{
  "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": 4.701911449432373,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 3.5703,
      "step": 5
    },
    {
      "epoch": 0.0055737476485752105,
      "grad_norm": 4.517332077026367,
      "learning_rate": 5e-06,
      "loss": 3.5228,
      "step": 10
    },
    {
      "epoch": 0.008360621472862817,
      "grad_norm": 4.366923809051514,
      "learning_rate": 7.777777777777777e-06,
      "loss": 3.4713,
      "step": 15
    },
    {
      "epoch": 0.011147495297150421,
      "grad_norm": 4.379341125488281,
      "learning_rate": 1.0555555555555555e-05,
      "loss": 3.5622,
      "step": 20
    },
    {
      "epoch": 0.013934369121438027,
      "grad_norm": 5.151899814605713,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 3.4422,
      "step": 25
    },
    {
      "epoch": 0.016721242945725634,
      "grad_norm": 4.80206298828125,
      "learning_rate": 1.6111111111111115e-05,
      "loss": 3.3216,
      "step": 30
    },
    {
      "epoch": 0.019508116770013236,
      "grad_norm": 3.687617778778076,
      "learning_rate": 1.888888888888889e-05,
      "loss": 2.942,
      "step": 35
    },
    {
      "epoch": 0.022294990594300842,
      "grad_norm": 3.1759581565856934,
      "learning_rate": 2.1666666666666667e-05,
      "loss": 2.6761,
      "step": 40
    },
    {
      "epoch": 0.025081864418588447,
      "grad_norm": 2.588658332824707,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 2.6441,
      "step": 45
    },
    {
      "epoch": 0.027868738242876053,
      "grad_norm": 2.1114368438720703,
      "learning_rate": 2.7222222222222223e-05,
      "loss": 2.4923,
      "step": 50
    },
    {
      "epoch": 0.03065561206716366,
      "grad_norm": 2.0633840560913086,
      "learning_rate": 3e-05,
      "loss": 2.2875,
      "step": 55
    },
    {
      "epoch": 0.03344248589145127,
      "grad_norm": 1.8759537935256958,
      "learning_rate": 3.277777777777778e-05,
      "loss": 2.408,
      "step": 60
    },
    {
      "epoch": 0.03622935971573887,
      "grad_norm": 1.9867842197418213,
      "learning_rate": 3.555555555555556e-05,
      "loss": 2.3362,
      "step": 65
    },
    {
      "epoch": 0.03901623354002647,
      "grad_norm": 1.781015157699585,
      "learning_rate": 3.8333333333333334e-05,
      "loss": 2.1793,
      "step": 70
    },
    {
      "epoch": 0.04180310736431408,
      "grad_norm": 2.1073060035705566,
      "learning_rate": 4.111111111111111e-05,
      "loss": 2.2342,
      "step": 75
    },
    {
      "epoch": 0.044589981188601684,
      "grad_norm": 1.8073269128799438,
      "learning_rate": 4.388888888888889e-05,
      "loss": 2.1447,
      "step": 80
    },
    {
      "epoch": 0.04737685501288929,
      "grad_norm": 1.8754196166992188,
      "learning_rate": 4.666666666666667e-05,
      "loss": 2.3169,
      "step": 85
    },
    {
      "epoch": 0.050163728837176895,
      "grad_norm": 1.9674766063690186,
      "learning_rate": 4.9444444444444446e-05,
      "loss": 2.1966,
      "step": 90
    },
    {
      "epoch": 0.0529506026614645,
      "grad_norm": 2.0507538318634033,
      "learning_rate": 5.222222222222223e-05,
      "loss": 2.1436,
      "step": 95
    },
    {
      "epoch": 0.055737476485752106,
      "grad_norm": 2.1478066444396973,
      "learning_rate": 5.500000000000001e-05,
      "loss": 2.0947,
      "step": 100
    },
    {
      "epoch": 0.05852435031003971,
      "grad_norm": 1.8122990131378174,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 2.1266,
      "step": 105
    },
    {
      "epoch": 0.06131122413432732,
      "grad_norm": 2.30039119720459,
      "learning_rate": 6.055555555555555e-05,
      "loss": 2.175,
      "step": 110
    },
    {
      "epoch": 0.06409809795861493,
      "grad_norm": 2.1259849071502686,
      "learning_rate": 6.333333333333333e-05,
      "loss": 2.0622,
      "step": 115
    },
    {
      "epoch": 0.06688497178290254,
      "grad_norm": 2.159100294113159,
      "learning_rate": 6.611111111111111e-05,
      "loss": 2.078,
      "step": 120
    },
    {
      "epoch": 0.06967184560719013,
      "grad_norm": 2.2886006832122803,
      "learning_rate": 6.88888888888889e-05,
      "loss": 2.1515,
      "step": 125
    },
    {
      "epoch": 0.07245871943147773,
      "grad_norm": 2.1963438987731934,
      "learning_rate": 7.166666666666667e-05,
      "loss": 2.0886,
      "step": 130
    },
    {
      "epoch": 0.07524559325576534,
      "grad_norm": 1.9774861335754395,
      "learning_rate": 7.444444444444444e-05,
      "loss": 2.0426,
      "step": 135
    },
    {
      "epoch": 0.07803246708005294,
      "grad_norm": 2.2124135494232178,
      "learning_rate": 7.722222222222223e-05,
      "loss": 2.0686,
      "step": 140
    },
    {
      "epoch": 0.08081934090434055,
      "grad_norm": 2.0250704288482666,
      "learning_rate": 8e-05,
      "loss": 2.1701,
      "step": 145
    },
    {
      "epoch": 0.08360621472862816,
      "grad_norm": 2.2717087268829346,
      "learning_rate": 8.277777777777778e-05,
      "loss": 2.1313,
      "step": 150
    },
    {
      "epoch": 0.08639308855291576,
      "grad_norm": 2.090191602706909,
      "learning_rate": 8.555555555555556e-05,
      "loss": 2.0572,
      "step": 155
    },
    {
      "epoch": 0.08917996237720337,
      "grad_norm": 2.09201979637146,
      "learning_rate": 8.833333333333333e-05,
      "loss": 1.982,
      "step": 160
    },
    {
      "epoch": 0.09196683620149097,
      "grad_norm": 2.206078290939331,
      "learning_rate": 9.111111111111112e-05,
      "loss": 2.0063,
      "step": 165
    },
    {
      "epoch": 0.09475371002577858,
      "grad_norm": 2.1212692260742188,
      "learning_rate": 9.388888888888889e-05,
      "loss": 2.0597,
      "step": 170
    },
    {
      "epoch": 0.09754058385006618,
      "grad_norm": 2.215811252593994,
      "learning_rate": 9.666666666666667e-05,
      "loss": 2.0566,
      "step": 175
    },
    {
      "epoch": 0.10032745767435379,
      "grad_norm": 2.1642961502075195,
      "learning_rate": 9.944444444444446e-05,
      "loss": 2.0843,
      "step": 180
    },
    {
      "epoch": 0.1031143314986414,
      "grad_norm": 2.315004348754883,
      "learning_rate": 9.999848639521432e-05,
      "loss": 2.0011,
      "step": 185
    },
    {
      "epoch": 0.105901205322929,
      "grad_norm": 2.2057459354400635,
      "learning_rate": 9.999233753283091e-05,
      "loss": 1.9906,
      "step": 190
    },
    {
      "epoch": 0.1086880791472166,
      "grad_norm": 2.516157865524292,
      "learning_rate": 9.998145939378577e-05,
      "loss": 1.9941,
      "step": 195
    },
    {
      "epoch": 0.11147495297150421,
      "grad_norm": 2.2229058742523193,
      "learning_rate": 9.996585300715116e-05,
      "loss": 1.9935,
      "step": 200
    },
    {
      "epoch": 0.11426182679579182,
      "grad_norm": 2.1316332817077637,
      "learning_rate": 9.994551984929175e-05,
      "loss": 1.9692,
      "step": 205
    },
    {
      "epoch": 0.11704870062007942,
      "grad_norm": 1.8723924160003662,
      "learning_rate": 9.992046184372492e-05,
      "loss": 2.0208,
      "step": 210
    },
    {
      "epoch": 0.11983557444436703,
      "grad_norm": 2.1844356060028076,
      "learning_rate": 9.989068136093873e-05,
      "loss": 1.9584,
      "step": 215
    },
    {
      "epoch": 0.12262244826865464,
      "grad_norm": 2.2411093711853027,
      "learning_rate": 9.985618121816779e-05,
      "loss": 2.0796,
      "step": 220
    },
    {
      "epoch": 0.12540932209294225,
      "grad_norm": 2.0734896659851074,
      "learning_rate": 9.981696467912664e-05,
      "loss": 1.9712,
      "step": 225
    },
    {
      "epoch": 0.12819619591722986,
      "grad_norm": 2.1242079734802246,
      "learning_rate": 9.97730354537011e-05,
      "loss": 1.9637,
      "step": 230
    },
    {
      "epoch": 0.13098306974151747,
      "grad_norm": 2.026093006134033,
      "learning_rate": 9.972439769759722e-05,
      "loss": 1.8615,
      "step": 235
    },
    {
      "epoch": 0.13376994356580507,
      "grad_norm": 2.029449701309204,
      "learning_rate": 9.967105601194823e-05,
      "loss": 2.11,
      "step": 240
    },
    {
      "epoch": 0.13655681739009268,
      "grad_norm": 2.0343263149261475,
      "learning_rate": 9.961301544287922e-05,
      "loss": 1.9325,
      "step": 245
    },
    {
      "epoch": 0.13934369121438026,
      "grad_norm": 2.2704412937164307,
      "learning_rate": 9.955028148102979e-05,
      "loss": 1.8849,
      "step": 250
    },
    {
      "epoch": 0.14213056503866786,
      "grad_norm": 2.563328742980957,
      "learning_rate": 9.948286006103466e-05,
      "loss": 2.1103,
      "step": 255
    },
    {
      "epoch": 0.14491743886295547,
      "grad_norm": 2.4252612590789795,
      "learning_rate": 9.941075756096226e-05,
      "loss": 1.9867,
      "step": 260
    },
    {
      "epoch": 0.14770431268724307,
      "grad_norm": 2.1968955993652344,
      "learning_rate": 9.933398080171123e-05,
      "loss": 1.998,
      "step": 265
    },
    {
      "epoch": 0.15049118651153068,
      "grad_norm": 2.1625869274139404,
      "learning_rate": 9.925253704636543e-05,
      "loss": 1.9043,
      "step": 270
    },
    {
      "epoch": 0.15327806033581828,
      "grad_norm": 2.349252700805664,
      "learning_rate": 9.916643399950656e-05,
      "loss": 1.9854,
      "step": 275
    },
    {
      "epoch": 0.1560649341601059,
      "grad_norm": 2.2877767086029053,
      "learning_rate": 9.907567980648549e-05,
      "loss": 1.9119,
      "step": 280
    },
    {
      "epoch": 0.1588518079843935,
      "grad_norm": 2.0312581062316895,
      "learning_rate": 9.898028305265169e-05,
      "loss": 2.0732,
      "step": 285
    },
    {
      "epoch": 0.1616386818086811,
      "grad_norm": 2.036317825317383,
      "learning_rate": 9.888025276254096e-05,
      "loss": 2.0147,
      "step": 290
    },
    {
      "epoch": 0.1644255556329687,
      "grad_norm": 2.345585823059082,
      "learning_rate": 9.877559839902184e-05,
      "loss": 1.9635,
      "step": 295
    },
    {
      "epoch": 0.1672124294572563,
      "grad_norm": 2.371412992477417,
      "learning_rate": 9.86663298624003e-05,
      "loss": 1.9034,
      "step": 300
    },
    {
      "epoch": 0.16999930328154392,
      "grad_norm": 2.3761579990386963,
      "learning_rate": 9.855245748948326e-05,
      "loss": 1.9687,
      "step": 305
    },
    {
      "epoch": 0.17278617710583152,
      "grad_norm": 2.1475725173950195,
      "learning_rate": 9.843399205260068e-05,
      "loss": 1.9903,
      "step": 310
    },
    {
      "epoch": 0.17557305093011913,
      "grad_norm": 2.40266752243042,
      "learning_rate": 9.831094475858652e-05,
      "loss": 1.9729,
      "step": 315
    },
    {
      "epoch": 0.17835992475440673,
      "grad_norm": 2.187901258468628,
      "learning_rate": 9.818332724771857e-05,
      "loss": 2.0002,
      "step": 320
    },
    {
      "epoch": 0.18114679857869434,
      "grad_norm": 2.374246597290039,
      "learning_rate": 9.805115159261726e-05,
      "loss": 1.8938,
      "step": 325
    },
    {
      "epoch": 0.18393367240298195,
      "grad_norm": 2.2676525115966797,
      "learning_rate": 9.791443029710361e-05,
      "loss": 1.8728,
      "step": 330
    },
    {
      "epoch": 0.18672054622726955,
      "grad_norm": 2.175401449203491,
      "learning_rate": 9.777317629501636e-05,
      "loss": 1.9195,
      "step": 335
    },
    {
      "epoch": 0.18950742005155716,
      "grad_norm": 2.104386329650879,
      "learning_rate": 9.762740294898846e-05,
      "loss": 1.8618,
      "step": 340
    },
    {
      "epoch": 0.19229429387584476,
      "grad_norm": 2.2644734382629395,
      "learning_rate": 9.747712404918286e-05,
      "loss": 1.8384,
      "step": 345
    },
    {
      "epoch": 0.19508116770013237,
      "grad_norm": 2.6918885707855225,
      "learning_rate": 9.732235381198813e-05,
      "loss": 1.9305,
      "step": 350
    },
    {
      "epoch": 0.19786804152441997,
      "grad_norm": 2.374101161956787,
      "learning_rate": 9.716310687867342e-05,
      "loss": 1.7964,
      "step": 355
    },
    {
      "epoch": 0.20065491534870758,
      "grad_norm": 2.401925802230835,
      "learning_rate": 9.699939831400351e-05,
      "loss": 2.0372,
      "step": 360
    },
    {
      "epoch": 0.20344178917299519,
      "grad_norm": 2.3377015590667725,
      "learning_rate": 9.683124360481364e-05,
      "loss": 1.9354,
      "step": 365
    },
    {
      "epoch": 0.2062286629972828,
      "grad_norm": 2.669875144958496,
      "learning_rate": 9.665865865854445e-05,
      "loss": 1.9243,
      "step": 370
    },
    {
      "epoch": 0.2090155368215704,
      "grad_norm": 2.251633882522583,
      "learning_rate": 9.648165980173712e-05,
      "loss": 1.9379,
      "step": 375
    },
    {
      "epoch": 0.211802410645858,
      "grad_norm": 2.413219928741455,
      "learning_rate": 9.630026377848892e-05,
      "loss": 1.9232,
      "step": 380
    },
    {
      "epoch": 0.2145892844701456,
      "grad_norm": 2.4033939838409424,
      "learning_rate": 9.611448774886924e-05,
      "loss": 1.9206,
      "step": 385
    },
    {
      "epoch": 0.2173761582944332,
      "grad_norm": 2.7085165977478027,
      "learning_rate": 9.592434928729616e-05,
      "loss": 1.965,
      "step": 390
    },
    {
      "epoch": 0.22016303211872082,
      "grad_norm": 2.440049409866333,
      "learning_rate": 9.572986638087396e-05,
      "loss": 1.938,
      "step": 395
    },
    {
      "epoch": 0.22294990594300843,
      "grad_norm": 2.2908005714416504,
      "learning_rate": 9.553105742769154e-05,
      "loss": 1.8841,
      "step": 400
    },
    {
      "epoch": 0.22573677976729603,
      "grad_norm": 2.0851612091064453,
      "learning_rate": 9.532794123508197e-05,
      "loss": 1.7699,
      "step": 405
    },
    {
      "epoch": 0.22852365359158364,
      "grad_norm": 2.466432809829712,
      "learning_rate": 9.512053701784329e-05,
      "loss": 1.9162,
      "step": 410
    },
    {
      "epoch": 0.23131052741587124,
      "grad_norm": 2.130993127822876,
      "learning_rate": 9.490886439642081e-05,
      "loss": 1.9023,
      "step": 415
    },
    {
      "epoch": 0.23409740124015885,
      "grad_norm": 2.220752716064453,
      "learning_rate": 9.469294339505098e-05,
      "loss": 1.975,
      "step": 420
    },
    {
      "epoch": 0.23688427506444645,
      "grad_norm": 2.1913466453552246,
      "learning_rate": 9.447279443986716e-05,
      "loss": 1.8632,
      "step": 425
    },
    {
      "epoch": 0.23967114888873406,
      "grad_norm": 2.321277379989624,
      "learning_rate": 9.424843835696724e-05,
      "loss": 1.9294,
      "step": 430
    },
    {
      "epoch": 0.24245802271302166,
      "grad_norm": 2.286395788192749,
      "learning_rate": 9.401989637044355e-05,
      "loss": 1.8428,
      "step": 435
    },
    {
      "epoch": 0.24524489653730927,
      "grad_norm": 2.3919999599456787,
      "learning_rate": 9.3787190100375e-05,
      "loss": 1.8813,
      "step": 440
    },
    {
      "epoch": 0.24803177036159688,
      "grad_norm": 2.1720130443573,
      "learning_rate": 9.355034156078188e-05,
      "loss": 1.8569,
      "step": 445
    },
    {
      "epoch": 0.2508186441858845,
      "grad_norm": 2.086071014404297,
      "learning_rate": 9.330937315754329e-05,
      "loss": 1.8796,
      "step": 450
    },
    {
      "epoch": 0.2536055180101721,
      "grad_norm": 2.447908639907837,
      "learning_rate": 9.306430768627753e-05,
      "loss": 1.851,
      "step": 455
    },
    {
      "epoch": 0.2563923918344597,
      "grad_norm": 2.631868362426758,
      "learning_rate": 9.281516833018571e-05,
      "loss": 1.8706,
      "step": 460
    },
    {
      "epoch": 0.2591792656587473,
      "grad_norm": 2.198009729385376,
      "learning_rate": 9.256197865785854e-05,
      "loss": 1.9986,
      "step": 465
    },
    {
      "epoch": 0.26196613948303493,
      "grad_norm": 2.151010513305664,
      "learning_rate": 9.230476262104677e-05,
      "loss": 1.8895,
      "step": 470
    },
    {
      "epoch": 0.26475301330732254,
      "grad_norm": 2.335463047027588,
      "learning_rate": 9.204354455239539e-05,
      "loss": 1.9415,
      "step": 475
    },
    {
      "epoch": 0.26753988713161014,
      "grad_norm": 2.4173576831817627,
      "learning_rate": 9.177834916314165e-05,
      "loss": 1.8676,
      "step": 480
    },
    {
      "epoch": 0.27032676095589775,
      "grad_norm": 2.409341812133789,
      "learning_rate": 9.150920154077754e-05,
      "loss": 1.879,
      "step": 485
    },
    {
      "epoch": 0.27311363478018535,
      "grad_norm": 2.496391534805298,
      "learning_rate": 9.123612714667634e-05,
      "loss": 1.9162,
      "step": 490
    },
    {
      "epoch": 0.2759005086044729,
      "grad_norm": 2.2236454486846924,
      "learning_rate": 9.095915181368412e-05,
      "loss": 1.9825,
      "step": 495
    },
    {
      "epoch": 0.2786873824287605,
      "grad_norm": 2.4536592960357666,
      "learning_rate": 9.067830174367586e-05,
      "loss": 1.933,
      "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": 2995820509261824.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}