{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.10143578835416113,
  "eval_steps": 500,
  "global_step": 3052,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001030311087476735,
      "grad_norm": 60.25547409057617,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 8.8455,
      "step": 31
    },
    {
      "epoch": 0.00206062217495347,
      "grad_norm": 15.669363975524902,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.1553,
      "step": 62
    },
    {
      "epoch": 0.003090933262430205,
      "grad_norm": 15.366345405578613,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 5.8784,
      "step": 93
    },
    {
      "epoch": 0.00412124434990694,
      "grad_norm": 36.30561828613281,
      "learning_rate": 4.062909567496723e-06,
      "loss": 4.7708,
      "step": 124
    },
    {
      "epoch": 0.005151555437383675,
      "grad_norm": 27.202678680419922,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.1629,
      "step": 155
    },
    {
      "epoch": 0.00618186652486041,
      "grad_norm": 24.30484962463379,
      "learning_rate": 6.094364351245085e-06,
      "loss": 3.867,
      "step": 186
    },
    {
      "epoch": 0.007212177612337145,
      "grad_norm": 19.916366577148438,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.6131,
      "step": 217
    },
    {
      "epoch": 0.00824248869981388,
      "grad_norm": 17.577274322509766,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.4772,
      "step": 248
    },
    {
      "epoch": 0.009272799787290615,
      "grad_norm": 12.133190155029297,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.3218,
      "step": 279
    },
    {
      "epoch": 0.01030311087476735,
      "grad_norm": 19.79263687133789,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.2055,
      "step": 310
    },
    {
      "epoch": 0.011333421962244085,
      "grad_norm": 16.38133430480957,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.1062,
      "step": 341
    },
    {
      "epoch": 0.01236373304972082,
      "grad_norm": 12.638299942016602,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.0106,
      "step": 372
    },
    {
      "epoch": 0.013394044137197554,
      "grad_norm": 9.46596908569336,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 2.924,
      "step": 403
    },
    {
      "epoch": 0.01442435522467429,
      "grad_norm": 10.945392608642578,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 2.844,
      "step": 434
    },
    {
      "epoch": 0.015454666312151024,
      "grad_norm": 8.474015235900879,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 2.7892,
      "step": 465
    },
    {
      "epoch": 0.01648497739962776,
      "grad_norm": 9.370804786682129,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.7509,
      "step": 496
    },
    {
      "epoch": 0.017515288487104493,
      "grad_norm": 11.63398551940918,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.6999,
      "step": 527
    },
    {
      "epoch": 0.01854559957458123,
      "grad_norm": 9.17713451385498,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.6459,
      "step": 558
    },
    {
      "epoch": 0.019575910662057962,
      "grad_norm": 7.119054794311523,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.603,
      "step": 589
    },
    {
      "epoch": 0.0206062217495347,
      "grad_norm": 6.653646945953369,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.5588,
      "step": 620
    },
    {
      "epoch": 0.021636532837011432,
      "grad_norm": 8.332653045654297,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.5357,
      "step": 651
    },
    {
      "epoch": 0.02266684392448817,
      "grad_norm": 6.4949116706848145,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.4967,
      "step": 682
    },
    {
      "epoch": 0.023697155011964902,
      "grad_norm": 9.41009521484375,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.4563,
      "step": 713
    },
    {
      "epoch": 0.02472746609944164,
      "grad_norm": 7.840345859527588,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.4383,
      "step": 744
    },
    {
      "epoch": 0.025757777186918372,
      "grad_norm": 6.116458415985107,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.3817,
      "step": 775
    },
    {
      "epoch": 0.02678808827439511,
      "grad_norm": 5.938300609588623,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.3508,
      "step": 806
    },
    {
      "epoch": 0.027818399361871842,
      "grad_norm": 5.4408345222473145,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.3325,
      "step": 837
    },
    {
      "epoch": 0.02884871044934858,
      "grad_norm": 5.375136375427246,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.3101,
      "step": 868
    },
    {
      "epoch": 0.029879021536825312,
      "grad_norm": 5.149726867675781,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.282,
      "step": 899
    },
    {
      "epoch": 0.03090933262430205,
      "grad_norm": 4.591221332550049,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.2427,
      "step": 930
    },
    {
      "epoch": 0.031939643711778785,
      "grad_norm": 4.977034091949463,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.2218,
      "step": 961
    },
    {
      "epoch": 0.03296995479925552,
      "grad_norm": 5.038781642913818,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.2044,
      "step": 992
    },
    {
      "epoch": 0.03400026588673225,
      "grad_norm": 4.872281551361084,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.1657,
      "step": 1023
    },
    {
      "epoch": 0.035030576974208985,
      "grad_norm": 4.370841979980469,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.1365,
      "step": 1054
    },
    {
      "epoch": 0.036060888061685725,
      "grad_norm": 4.087072849273682,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.1253,
      "step": 1085
    },
    {
      "epoch": 0.03709119914916246,
      "grad_norm": 4.113957405090332,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.0973,
      "step": 1116
    },
    {
      "epoch": 0.03812151023663919,
      "grad_norm": 4.0119733810424805,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.1024,
      "step": 1147
    },
    {
      "epoch": 0.039151821324115925,
      "grad_norm": 4.247573375701904,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.0722,
      "step": 1178
    },
    {
      "epoch": 0.04018213241159266,
      "grad_norm": 3.5575129985809326,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.056,
      "step": 1209
    },
    {
      "epoch": 0.0412124434990694,
      "grad_norm": 3.8885862827301025,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.0389,
      "step": 1240
    },
    {
      "epoch": 0.04224275458654613,
      "grad_norm": 3.680628538131714,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.0385,
      "step": 1271
    },
    {
      "epoch": 0.043273065674022865,
      "grad_norm": 3.780876874923706,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.0097,
      "step": 1302
    },
    {
      "epoch": 0.0443033767614996,
      "grad_norm": 4.235328674316406,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.0024,
      "step": 1333
    },
    {
      "epoch": 0.04533368784897634,
      "grad_norm": 3.326941967010498,
      "learning_rate": 4.469200524246396e-05,
      "loss": 1.9953,
      "step": 1364
    },
    {
      "epoch": 0.04636399893645307,
      "grad_norm": 3.28456449508667,
      "learning_rate": 4.570773263433814e-05,
      "loss": 1.9579,
      "step": 1395
    },
    {
      "epoch": 0.047394310023929805,
      "grad_norm": 16.107433319091797,
      "learning_rate": 4.672346002621232e-05,
      "loss": 1.9701,
      "step": 1426
    },
    {
      "epoch": 0.04842462111140654,
      "grad_norm": 3.5708224773406982,
      "learning_rate": 4.77391874180865e-05,
      "loss": 1.9621,
      "step": 1457
    },
    {
      "epoch": 0.04945493219888328,
      "grad_norm": 2.9053499698638916,
      "learning_rate": 4.875491480996068e-05,
      "loss": 1.9458,
      "step": 1488
    },
    {
      "epoch": 0.05048524328636001,
      "grad_norm": 3.0863258838653564,
      "learning_rate": 4.977064220183487e-05,
      "loss": 1.9483,
      "step": 1519
    },
    {
      "epoch": 0.051515554373836744,
      "grad_norm": 2.9012269973754883,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 1.928,
      "step": 1550
    },
    {
      "epoch": 0.05254586546131348,
      "grad_norm": 3.0949041843414307,
      "learning_rate": 4.999955597496219e-05,
      "loss": 1.9229,
      "step": 1581
    },
    {
      "epoch": 0.05357617654879022,
      "grad_norm": 2.8687901496887207,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 1.915,
      "step": 1612
    },
    {
      "epoch": 0.05460648763626695,
      "grad_norm": 3.2136878967285156,
      "learning_rate": 4.999799067923527e-05,
      "loss": 1.9197,
      "step": 1643
    },
    {
      "epoch": 0.055636798723743684,
      "grad_norm": 2.590843677520752,
      "learning_rate": 4.999678487776908e-05,
      "loss": 1.8756,
      "step": 1674
    },
    {
      "epoch": 0.05666710981122042,
      "grad_norm": 2.64634108543396,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 1.9033,
      "step": 1705
    },
    {
      "epoch": 0.05769742089869716,
      "grad_norm": 3.0151331424713135,
      "learning_rate": 4.999352703566763e-05,
      "loss": 1.8883,
      "step": 1736
    },
    {
      "epoch": 0.05872773198617389,
      "grad_norm": 2.526806354522705,
      "learning_rate": 4.999147503179668e-05,
      "loss": 1.8666,
      "step": 1767
    },
    {
      "epoch": 0.059758043073650624,
      "grad_norm": 2.510300397872925,
      "learning_rate": 4.998914100252672e-05,
      "loss": 1.854,
      "step": 1798
    },
    {
      "epoch": 0.06078835416112736,
      "grad_norm": 2.4867682456970215,
      "learning_rate": 4.998652497419696e-05,
      "loss": 1.8548,
      "step": 1829
    },
    {
      "epoch": 0.0618186652486041,
      "grad_norm": 2.3920586109161377,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 1.8495,
      "step": 1860
    },
    {
      "epoch": 0.06284897633608083,
      "grad_norm": 2.714177370071411,
      "learning_rate": 4.998044704162613e-05,
      "loss": 1.8433,
      "step": 1891
    },
    {
      "epoch": 0.06387928742355757,
      "grad_norm": 2.3094465732574463,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 1.8382,
      "step": 1922
    },
    {
      "epoch": 0.0649095985110343,
      "grad_norm": 2.47184419631958,
      "learning_rate": 4.997324150843799e-05,
      "loss": 1.8464,
      "step": 1953
    },
    {
      "epoch": 0.06593990959851104,
      "grad_norm": 2.391841411590576,
      "learning_rate": 4.99692159912661e-05,
      "loss": 1.8179,
      "step": 1984
    },
    {
      "epoch": 0.06697022068598776,
      "grad_norm": 2.2471864223480225,
      "learning_rate": 4.996490869988546e-05,
      "loss": 1.8149,
      "step": 2015
    },
    {
      "epoch": 0.0680005317734645,
      "grad_norm": 2.5497376918792725,
      "learning_rate": 4.996031968290326e-05,
      "loss": 1.8099,
      "step": 2046
    },
    {
      "epoch": 0.06903084286094124,
      "grad_norm": 2.330463409423828,
      "learning_rate": 4.995544899210594e-05,
      "loss": 1.8267,
      "step": 2077
    },
    {
      "epoch": 0.07006115394841797,
      "grad_norm": 2.3259341716766357,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 1.7801,
      "step": 2108
    },
    {
      "epoch": 0.07109146503589471,
      "grad_norm": 2.1711952686309814,
      "learning_rate": 4.994486281210429e-05,
      "loss": 1.7961,
      "step": 2139
    },
    {
      "epoch": 0.07212177612337145,
      "grad_norm": 2.1808884143829346,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 1.8109,
      "step": 2170
    },
    {
      "epoch": 0.07315208721084818,
      "grad_norm": 2.089256525039673,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 1.8026,
      "step": 2201
    },
    {
      "epoch": 0.07418239829832492,
      "grad_norm": 2.0864951610565186,
      "learning_rate": 4.992687246588743e-05,
      "loss": 1.7753,
      "step": 2232
    },
    {
      "epoch": 0.07521270938580164,
      "grad_norm": 2.36157488822937,
      "learning_rate": 4.992031299767347e-05,
      "loss": 1.7746,
      "step": 2263
    },
    {
      "epoch": 0.07624302047327838,
      "grad_norm": 2.5334439277648926,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 1.7927,
      "step": 2294
    },
    {
      "epoch": 0.07727333156075512,
      "grad_norm": 2.2565715312957764,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 1.7668,
      "step": 2325
    },
    {
      "epoch": 0.07830364264823185,
      "grad_norm": 2.1043128967285156,
      "learning_rate": 4.989894757091861e-05,
      "loss": 1.7771,
      "step": 2356
    },
    {
      "epoch": 0.07933395373570859,
      "grad_norm": 1.9659819602966309,
      "learning_rate": 4.989126368919158e-05,
      "loss": 1.7666,
      "step": 2387
    },
    {
      "epoch": 0.08036426482318532,
      "grad_norm": 2.0778403282165527,
      "learning_rate": 4.988329891293693e-05,
      "loss": 1.7405,
      "step": 2418
    },
    {
      "epoch": 0.08139457591066206,
      "grad_norm": 2.1767923831939697,
      "learning_rate": 4.987505333203608e-05,
      "loss": 1.7495,
      "step": 2449
    },
    {
      "epoch": 0.0824248869981388,
      "grad_norm": 2.260143280029297,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.7504,
      "step": 2480
    },
    {
      "epoch": 0.08345519808561552,
      "grad_norm": 2.18271803855896,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 1.7456,
      "step": 2511
    },
    {
      "epoch": 0.08448550917309226,
      "grad_norm": 2.209594964981079,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 1.7528,
      "step": 2542
    },
    {
      "epoch": 0.085515820260569,
      "grad_norm": 2.0666229724884033,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.7517,
      "step": 2573
    },
    {
      "epoch": 0.08654613134804573,
      "grad_norm": 2.1070454120635986,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.7474,
      "step": 2604
    },
    {
      "epoch": 0.08757644243552247,
      "grad_norm": 1.9430303573608398,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.7348,
      "step": 2635
    },
    {
      "epoch": 0.0886067535229992,
      "grad_norm": 1.9638925790786743,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.7202,
      "step": 2666
    },
    {
      "epoch": 0.08963706461047594,
      "grad_norm": 1.8845652341842651,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.7375,
      "step": 2697
    },
    {
      "epoch": 0.09066737569795268,
      "grad_norm": 5.712058067321777,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.7379,
      "step": 2728
    },
    {
      "epoch": 0.0916976867854294,
      "grad_norm": 1.9520670175552368,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.7302,
      "step": 2759
    },
    {
      "epoch": 0.09272799787290614,
      "grad_norm": 1.8802224397659302,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.74,
      "step": 2790
    },
    {
      "epoch": 0.09375830896038288,
      "grad_norm": 2.1098153591156006,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.7024,
      "step": 2821
    },
    {
      "epoch": 0.09478862004785961,
      "grad_norm": 2.1568291187286377,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.7131,
      "step": 2852
    },
    {
      "epoch": 0.09581893113533635,
      "grad_norm": 1.8769980669021606,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.7008,
      "step": 2883
    },
    {
      "epoch": 0.09684924222281308,
      "grad_norm": 1.8325533866882324,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.7058,
      "step": 2914
    },
    {
      "epoch": 0.09787955331028982,
      "grad_norm": 2.086416721343994,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.711,
      "step": 2945
    },
    {
      "epoch": 0.09890986439776656,
      "grad_norm": 1.8294793367385864,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.6998,
      "step": 2976
    },
    {
      "epoch": 0.09994017548524328,
      "grad_norm": 2.0608153343200684,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.6948,
      "step": 3007
    },
    {
      "epoch": 0.10097048657272002,
      "grad_norm": 3.2166008949279785,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.6788,
      "step": 3038
    }
  ],
  "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": 2.263945054022271e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}