{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 37460,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13347570742124934,
      "grad_norm": 3.1176345348358154,
      "learning_rate": 4.933395621996797e-05,
      "loss": 1.6549,
      "step": 500
    },
    {
      "epoch": 0.2669514148424987,
      "grad_norm": 2.8823812007904053,
      "learning_rate": 4.866657768286172e-05,
      "loss": 1.3869,
      "step": 1000
    },
    {
      "epoch": 0.400427122263748,
      "grad_norm": 3.436349630355835,
      "learning_rate": 4.7999199145755475e-05,
      "loss": 1.31,
      "step": 1500
    },
    {
      "epoch": 0.5339028296849974,
      "grad_norm": 2.7581417560577393,
      "learning_rate": 4.733182060864923e-05,
      "loss": 1.2632,
      "step": 2000
    },
    {
      "epoch": 0.6673785371062466,
      "grad_norm": 3.0282974243164062,
      "learning_rate": 4.666444207154298e-05,
      "loss": 1.2197,
      "step": 2500
    },
    {
      "epoch": 0.800854244527496,
      "grad_norm": 2.7351794242858887,
      "learning_rate": 4.599706353443674e-05,
      "loss": 1.1942,
      "step": 3000
    },
    {
      "epoch": 0.9343299519487454,
      "grad_norm": 2.228116035461426,
      "learning_rate": 4.532968499733049e-05,
      "loss": 1.1273,
      "step": 3500
    },
    {
      "epoch": 1.0678056593699947,
      "grad_norm": 3.064039707183838,
      "learning_rate": 4.466230646022424e-05,
      "loss": 1.0261,
      "step": 4000
    },
    {
      "epoch": 1.201281366791244,
      "grad_norm": 2.9547019004821777,
      "learning_rate": 4.3994927923117995e-05,
      "loss": 0.9262,
      "step": 4500
    },
    {
      "epoch": 1.3347570742124932,
      "grad_norm": 2.8099584579467773,
      "learning_rate": 4.332754938601175e-05,
      "loss": 0.9193,
      "step": 5000
    },
    {
      "epoch": 1.4682327816337426,
      "grad_norm": 2.4523322582244873,
      "learning_rate": 4.26601708489055e-05,
      "loss": 0.939,
      "step": 5500
    },
    {
      "epoch": 1.601708489054992,
      "grad_norm": 2.4100770950317383,
      "learning_rate": 4.199279231179926e-05,
      "loss": 0.9084,
      "step": 6000
    },
    {
      "epoch": 1.7351841964762413,
      "grad_norm": 2.4628851413726807,
      "learning_rate": 4.1325413774693004e-05,
      "loss": 0.9303,
      "step": 6500
    },
    {
      "epoch": 1.8686599038974907,
      "grad_norm": 2.522763967514038,
      "learning_rate": 4.0658035237586763e-05,
      "loss": 0.8988,
      "step": 7000
    },
    {
      "epoch": 2.00213561131874,
      "grad_norm": 1.8565527200698853,
      "learning_rate": 3.9990656700480516e-05,
      "loss": 0.9408,
      "step": 7500
    },
    {
      "epoch": 2.1356113187399894,
      "grad_norm": 3.084707498550415,
      "learning_rate": 3.932327816337427e-05,
      "loss": 0.7548,
      "step": 8000
    },
    {
      "epoch": 2.269087026161239,
      "grad_norm": 1.836754322052002,
      "learning_rate": 3.865589962626802e-05,
      "loss": 0.7447,
      "step": 8500
    },
    {
      "epoch": 2.402562733582488,
      "grad_norm": 2.452847719192505,
      "learning_rate": 3.798852108916178e-05,
      "loss": 0.7293,
      "step": 9000
    },
    {
      "epoch": 2.536038441003737,
      "grad_norm": 3.363875150680542,
      "learning_rate": 3.7321142552055525e-05,
      "loss": 0.7437,
      "step": 9500
    },
    {
      "epoch": 2.6695141484249865,
      "grad_norm": 2.087559700012207,
      "learning_rate": 3.6653764014949284e-05,
      "loss": 0.7217,
      "step": 10000
    },
    {
      "epoch": 2.802989855846236,
      "grad_norm": 2.1554408073425293,
      "learning_rate": 3.598638547784303e-05,
      "loss": 0.7459,
      "step": 10500
    },
    {
      "epoch": 2.936465563267485,
      "grad_norm": 1.8922470808029175,
      "learning_rate": 3.531900694073679e-05,
      "loss": 0.7586,
      "step": 11000
    },
    {
      "epoch": 3.0699412706887346,
      "grad_norm": 2.0937387943267822,
      "learning_rate": 3.465162840363054e-05,
      "loss": 0.6552,
      "step": 11500
    },
    {
      "epoch": 3.203416978109984,
      "grad_norm": 2.954742908477783,
      "learning_rate": 3.398424986652429e-05,
      "loss": 0.6012,
      "step": 12000
    },
    {
      "epoch": 3.3368926855312333,
      "grad_norm": 2.8236591815948486,
      "learning_rate": 3.3316871329418045e-05,
      "loss": 0.607,
      "step": 12500
    },
    {
      "epoch": 3.4703683929524827,
      "grad_norm": 2.4959943294525146,
      "learning_rate": 3.2649492792311804e-05,
      "loss": 0.6181,
      "step": 13000
    },
    {
      "epoch": 3.603844100373732,
      "grad_norm": 3.4247825145721436,
      "learning_rate": 3.198211425520555e-05,
      "loss": 0.6029,
      "step": 13500
    },
    {
      "epoch": 3.7373198077949814,
      "grad_norm": 2.6698663234710693,
      "learning_rate": 3.131473571809931e-05,
      "loss": 0.6032,
      "step": 14000
    },
    {
      "epoch": 3.8707955152162308,
      "grad_norm": 1.9999539852142334,
      "learning_rate": 3.064735718099306e-05,
      "loss": 0.6241,
      "step": 14500
    },
    {
      "epoch": 4.00427122263748,
      "grad_norm": 2.449364185333252,
      "learning_rate": 2.9979978643886814e-05,
      "loss": 0.6183,
      "step": 15000
    },
    {
      "epoch": 4.1377469300587295,
      "grad_norm": 4.328239917755127,
      "learning_rate": 2.931260010678057e-05,
      "loss": 0.4832,
      "step": 15500
    },
    {
      "epoch": 4.271222637479979,
      "grad_norm": 1.9375187158584595,
      "learning_rate": 2.8645221569674318e-05,
      "loss": 0.4963,
      "step": 16000
    },
    {
      "epoch": 4.404698344901228,
      "grad_norm": 2.9625084400177,
      "learning_rate": 2.7977843032568074e-05,
      "loss": 0.5162,
      "step": 16500
    },
    {
      "epoch": 4.538174052322478,
      "grad_norm": 3.514462947845459,
      "learning_rate": 2.731046449546183e-05,
      "loss": 0.4913,
      "step": 17000
    },
    {
      "epoch": 4.671649759743727,
      "grad_norm": 2.035665988922119,
      "learning_rate": 2.664308595835558e-05,
      "loss": 0.5062,
      "step": 17500
    },
    {
      "epoch": 4.805125467164976,
      "grad_norm": 3.025426149368286,
      "learning_rate": 2.5975707421249334e-05,
      "loss": 0.5167,
      "step": 18000
    },
    {
      "epoch": 4.938601174586225,
      "grad_norm": 3.0361483097076416,
      "learning_rate": 2.530832888414309e-05,
      "loss": 0.4961,
      "step": 18500
    },
    {
      "epoch": 5.072076882007474,
      "grad_norm": 2.769930124282837,
      "learning_rate": 2.464095034703684e-05,
      "loss": 0.4582,
      "step": 19000
    },
    {
      "epoch": 5.205552589428724,
      "grad_norm": 2.209397077560425,
      "learning_rate": 2.3973571809930594e-05,
      "loss": 0.3926,
      "step": 19500
    },
    {
      "epoch": 5.339028296849973,
      "grad_norm": 2.680094003677368,
      "learning_rate": 2.3306193272824347e-05,
      "loss": 0.4187,
      "step": 20000
    },
    {
      "epoch": 5.472504004271222,
      "grad_norm": 2.425154447555542,
      "learning_rate": 2.26388147357181e-05,
      "loss": 0.4182,
      "step": 20500
    },
    {
      "epoch": 5.605979711692472,
      "grad_norm": 4.719789505004883,
      "learning_rate": 2.197143619861185e-05,
      "loss": 0.4257,
      "step": 21000
    },
    {
      "epoch": 5.739455419113721,
      "grad_norm": 1.3763984441757202,
      "learning_rate": 2.1304057661505607e-05,
      "loss": 0.4222,
      "step": 21500
    },
    {
      "epoch": 5.87293112653497,
      "grad_norm": 2.8332626819610596,
      "learning_rate": 2.063667912439936e-05,
      "loss": 0.4247,
      "step": 22000
    },
    {
      "epoch": 6.00640683395622,
      "grad_norm": 1.990326166152954,
      "learning_rate": 1.996930058729311e-05,
      "loss": 0.4149,
      "step": 22500
    },
    {
      "epoch": 6.139882541377469,
      "grad_norm": 1.9501370191574097,
      "learning_rate": 1.9301922050186867e-05,
      "loss": 0.3451,
      "step": 23000
    },
    {
      "epoch": 6.2733582487987185,
      "grad_norm": 2.518463611602783,
      "learning_rate": 1.863454351308062e-05,
      "loss": 0.3338,
      "step": 23500
    },
    {
      "epoch": 6.406833956219968,
      "grad_norm": 2.567760705947876,
      "learning_rate": 1.7967164975974375e-05,
      "loss": 0.3537,
      "step": 24000
    },
    {
      "epoch": 6.540309663641217,
      "grad_norm": 3.4953105449676514,
      "learning_rate": 1.7299786438868128e-05,
      "loss": 0.3463,
      "step": 24500
    },
    {
      "epoch": 6.673785371062467,
      "grad_norm": 4.280869007110596,
      "learning_rate": 1.663240790176188e-05,
      "loss": 0.3794,
      "step": 25000
    },
    {
      "epoch": 6.807261078483716,
      "grad_norm": 1.3708416223526,
      "learning_rate": 1.5965029364655636e-05,
      "loss": 0.3595,
      "step": 25500
    },
    {
      "epoch": 6.940736785904965,
      "grad_norm": 1.9402921199798584,
      "learning_rate": 1.5297650827549388e-05,
      "loss": 0.3567,
      "step": 26000
    },
    {
      "epoch": 7.074212493326215,
      "grad_norm": 3.4459097385406494,
      "learning_rate": 1.463027229044314e-05,
      "loss": 0.3271,
      "step": 26500
    },
    {
      "epoch": 7.207688200747464,
      "grad_norm": 3.0598597526550293,
      "learning_rate": 1.3962893753336892e-05,
      "loss": 0.3053,
      "step": 27000
    },
    {
      "epoch": 7.3411639081687134,
      "grad_norm": 1.887160062789917,
      "learning_rate": 1.3295515216230648e-05,
      "loss": 0.2941,
      "step": 27500
    },
    {
      "epoch": 7.474639615589963,
      "grad_norm": 2.143535852432251,
      "learning_rate": 1.26281366791244e-05,
      "loss": 0.2981,
      "step": 28000
    },
    {
      "epoch": 7.608115323011212,
      "grad_norm": 4.371819019317627,
      "learning_rate": 1.1960758142018154e-05,
      "loss": 0.2986,
      "step": 28500
    },
    {
      "epoch": 7.7415910304324616,
      "grad_norm": 3.378239631652832,
      "learning_rate": 1.1293379604911907e-05,
      "loss": 0.305,
      "step": 29000
    },
    {
      "epoch": 7.875066737853711,
      "grad_norm": 2.8271336555480957,
      "learning_rate": 1.062600106780566e-05,
      "loss": 0.3193,
      "step": 29500
    },
    {
      "epoch": 8.00854244527496,
      "grad_norm": 2.825468063354492,
      "learning_rate": 9.958622530699413e-06,
      "loss": 0.3004,
      "step": 30000
    },
    {
      "epoch": 8.14201815269621,
      "grad_norm": 2.4525914192199707,
      "learning_rate": 9.291243993593167e-06,
      "loss": 0.2669,
      "step": 30500
    },
    {
      "epoch": 8.275493860117459,
      "grad_norm": 4.695706844329834,
      "learning_rate": 8.62386545648692e-06,
      "loss": 0.2737,
      "step": 31000
    },
    {
      "epoch": 8.408969567538708,
      "grad_norm": 3.336611270904541,
      "learning_rate": 7.956486919380673e-06,
      "loss": 0.2697,
      "step": 31500
    },
    {
      "epoch": 8.542445274959958,
      "grad_norm": 3.6071436405181885,
      "learning_rate": 7.289108382274426e-06,
      "loss": 0.2644,
      "step": 32000
    },
    {
      "epoch": 8.675920982381207,
      "grad_norm": 3.1964738368988037,
      "learning_rate": 6.62172984516818e-06,
      "loss": 0.2669,
      "step": 32500
    },
    {
      "epoch": 8.809396689802456,
      "grad_norm": 2.115790843963623,
      "learning_rate": 5.9543513080619334e-06,
      "loss": 0.2638,
      "step": 33000
    },
    {
      "epoch": 8.942872397223706,
      "grad_norm": 1.7221794128417969,
      "learning_rate": 5.286972770955687e-06,
      "loss": 0.282,
      "step": 33500
    },
    {
      "epoch": 9.076348104644955,
      "grad_norm": 2.4076969623565674,
      "learning_rate": 4.61959423384944e-06,
      "loss": 0.2601,
      "step": 34000
    },
    {
      "epoch": 9.209823812066205,
      "grad_norm": 1.874233603477478,
      "learning_rate": 3.952215696743193e-06,
      "loss": 0.2474,
      "step": 34500
    },
    {
      "epoch": 9.343299519487454,
      "grad_norm": 2.8450875282287598,
      "learning_rate": 3.2848371596369464e-06,
      "loss": 0.2405,
      "step": 35000
    },
    {
      "epoch": 9.476775226908703,
      "grad_norm": 2.4307425022125244,
      "learning_rate": 2.6174586225306996e-06,
      "loss": 0.244,
      "step": 35500
    },
    {
      "epoch": 9.610250934329953,
      "grad_norm": 1.2817405462265015,
      "learning_rate": 1.9500800854244527e-06,
      "loss": 0.2457,
      "step": 36000
    },
    {
      "epoch": 9.743726641751202,
      "grad_norm": 2.613475799560547,
      "learning_rate": 1.282701548318206e-06,
      "loss": 0.2446,
      "step": 36500
    },
    {
      "epoch": 9.877202349172451,
      "grad_norm": 1.7706152200698853,
      "learning_rate": 6.153230112119594e-07,
      "loss": 0.2464,
      "step": 37000
    },
    {
      "epoch": 10.0,
      "step": 37460,
      "total_flos": 1.5022851621224448e+16,
      "train_loss": 0.5662414055868578,
      "train_runtime": 9241.9444,
      "train_samples_per_second": 16.213,
      "train_steps_per_second": 4.053
    }
  ],
  "logging_steps": 500,
  "max_steps": 37460,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5022851621224448e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}