "epoch": 0.03180282078692369,
      "grad_norm": 1.2338181734085083,
      "learning_rate": 4.931639196334338e-05,
      "loss": 0.8089,
      "step": 3689
    },
    {
      "epoch": 0.032070071381771786,
      "grad_norm": 1.1462433338165283,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 0.8216,
      "step": 3720
    },
    {
      "epoch": 0.032337321976619884,
      "grad_norm": 1.1631278991699219,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 0.8272,
      "step": 3751
    },
    {
      "epoch": 0.03260457257146798,
      "grad_norm": 0.903836727142334,
      "learning_rate": 4.925664290937196e-05,
      "loss": 0.816,
      "step": 3782
    },
    {
      "epoch": 0.03287182316631608,
      "grad_norm": 0.944284975528717,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 0.8031,
      "step": 3813
    },
    {
      "epoch": 0.03313907376116418,
      "grad_norm": 1.1210039854049683,
      "learning_rate": 4.921544116753962e-05,
      "loss": 0.8096,
      "step": 3844
    },
    {
      "epoch": 0.033406324356012276,
      "grad_norm": 1.0448529720306396,
      "learning_rate": 4.919443027766935e-05,
      "loss": 0.8059,
      "step": 3875
    },
    {
      "epoch": 0.033673574950860374,
      "grad_norm": 1.0655525922775269,
      "learning_rate": 4.91731463569418e-05,
      "loss": 0.81,
      "step": 3906
    },
    {
      "epoch": 0.03394082554570847,
      "grad_norm": 1.223928451538086,
      "learning_rate": 4.915158964554312e-05,
      "loss": 0.8175,
      "step": 3937
    },
    {
      "epoch": 0.03420807614055657,
      "grad_norm": 1.0895342826843262,
      "learning_rate": 4.912976038673786e-05,
      "loss": 0.7992,
      "step": 3968
    },
    {
      "epoch": 0.03447532673540467,
      "grad_norm": 1.1117792129516602,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 0.8122,
      "step": 3999
    },
    {
      "epoch": 0.034742577330252766,
      "grad_norm": 1.0806598663330078,
      "learning_rate": 4.908528521534139e-05,
      "loss": 0.811,
      "step": 4030
    },
    {
      "epoch": 0.035009827925100864,
      "grad_norm": 0.7528677582740784,
      "learning_rate": 4.906263980464644e-05,
      "loss": 0.8036,
      "step": 4061
    },
    {
      "epoch": 0.03527707851994896,
      "grad_norm": 1.1509230136871338,
      "learning_rate": 4.903972285033178e-05,
      "loss": 0.8047,
      "step": 4092
    },
    {
      "epoch": 0.03554432911479706,
      "grad_norm": 1.1190931797027588,
      "learning_rate": 4.901653461101213e-05,
      "loss": 0.8123,
      "step": 4123
    },
    {
      "epoch": 0.03581157970964516,
      "grad_norm": 0.8739308714866638,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 0.8015,
      "step": 4154
    },
    {
      "epoch": 0.036078830304493256,
      "grad_norm": 0.9889201521873474,
      "learning_rate": 4.896934532712084e-05,
      "loss": 0.8056,
      "step": 4185
    },
    {
      "epoch": 0.036346080899341354,
      "grad_norm": 1.1964142322540283,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 0.8083,
      "step": 4216
    },
    {
      "epoch": 0.03661333149418945,
      "grad_norm": 1.0793588161468506,
      "learning_rate": 4.892107408306516e-05,
      "loss": 0.7976,
      "step": 4247
    },
    {
      "epoch": 0.03688058208903755,
      "grad_norm": 1.2313084602355957,
      "learning_rate": 4.889653340498669e-05,
      "loss": 0.7954,
      "step": 4278
    },
    {
      "epoch": 0.03714783268388565,
      "grad_norm": 1.1145700216293335,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 0.7947,
      "step": 4309
    },
    {
      "epoch": 0.037415083278733746,
      "grad_norm": 0.9714220762252808,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 0.7973,
      "step": 4340
    },
    {
      "epoch": 0.03768233387358185,
      "grad_norm": 1.0865004062652588,
      "learning_rate": 4.882129447892753e-05,
      "loss": 0.7918,
      "step": 4371
    },
    {
      "epoch": 0.03794958446842995,
      "grad_norm": 0.8420351147651672,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 0.805,
      "step": 4402
    },
    {
      "epoch": 0.03821683506327805,
      "grad_norm": 1.0622302293777466,
      "learning_rate": 4.876979062282995e-05,
      "loss": 0.7977,
      "step": 4433
    },
    {
      "epoch": 0.038484085658126145,
      "grad_norm": 0.8536683320999146,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 0.7946,
      "step": 4464
    },
    {
      "epoch": 0.03875133625297424,
      "grad_norm": 0.9905999302864075,
      "learning_rate": 4.871721381433344e-05,
      "loss": 0.7932,
      "step": 4495
    },
    {
      "epoch": 0.03901858684782234,
      "grad_norm": 1.005560278892517,
      "learning_rate": 4.869052379269719e-05,
      "loss": 0.7956,
      "step": 4526
    },
    {
      "epoch": 0.03928583744267044,
      "grad_norm": 0.8386696577072144,
      "learning_rate": 4.866356642671985e-05,
      "loss": 0.7865,
      "step": 4557
    },
    {
      "epoch": 0.03955308803751854,
      "grad_norm": 0.8390759229660034,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 0.7925,
      "step": 4588
    },
    {
      "epoch": 0.039820338632366635,
      "grad_norm": 0.7457160949707031,
      "learning_rate": 4.860885088159626e-05,
      "loss": 0.7911,
      "step": 4619
    },
    {
      "epoch": 0.04008758922721473,
      "grad_norm": 0.9717801213264465,
      "learning_rate": 4.858109331990751e-05,
      "loss": 0.7917,
      "step": 4650
    },
    {
      "epoch": 0.04035483982206283,
      "grad_norm": 0.8517304062843323,
      "learning_rate": 4.855306964878567e-05,
      "loss": 0.7966,
      "step": 4681
    },
    {
      "epoch": 0.04062209041691093,
      "grad_norm": 0.7597792744636536,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 0.7798,
      "step": 4712
    },
    {
      "epoch": 0.04088934101175903,
      "grad_norm": 0.9633410573005676,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 0.784,
      "step": 4743
    },
    {
      "epoch": 0.041156591606607125,
      "grad_norm": 0.8668884634971619,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 0.7925,
      "step": 4774
    },
    {
      "epoch": 0.04142384220145522,
      "grad_norm": 1.348170280456543,
      "learning_rate": 4.843832023980392e-05,
      "loss": 0.7882,
      "step": 4805
    },
    {
      "epoch": 0.04169109279630332,
      "grad_norm": 0.9085477590560913,
      "learning_rate": 4.840897082510106e-05,
      "loss": 0.7834,
      "step": 4836
    },
    {
      "epoch": 0.04195834339115142,
      "grad_norm": 1.2533594369888306,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 0.7968,
      "step": 4867
    },
    {
      "epoch": 0.04222559398599952,
      "grad_norm": 1.0250332355499268,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 0.7755,
      "step": 4898
    },
    {
      "epoch": 0.042492844580847615,
      "grad_norm": 0.8610180616378784,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 0.7819,
      "step": 4929
    },
    {
      "epoch": 0.04276009517569571,
      "grad_norm": 0.8235286474227905,
      "learning_rate": 4.828893485248369e-05,
      "loss": 0.7834,
      "step": 4960
    },
    {
      "epoch": 0.04302734577054381,
      "grad_norm": 0.95796799659729,
      "learning_rate": 4.825826797411682e-05,
      "loss": 0.7822,
      "step": 4991
    },
    {
      "epoch": 0.04329459636539191,
      "grad_norm": 1.022430419921875,
      "learning_rate": 4.822733862935702e-05,
      "loss": 0.781,
      "step": 5022
    },
    {
      "epoch": 0.04356184696024001,
      "grad_norm": 0.9080584645271301,
      "learning_rate": 4.819614716723775e-05,
      "loss": 0.783,
      "step": 5053
    },
    {
      "epoch": 0.043829097555088105,
      "grad_norm": 0.9808719158172607,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 0.7838,
      "step": 5084
    },
    {
      "epoch": 0.0440963481499362,
      "grad_norm": 0.8997575640678406,
      "learning_rate": 4.813297930184042e-05,
      "loss": 0.7816,
      "step": 5115
    },
    {
      "epoch": 0.0443635987447843,
      "grad_norm": 0.8737884163856506,
      "learning_rate": 4.810100361140314e-05,
      "loss": 0.7785,
      "step": 5146
    },
    {
      "epoch": 0.0446308493396324,
      "grad_norm": 0.7924982309341431,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 0.7752,
      "step": 5177
    },
    {
      "epoch": 0.0448980999344805,
      "grad_norm": 1.1685878038406372,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 0.7881,
      "step": 5208
    },
    {
      "epoch": 0.045165350529328595,
      "grad_norm": 1.0199975967407227,
      "learning_rate": 4.8003513848046e-05,
      "loss": 0.7663,
      "step": 5239
    },
    {
      "epoch": 0.04543260112417669,
      "grad_norm": 0.8155254125595093,
      "learning_rate": 4.79704975853109e-05,
      "loss": 0.7775,
      "step": 5270
    },
    {
      "epoch": 0.04569985171902479,
      "grad_norm": 0.976006805896759,
      "learning_rate": 4.793722210363262e-05,
      "loss": 0.7797,
      "step": 5301
    },
    {
      "epoch": 0.04596710231387289,
      "grad_norm": 1.1252824068069458,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 0.7824,
      "step": 5332
    },
    {
      "epoch": 0.046234352908720994,
      "grad_norm": 0.7603053450584412,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 0.7759,
      "step": 5363
    },
    {
      "epoch": 0.04650160350356909,
      "grad_norm": 1.007767915725708,
      "learning_rate": 4.783584411463221e-05,
      "loss": 0.7832,
      "step": 5394
    },
    {
      "epoch": 0.04676885409841719,
      "grad_norm": 1.1207255125045776,
      "learning_rate": 4.780153554146274e-05,
      "loss": 0.7849,
      "step": 5425
    },
    {
      "epoch": 0.04703610469326529,
      "grad_norm": 1.0699808597564697,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 0.7814,
      "step": 5456
    },
    {
      "epoch": 0.047303355288113386,
      "grad_norm": 0.8154749274253845,
      "learning_rate": 4.773214684850662e-05,
      "loss": 0.7806,
      "step": 5487
    },
    {
      "epoch": 0.047570605882961484,
      "grad_norm": 0.7383331060409546,
      "learning_rate": 4.769706751176193e-05,
      "loss": 0.7765,
      "step": 5518
    },
    {
      "epoch": 0.04783785647780958,
      "grad_norm": 0.9729552865028381,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 0.7754,
      "step": 5549
    },
    {
      "epoch": 0.04810510707265768,
      "grad_norm": 0.9568310379981995,
      "learning_rate": 4.762614083706258e-05,
      "loss": 0.7733,
      "step": 5580
    },
    {
      "epoch": 0.04837235766750578,
      "grad_norm": 1.0747005939483643,
      "learning_rate": 4.759029429950581e-05,
      "loss": 0.7764,
      "step": 5611
    },
    {
      "epoch": 0.048639608262353876,
      "grad_norm": 1.2053782939910889,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 0.7725,
      "step": 5642
    },
    {
      "epoch": 0.048906858857201974,
      "grad_norm": 0.8168059587478638,
      "learning_rate": 4.751783684659e-05,
      "loss": 0.7779,
      "step": 5673
    },
    {
      "epoch": 0.04917410945205007,
      "grad_norm": 0.812278151512146,
      "learning_rate": 4.748122674890348e-05,
      "loss": 0.7884,
      "step": 5704
    },
    {
      "epoch": 0.04944136004689817,
      "grad_norm": 0.9640679359436035,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 0.7794,
      "step": 5735
    },
    {
      "epoch": 0.04970861064174627,
      "grad_norm": 0.8488368988037109,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 0.7705,
      "step": 5766
    },
    {
      "epoch": 0.049975861236594366,
      "grad_norm": 0.7826496362686157,
      "learning_rate": 4.736987593718397e-05,
      "loss": 0.7665,
      "step": 5797
    },
    {
      "epoch": 0.050243111831442464,
      "grad_norm": 0.7388876676559448,
      "learning_rate": 4.733225355658999e-05,
      "loss": 0.7715,
      "step": 5828
    },
    {
      "epoch": 0.05051036242629056,
      "grad_norm": 0.7776780128479004,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 0.7777,
      "step": 5859
    },
    {
      "epoch": 0.05077761302113866,
      "grad_norm": 1.1687864065170288,
      "learning_rate": 4.725625317347119e-05,
      "loss": 0.7698,
      "step": 5890
    },
    {
      "epoch": 0.05104486361598676,
      "grad_norm": 1.069354772567749,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 0.7671,
      "step": 5921
    },
    {
      "epoch": 0.051312114210834855,
      "grad_norm": 0.955895185470581,
      "learning_rate": 4.717924815801832e-05,
      "loss": 0.7641,
      "step": 5952
    },
    {
      "epoch": 0.05157936480568295,
      "grad_norm": 0.8063961267471313,
      "learning_rate": 4.714036999763532e-05,
      "loss": 0.772,
      "step": 5983
    },
    {
      "epoch": 0.05184661540053105,
      "grad_norm": 0.9584303498268127,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 0.762,
      "step": 6014
    },
    {
      "epoch": 0.05211386599537915,
      "grad_norm": 0.9711636900901794,
      "learning_rate": 4.7061864565225e-05,
      "loss": 0.7742,
      "step": 6045
    },
    {
      "epoch": 0.05238111659022725,
      "grad_norm": 0.8177659511566162,
      "learning_rate": 4.702223817912081e-05,
      "loss": 0.7532,
      "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": 4.24270815620078e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}