{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.977777777777778,
  "eval_steps": 500,
  "global_step": 560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 40.8252015336812,
      "learning_rate": 4.4642857142857147e-07,
      "loss": 2.7614,
      "step": 5
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 41.40766383255297,
      "learning_rate": 8.928571428571429e-07,
      "loss": 2.5209,
      "step": 10
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 19.36067123800515,
      "learning_rate": 1.3392857142857143e-06,
      "loss": 2.3634,
      "step": 15
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 13.847149981096214,
      "learning_rate": 1.7857142857142859e-06,
      "loss": 2.1595,
      "step": 20
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 9.243065607906033,
      "learning_rate": 2.2321428571428573e-06,
      "loss": 1.8998,
      "step": 25
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 9.15946624628624,
      "learning_rate": 2.6785714285714285e-06,
      "loss": 1.8978,
      "step": 30
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 7.049829158102999,
      "learning_rate": 3.125e-06,
      "loss": 1.8639,
      "step": 35
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 7.076253631073926,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 1.7955,
      "step": 40
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.057673864442862,
      "learning_rate": 4.017857142857143e-06,
      "loss": 1.7781,
      "step": 45
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 5.958183563465681,
      "learning_rate": 4.464285714285715e-06,
      "loss": 1.7338,
      "step": 50
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 5.866897144123113,
      "learning_rate": 4.910714285714286e-06,
      "loss": 1.67,
      "step": 55
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 5.869994369951594,
      "learning_rate": 5.357142857142857e-06,
      "loss": 1.6055,
      "step": 60
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 5.544599606757217,
      "learning_rate": 5.8035714285714295e-06,
      "loss": 1.621,
      "step": 65
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 5.922951541902071,
      "learning_rate": 6.25e-06,
      "loss": 1.5503,
      "step": 70
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 5.660948616961859,
      "learning_rate": 6.696428571428571e-06,
      "loss": 1.5419,
      "step": 75
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 5.515037072569432,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 1.6159,
      "step": 80
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 5.558602833573966,
      "learning_rate": 7.589285714285714e-06,
      "loss": 1.5742,
      "step": 85
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.305700239963537,
      "learning_rate": 8.035714285714286e-06,
      "loss": 1.5467,
      "step": 90
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 5.004136205117595,
      "learning_rate": 8.482142857142858e-06,
      "loss": 1.5673,
      "step": 95
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 5.1781341363875875,
      "learning_rate": 8.92857142857143e-06,
      "loss": 1.5419,
      "step": 100
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 5.2348081332809855,
      "learning_rate": 9.375000000000001e-06,
      "loss": 1.5291,
      "step": 105
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 5.599326305182384,
      "learning_rate": 9.821428571428573e-06,
      "loss": 1.4805,
      "step": 110
    },
    {
      "epoch": 1.0222222222222221,
      "grad_norm": 4.95489443166838,
      "learning_rate": 9.933035714285715e-06,
      "loss": 1.4025,
      "step": 115
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 5.441338497159552,
      "learning_rate": 9.821428571428573e-06,
      "loss": 1.2921,
      "step": 120
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 4.920817976988318,
      "learning_rate": 9.70982142857143e-06,
      "loss": 1.3219,
      "step": 125
    },
    {
      "epoch": 1.1555555555555554,
      "grad_norm": 5.939277340856136,
      "learning_rate": 9.598214285714287e-06,
      "loss": 1.242,
      "step": 130
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.335319671954436,
      "learning_rate": 9.486607142857144e-06,
      "loss": 1.2232,
      "step": 135
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 6.183455759561233,
      "learning_rate": 9.375000000000001e-06,
      "loss": 1.2734,
      "step": 140
    },
    {
      "epoch": 1.2888888888888888,
      "grad_norm": 6.122390387016121,
      "learning_rate": 9.263392857142858e-06,
      "loss": 1.2504,
      "step": 145
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 5.649941977788642,
      "learning_rate": 9.151785714285715e-06,
      "loss": 1.2913,
      "step": 150
    },
    {
      "epoch": 1.3777777777777778,
      "grad_norm": 6.2118082348430335,
      "learning_rate": 9.040178571428572e-06,
      "loss": 1.1898,
      "step": 155
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 6.125925282923714,
      "learning_rate": 8.92857142857143e-06,
      "loss": 1.2446,
      "step": 160
    },
    {
      "epoch": 1.4666666666666668,
      "grad_norm": 5.882840215119527,
      "learning_rate": 8.816964285714286e-06,
      "loss": 1.2967,
      "step": 165
    },
    {
      "epoch": 1.511111111111111,
      "grad_norm": 6.244265409736289,
      "learning_rate": 8.705357142857143e-06,
      "loss": 1.2507,
      "step": 170
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 6.021345396953319,
      "learning_rate": 8.59375e-06,
      "loss": 1.2116,
      "step": 175
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.289159755622968,
      "learning_rate": 8.482142857142858e-06,
      "loss": 1.2572,
      "step": 180
    },
    {
      "epoch": 1.6444444444444444,
      "grad_norm": 5.60112064803209,
      "learning_rate": 8.370535714285715e-06,
      "loss": 1.2467,
      "step": 185
    },
    {
      "epoch": 1.6888888888888889,
      "grad_norm": 5.766305702978478,
      "learning_rate": 8.258928571428572e-06,
      "loss": 1.2423,
      "step": 190
    },
    {
      "epoch": 1.7333333333333334,
      "grad_norm": 6.535746848002774,
      "learning_rate": 8.147321428571429e-06,
      "loss": 1.1516,
      "step": 195
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 5.8246000440527,
      "learning_rate": 8.035714285714286e-06,
      "loss": 1.2577,
      "step": 200
    },
    {
      "epoch": 1.8222222222222222,
      "grad_norm": 5.708942474707697,
      "learning_rate": 7.924107142857143e-06,
      "loss": 1.2754,
      "step": 205
    },
    {
      "epoch": 1.8666666666666667,
      "grad_norm": 5.561728988256993,
      "learning_rate": 7.8125e-06,
      "loss": 1.2149,
      "step": 210
    },
    {
      "epoch": 1.911111111111111,
      "grad_norm": 5.5782199002239565,
      "learning_rate": 7.700892857142857e-06,
      "loss": 1.2308,
      "step": 215
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 5.39044075769101,
      "learning_rate": 7.589285714285714e-06,
      "loss": 1.2042,
      "step": 220
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.532219423056963,
      "learning_rate": 7.4776785714285714e-06,
      "loss": 1.2174,
      "step": 225
    },
    {
      "epoch": 2.0444444444444443,
      "grad_norm": 4.883317970060585,
      "learning_rate": 7.366071428571429e-06,
      "loss": 0.8016,
      "step": 230
    },
    {
      "epoch": 2.088888888888889,
      "grad_norm": 6.16427200874355,
      "learning_rate": 7.2544642857142865e-06,
      "loss": 0.7494,
      "step": 235
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 6.874292360395087,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.7405,
      "step": 240
    },
    {
      "epoch": 2.1777777777777776,
      "grad_norm": 7.121149956084331,
      "learning_rate": 7.031250000000001e-06,
      "loss": 0.7379,
      "step": 245
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.81637228279681,
      "learning_rate": 6.919642857142858e-06,
      "loss": 0.7584,
      "step": 250
    },
    {
      "epoch": 2.2666666666666666,
      "grad_norm": 5.754815424689757,
      "learning_rate": 6.808035714285715e-06,
      "loss": 0.7425,
      "step": 255
    },
    {
      "epoch": 2.311111111111111,
      "grad_norm": 7.091048722340656,
      "learning_rate": 6.696428571428571e-06,
      "loss": 0.7495,
      "step": 260
    },
    {
      "epoch": 2.3555555555555556,
      "grad_norm": 6.094106982847032,
      "learning_rate": 6.58482142857143e-06,
      "loss": 0.7427,
      "step": 265
    },
    {
      "epoch": 2.4,
      "grad_norm": 6.414692728098821,
      "learning_rate": 6.473214285714287e-06,
      "loss": 0.7319,
      "step": 270
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 6.708504578115011,
      "learning_rate": 6.361607142857143e-06,
      "loss": 0.7342,
      "step": 275
    },
    {
      "epoch": 2.488888888888889,
      "grad_norm": 7.165696647143685,
      "learning_rate": 6.25e-06,
      "loss": 0.787,
      "step": 280
    },
    {
      "epoch": 2.533333333333333,
      "grad_norm": 5.95046471694673,
      "learning_rate": 6.138392857142857e-06,
      "loss": 0.685,
      "step": 285
    },
    {
      "epoch": 2.5777777777777775,
      "grad_norm": 6.168087419964014,
      "learning_rate": 6.0267857142857145e-06,
      "loss": 0.7947,
      "step": 290
    },
    {
      "epoch": 2.6222222222222222,
      "grad_norm": 8.12776947225484,
      "learning_rate": 5.9151785714285716e-06,
      "loss": 0.7513,
      "step": 295
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 6.433781125784366,
      "learning_rate": 5.8035714285714295e-06,
      "loss": 0.7135,
      "step": 300
    },
    {
      "epoch": 2.7111111111111112,
      "grad_norm": 6.384935891420059,
      "learning_rate": 5.691964285714287e-06,
      "loss": 0.7493,
      "step": 305
    },
    {
      "epoch": 2.7555555555555555,
      "grad_norm": 6.433040870475973,
      "learning_rate": 5.580357142857144e-06,
      "loss": 0.727,
      "step": 310
    },
    {
      "epoch": 2.8,
      "grad_norm": 6.571634780231645,
      "learning_rate": 5.468750000000001e-06,
      "loss": 0.7117,
      "step": 315
    },
    {
      "epoch": 2.8444444444444446,
      "grad_norm": 6.181645383932599,
      "learning_rate": 5.357142857142857e-06,
      "loss": 0.7814,
      "step": 320
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 5.993415912788299,
      "learning_rate": 5.245535714285714e-06,
      "loss": 0.7382,
      "step": 325
    },
    {
      "epoch": 2.9333333333333336,
      "grad_norm": 7.296220837415293,
      "learning_rate": 5.133928571428571e-06,
      "loss": 0.7704,
      "step": 330
    },
    {
      "epoch": 2.977777777777778,
      "grad_norm": 6.26275860113005,
      "learning_rate": 5.022321428571429e-06,
      "loss": 0.7475,
      "step": 335
    },
    {
      "epoch": 3.022222222222222,
      "grad_norm": 5.547644012720461,
      "learning_rate": 4.910714285714286e-06,
      "loss": 0.5714,
      "step": 340
    },
    {
      "epoch": 3.066666666666667,
      "grad_norm": 7.45609250503262,
      "learning_rate": 4.799107142857143e-06,
      "loss": 0.3736,
      "step": 345
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 9.220779587965055,
      "learning_rate": 4.6875000000000004e-06,
      "loss": 0.379,
      "step": 350
    },
    {
      "epoch": 3.1555555555555554,
      "grad_norm": 5.903988909331705,
      "learning_rate": 4.5758928571428575e-06,
      "loss": 0.3091,
      "step": 355
    },
    {
      "epoch": 3.2,
      "grad_norm": 6.770193918782061,
      "learning_rate": 4.464285714285715e-06,
      "loss": 0.333,
      "step": 360
    },
    {
      "epoch": 3.2444444444444445,
      "grad_norm": 6.276791425358006,
      "learning_rate": 4.352678571428572e-06,
      "loss": 0.3177,
      "step": 365
    },
    {
      "epoch": 3.2888888888888888,
      "grad_norm": 7.193401410169643,
      "learning_rate": 4.241071428571429e-06,
      "loss": 0.3509,
      "step": 370
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 7.049225730521727,
      "learning_rate": 4.129464285714286e-06,
      "loss": 0.3327,
      "step": 375
    },
    {
      "epoch": 3.3777777777777778,
      "grad_norm": 6.890952439455029,
      "learning_rate": 4.017857142857143e-06,
      "loss": 0.3014,
      "step": 380
    },
    {
      "epoch": 3.422222222222222,
      "grad_norm": 6.450392876649866,
      "learning_rate": 3.90625e-06,
      "loss": 0.3134,
      "step": 385
    },
    {
      "epoch": 3.466666666666667,
      "grad_norm": 6.50857632933258,
      "learning_rate": 3.794642857142857e-06,
      "loss": 0.3281,
      "step": 390
    },
    {
      "epoch": 3.511111111111111,
      "grad_norm": 7.00409851180514,
      "learning_rate": 3.6830357142857147e-06,
      "loss": 0.3032,
      "step": 395
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 7.985948331505739,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 0.3148,
      "step": 400
    },
    {
      "epoch": 3.6,
      "grad_norm": 7.300806552749991,
      "learning_rate": 3.459821428571429e-06,
      "loss": 0.3393,
      "step": 405
    },
    {
      "epoch": 3.6444444444444444,
      "grad_norm": 6.768267789836914,
      "learning_rate": 3.3482142857142855e-06,
      "loss": 0.3094,
      "step": 410
    },
    {
      "epoch": 3.688888888888889,
      "grad_norm": 6.849516994792115,
      "learning_rate": 3.2366071428571435e-06,
      "loss": 0.308,
      "step": 415
    },
    {
      "epoch": 3.7333333333333334,
      "grad_norm": 6.895514738982963,
      "learning_rate": 3.125e-06,
      "loss": 0.3151,
      "step": 420
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 7.278283240180865,
      "learning_rate": 3.0133928571428572e-06,
      "loss": 0.311,
      "step": 425
    },
    {
      "epoch": 3.822222222222222,
      "grad_norm": 7.770700236659141,
      "learning_rate": 2.9017857142857148e-06,
      "loss": 0.3261,
      "step": 430
    },
    {
      "epoch": 3.8666666666666667,
      "grad_norm": 6.82699249519738,
      "learning_rate": 2.790178571428572e-06,
      "loss": 0.3286,
      "step": 435
    },
    {
      "epoch": 3.911111111111111,
      "grad_norm": 6.058725073774486,
      "learning_rate": 2.6785714285714285e-06,
      "loss": 0.3055,
      "step": 440
    },
    {
      "epoch": 3.9555555555555557,
      "grad_norm": 7.923771994803603,
      "learning_rate": 2.5669642857142856e-06,
      "loss": 0.3475,
      "step": 445
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.583308434767343,
      "learning_rate": 2.455357142857143e-06,
      "loss": 0.3072,
      "step": 450
    },
    {
      "epoch": 4.044444444444444,
      "grad_norm": 4.888668375327939,
      "learning_rate": 2.3437500000000002e-06,
      "loss": 0.123,
      "step": 455
    },
    {
      "epoch": 4.088888888888889,
      "grad_norm": 5.30026249113343,
      "learning_rate": 2.2321428571428573e-06,
      "loss": 0.1172,
      "step": 460
    },
    {
      "epoch": 4.133333333333334,
      "grad_norm": 5.7543964521851185,
      "learning_rate": 2.1205357142857144e-06,
      "loss": 0.1063,
      "step": 465
    },
    {
      "epoch": 4.177777777777778,
      "grad_norm": 6.961825045250998,
      "learning_rate": 2.0089285714285715e-06,
      "loss": 0.0987,
      "step": 470
    },
    {
      "epoch": 4.222222222222222,
      "grad_norm": 6.441588539096942,
      "learning_rate": 1.8973214285714286e-06,
      "loss": 0.0938,
      "step": 475
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 4.658464773050257,
      "learning_rate": 1.7857142857142859e-06,
      "loss": 0.0875,
      "step": 480
    },
    {
      "epoch": 4.311111111111111,
      "grad_norm": 6.424114535270574,
      "learning_rate": 1.6741071428571428e-06,
      "loss": 0.1084,
      "step": 485
    },
    {
      "epoch": 4.355555555555555,
      "grad_norm": 4.590840131788235,
      "learning_rate": 1.5625e-06,
      "loss": 0.0878,
      "step": 490
    },
    {
      "epoch": 4.4,
      "grad_norm": 4.87276433451291,
      "learning_rate": 1.4508928571428574e-06,
      "loss": 0.0934,
      "step": 495
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 5.392164301915455,
      "learning_rate": 1.3392857142857143e-06,
      "loss": 0.1087,
      "step": 500
    },
    {
      "epoch": 4.488888888888889,
      "grad_norm": 4.187721869186243,
      "learning_rate": 1.2276785714285716e-06,
      "loss": 0.088,
      "step": 505
    },
    {
      "epoch": 4.533333333333333,
      "grad_norm": 5.2715420146235665,
      "learning_rate": 1.1160714285714287e-06,
      "loss": 0.1148,
      "step": 510
    },
    {
      "epoch": 4.5777777777777775,
      "grad_norm": 5.618347246851232,
      "learning_rate": 1.0044642857142857e-06,
      "loss": 0.0956,
      "step": 515
    },
    {
      "epoch": 4.622222222222222,
      "grad_norm": 6.752957019393521,
      "learning_rate": 8.928571428571429e-07,
      "loss": 0.1017,
      "step": 520
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 4.927856784257147,
      "learning_rate": 7.8125e-07,
      "loss": 0.1014,
      "step": 525
    },
    {
      "epoch": 4.711111111111111,
      "grad_norm": 5.446418154433167,
      "learning_rate": 6.696428571428571e-07,
      "loss": 0.101,
      "step": 530
    },
    {
      "epoch": 4.7555555555555555,
      "grad_norm": 5.965809303232395,
      "learning_rate": 5.580357142857143e-07,
      "loss": 0.0991,
      "step": 535
    },
    {
      "epoch": 4.8,
      "grad_norm": 6.186953872684265,
      "learning_rate": 4.4642857142857147e-07,
      "loss": 0.0835,
      "step": 540
    },
    {
      "epoch": 4.844444444444444,
      "grad_norm": 5.017392231671649,
      "learning_rate": 3.3482142857142856e-07,
      "loss": 0.0863,
      "step": 545
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 5.0701043320958625,
      "learning_rate": 2.2321428571428574e-07,
      "loss": 0.1016,
      "step": 550
    },
    {
      "epoch": 4.933333333333334,
      "grad_norm": 4.872193534628404,
      "learning_rate": 1.1160714285714287e-07,
      "loss": 0.0958,
      "step": 555
    },
    {
      "epoch": 4.977777777777778,
      "grad_norm": 6.017534872316958,
      "learning_rate": 0.0,
      "loss": 0.0917,
      "step": 560
    },
    {
      "epoch": 4.977777777777778,
      "step": 560,
      "total_flos": 1995521605632.0,
      "train_loss": 0.8461004374814886,
      "train_runtime": 1804.2136,
      "train_samples_per_second": 9.977,
      "train_steps_per_second": 0.31
    }
  ],
  "logging_steps": 5,
  "max_steps": 560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1995521605632.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}