{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2786873824287605,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0027868738242876052,
      "grad_norm": 3.5925259590148926,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 2.5484,
      "step": 5
    },
    {
      "epoch": 0.0055737476485752105,
      "grad_norm": 3.7384228706359863,
      "learning_rate": 5e-06,
      "loss": 2.5036,
      "step": 10
    },
    {
      "epoch": 0.008360621472862817,
      "grad_norm": 3.7028982639312744,
      "learning_rate": 7.777777777777777e-06,
      "loss": 2.5177,
      "step": 15
    },
    {
      "epoch": 0.011147495297150421,
      "grad_norm": 3.1780974864959717,
      "learning_rate": 1.0555555555555555e-05,
      "loss": 2.4746,
      "step": 20
    },
    {
      "epoch": 0.013934369121438027,
      "grad_norm": 2.8278558254241943,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.5083,
      "step": 25
    },
    {
      "epoch": 0.016721242945725634,
      "grad_norm": 2.5323452949523926,
      "learning_rate": 1.6111111111111115e-05,
      "loss": 2.4383,
      "step": 30
    },
    {
      "epoch": 0.019508116770013236,
      "grad_norm": 1.2514830827713013,
      "learning_rate": 1.888888888888889e-05,
      "loss": 2.1978,
      "step": 35
    },
    {
      "epoch": 0.022294990594300842,
      "grad_norm": 1.1612293720245361,
      "learning_rate": 2.1666666666666667e-05,
      "loss": 2.0549,
      "step": 40
    },
    {
      "epoch": 0.025081864418588447,
      "grad_norm": 0.9937079548835754,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 2.1649,
      "step": 45
    },
    {
      "epoch": 0.027868738242876053,
      "grad_norm": 1.056492567062378,
      "learning_rate": 2.7222222222222223e-05,
      "loss": 2.0852,
      "step": 50
    },
    {
      "epoch": 0.03065561206716366,
      "grad_norm": 0.9864413142204285,
      "learning_rate": 3e-05,
      "loss": 1.934,
      "step": 55
    },
    {
      "epoch": 0.03344248589145127,
      "grad_norm": 0.9184162616729736,
      "learning_rate": 3.277777777777778e-05,
      "loss": 2.0746,
      "step": 60
    },
    {
      "epoch": 0.03622935971573887,
      "grad_norm": 1.044710397720337,
      "learning_rate": 3.555555555555556e-05,
      "loss": 1.9812,
      "step": 65
    },
    {
      "epoch": 0.03901623354002647,
      "grad_norm": 0.9160545468330383,
      "learning_rate": 3.8333333333333334e-05,
      "loss": 1.8794,
      "step": 70
    },
    {
      "epoch": 0.04180310736431408,
      "grad_norm": 0.9865038990974426,
      "learning_rate": 4.111111111111111e-05,
      "loss": 1.9266,
      "step": 75
    },
    {
      "epoch": 0.044589981188601684,
      "grad_norm": 0.9860550165176392,
      "learning_rate": 4.388888888888889e-05,
      "loss": 1.8417,
      "step": 80
    },
    {
      "epoch": 0.04737685501288929,
      "grad_norm": 0.9826656579971313,
      "learning_rate": 4.666666666666667e-05,
      "loss": 1.9756,
      "step": 85
    },
    {
      "epoch": 0.050163728837176895,
      "grad_norm": 0.9479106664657593,
      "learning_rate": 4.9444444444444446e-05,
      "loss": 1.8803,
      "step": 90
    },
    {
      "epoch": 0.0529506026614645,
      "grad_norm": 1.005937933921814,
      "learning_rate": 5.222222222222223e-05,
      "loss": 1.8677,
      "step": 95
    },
    {
      "epoch": 0.055737476485752106,
      "grad_norm": 1.1173028945922852,
      "learning_rate": 5.500000000000001e-05,
      "loss": 1.8416,
      "step": 100
    },
    {
      "epoch": 0.05852435031003971,
      "grad_norm": 1.0000165700912476,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 1.8607,
      "step": 105
    },
    {
      "epoch": 0.06131122413432732,
      "grad_norm": 1.3015921115875244,
      "learning_rate": 6.055555555555555e-05,
      "loss": 1.9752,
      "step": 110
    },
    {
      "epoch": 0.06409809795861493,
      "grad_norm": 1.0431042909622192,
      "learning_rate": 6.333333333333333e-05,
      "loss": 1.8514,
      "step": 115
    },
    {
      "epoch": 0.06688497178290254,
      "grad_norm": 1.0463422536849976,
      "learning_rate": 6.611111111111111e-05,
      "loss": 1.8656,
      "step": 120
    },
    {
      "epoch": 0.06967184560719013,
      "grad_norm": 1.0947015285491943,
      "learning_rate": 6.88888888888889e-05,
      "loss": 1.9479,
      "step": 125
    },
    {
      "epoch": 0.07245871943147773,
      "grad_norm": 0.9953688979148865,
      "learning_rate": 7.166666666666667e-05,
      "loss": 1.8627,
      "step": 130
    },
    {
      "epoch": 0.07524559325576534,
      "grad_norm": 1.0779668092727661,
      "learning_rate": 7.444444444444444e-05,
      "loss": 1.8262,
      "step": 135
    },
    {
      "epoch": 0.07803246708005294,
      "grad_norm": 1.2166972160339355,
      "learning_rate": 7.722222222222223e-05,
      "loss": 1.8461,
      "step": 140
    },
    {
      "epoch": 0.08081934090434055,
      "grad_norm": 1.055553674697876,
      "learning_rate": 8e-05,
      "loss": 1.9684,
      "step": 145
    },
    {
      "epoch": 0.08360621472862816,
      "grad_norm": 1.0838943719863892,
      "learning_rate": 8.277777777777778e-05,
      "loss": 1.8891,
      "step": 150
    },
    {
      "epoch": 0.08639308855291576,
      "grad_norm": 0.9238198399543762,
      "learning_rate": 8.555555555555556e-05,
      "loss": 1.8507,
      "step": 155
    },
    {
      "epoch": 0.08917996237720337,
      "grad_norm": 1.186009168624878,
      "learning_rate": 8.833333333333333e-05,
      "loss": 1.8269,
      "step": 160
    },
    {
      "epoch": 0.09196683620149097,
      "grad_norm": 1.122663140296936,
      "learning_rate": 9.111111111111112e-05,
      "loss": 1.7877,
      "step": 165
    },
    {
      "epoch": 0.09475371002577858,
      "grad_norm": 1.176620364189148,
      "learning_rate": 9.388888888888889e-05,
      "loss": 1.8373,
      "step": 170
    },
    {
      "epoch": 0.09754058385006618,
      "grad_norm": 1.0495474338531494,
      "learning_rate": 9.666666666666667e-05,
      "loss": 1.838,
      "step": 175
    },
    {
      "epoch": 0.10032745767435379,
      "grad_norm": 1.0859992504119873,
      "learning_rate": 9.944444444444446e-05,
      "loss": 1.845,
      "step": 180
    },
    {
      "epoch": 0.1031143314986414,
      "grad_norm": 1.1708000898361206,
      "learning_rate": 9.999848639521432e-05,
      "loss": 1.8148,
      "step": 185
    },
    {
      "epoch": 0.105901205322929,
      "grad_norm": 1.0338629484176636,
      "learning_rate": 9.999233753283091e-05,
      "loss": 1.7982,
      "step": 190
    },
    {
      "epoch": 0.1086880791472166,
      "grad_norm": 1.2173702716827393,
      "learning_rate": 9.998145939378577e-05,
      "loss": 1.792,
      "step": 195
    },
    {
      "epoch": 0.11147495297150421,
      "grad_norm": 1.1683762073516846,
      "learning_rate": 9.996585300715116e-05,
      "loss": 1.8381,
      "step": 200
    },
    {
      "epoch": 0.11426182679579182,
      "grad_norm": 1.0897382497787476,
      "learning_rate": 9.994551984929175e-05,
      "loss": 1.8142,
      "step": 205
    },
    {
      "epoch": 0.11704870062007942,
      "grad_norm": 0.9195744395256042,
      "learning_rate": 9.992046184372492e-05,
      "loss": 1.8362,
      "step": 210
    },
    {
      "epoch": 0.11983557444436703,
      "grad_norm": 1.039374589920044,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.7565,
      "step": 215
    },
    {
      "epoch": 0.12262244826865464,
      "grad_norm": 1.0218507051467896,
      "learning_rate": 9.985618121816779e-05,
      "loss": 1.8728,
      "step": 220
    },
    {
      "epoch": 0.12540932209294225,
      "grad_norm": 1.0597004890441895,
      "learning_rate": 9.981696467912664e-05,
      "loss": 1.7938,
      "step": 225
    },
    {
      "epoch": 0.12819619591722986,
      "grad_norm": 1.1160184144973755,
      "learning_rate": 9.97730354537011e-05,
      "loss": 1.8156,
      "step": 230
    },
    {
      "epoch": 0.13098306974151747,
      "grad_norm": 0.99184250831604,
      "learning_rate": 9.972439769759722e-05,
      "loss": 1.7255,
      "step": 235
    },
    {
      "epoch": 0.13376994356580507,
      "grad_norm": 1.0527223348617554,
      "learning_rate": 9.967105601194823e-05,
      "loss": 1.8942,
      "step": 240
    },
    {
      "epoch": 0.13655681739009268,
      "grad_norm": 1.0618565082550049,
      "learning_rate": 9.961301544287922e-05,
      "loss": 1.7666,
      "step": 245
    },
    {
      "epoch": 0.13934369121438026,
      "grad_norm": 1.1174343824386597,
      "learning_rate": 9.955028148102979e-05,
      "loss": 1.7274,
      "step": 250
    },
    {
      "epoch": 0.14213056503866786,
      "grad_norm": 1.045905351638794,
      "learning_rate": 9.948286006103466e-05,
      "loss": 1.9156,
      "step": 255
    },
    {
      "epoch": 0.14491743886295547,
      "grad_norm": 1.1782068014144897,
      "learning_rate": 9.941075756096226e-05,
      "loss": 1.8142,
      "step": 260
    },
    {
      "epoch": 0.14770431268724307,
      "grad_norm": 1.0786527395248413,
      "learning_rate": 9.933398080171123e-05,
      "loss": 1.8206,
      "step": 265
    },
    {
      "epoch": 0.15049118651153068,
      "grad_norm": 1.0150487422943115,
      "learning_rate": 9.925253704636543e-05,
      "loss": 1.7779,
      "step": 270
    },
    {
      "epoch": 0.15327806033581828,
      "grad_norm": 1.0553853511810303,
      "learning_rate": 9.916643399950656e-05,
      "loss": 1.8223,
      "step": 275
    },
    {
      "epoch": 0.1560649341601059,
      "grad_norm": 1.0968186855316162,
      "learning_rate": 9.907567980648549e-05,
      "loss": 1.7834,
      "step": 280
    },
    {
      "epoch": 0.1588518079843935,
      "grad_norm": 0.9188632369041443,
      "learning_rate": 9.898028305265169e-05,
      "loss": 1.8808,
      "step": 285
    },
    {
      "epoch": 0.1616386818086811,
      "grad_norm": 0.9829431772232056,
      "learning_rate": 9.888025276254096e-05,
      "loss": 1.8455,
      "step": 290
    },
    {
      "epoch": 0.1644255556329687,
      "grad_norm": 1.034116268157959,
      "learning_rate": 9.877559839902184e-05,
      "loss": 1.8042,
      "step": 295
    },
    {
      "epoch": 0.1672124294572563,
      "grad_norm": 1.0451691150665283,
      "learning_rate": 9.86663298624003e-05,
      "loss": 1.7696,
      "step": 300
    },
    {
      "epoch": 0.16999930328154392,
      "grad_norm": 1.064620852470398,
      "learning_rate": 9.855245748948326e-05,
      "loss": 1.8668,
      "step": 305
    },
    {
      "epoch": 0.17278617710583152,
      "grad_norm": 1.0409430265426636,
      "learning_rate": 9.843399205260068e-05,
      "loss": 1.8274,
      "step": 310
    },
    {
      "epoch": 0.17557305093011913,
      "grad_norm": 1.0153954029083252,
      "learning_rate": 9.831094475858652e-05,
      "loss": 1.8084,
      "step": 315
    },
    {
      "epoch": 0.17835992475440673,
      "grad_norm": 0.9218955039978027,
      "learning_rate": 9.818332724771857e-05,
      "loss": 1.8457,
      "step": 320
    },
    {
      "epoch": 0.18114679857869434,
      "grad_norm": 0.9664631485939026,
      "learning_rate": 9.805115159261726e-05,
      "loss": 1.772,
      "step": 325
    },
    {
      "epoch": 0.18393367240298195,
      "grad_norm": 1.0060662031173706,
      "learning_rate": 9.791443029710361e-05,
      "loss": 1.727,
      "step": 330
    },
    {
      "epoch": 0.18672054622726955,
      "grad_norm": 1.0184046030044556,
      "learning_rate": 9.777317629501636e-05,
      "loss": 1.7638,
      "step": 335
    },
    {
      "epoch": 0.18950742005155716,
      "grad_norm": 0.9374409317970276,
      "learning_rate": 9.762740294898846e-05,
      "loss": 1.7323,
      "step": 340
    },
    {
      "epoch": 0.19229429387584476,
      "grad_norm": 0.9715288877487183,
      "learning_rate": 9.747712404918286e-05,
      "loss": 1.7169,
      "step": 345
    },
    {
      "epoch": 0.19508116770013237,
      "grad_norm": 1.088680624961853,
      "learning_rate": 9.732235381198813e-05,
      "loss": 1.8065,
      "step": 350
    },
    {
      "epoch": 0.19786804152441997,
      "grad_norm": 0.9513164758682251,
      "learning_rate": 9.716310687867342e-05,
      "loss": 1.6545,
      "step": 355
    },
    {
      "epoch": 0.20065491534870758,
      "grad_norm": 0.9501610398292542,
      "learning_rate": 9.699939831400351e-05,
      "loss": 1.8605,
      "step": 360
    },
    {
      "epoch": 0.20344178917299519,
      "grad_norm": 0.9788233041763306,
      "learning_rate": 9.683124360481364e-05,
      "loss": 1.7838,
      "step": 365
    },
    {
      "epoch": 0.2062286629972828,
      "grad_norm": 1.0686471462249756,
      "learning_rate": 9.665865865854445e-05,
      "loss": 1.7832,
      "step": 370
    },
    {
      "epoch": 0.2090155368215704,
      "grad_norm": 0.9658113121986389,
      "learning_rate": 9.648165980173712e-05,
      "loss": 1.7721,
      "step": 375
    },
    {
      "epoch": 0.211802410645858,
      "grad_norm": 1.0429141521453857,
      "learning_rate": 9.630026377848892e-05,
      "loss": 1.7741,
      "step": 380
    },
    {
      "epoch": 0.2145892844701456,
      "grad_norm": 0.9593327641487122,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.7534,
      "step": 385
    },
    {
      "epoch": 0.2173761582944332,
      "grad_norm": 0.9077600240707397,
      "learning_rate": 9.592434928729616e-05,
      "loss": 1.8398,
      "step": 390
    },
    {
      "epoch": 0.22016303211872082,
      "grad_norm": 1.048581600189209,
      "learning_rate": 9.572986638087396e-05,
      "loss": 1.8134,
      "step": 395
    },
    {
      "epoch": 0.22294990594300843,
      "grad_norm": 0.9498254060745239,
      "learning_rate": 9.553105742769154e-05,
      "loss": 1.7497,
      "step": 400
    },
    {
      "epoch": 0.22573677976729603,
      "grad_norm": 1.0249079465866089,
      "learning_rate": 9.532794123508197e-05,
      "loss": 1.6412,
      "step": 405
    },
    {
      "epoch": 0.22852365359158364,
      "grad_norm": 0.9723719358444214,
      "learning_rate": 9.512053701784329e-05,
      "loss": 1.7741,
      "step": 410
    },
    {
      "epoch": 0.23131052741587124,
      "grad_norm": 0.954607367515564,
      "learning_rate": 9.490886439642081e-05,
      "loss": 1.7865,
      "step": 415
    },
    {
      "epoch": 0.23409740124015885,
      "grad_norm": 0.9485211372375488,
      "learning_rate": 9.469294339505098e-05,
      "loss": 1.813,
      "step": 420
    },
    {
      "epoch": 0.23688427506444645,
      "grad_norm": 0.9606624841690063,
      "learning_rate": 9.447279443986716e-05,
      "loss": 1.7426,
      "step": 425
    },
    {
      "epoch": 0.23967114888873406,
      "grad_norm": 0.9899424910545349,
      "learning_rate": 9.424843835696724e-05,
      "loss": 1.7802,
      "step": 430
    },
    {
      "epoch": 0.24245802271302166,
      "grad_norm": 1.0416887998580933,
      "learning_rate": 9.401989637044355e-05,
      "loss": 1.7418,
      "step": 435
    },
    {
      "epoch": 0.24524489653730927,
      "grad_norm": 0.958951473236084,
      "learning_rate": 9.3787190100375e-05,
      "loss": 1.7791,
      "step": 440
    },
    {
      "epoch": 0.24803177036159688,
      "grad_norm": 0.993934154510498,
      "learning_rate": 9.355034156078188e-05,
      "loss": 1.739,
      "step": 445
    },
    {
      "epoch": 0.2508186441858845,
      "grad_norm": 0.9372468590736389,
      "learning_rate": 9.330937315754329e-05,
      "loss": 1.7406,
      "step": 450
    },
    {
      "epoch": 0.2536055180101721,
      "grad_norm": 0.9601876735687256,
      "learning_rate": 9.306430768627753e-05,
      "loss": 1.7357,
      "step": 455
    },
    {
      "epoch": 0.2563923918344597,
      "grad_norm": 0.9922548532485962,
      "learning_rate": 9.281516833018571e-05,
      "loss": 1.7677,
      "step": 460
    },
    {
      "epoch": 0.2591792656587473,
      "grad_norm": 0.9353616833686829,
      "learning_rate": 9.256197865785854e-05,
      "loss": 1.8202,
      "step": 465
    },
    {
      "epoch": 0.26196613948303493,
      "grad_norm": 0.8890717625617981,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.7473,
      "step": 470
    },
    {
      "epoch": 0.26475301330732254,
      "grad_norm": 1.0059962272644043,
      "learning_rate": 9.204354455239539e-05,
      "loss": 1.7978,
      "step": 475
    },
    {
      "epoch": 0.26753988713161014,
      "grad_norm": 0.9888349771499634,
      "learning_rate": 9.177834916314165e-05,
      "loss": 1.7481,
      "step": 480
    },
    {
      "epoch": 0.27032676095589775,
      "grad_norm": 0.9407850503921509,
      "learning_rate": 9.150920154077754e-05,
      "loss": 1.7667,
      "step": 485
    },
    {
      "epoch": 0.27311363478018535,
      "grad_norm": 1.0001949071884155,
      "learning_rate": 9.123612714667634e-05,
      "loss": 1.7742,
      "step": 490
    },
    {
      "epoch": 0.2759005086044729,
      "grad_norm": 0.9057056307792664,
      "learning_rate": 9.095915181368412e-05,
      "loss": 1.816,
      "step": 495
    },
    {
      "epoch": 0.2786873824287605,
      "grad_norm": 1.0183491706848145,
      "learning_rate": 9.067830174367586e-05,
      "loss": 1.8137,
      "step": 500
    }
  ],
  "logging_steps": 5,
  "max_steps": 1795,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2509451155603456e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}