{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1171.4285714285713,
  "eval_steps": 500,
  "global_step": 8200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 7.14,
      "learning_rate": 9.947619047619048e-05,
      "loss": 0.6717,
      "step": 50
    },
    {
      "epoch": 14.29,
      "learning_rate": 9.888095238095238e-05,
      "loss": 0.2784,
      "step": 100
    },
    {
      "epoch": 21.43,
      "learning_rate": 9.828571428571429e-05,
      "loss": 0.145,
      "step": 150
    },
    {
      "epoch": 28.57,
      "learning_rate": 9.76904761904762e-05,
      "loss": 0.1728,
      "step": 200
    },
    {
      "epoch": 35.71,
      "learning_rate": 9.70952380952381e-05,
      "loss": 0.1473,
      "step": 250
    },
    {
      "epoch": 42.86,
      "learning_rate": 9.65e-05,
      "loss": 0.1465,
      "step": 300
    },
    {
      "epoch": 50.0,
      "learning_rate": 9.59047619047619e-05,
      "loss": 0.1149,
      "step": 350
    },
    {
      "epoch": 57.14,
      "learning_rate": 9.530952380952381e-05,
      "loss": 0.1282,
      "step": 400
    },
    {
      "epoch": 64.29,
      "learning_rate": 9.471428571428573e-05,
      "loss": 0.1007,
      "step": 450
    },
    {
      "epoch": 71.43,
      "learning_rate": 9.411904761904763e-05,
      "loss": 0.1021,
      "step": 500
    },
    {
      "epoch": 78.57,
      "learning_rate": 9.352380952380952e-05,
      "loss": 0.1168,
      "step": 550
    },
    {
      "epoch": 85.71,
      "learning_rate": 9.292857142857143e-05,
      "loss": 0.1142,
      "step": 600
    },
    {
      "epoch": 92.86,
      "learning_rate": 9.233333333333333e-05,
      "loss": 0.1003,
      "step": 650
    },
    {
      "epoch": 100.0,
      "learning_rate": 9.173809523809525e-05,
      "loss": 0.1018,
      "step": 700
    },
    {
      "epoch": 107.14,
      "learning_rate": 9.114285714285716e-05,
      "loss": 0.1016,
      "step": 750
    },
    {
      "epoch": 114.29,
      "learning_rate": 9.054761904761905e-05,
      "loss": 0.0862,
      "step": 800
    },
    {
      "epoch": 121.43,
      "learning_rate": 8.995238095238095e-05,
      "loss": 0.0949,
      "step": 850
    },
    {
      "epoch": 128.57,
      "learning_rate": 8.935714285714285e-05,
      "loss": 0.0958,
      "step": 900
    },
    {
      "epoch": 135.71,
      "learning_rate": 8.876190476190477e-05,
      "loss": 0.0988,
      "step": 950
    },
    {
      "epoch": 142.86,
      "learning_rate": 8.816666666666668e-05,
      "loss": 0.1019,
      "step": 1000
    },
    {
      "epoch": 150.0,
      "learning_rate": 8.757142857142857e-05,
      "loss": 0.0828,
      "step": 1050
    },
    {
      "epoch": 157.14,
      "learning_rate": 8.697619047619047e-05,
      "loss": 0.1038,
      "step": 1100
    },
    {
      "epoch": 164.29,
      "learning_rate": 8.638095238095239e-05,
      "loss": 0.0786,
      "step": 1150
    },
    {
      "epoch": 171.43,
      "learning_rate": 8.57857142857143e-05,
      "loss": 0.0899,
      "step": 1200
    },
    {
      "epoch": 178.57,
      "learning_rate": 8.51904761904762e-05,
      "loss": 0.0902,
      "step": 1250
    },
    {
      "epoch": 185.71,
      "learning_rate": 8.459523809523809e-05,
      "loss": 0.092,
      "step": 1300
    },
    {
      "epoch": 192.86,
      "learning_rate": 8.4e-05,
      "loss": 0.1105,
      "step": 1350
    },
    {
      "epoch": 200.0,
      "learning_rate": 8.340476190476191e-05,
      "loss": 0.1127,
      "step": 1400
    },
    {
      "epoch": 207.14,
      "learning_rate": 8.280952380952382e-05,
      "loss": 0.0982,
      "step": 1450
    },
    {
      "epoch": 214.29,
      "learning_rate": 8.221428571428572e-05,
      "loss": 0.0833,
      "step": 1500
    },
    {
      "epoch": 221.43,
      "learning_rate": 8.161904761904763e-05,
      "loss": 0.0933,
      "step": 1550
    },
    {
      "epoch": 228.57,
      "learning_rate": 8.102380952380952e-05,
      "loss": 0.0783,
      "step": 1600
    },
    {
      "epoch": 235.71,
      "learning_rate": 8.042857142857144e-05,
      "loss": 0.0984,
      "step": 1650
    },
    {
      "epoch": 242.86,
      "learning_rate": 7.983333333333334e-05,
      "loss": 0.0915,
      "step": 1700
    },
    {
      "epoch": 250.0,
      "learning_rate": 7.923809523809524e-05,
      "loss": 0.0815,
      "step": 1750
    },
    {
      "epoch": 257.14,
      "learning_rate": 7.864285714285715e-05,
      "loss": 0.1008,
      "step": 1800
    },
    {
      "epoch": 264.29,
      "learning_rate": 7.804761904761905e-05,
      "loss": 0.0739,
      "step": 1850
    },
    {
      "epoch": 271.43,
      "learning_rate": 7.745238095238096e-05,
      "loss": 0.0777,
      "step": 1900
    },
    {
      "epoch": 278.57,
      "learning_rate": 7.685714285714286e-05,
      "loss": 0.0946,
      "step": 1950
    },
    {
      "epoch": 285.71,
      "learning_rate": 7.626190476190477e-05,
      "loss": 0.0688,
      "step": 2000
    },
    {
      "epoch": 292.86,
      "learning_rate": 7.567857142857143e-05,
      "loss": 0.0775,
      "step": 2050
    },
    {
      "epoch": 300.0,
      "learning_rate": 7.508333333333333e-05,
      "loss": 0.0716,
      "step": 2100
    },
    {
      "epoch": 307.14,
      "learning_rate": 7.448809523809525e-05,
      "loss": 0.0658,
      "step": 2150
    },
    {
      "epoch": 314.29,
      "learning_rate": 7.389285714285714e-05,
      "loss": 0.0853,
      "step": 2200
    },
    {
      "epoch": 321.43,
      "learning_rate": 7.329761904761905e-05,
      "loss": 0.0672,
      "step": 2250
    },
    {
      "epoch": 328.57,
      "learning_rate": 7.270238095238095e-05,
      "loss": 0.0764,
      "step": 2300
    },
    {
      "epoch": 335.71,
      "learning_rate": 7.210714285714286e-05,
      "loss": 0.1021,
      "step": 2350
    },
    {
      "epoch": 342.86,
      "learning_rate": 7.151190476190477e-05,
      "loss": 0.0771,
      "step": 2400
    },
    {
      "epoch": 350.0,
      "learning_rate": 7.091666666666666e-05,
      "loss": 0.0682,
      "step": 2450
    },
    {
      "epoch": 357.14,
      "learning_rate": 7.032142857142857e-05,
      "loss": 0.0714,
      "step": 2500
    },
    {
      "epoch": 364.29,
      "learning_rate": 6.972619047619047e-05,
      "loss": 0.073,
      "step": 2550
    },
    {
      "epoch": 371.43,
      "learning_rate": 6.913095238095239e-05,
      "loss": 0.0627,
      "step": 2600
    },
    {
      "epoch": 378.57,
      "learning_rate": 6.85357142857143e-05,
      "loss": 0.0748,
      "step": 2650
    },
    {
      "epoch": 385.71,
      "learning_rate": 6.79404761904762e-05,
      "loss": 0.0668,
      "step": 2700
    },
    {
      "epoch": 392.86,
      "learning_rate": 6.734523809523809e-05,
      "loss": 0.0668,
      "step": 2750
    },
    {
      "epoch": 400.0,
      "learning_rate": 6.675e-05,
      "loss": 0.0813,
      "step": 2800
    },
    {
      "epoch": 407.14,
      "learning_rate": 6.615476190476191e-05,
      "loss": 0.0764,
      "step": 2850
    },
    {
      "epoch": 414.29,
      "learning_rate": 6.555952380952382e-05,
      "loss": 0.0671,
      "step": 2900
    },
    {
      "epoch": 421.43,
      "learning_rate": 6.496428571428572e-05,
      "loss": 0.0523,
      "step": 2950
    },
    {
      "epoch": 428.57,
      "learning_rate": 6.436904761904761e-05,
      "loss": 0.051,
      "step": 3000
    },
    {
      "epoch": 435.71,
      "learning_rate": 6.377380952380952e-05,
      "loss": 0.0612,
      "step": 3050
    },
    {
      "epoch": 442.86,
      "learning_rate": 6.317857142857144e-05,
      "loss": 0.0767,
      "step": 3100
    },
    {
      "epoch": 450.0,
      "learning_rate": 6.258333333333334e-05,
      "loss": 0.062,
      "step": 3150
    },
    {
      "epoch": 457.14,
      "learning_rate": 6.198809523809525e-05,
      "loss": 0.0704,
      "step": 3200
    },
    {
      "epoch": 464.29,
      "learning_rate": 6.139285714285714e-05,
      "loss": 0.0691,
      "step": 3250
    },
    {
      "epoch": 471.43,
      "learning_rate": 6.079761904761905e-05,
      "loss": 0.0654,
      "step": 3300
    },
    {
      "epoch": 478.57,
      "learning_rate": 6.020238095238095e-05,
      "loss": 0.055,
      "step": 3350
    },
    {
      "epoch": 485.71,
      "learning_rate": 5.9607142857142864e-05,
      "loss": 0.0572,
      "step": 3400
    },
    {
      "epoch": 492.86,
      "learning_rate": 5.901190476190477e-05,
      "loss": 0.0514,
      "step": 3450
    },
    {
      "epoch": 500.0,
      "learning_rate": 5.8416666666666666e-05,
      "loss": 0.0588,
      "step": 3500
    },
    {
      "epoch": 507.14,
      "learning_rate": 5.782142857142857e-05,
      "loss": 0.0612,
      "step": 3550
    },
    {
      "epoch": 514.29,
      "learning_rate": 5.7226190476190475e-05,
      "loss": 0.0629,
      "step": 3600
    },
    {
      "epoch": 521.43,
      "learning_rate": 5.6630952380952386e-05,
      "loss": 0.0558,
      "step": 3650
    },
    {
      "epoch": 528.57,
      "learning_rate": 5.603571428571429e-05,
      "loss": 0.0515,
      "step": 3700
    },
    {
      "epoch": 535.71,
      "learning_rate": 5.5440476190476195e-05,
      "loss": 0.0532,
      "step": 3750
    },
    {
      "epoch": 542.86,
      "learning_rate": 5.484523809523809e-05,
      "loss": 0.051,
      "step": 3800
    },
    {
      "epoch": 550.0,
      "learning_rate": 5.4250000000000004e-05,
      "loss": 0.0525,
      "step": 3850
    },
    {
      "epoch": 557.14,
      "learning_rate": 5.365476190476191e-05,
      "loss": 0.0465,
      "step": 3900
    },
    {
      "epoch": 564.29,
      "learning_rate": 5.305952380952381e-05,
      "loss": 0.0554,
      "step": 3950
    },
    {
      "epoch": 571.43,
      "learning_rate": 5.2464285714285724e-05,
      "loss": 0.0546,
      "step": 4000
    },
    {
      "epoch": 578.57,
      "learning_rate": 5.1880952380952386e-05,
      "loss": 0.0527,
      "step": 4050
    },
    {
      "epoch": 585.71,
      "learning_rate": 5.128571428571429e-05,
      "loss": 0.0565,
      "step": 4100
    },
    {
      "epoch": 592.86,
      "learning_rate": 5.069047619047619e-05,
      "loss": 0.051,
      "step": 4150
    },
    {
      "epoch": 600.0,
      "learning_rate": 5.009523809523809e-05,
      "loss": 0.0517,
      "step": 4200
    },
    {
      "epoch": 607.14,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.0466,
      "step": 4250
    },
    {
      "epoch": 614.29,
      "learning_rate": 4.890476190476191e-05,
      "loss": 0.0558,
      "step": 4300
    },
    {
      "epoch": 621.43,
      "learning_rate": 4.830952380952381e-05,
      "loss": 0.055,
      "step": 4350
    },
    {
      "epoch": 628.57,
      "learning_rate": 4.771428571428572e-05,
      "loss": 0.0587,
      "step": 4400
    },
    {
      "epoch": 635.71,
      "learning_rate": 4.711904761904762e-05,
      "loss": 0.0545,
      "step": 4450
    },
    {
      "epoch": 642.86,
      "learning_rate": 4.6523809523809526e-05,
      "loss": 0.0454,
      "step": 4500
    },
    {
      "epoch": 650.0,
      "learning_rate": 4.592857142857143e-05,
      "loss": 0.0465,
      "step": 4550
    },
    {
      "epoch": 657.14,
      "learning_rate": 4.5333333333333335e-05,
      "loss": 0.0528,
      "step": 4600
    },
    {
      "epoch": 664.29,
      "learning_rate": 4.473809523809524e-05,
      "loss": 0.0584,
      "step": 4650
    },
    {
      "epoch": 671.43,
      "learning_rate": 4.4142857142857144e-05,
      "loss": 0.0466,
      "step": 4700
    },
    {
      "epoch": 678.57,
      "learning_rate": 4.354761904761905e-05,
      "loss": 0.0576,
      "step": 4750
    },
    {
      "epoch": 685.71,
      "learning_rate": 4.295238095238095e-05,
      "loss": 0.0496,
      "step": 4800
    },
    {
      "epoch": 692.86,
      "learning_rate": 4.2357142857142864e-05,
      "loss": 0.0401,
      "step": 4850
    },
    {
      "epoch": 700.0,
      "learning_rate": 4.176190476190476e-05,
      "loss": 0.0655,
      "step": 4900
    },
    {
      "epoch": 707.14,
      "learning_rate": 4.116666666666667e-05,
      "loss": 0.0533,
      "step": 4950
    },
    {
      "epoch": 714.29,
      "learning_rate": 4.057142857142857e-05,
      "loss": 0.0417,
      "step": 5000
    },
    {
      "epoch": 721.43,
      "learning_rate": 3.9976190476190476e-05,
      "loss": 0.0461,
      "step": 5050
    },
    {
      "epoch": 728.57,
      "learning_rate": 3.938095238095239e-05,
      "loss": 0.0434,
      "step": 5100
    },
    {
      "epoch": 735.71,
      "learning_rate": 3.8785714285714285e-05,
      "loss": 0.0409,
      "step": 5150
    },
    {
      "epoch": 742.86,
      "learning_rate": 3.8190476190476196e-05,
      "loss": 0.0404,
      "step": 5200
    },
    {
      "epoch": 750.0,
      "learning_rate": 3.7595238095238094e-05,
      "loss": 0.0494,
      "step": 5250
    },
    {
      "epoch": 757.14,
      "learning_rate": 3.7e-05,
      "loss": 0.0473,
      "step": 5300
    },
    {
      "epoch": 764.29,
      "learning_rate": 3.640476190476191e-05,
      "loss": 0.0449,
      "step": 5350
    },
    {
      "epoch": 771.43,
      "learning_rate": 3.580952380952381e-05,
      "loss": 0.044,
      "step": 5400
    },
    {
      "epoch": 778.57,
      "learning_rate": 3.521428571428572e-05,
      "loss": 0.0346,
      "step": 5450
    },
    {
      "epoch": 785.71,
      "learning_rate": 3.461904761904762e-05,
      "loss": 0.0436,
      "step": 5500
    },
    {
      "epoch": 792.86,
      "learning_rate": 3.402380952380953e-05,
      "loss": 0.0516,
      "step": 5550
    },
    {
      "epoch": 800.0,
      "learning_rate": 3.342857142857143e-05,
      "loss": 0.043,
      "step": 5600
    },
    {
      "epoch": 807.14,
      "learning_rate": 3.283333333333333e-05,
      "loss": 0.0494,
      "step": 5650
    },
    {
      "epoch": 814.29,
      "learning_rate": 3.223809523809524e-05,
      "loss": 0.0417,
      "step": 5700
    },
    {
      "epoch": 821.43,
      "learning_rate": 3.1642857142857145e-05,
      "loss": 0.0407,
      "step": 5750
    },
    {
      "epoch": 828.57,
      "learning_rate": 3.104761904761905e-05,
      "loss": 0.0387,
      "step": 5800
    },
    {
      "epoch": 835.71,
      "learning_rate": 3.0452380952380954e-05,
      "loss": 0.0418,
      "step": 5850
    },
    {
      "epoch": 842.86,
      "learning_rate": 2.9857142857142862e-05,
      "loss": 0.0407,
      "step": 5900
    },
    {
      "epoch": 850.0,
      "learning_rate": 2.9261904761904763e-05,
      "loss": 0.0348,
      "step": 5950
    },
    {
      "epoch": 857.14,
      "learning_rate": 2.8666666666666668e-05,
      "loss": 0.037,
      "step": 6000
    },
    {
      "epoch": 864.29,
      "learning_rate": 2.8083333333333333e-05,
      "loss": 0.041,
      "step": 6050
    },
    {
      "epoch": 871.43,
      "learning_rate": 2.748809523809524e-05,
      "loss": 0.0378,
      "step": 6100
    },
    {
      "epoch": 878.57,
      "learning_rate": 2.689285714285714e-05,
      "loss": 0.0364,
      "step": 6150
    },
    {
      "epoch": 885.71,
      "learning_rate": 2.629761904761905e-05,
      "loss": 0.0361,
      "step": 6200
    },
    {
      "epoch": 892.86,
      "learning_rate": 2.5702380952380954e-05,
      "loss": 0.0377,
      "step": 6250
    },
    {
      "epoch": 900.0,
      "learning_rate": 2.510714285714286e-05,
      "loss": 0.0353,
      "step": 6300
    },
    {
      "epoch": 907.14,
      "learning_rate": 2.4511904761904763e-05,
      "loss": 0.0323,
      "step": 6350
    },
    {
      "epoch": 914.29,
      "learning_rate": 2.3916666666666668e-05,
      "loss": 0.0413,
      "step": 6400
    },
    {
      "epoch": 921.43,
      "learning_rate": 2.3321428571428572e-05,
      "loss": 0.0358,
      "step": 6450
    },
    {
      "epoch": 928.57,
      "learning_rate": 2.2726190476190476e-05,
      "loss": 0.0367,
      "step": 6500
    },
    {
      "epoch": 935.71,
      "learning_rate": 2.2130952380952384e-05,
      "loss": 0.0401,
      "step": 6550
    },
    {
      "epoch": 942.86,
      "learning_rate": 2.1535714285714285e-05,
      "loss": 0.0344,
      "step": 6600
    },
    {
      "epoch": 950.0,
      "learning_rate": 2.094047619047619e-05,
      "loss": 0.0297,
      "step": 6650
    },
    {
      "epoch": 957.14,
      "learning_rate": 2.0345238095238094e-05,
      "loss": 0.0292,
      "step": 6700
    },
    {
      "epoch": 964.29,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 0.0334,
      "step": 6750
    },
    {
      "epoch": 971.43,
      "learning_rate": 1.9154761904761907e-05,
      "loss": 0.0378,
      "step": 6800
    },
    {
      "epoch": 978.57,
      "learning_rate": 1.855952380952381e-05,
      "loss": 0.0386,
      "step": 6850
    },
    {
      "epoch": 985.71,
      "learning_rate": 1.7964285714285712e-05,
      "loss": 0.0284,
      "step": 6900
    },
    {
      "epoch": 992.86,
      "learning_rate": 1.736904761904762e-05,
      "loss": 0.0349,
      "step": 6950
    },
    {
      "epoch": 1000.0,
      "learning_rate": 1.6773809523809525e-05,
      "loss": 0.0296,
      "step": 7000
    },
    {
      "epoch": 1007.14,
      "learning_rate": 1.617857142857143e-05,
      "loss": 0.0302,
      "step": 7050
    },
    {
      "epoch": 1014.29,
      "learning_rate": 1.5583333333333334e-05,
      "loss": 0.0265,
      "step": 7100
    },
    {
      "epoch": 1021.43,
      "learning_rate": 1.498809523809524e-05,
      "loss": 0.0288,
      "step": 7150
    },
    {
      "epoch": 1028.57,
      "learning_rate": 1.4392857142857144e-05,
      "loss": 0.0318,
      "step": 7200
    },
    {
      "epoch": 1035.71,
      "learning_rate": 1.3797619047619049e-05,
      "loss": 0.0286,
      "step": 7250
    },
    {
      "epoch": 1042.86,
      "learning_rate": 1.3202380952380952e-05,
      "loss": 0.0328,
      "step": 7300
    },
    {
      "epoch": 1050.0,
      "learning_rate": 1.260714285714286e-05,
      "loss": 0.027,
      "step": 7350
    },
    {
      "epoch": 1057.14,
      "learning_rate": 1.2011904761904762e-05,
      "loss": 0.0266,
      "step": 7400
    },
    {
      "epoch": 1064.29,
      "learning_rate": 1.1416666666666667e-05,
      "loss": 0.0271,
      "step": 7450
    },
    {
      "epoch": 1071.43,
      "learning_rate": 1.0821428571428573e-05,
      "loss": 0.0252,
      "step": 7500
    },
    {
      "epoch": 1078.57,
      "learning_rate": 1.0226190476190476e-05,
      "loss": 0.0264,
      "step": 7550
    },
    {
      "epoch": 1085.71,
      "learning_rate": 9.63095238095238e-06,
      "loss": 0.0238,
      "step": 7600
    },
    {
      "epoch": 1092.86,
      "learning_rate": 9.035714285714287e-06,
      "loss": 0.0276,
      "step": 7650
    },
    {
      "epoch": 1100.0,
      "learning_rate": 8.440476190476191e-06,
      "loss": 0.026,
      "step": 7700
    },
    {
      "epoch": 1107.14,
      "learning_rate": 7.845238095238096e-06,
      "loss": 0.0253,
      "step": 7750
    },
    {
      "epoch": 1114.29,
      "learning_rate": 7.25e-06,
      "loss": 0.022,
      "step": 7800
    },
    {
      "epoch": 1121.43,
      "learning_rate": 6.654761904761905e-06,
      "loss": 0.0219,
      "step": 7850
    },
    {
      "epoch": 1128.57,
      "learning_rate": 6.05952380952381e-06,
      "loss": 0.0212,
      "step": 7900
    },
    {
      "epoch": 1135.71,
      "learning_rate": 5.464285714285715e-06,
      "loss": 0.0249,
      "step": 7950
    },
    {
      "epoch": 1142.86,
      "learning_rate": 4.86904761904762e-06,
      "loss": 0.0244,
      "step": 8000
    },
    {
      "epoch": 1150.0,
      "learning_rate": 4.273809523809523e-06,
      "loss": 0.023,
      "step": 8050
    },
    {
      "epoch": 1157.14,
      "learning_rate": 3.690476190476191e-06,
      "loss": 0.0227,
      "step": 8100
    },
    {
      "epoch": 1164.29,
      "learning_rate": 3.0952380952380953e-06,
      "loss": 0.0244,
      "step": 8150
    },
    {
      "epoch": 1171.43,
      "learning_rate": 2.5e-06,
      "loss": 0.0205,
      "step": 8200
    }
  ],
  "logging_steps": 50,
  "max_steps": 8400,
  "num_train_epochs": 1200,
  "save_steps": 200,
  "total_flos": 6.852747656247149e+18,
  "trial_name": null,
  "trial_params": null
}