{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9963379544860058,
  "eval_steps": 500,
  "global_step": 954,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020925974365681402,
      "grad_norm": 0.5565130669449413,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.0288,
      "step": 10
    },
    {
      "epoch": 0.041851948731362804,
      "grad_norm": 0.18606951178093903,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0118,
      "step": 20
    },
    {
      "epoch": 0.0627779230970442,
      "grad_norm": 0.03950245765434584,
      "learning_rate": 6.25e-06,
      "loss": 0.0039,
      "step": 30
    },
    {
      "epoch": 0.08370389746272561,
      "grad_norm": 0.03819057800914349,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0029,
      "step": 40
    },
    {
      "epoch": 0.10462987182840701,
      "grad_norm": 0.0544544775809417,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.0028,
      "step": 50
    },
    {
      "epoch": 0.1255558461940884,
      "grad_norm": 0.017750069891728056,
      "learning_rate": 1.25e-05,
      "loss": 0.0024,
      "step": 60
    },
    {
      "epoch": 0.14648182055976983,
      "grad_norm": 0.025688827192495198,
      "learning_rate": 1.4583333333333333e-05,
      "loss": 0.0021,
      "step": 70
    },
    {
      "epoch": 0.16740779492545121,
      "grad_norm": 0.009892162090487124,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0013,
      "step": 80
    },
    {
      "epoch": 0.18833376929113263,
      "grad_norm": 0.013992833360284824,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0006,
      "step": 90
    },
    {
      "epoch": 0.20925974365681402,
      "grad_norm": 0.009570194364057852,
      "learning_rate": 1.9998927475076107e-05,
      "loss": 0.0029,
      "step": 100
    },
    {
      "epoch": 0.23018571802249543,
      "grad_norm": 0.011885486768552911,
      "learning_rate": 1.998686421164407e-05,
      "loss": 0.0019,
      "step": 110
    },
    {
      "epoch": 0.2511116923881768,
      "grad_norm": 0.00535124159298834,
      "learning_rate": 1.9961413253717214e-05,
      "loss": 0.0007,
      "step": 120
    },
    {
      "epoch": 0.2720376667538582,
      "grad_norm": 0.014087761849564343,
      "learning_rate": 1.9922608719076874e-05,
      "loss": 0.002,
      "step": 130
    },
    {
      "epoch": 0.29296364111953965,
      "grad_norm": 0.014243222134916782,
      "learning_rate": 1.9870502626379127e-05,
      "loss": 0.0011,
      "step": 140
    },
    {
      "epoch": 0.31388961548522104,
      "grad_norm": 0.02081790679071402,
      "learning_rate": 1.980516482542224e-05,
      "loss": 0.0019,
      "step": 150
    },
    {
      "epoch": 0.33481558985090243,
      "grad_norm": 0.017403596614001808,
      "learning_rate": 1.972668290351084e-05,
      "loss": 0.0022,
      "step": 160
    },
    {
      "epoch": 0.3557415642165838,
      "grad_norm": 0.014181950638291016,
      "learning_rate": 1.9635162068042547e-05,
      "loss": 0.0013,
      "step": 170
    },
    {
      "epoch": 0.37666753858226526,
      "grad_norm": 0.004266630945852344,
      "learning_rate": 1.9530725005474195e-05,
      "loss": 0.0016,
      "step": 180
    },
    {
      "epoch": 0.39759351294794665,
      "grad_norm": 0.003584919486125673,
      "learning_rate": 1.9413511716856973e-05,
      "loss": 0.0017,
      "step": 190
    },
    {
      "epoch": 0.41851948731362804,
      "grad_norm": 0.010407187744097775,
      "learning_rate": 1.9283679330160726e-05,
      "loss": 0.0006,
      "step": 200
    },
    {
      "epoch": 0.4394454616793094,
      "grad_norm": 0.011994286058888401,
      "learning_rate": 1.9141401889639167e-05,
      "loss": 0.004,
      "step": 210
    },
    {
      "epoch": 0.46037143604499087,
      "grad_norm": 0.002559284420598525,
      "learning_rate": 1.898687012251826e-05,
      "loss": 0.0014,
      "step": 220
    },
    {
      "epoch": 0.48129741041067226,
      "grad_norm": 0.01635207542112704,
      "learning_rate": 1.8820291183320602e-05,
      "loss": 0.0035,
      "step": 230
    },
    {
      "epoch": 0.5022233847763536,
      "grad_norm": 0.019845090979002694,
      "learning_rate": 1.8641888376168483e-05,
      "loss": 0.0004,
      "step": 240
    },
    {
      "epoch": 0.5231493591420351,
      "grad_norm": 0.037240037127226344,
      "learning_rate": 1.845190085543795e-05,
      "loss": 0.0032,
      "step": 250
    },
    {
      "epoch": 0.5440753335077164,
      "grad_norm": 0.006653507161649646,
      "learning_rate": 1.8250583305165098e-05,
      "loss": 0.0009,
      "step": 260
    },
    {
      "epoch": 0.5650013078733979,
      "grad_norm": 0.6020150741013875,
      "learning_rate": 1.8038205597634392e-05,
      "loss": 0.0011,
      "step": 270
    },
    {
      "epoch": 0.5859272822390793,
      "grad_norm": 0.05961677548525192,
      "learning_rate": 1.7815052431606702e-05,
      "loss": 0.0042,
      "step": 280
    },
    {
      "epoch": 0.6068532566047606,
      "grad_norm": 0.03667937697835753,
      "learning_rate": 1.7581422950671942e-05,
      "loss": 0.0029,
      "step": 290
    },
    {
      "epoch": 0.6277792309704421,
      "grad_norm": 0.02203193747208106,
      "learning_rate": 1.733763034223804e-05,
      "loss": 0.0017,
      "step": 300
    },
    {
      "epoch": 0.6487052053361234,
      "grad_norm": 0.03047813653712161,
      "learning_rate": 1.7084001417693702e-05,
      "loss": 0.0013,
      "step": 310
    },
    {
      "epoch": 0.6696311797018049,
      "grad_norm": 0.02403172913701344,
      "learning_rate": 1.682087617430782e-05,
      "loss": 0.001,
      "step": 320
    },
    {
      "epoch": 0.6905571540674863,
      "grad_norm": 0.020885112433327038,
      "learning_rate": 1.6548607339452853e-05,
      "loss": 0.0009,
      "step": 330
    },
    {
      "epoch": 0.7114831284331676,
      "grad_norm": 0.02143325180735118,
      "learning_rate": 1.626755989776303e-05,
      "loss": 0.0007,
      "step": 340
    },
    {
      "epoch": 0.7324091027988491,
      "grad_norm": 0.028502523632724423,
      "learning_rate": 1.5978110601861408e-05,
      "loss": 0.0016,
      "step": 350
    },
    {
      "epoch": 0.7533350771645305,
      "grad_norm": 0.00832352415812105,
      "learning_rate": 1.568064746731156e-05,
      "loss": 0.0009,
      "step": 360
    },
    {
      "epoch": 0.7742610515302119,
      "grad_norm": 0.009527918206519135,
      "learning_rate": 1.5375569252470897e-05,
      "loss": 0.0018,
      "step": 370
    },
    {
      "epoch": 0.7951870258958933,
      "grad_norm": 0.025970668054977514,
      "learning_rate": 1.506328492394303e-05,
      "loss": 0.0006,
      "step": 380
    },
    {
      "epoch": 0.8161130002615746,
      "grad_norm": 0.026371283237352123,
      "learning_rate": 1.4744213108345605e-05,
      "loss": 0.0023,
      "step": 390
    },
    {
      "epoch": 0.8370389746272561,
      "grad_norm": 0.030148449665476192,
      "learning_rate": 1.4418781531128636e-05,
      "loss": 0.0021,
      "step": 400
    },
    {
      "epoch": 0.8579649489929375,
      "grad_norm": 0.009123367946997817,
      "learning_rate": 1.4087426443195549e-05,
      "loss": 0.0009,
      "step": 410
    },
    {
      "epoch": 0.8788909233586188,
      "grad_norm": 0.033901495227343266,
      "learning_rate": 1.375059203609562e-05,
      "loss": 0.0023,
      "step": 420
    },
    {
      "epoch": 0.8998168977243003,
      "grad_norm": 0.007725818014371622,
      "learning_rate": 1.3408729846571716e-05,
      "loss": 0.0006,
      "step": 430
    },
    {
      "epoch": 0.9207428720899817,
      "grad_norm": 0.01076304174801131,
      "learning_rate": 1.3062298151261592e-05,
      "loss": 0.0009,
      "step": 440
    },
    {
      "epoch": 0.9416688464556631,
      "grad_norm": 0.08792335870716929,
      "learning_rate": 1.2711761352364172e-05,
      "loss": 0.0016,
      "step": 450
    },
    {
      "epoch": 0.9625948208213445,
      "grad_norm": 0.019193508970073957,
      "learning_rate": 1.2357589355094275e-05,
      "loss": 0.0005,
      "step": 460
    },
    {
      "epoch": 0.9835207951870258,
      "grad_norm": 0.008676099200769616,
      "learning_rate": 1.2000256937760446e-05,
      "loss": 0.0019,
      "step": 470
    },
    {
      "epoch": 1.0044467695527073,
      "grad_norm": 0.0014044382775048332,
      "learning_rate": 1.1640243115310219e-05,
      "loss": 0.0013,
      "step": 480
    },
    {
      "epoch": 1.0253727439183886,
      "grad_norm": 0.0047799061066043395,
      "learning_rate": 1.127803049719605e-05,
      "loss": 0.0008,
      "step": 490
    },
    {
      "epoch": 1.0462987182840702,
      "grad_norm": 0.006651240484213586,
      "learning_rate": 1.091410464042268e-05,
      "loss": 0.0004,
      "step": 500
    },
    {
      "epoch": 1.0672246926497515,
      "grad_norm": 0.007370492651937019,
      "learning_rate": 1.0548953398643276e-05,
      "loss": 0.0002,
      "step": 510
    },
    {
      "epoch": 1.0881506670154328,
      "grad_norm": 0.004999091099852188,
      "learning_rate": 1.0183066268176775e-05,
      "loss": 0.0013,
      "step": 520
    },
    {
      "epoch": 1.1090766413811144,
      "grad_norm": 0.0067648135423557494,
      "learning_rate": 9.81693373182323e-06,
      "loss": 0.0004,
      "step": 530
    },
    {
      "epoch": 1.1300026157467957,
      "grad_norm": 0.006355010008632378,
      "learning_rate": 9.451046601356725e-06,
      "loss": 0.0005,
      "step": 540
    },
    {
      "epoch": 1.150928590112477,
      "grad_norm": 0.0019993701626922662,
      "learning_rate": 9.085895359577324e-06,
      "loss": 0.0012,
      "step": 550
    },
    {
      "epoch": 1.1718545644781586,
      "grad_norm": 0.07841604107285226,
      "learning_rate": 8.721969502803954e-06,
      "loss": 0.0007,
      "step": 560
    },
    {
      "epoch": 1.19278053884384,
      "grad_norm": 0.008878084655677647,
      "learning_rate": 8.359756884689785e-06,
      "loss": 0.0006,
      "step": 570
    },
    {
      "epoch": 1.2137065132095213,
      "grad_norm": 0.022433859996456522,
      "learning_rate": 7.999743062239557e-06,
      "loss": 0.001,
      "step": 580
    },
    {
      "epoch": 1.2346324875752028,
      "grad_norm": 0.007321094493119974,
      "learning_rate": 7.642410644905726e-06,
      "loss": 0.0002,
      "step": 590
    },
    {
      "epoch": 1.2555584619408842,
      "grad_norm": 0.004130666592411099,
      "learning_rate": 7.2882386476358304e-06,
      "loss": 0.0008,
      "step": 600
    },
    {
      "epoch": 1.2764844363065655,
      "grad_norm": 0.009136384794392362,
      "learning_rate": 6.937701848738407e-06,
      "loss": 0.0005,
      "step": 610
    },
    {
      "epoch": 1.297410410672247,
      "grad_norm": 0.014936959834984399,
      "learning_rate": 6.591270153428288e-06,
      "loss": 0.0015,
      "step": 620
    },
    {
      "epoch": 1.3183363850379284,
      "grad_norm": 0.00995968283183629,
      "learning_rate": 6.249407963904381e-06,
      "loss": 0.0003,
      "step": 630
    },
    {
      "epoch": 1.3392623594036097,
      "grad_norm": 0.007922264005744792,
      "learning_rate": 5.912573556804453e-06,
      "loss": 0.002,
      "step": 640
    },
    {
      "epoch": 1.3601883337692913,
      "grad_norm": 0.009732474218704226,
      "learning_rate": 5.581218468871365e-06,
      "loss": 0.0015,
      "step": 650
    },
    {
      "epoch": 1.3811143081349726,
      "grad_norm": 0.029322482465715245,
      "learning_rate": 5.2557868916543996e-06,
      "loss": 0.0004,
      "step": 660
    },
    {
      "epoch": 1.402040282500654,
      "grad_norm": 0.002219147933917833,
      "learning_rate": 4.9367150760569746e-06,
      "loss": 0.0013,
      "step": 670
    },
    {
      "epoch": 1.4229662568663353,
      "grad_norm": 0.016601760824032873,
      "learning_rate": 4.6244307475291025e-06,
      "loss": 0.0007,
      "step": 680
    },
    {
      "epoch": 1.4438922312320168,
      "grad_norm": 0.03618373662395389,
      "learning_rate": 4.319352532688444e-06,
      "loss": 0.0016,
      "step": 690
    },
    {
      "epoch": 1.4648182055976982,
      "grad_norm": 0.008705175714719602,
      "learning_rate": 4.0218893981385935e-06,
      "loss": 0.0009,
      "step": 700
    },
    {
      "epoch": 1.4857441799633795,
      "grad_norm": 0.0029313903115019873,
      "learning_rate": 3.732440102236975e-06,
      "loss": 0.0002,
      "step": 710
    },
    {
      "epoch": 1.5066701543290608,
      "grad_norm": 0.020710402409599028,
      "learning_rate": 3.4513926605471504e-06,
      "loss": 0.0007,
      "step": 720
    },
    {
      "epoch": 1.5275961286947424,
      "grad_norm": 0.03507089834475249,
      "learning_rate": 3.1791238256921785e-06,
      "loss": 0.0021,
      "step": 730
    },
    {
      "epoch": 1.5485221030604237,
      "grad_norm": 0.07040984119700786,
      "learning_rate": 2.9159985823062997e-06,
      "loss": 0.0012,
      "step": 740
    },
    {
      "epoch": 1.569448077426105,
      "grad_norm": 0.0046045007369338465,
      "learning_rate": 2.662369657761963e-06,
      "loss": 0.0005,
      "step": 750
    },
    {
      "epoch": 1.5903740517917866,
      "grad_norm": 0.044927829447691095,
      "learning_rate": 2.418577049328058e-06,
      "loss": 0.001,
      "step": 760
    },
    {
      "epoch": 1.611300026157468,
      "grad_norm": 0.005367586480852439,
      "learning_rate": 2.1849475683932996e-06,
      "loss": 0.0005,
      "step": 770
    },
    {
      "epoch": 1.6322260005231493,
      "grad_norm": 0.014755038810438978,
      "learning_rate": 1.961794402365611e-06,
      "loss": 0.0012,
      "step": 780
    },
    {
      "epoch": 1.6531519748888308,
      "grad_norm": 0.007586271120637286,
      "learning_rate": 1.7494166948349057e-06,
      "loss": 0.0023,
      "step": 790
    },
    {
      "epoch": 1.6740779492545121,
      "grad_norm": 0.01450560934843513,
      "learning_rate": 1.5480991445620541e-06,
      "loss": 0.001,
      "step": 800
    },
    {
      "epoch": 1.6950039236201935,
      "grad_norm": 0.007197310242295822,
      "learning_rate": 1.3581116238315194e-06,
      "loss": 0.0004,
      "step": 810
    },
    {
      "epoch": 1.715929897985875,
      "grad_norm": 0.011429665649141593,
      "learning_rate": 1.1797088166794002e-06,
      "loss": 0.0009,
      "step": 820
    },
    {
      "epoch": 1.7368558723515564,
      "grad_norm": 0.005479589436376736,
      "learning_rate": 1.013129877481741e-06,
      "loss": 0.0015,
      "step": 830
    },
    {
      "epoch": 1.7577818467172377,
      "grad_norm": 0.010284806366489312,
      "learning_rate": 8.585981103608343e-07,
      "loss": 0.0003,
      "step": 840
    },
    {
      "epoch": 1.7787078210829192,
      "grad_norm": 0.004805391506128972,
      "learning_rate": 7.163206698392744e-07,
      "loss": 0.0007,
      "step": 850
    },
    {
      "epoch": 1.7996337954486006,
      "grad_norm": 0.00832006380541767,
      "learning_rate": 5.864882831430274e-07,
      "loss": 0.0013,
      "step": 860
    },
    {
      "epoch": 1.820559769814282,
      "grad_norm": 0.0016488023765299498,
      "learning_rate": 4.6927499452580574e-07,
      "loss": 0.0004,
      "step": 870
    },
    {
      "epoch": 1.8414857441799635,
      "grad_norm": 0.010828047107950923,
      "learning_rate": 3.6483793195745686e-07,
      "loss": 0.002,
      "step": 880
    },
    {
      "epoch": 1.8624117185456448,
      "grad_norm": 0.00029272379323776047,
      "learning_rate": 2.733170964891607e-07,
      "loss": 0.0008,
      "step": 890
    },
    {
      "epoch": 1.8833376929113261,
      "grad_norm": 0.011245028659871065,
      "learning_rate": 1.9483517457776436e-07,
      "loss": 0.0013,
      "step": 900
    },
    {
      "epoch": 1.9042636672770077,
      "grad_norm": 0.005356139523384608,
      "learning_rate": 1.2949737362087156e-07,
      "loss": 0.0009,
      "step": 910
    },
    {
      "epoch": 1.925189641642689,
      "grad_norm": 0.006490951580098314,
      "learning_rate": 7.73912809231292e-08,
      "loss": 0.0005,
      "step": 920
    },
    {
      "epoch": 1.9461156160083704,
      "grad_norm": 0.003383501070683301,
      "learning_rate": 3.858674628278825e-08,
      "loss": 0.003,
      "step": 930
    },
    {
      "epoch": 1.967041590374052,
      "grad_norm": 0.012526402652665877,
      "learning_rate": 1.3135788355934652e-08,
      "loss": 0.0004,
      "step": 940
    },
    {
      "epoch": 1.9879675647397332,
      "grad_norm": 0.016656680718419487,
      "learning_rate": 1.0725249238940916e-09,
      "loss": 0.0006,
      "step": 950
    },
    {
      "epoch": 1.9963379544860058,
      "step": 954,
      "total_flos": 841298254757888.0,
      "train_loss": 0.001730952451793395,
      "train_runtime": 49706.0079,
      "train_samples_per_second": 2.461,
      "train_steps_per_second": 0.019
    }
  ],
  "logging_steps": 10,
  "max_steps": 954,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 841298254757888.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}