{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6144764723689078,
  "eval_steps": 500,
  "global_step": 12000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 4.9997567688496474e-05,
      "loss": 2.8699,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9990172715142793e-05,
      "loss": 0.0389,
      "step": 200
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.997781629993153e-05,
      "loss": 0.0301,
      "step": 300
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9960500896052476e-05,
      "loss": 0.0244,
      "step": 400
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.993822994123172e-05,
      "loss": 0.0151,
      "step": 500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.9911007857049264e-05,
      "loss": 0.012,
      "step": 600
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.987884004806111e-05,
      "loss": 0.0268,
      "step": 700
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.984173290072626e-05,
      "loss": 0.0187,
      "step": 800
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.979969378213884e-05,
      "loss": 0.0166,
      "step": 900
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.975273103856537e-05,
      "loss": 0.022,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.970085399378785e-05,
      "loss": 0.0194,
      "step": 1100
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.964407294725254e-05,
      "loss": 0.0182,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.958239917202523e-05,
      "loss": 0.0204,
      "step": 1300
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.9515844912553106e-05,
      "loss": 0.013,
      "step": 1400
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.944442338223378e-05,
      "loss": 0.0107,
      "step": 1500
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.9368148760792e-05,
      "loss": 0.0156,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.9287036191464414e-05,
      "loss": 0.0141,
      "step": 1700
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.9201101777993116e-05,
      "loss": 0.0104,
      "step": 1800
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.911036258142853e-05,
      "loss": 0.0087,
      "step": 1900
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.9014836616742065e-05,
      "loss": 0.013,
      "step": 2000
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.891454284924965e-05,
      "loss": 0.0194,
      "step": 2100
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.8809501190846284e-05,
      "loss": 0.0131,
      "step": 2200
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.8699732496052965e-05,
      "loss": 0.0135,
      "step": 2300
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.858525855787619e-05,
      "loss": 0.0106,
      "step": 2400
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.846610210348137e-05,
      "loss": 0.0089,
      "step": 2500
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.8342286789680605e-05,
      "loss": 0.0086,
      "step": 2600
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.8213837198235996e-05,
      "loss": 0.0035,
      "step": 2700
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.8080778830979276e-05,
      "loss": 0.0121,
      "step": 2800
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.794313810474878e-05,
      "loss": 0.0146,
      "step": 2900
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.780094234614479e-05,
      "loss": 0.0047,
      "step": 3000
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.765421978610418e-05,
      "loss": 0.0096,
      "step": 3100
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.750299955429562e-05,
      "loss": 0.0116,
      "step": 3200
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.734731167333625e-05,
      "loss": 0.0192,
      "step": 3300
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.718718705283113e-05,
      "loss": 0.0111,
      "step": 3400
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.702265748323661e-05,
      "loss": 0.0114,
      "step": 3500
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.685375562954876e-05,
      "loss": 0.0113,
      "step": 3600
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.668051502481818e-05,
      "loss": 0.0097,
      "step": 3700
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.650297006349257e-05,
      "loss": 0.0094,
      "step": 3800
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.63211559945881e-05,
      "loss": 0.0059,
      "step": 3900
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.613699021830524e-05,
      "loss": 0.0195,
      "step": 4000
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.59467888395669e-05,
      "loss": 0.0088,
      "step": 4100
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.575242877508777e-05,
      "loss": 0.0087,
      "step": 4200
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.5553948612282607e-05,
      "loss": 0.0088,
      "step": 4300
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.5351387756552846e-05,
      "loss": 0.0122,
      "step": 4400
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.51447864234632e-05,
      "loss": 0.0081,
      "step": 4500
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.4934185630757484e-05,
      "loss": 0.0079,
      "step": 4600
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.4719627190215064e-05,
      "loss": 0.011,
      "step": 4700
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.450115369934976e-05,
      "loss": 0.0146,
      "step": 4800
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.427880853295274e-05,
      "loss": 0.0057,
      "step": 4900
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.4052635834481025e-05,
      "loss": 0.0051,
      "step": 5000
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.3822680507293455e-05,
      "loss": 0.006,
      "step": 5100
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.358898820573581e-05,
      "loss": 0.0127,
      "step": 5200
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.3351605326076724e-05,
      "loss": 0.0062,
      "step": 5300
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.3110578997296416e-05,
      "loss": 0.0157,
      "step": 5400
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.286595707172986e-05,
      "loss": 0.0111,
      "step": 5500
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.261778811556646e-05,
      "loss": 0.0141,
      "step": 5600
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.236612139920786e-05,
      "loss": 0.0093,
      "step": 5700
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.2111006887486035e-05,
      "loss": 0.0091,
      "step": 5800
    },
    {
      "epoch": 0.79,
      "learning_rate": 4.185249522974346e-05,
      "loss": 0.0043,
      "step": 5900
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.159063774977748e-05,
      "loss": 0.0072,
      "step": 6000
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.1325486435650625e-05,
      "loss": 0.0063,
      "step": 6100
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.105709392936914e-05,
      "loss": 0.005,
      "step": 6200
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.0785513516431705e-05,
      "loss": 0.0167,
      "step": 6300
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.051079911525031e-05,
      "loss": 0.0068,
      "step": 6400
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.023300526644557e-05,
      "loss": 0.0076,
      "step": 6500
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.9952187122018435e-05,
      "loss": 0.0123,
      "step": 6600
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.966840043440054e-05,
      "loss": 0.0074,
      "step": 6700
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.9381701545385384e-05,
      "loss": 0.0045,
      "step": 6800
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.9092147374942454e-05,
      "loss": 0.0141,
      "step": 6900
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.87997954099166e-05,
      "loss": 0.0036,
      "step": 7000
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.8504703692614795e-05,
      "loss": 0.0077,
      "step": 7100
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.820693080928271e-05,
      "loss": 0.0063,
      "step": 7200
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.790653587847325e-05,
      "loss": 0.0078,
      "step": 7300
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.760357853930936e-05,
      "loss": 0.004,
      "step": 7400
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.72981189396436e-05,
      "loss": 0.003,
      "step": 7500
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.699021772411661e-05,
      "loss": 0.0064,
      "step": 7600
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.667993602211695e-05,
      "loss": 0.0047,
      "step": 7700
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.636733543564481e-05,
      "loss": 0.0031,
      "step": 7800
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.605247802708175e-05,
      "loss": 0.0015,
      "step": 7900
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.5735426306869135e-05,
      "loss": 0.0103,
      "step": 8000
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.5416243221097553e-05,
      "loss": 0.0062,
      "step": 8100
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.509499213900977e-05,
      "loss": 0.0046,
      "step": 8200
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.477497910387336e-05,
      "loss": 0.0082,
      "step": 8300
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.44498028479363e-05,
      "loss": 0.0065,
      "step": 8400
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.412275046861546e-05,
      "loss": 0.0025,
      "step": 8500
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.379388689748946e-05,
      "loss": 0.0047,
      "step": 8600
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.346327742572318e-05,
      "loss": 0.0028,
      "step": 8700
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.3130987691105155e-05,
      "loss": 0.0037,
      "step": 8800
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.279708366501608e-05,
      "loss": 0.0067,
      "step": 8900
    },
    {
      "epoch": 1.21,
      "learning_rate": 3.246163163933119e-05,
      "loss": 0.0045,
      "step": 9000
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.2124698213258904e-05,
      "loss": 0.0038,
      "step": 9100
    },
    {
      "epoch": 1.24,
      "learning_rate": 3.178635028011855e-05,
      "loss": 0.0103,
      "step": 9200
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.144665501405963e-05,
      "loss": 0.0061,
      "step": 9300
    },
    {
      "epoch": 1.26,
      "learning_rate": 3.110567985672535e-05,
      "loss": 0.0042,
      "step": 9400
    },
    {
      "epoch": 1.28,
      "learning_rate": 3.076349250386302e-05,
      "loss": 0.0058,
      "step": 9500
    },
    {
      "epoch": 1.29,
      "learning_rate": 3.0420160891884115e-05,
      "loss": 0.0077,
      "step": 9600
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.0075753184376354e-05,
      "loss": 0.0036,
      "step": 9700
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.9730337758570913e-05,
      "loss": 0.007,
      "step": 9800
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.938398319176699e-05,
      "loss": 0.0073,
      "step": 9900
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.903675824771684e-05,
      "loss": 0.0053,
      "step": 10000
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.8688731862973677e-05,
      "loss": 0.0074,
      "step": 10100
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.833997313320535e-05,
      "loss": 0.004,
      "step": 10200
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.799055129947631e-05,
      "loss": 0.003,
      "step": 10300
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.7644038601038813e-05,
      "loss": 0.0061,
      "step": 10400
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.7293503693483008e-05,
      "loss": 0.0044,
      "step": 10500
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.6942513443527008e-05,
      "loss": 0.0029,
      "step": 10600
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.6591137535271742e-05,
      "loss": 0.001,
      "step": 10700
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.6239445729385104e-05,
      "loss": 0.0047,
      "step": 10800
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.5887507849251914e-05,
      "loss": 0.005,
      "step": 10900
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5535393767111533e-05,
      "loss": 0.0023,
      "step": 11000
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5183173390185736e-05,
      "loss": 0.003,
      "step": 11100
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.483091664679959e-05,
      "loss": 0.0016,
      "step": 11200
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.44786934724982e-05,
      "loss": 0.0027,
      "step": 11300
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4126573796162015e-05,
      "loss": 0.003,
      "step": 11400
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.377462752612344e-05,
      "loss": 0.002,
      "step": 11500
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.3422924536287548e-05,
      "loss": 0.0026,
      "step": 11600
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.307153465225958e-05,
      "loss": 0.0037,
      "step": 11700
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.272052763748209e-05,
      "loss": 0.0063,
      "step": 11800
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.2369973179384445e-05,
      "loss": 0.0046,
      "step": 11900
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.201994087554733e-05,
      "loss": 0.0094,
      "step": 12000
    }
  ],
  "logging_steps": 100,
  "max_steps": 22296,
  "num_train_epochs": 3,
  "save_steps": 3000,
  "total_flos": 1.9310189130989568e+18,
  "trial_name": null,
  "trial_params": null
}