{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 300000000000000000,
  "global_step": 4230,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02364066193853428,
      "grad_norm": 4265.77734375,
      "learning_rate": 2.695035460992908e-06,
      "loss": 19.7865,
      "step": 50
    },
    {
      "epoch": 0.04728132387706856,
      "grad_norm": Infinity,
      "learning_rate": 6.2411347517730495e-06,
      "loss": 9.5661,
      "step": 100
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 1082.762939453125,
      "learning_rate": 9.645390070921986e-06,
      "loss": 2.3602,
      "step": 150
    },
    {
      "epoch": 0.09456264775413711,
      "grad_norm": 249.59776306152344,
      "learning_rate": 1.3191489361702129e-05,
      "loss": 1.7304,
      "step": 200
    },
    {
      "epoch": 0.1182033096926714,
      "grad_norm": 171.15611267089844,
      "learning_rate": 1.673758865248227e-05,
      "loss": 1.6707,
      "step": 250
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 64.97687530517578,
      "learning_rate": 2.028368794326241e-05,
      "loss": 1.5563,
      "step": 300
    },
    {
      "epoch": 0.16548463356973994,
      "grad_norm": 22.22820472717285,
      "learning_rate": 2.3829787234042553e-05,
      "loss": 1.4809,
      "step": 350
    },
    {
      "epoch": 0.18912529550827423,
      "grad_norm": 17.94223403930664,
      "learning_rate": 2.7375886524822697e-05,
      "loss": 1.3877,
      "step": 400
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 6.133880615234375,
      "learning_rate": 2.999913686685402e-05,
      "loss": 1.2877,
      "step": 450
    },
    {
      "epoch": 0.2364066193853428,
      "grad_norm": 5.359964370727539,
      "learning_rate": 2.9979733510221677e-05,
      "loss": 1.1985,
      "step": 500
    },
    {
      "epoch": 0.26004728132387706,
      "grad_norm": 8.637283325195312,
      "learning_rate": 2.9934831545542617e-05,
      "loss": 1.1513,
      "step": 550
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 13.667312622070312,
      "learning_rate": 2.986450740525784e-05,
      "loss": 1.0896,
      "step": 600
    },
    {
      "epoch": 0.3073286052009456,
      "grad_norm": 6.358010292053223,
      "learning_rate": 2.9768880795615002e-05,
      "loss": 1.0343,
      "step": 650
    },
    {
      "epoch": 0.3309692671394799,
      "grad_norm": 4.579405784606934,
      "learning_rate": 2.9648114492903583e-05,
      "loss": 0.9778,
      "step": 700
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 4.688839435577393,
      "learning_rate": 2.950241406637593e-05,
      "loss": 0.9688,
      "step": 750
    },
    {
      "epoch": 0.37825059101654845,
      "grad_norm": 5.001623153686523,
      "learning_rate": 2.9332027528325834e-05,
      "loss": 0.9246,
      "step": 800
    },
    {
      "epoch": 0.40189125295508277,
      "grad_norm": 5.845919609069824,
      "learning_rate": 2.9137244911920255e-05,
      "loss": 0.9075,
      "step": 850
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 6.357454299926758,
      "learning_rate": 2.891839777750281e-05,
      "loss": 0.9032,
      "step": 900
    },
    {
      "epoch": 0.4491725768321513,
      "grad_norm": 4.689474105834961,
      "learning_rate": 2.8675858648209442e-05,
      "loss": 0.8833,
      "step": 950
    },
    {
      "epoch": 0.4728132387706856,
      "grad_norm": 5.220682621002197,
      "learning_rate": 2.841004037585688e-05,
      "loss": 0.8512,
      "step": 1000
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 6.682931423187256,
      "learning_rate": 2.8121395438183372e-05,
      "loss": 0.8526,
      "step": 1050
    },
    {
      "epoch": 0.5200945626477541,
      "grad_norm": 5.913801670074463,
      "learning_rate": 2.7810415168637912e-05,
      "loss": 0.8187,
      "step": 1100
    },
    {
      "epoch": 0.5437352245862884,
      "grad_norm": 5.719775676727295,
      "learning_rate": 2.7477628920028935e-05,
      "loss": 0.8174,
      "step": 1150
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 7.033783912658691,
      "learning_rate": 2.712360316345627e-05,
      "loss": 0.8378,
      "step": 1200
    },
    {
      "epoch": 0.5910165484633569,
      "grad_norm": 10.351150512695312,
      "learning_rate": 2.6748940524060027e-05,
      "loss": 0.81,
      "step": 1250
    },
    {
      "epoch": 0.6146572104018913,
      "grad_norm": 6.334929466247559,
      "learning_rate": 2.6354278755227802e-05,
      "loss": 0.8019,
      "step": 1300
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 4.533697605133057,
      "learning_rate": 2.5940289653006427e-05,
      "loss": 0.7874,
      "step": 1350
    },
    {
      "epoch": 0.6619385342789598,
      "grad_norm": 5.18574857711792,
      "learning_rate": 2.550767791256593e-05,
      "loss": 0.7758,
      "step": 1400
    },
    {
      "epoch": 0.6855791962174941,
      "grad_norm": 5.3152923583984375,
      "learning_rate": 2.5057179928662506e-05,
      "loss": 0.7811,
      "step": 1450
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 6.021467685699463,
      "learning_rate": 2.458956254214211e-05,
      "loss": 0.7849,
      "step": 1500
    },
    {
      "epoch": 0.7328605200945626,
      "grad_norm": 9.204541206359863,
      "learning_rate": 2.4105621734618613e-05,
      "loss": 0.7653,
      "step": 1550
    },
    {
      "epoch": 0.7565011820330969,
      "grad_norm": 7.0010504722595215,
      "learning_rate": 2.3606181273548253e-05,
      "loss": 0.7578,
      "step": 1600
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 4.713606357574463,
      "learning_rate": 2.309209131000687e-05,
      "loss": 0.743,
      "step": 1650
    },
    {
      "epoch": 0.8037825059101655,
      "grad_norm": 8.350544929504395,
      "learning_rate": 2.256422693155675e-05,
      "loss": 0.7327,
      "step": 1700
    },
    {
      "epoch": 0.8274231678486997,
      "grad_norm": 5.2699713706970215,
      "learning_rate": 2.2023486672666385e-05,
      "loss": 0.7252,
      "step": 1750
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 5.1293625831604,
      "learning_rate": 2.1470790985218804e-05,
      "loss": 0.7169,
      "step": 1800
    },
    {
      "epoch": 0.8747044917257684,
      "grad_norm": 5.4763407707214355,
      "learning_rate": 2.0907080671711832e-05,
      "loss": 0.7208,
      "step": 1850
    },
    {
      "epoch": 0.8983451536643026,
      "grad_norm": 13.955887794494629,
      "learning_rate": 2.0333315283817486e-05,
      "loss": 0.7309,
      "step": 1900
    },
    {
      "epoch": 0.9219858156028369,
      "grad_norm": 8.627326011657715,
      "learning_rate": 1.975047148902632e-05,
      "loss": 0.7155,
      "step": 1950
    },
    {
      "epoch": 0.9456264775413712,
      "grad_norm": 8.94743537902832,
      "learning_rate": 1.9159541408157158e-05,
      "loss": 0.7068,
      "step": 2000
    },
    {
      "epoch": 0.9692671394799054,
      "grad_norm": 5.135090351104736,
      "learning_rate": 1.8561530926562023e-05,
      "loss": 0.6918,
      "step": 2050
    },
    {
      "epoch": 0.9929078014184397,
      "grad_norm": 38.368629455566406,
      "learning_rate": 1.795745798190099e-05,
      "loss": 0.6818,
      "step": 2100
    },
    {
      "epoch": 1.016548463356974,
      "grad_norm": 25.27336311340332,
      "learning_rate": 1.734835083140153e-05,
      "loss": 0.685,
      "step": 2150
    },
    {
      "epoch": 1.0401891252955082,
      "grad_norm": 7.319797992706299,
      "learning_rate": 1.6735246301551825e-05,
      "loss": 0.6739,
      "step": 2200
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 10.93574333190918,
      "learning_rate": 1.6119188023207348e-05,
      "loss": 0.6828,
      "step": 2250
    },
    {
      "epoch": 1.0874704491725768,
      "grad_norm": 51.116207122802734,
      "learning_rate": 1.5501224655115118e-05,
      "loss": 0.6514,
      "step": 2300
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 17.643468856811523,
      "learning_rate": 1.4882408098879367e-05,
      "loss": 0.6708,
      "step": 2350
    },
    {
      "epoch": 1.1347517730496455,
      "grad_norm": 24.42500114440918,
      "learning_rate": 1.426379170840718e-05,
      "loss": 0.6595,
      "step": 2400
    },
    {
      "epoch": 1.1583924349881798,
      "grad_norm": 7.433150291442871,
      "learning_rate": 1.364642849688209e-05,
      "loss": 0.6545,
      "step": 2450
    },
    {
      "epoch": 1.1820330969267139,
      "grad_norm": 12.09028148651123,
      "learning_rate": 1.3031369344317569e-05,
      "loss": 0.6533,
      "step": 2500
    },
    {
      "epoch": 1.2056737588652482,
      "grad_norm": 8.16505241394043,
      "learning_rate": 1.2419661208741687e-05,
      "loss": 0.6466,
      "step": 2550
    },
    {
      "epoch": 1.2293144208037825,
      "grad_norm": 8.545854568481445,
      "learning_rate": 1.181234534405775e-05,
      "loss": 0.6311,
      "step": 2600
    },
    {
      "epoch": 1.2529550827423168,
      "grad_norm": 7.951985836029053,
      "learning_rate": 1.1210455527614574e-05,
      "loss": 0.6338,
      "step": 2650
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 17.269718170166016,
      "learning_rate": 1.061501630050338e-05,
      "loss": 0.6214,
      "step": 2700
    },
    {
      "epoch": 1.3002364066193852,
      "grad_norm": 14.052486419677734,
      "learning_rate": 1.0027041223576735e-05,
      "loss": 0.6307,
      "step": 2750
    },
    {
      "epoch": 1.3238770685579198,
      "grad_norm": 7.13311767578125,
      "learning_rate": 9.447531152158089e-06,
      "loss": 0.6273,
      "step": 2800
    },
    {
      "epoch": 1.3475177304964538,
      "grad_norm": 9.32107162475586,
      "learning_rate": 8.877472532378836e-06,
      "loss": 0.6218,
      "step": 2850
    },
    {
      "epoch": 1.3711583924349882,
      "grad_norm": 6.4954118728637695,
      "learning_rate": 8.317835722042693e-06,
      "loss": 0.6085,
      "step": 2900
    },
    {
      "epoch": 1.3947990543735225,
      "grad_norm": 7.570118427276611,
      "learning_rate": 7.769573338875851e-06,
      "loss": 0.6007,
      "step": 2950
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 18.045612335205078,
      "learning_rate": 7.2336186389743095e-06,
      "loss": 0.6177,
      "step": 3000
    },
    {
      "epoch": 1.442080378250591,
      "grad_norm": 10.485590934753418,
      "learning_rate": 6.710883928208835e-06,
      "loss": 0.5859,
      "step": 3050
    },
    {
      "epoch": 1.4657210401891252,
      "grad_norm": 13.521890640258789,
      "learning_rate": 6.202259009291401e-06,
      "loss": 0.6117,
      "step": 3100
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 10.763672828674316,
      "learning_rate": 5.708609667146788e-06,
      "loss": 0.5647,
      "step": 3150
    },
    {
      "epoch": 1.5130023640661938,
      "grad_norm": 13.315629959106445,
      "learning_rate": 5.2307761951673425e-06,
      "loss": 0.6115,
      "step": 3200
    },
    {
      "epoch": 1.5366430260047281,
      "grad_norm": 13.952415466308594,
      "learning_rate": 4.769571964859664e-06,
      "loss": 0.594,
      "step": 3250
    },
    {
      "epoch": 1.5602836879432624,
      "grad_norm": 9.341636657714844,
      "learning_rate": 4.325782041317874e-06,
      "loss": 0.5803,
      "step": 3300
    },
    {
      "epoch": 1.5839243498817965,
      "grad_norm": 23.52338218688965,
      "learning_rate": 3.900161846880281e-06,
      "loss": 0.5648,
      "step": 3350
    },
    {
      "epoch": 1.607565011820331,
      "grad_norm": 22.669052124023438,
      "learning_rate": 3.4934358752441315e-06,
      "loss": 0.5588,
      "step": 3400
    },
    {
      "epoch": 1.6312056737588652,
      "grad_norm": 50.38961410522461,
      "learning_rate": 3.106296458227363e-06,
      "loss": 0.5777,
      "step": 3450
    },
    {
      "epoch": 1.6548463356973995,
      "grad_norm": 6.503214359283447,
      "learning_rate": 2.7394025872764556e-06,
      "loss": 0.5688,
      "step": 3500
    },
    {
      "epoch": 1.6784869976359338,
      "grad_norm": 81.57908630371094,
      "learning_rate": 2.39337879172658e-06,
      "loss": 0.5849,
      "step": 3550
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 10.680161476135254,
      "learning_rate": 2.0688140757233428e-06,
      "loss": 0.5761,
      "step": 3600
    },
    {
      "epoch": 1.7257683215130024,
      "grad_norm": 6.659136772155762,
      "learning_rate": 1.7662609156157749e-06,
      "loss": 0.566,
      "step": 3650
    },
    {
      "epoch": 1.7494089834515365,
      "grad_norm": 13.552983283996582,
      "learning_rate": 1.486234319527186e-06,
      "loss": 0.5625,
      "step": 3700
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 22.759199142456055,
      "learning_rate": 1.2292109507047273e-06,
      "loss": 0.557,
      "step": 3750
    },
    {
      "epoch": 1.7966903073286051,
      "grad_norm": 36.262474060058594,
      "learning_rate": 9.956283161398172e-07,
      "loss": 0.5561,
      "step": 3800
    },
    {
      "epoch": 1.8203309692671394,
      "grad_norm": 8.722681999206543,
      "learning_rate": 7.858840218406921e-07,
      "loss": 0.5624,
      "step": 3850
    },
    {
      "epoch": 1.8439716312056738,
      "grad_norm": 6.087028980255127,
      "learning_rate": 6.00335096024619e-07,
      "loss": 0.5747,
      "step": 3900
    },
    {
      "epoch": 1.867612293144208,
      "grad_norm": 9.753291130065918,
      "learning_rate": 4.3929738138196787e-07,
      "loss": 0.5629,
      "step": 3950
    },
    {
      "epoch": 1.8912529550827424,
      "grad_norm": 13.233976364135742,
      "learning_rate": 3.030449974465471e-07,
      "loss": 0.5588,
      "step": 4000
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 23.338430404663086,
      "learning_rate": 1.9180987398740358e-07,
      "loss": 0.5632,
      "step": 4050
    },
    {
      "epoch": 1.938534278959811,
      "grad_norm": 30.731971740722656,
      "learning_rate": 1.0578135621633178e-07,
      "loss": 0.5566,
      "step": 4100
    },
    {
      "epoch": 1.962174940898345,
      "grad_norm": 12.004512786865234,
      "learning_rate": 4.5105882483119643e-08,
      "loss": 0.5446,
      "step": 4150
    },
    {
      "epoch": 1.9858156028368794,
      "grad_norm": 18.55535125732422,
      "learning_rate": 9.886735007152425e-09,
      "loss": 0.5539,
      "step": 4200
    }
  ],
  "logging_steps": 50,
  "max_steps": 4230,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.480992965002527e+17,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}