{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1925,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005194805194805195,
      "grad_norm": 1.3003292727167528,
      "learning_rate": 2.8735632183908046e-06,
      "loss": 0.5747,
      "step": 10
    },
    {
      "epoch": 0.01038961038961039,
      "grad_norm": 0.8164205391130951,
      "learning_rate": 5.747126436781609e-06,
      "loss": 0.5181,
      "step": 20
    },
    {
      "epoch": 0.015584415584415584,
      "grad_norm": 0.488946468114722,
      "learning_rate": 8.620689655172414e-06,
      "loss": 0.4525,
      "step": 30
    },
    {
      "epoch": 0.02077922077922078,
      "grad_norm": 0.5240257767228009,
      "learning_rate": 1.1494252873563218e-05,
      "loss": 0.4153,
      "step": 40
    },
    {
      "epoch": 0.025974025974025976,
      "grad_norm": 0.3878833187137561,
      "learning_rate": 1.4367816091954022e-05,
      "loss": 0.3917,
      "step": 50
    },
    {
      "epoch": 0.03116883116883117,
      "grad_norm": 0.3932438541306065,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 0.3707,
      "step": 60
    },
    {
      "epoch": 0.03636363636363636,
      "grad_norm": 0.4466114932768719,
      "learning_rate": 2.0114942528735632e-05,
      "loss": 0.3605,
      "step": 70
    },
    {
      "epoch": 0.04155844155844156,
      "grad_norm": 0.8032738854038463,
      "learning_rate": 2.2988505747126437e-05,
      "loss": 0.3591,
      "step": 80
    },
    {
      "epoch": 0.046753246753246755,
      "grad_norm": 0.543812557370986,
      "learning_rate": 2.5862068965517244e-05,
      "loss": 0.3472,
      "step": 90
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 0.5733743111451876,
      "learning_rate": 2.8735632183908045e-05,
      "loss": 0.3341,
      "step": 100
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 0.9136032757030011,
      "learning_rate": 3.160919540229885e-05,
      "loss": 0.34,
      "step": 110
    },
    {
      "epoch": 0.06233766233766234,
      "grad_norm": 0.8015940010159968,
      "learning_rate": 3.4482758620689657e-05,
      "loss": 0.3346,
      "step": 120
    },
    {
      "epoch": 0.06753246753246753,
      "grad_norm": 0.5317651402718742,
      "learning_rate": 3.735632183908046e-05,
      "loss": 0.3205,
      "step": 130
    },
    {
      "epoch": 0.07272727272727272,
      "grad_norm": 0.9794116224334949,
      "learning_rate": 4.0229885057471265e-05,
      "loss": 0.3247,
      "step": 140
    },
    {
      "epoch": 0.07792207792207792,
      "grad_norm": 0.5926607382301553,
      "learning_rate": 4.3103448275862066e-05,
      "loss": 0.3137,
      "step": 150
    },
    {
      "epoch": 0.08311688311688312,
      "grad_norm": 0.501504811182817,
      "learning_rate": 4.597701149425287e-05,
      "loss": 0.3196,
      "step": 160
    },
    {
      "epoch": 0.08831168831168831,
      "grad_norm": 0.4844466938932139,
      "learning_rate": 4.885057471264368e-05,
      "loss": 0.3185,
      "step": 170
    },
    {
      "epoch": 0.09350649350649351,
      "grad_norm": 0.8792222600261355,
      "learning_rate": 4.999985842691236e-05,
      "loss": 0.317,
      "step": 180
    },
    {
      "epoch": 0.0987012987012987,
      "grad_norm": 0.538260729358137,
      "learning_rate": 4.999899326385009e-05,
      "loss": 0.3122,
      "step": 190
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 0.5532332343125024,
      "learning_rate": 4.99973416166265e-05,
      "loss": 0.308,
      "step": 200
    },
    {
      "epoch": 0.10909090909090909,
      "grad_norm": 0.5065931571721304,
      "learning_rate": 4.999490353720347e-05,
      "loss": 0.305,
      "step": 210
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 0.4661004020519144,
      "learning_rate": 4.9991679102284494e-05,
      "loss": 0.3031,
      "step": 220
    },
    {
      "epoch": 0.11948051948051948,
      "grad_norm": 0.4350447913864305,
      "learning_rate": 4.998766841331236e-05,
      "loss": 0.2979,
      "step": 230
    },
    {
      "epoch": 0.12467532467532468,
      "grad_norm": 0.3782265622233307,
      "learning_rate": 4.998287159646586e-05,
      "loss": 0.3035,
      "step": 240
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.3863507226733708,
      "learning_rate": 4.997728880265592e-05,
      "loss": 0.3024,
      "step": 250
    },
    {
      "epoch": 0.13506493506493505,
      "grad_norm": 0.45465564793802865,
      "learning_rate": 4.9970920207520756e-05,
      "loss": 0.2984,
      "step": 260
    },
    {
      "epoch": 0.14025974025974025,
      "grad_norm": 0.331617769411873,
      "learning_rate": 4.9963766011420394e-05,
      "loss": 0.2947,
      "step": 270
    },
    {
      "epoch": 0.14545454545454545,
      "grad_norm": 0.4323281120205803,
      "learning_rate": 4.9955826439430384e-05,
      "loss": 0.2885,
      "step": 280
    },
    {
      "epoch": 0.15064935064935064,
      "grad_norm": 0.3828139513983012,
      "learning_rate": 4.994710174133469e-05,
      "loss": 0.2957,
      "step": 290
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 0.45442672024966796,
      "learning_rate": 4.9937592191617846e-05,
      "loss": 0.2929,
      "step": 300
    },
    {
      "epoch": 0.16103896103896104,
      "grad_norm": 0.4324048943234123,
      "learning_rate": 4.992729808945629e-05,
      "loss": 0.287,
      "step": 310
    },
    {
      "epoch": 0.16623376623376623,
      "grad_norm": 0.3358736775493111,
      "learning_rate": 4.991621975870901e-05,
      "loss": 0.2831,
      "step": 320
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 0.33113790181444713,
      "learning_rate": 4.990435754790731e-05,
      "loss": 0.2868,
      "step": 330
    },
    {
      "epoch": 0.17662337662337663,
      "grad_norm": 0.3207965425558816,
      "learning_rate": 4.9891711830243845e-05,
      "loss": 0.2911,
      "step": 340
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.361708114933415,
      "learning_rate": 4.987828300356091e-05,
      "loss": 0.2857,
      "step": 350
    },
    {
      "epoch": 0.18701298701298702,
      "grad_norm": 0.34168626805695096,
      "learning_rate": 4.9864071490337896e-05,
      "loss": 0.2849,
      "step": 360
    },
    {
      "epoch": 0.19220779220779222,
      "grad_norm": 0.2908926746764898,
      "learning_rate": 4.9849077737678e-05,
      "loss": 0.2794,
      "step": 370
    },
    {
      "epoch": 0.1974025974025974,
      "grad_norm": 0.5087944109212423,
      "learning_rate": 4.983330221729419e-05,
      "loss": 0.2787,
      "step": 380
    },
    {
      "epoch": 0.2025974025974026,
      "grad_norm": 0.27804001121817934,
      "learning_rate": 4.9816745425494326e-05,
      "loss": 0.2759,
      "step": 390
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 0.4013635528070106,
      "learning_rate": 4.979940788316556e-05,
      "loss": 0.2817,
      "step": 400
    },
    {
      "epoch": 0.21298701298701297,
      "grad_norm": 0.25511672135848257,
      "learning_rate": 4.978129013575796e-05,
      "loss": 0.2785,
      "step": 410
    },
    {
      "epoch": 0.21818181818181817,
      "grad_norm": 0.298813003536876,
      "learning_rate": 4.976239275326733e-05,
      "loss": 0.2803,
      "step": 420
    },
    {
      "epoch": 0.22337662337662337,
      "grad_norm": 0.25617082862388485,
      "learning_rate": 4.974271633021729e-05,
      "loss": 0.2736,
      "step": 430
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.29247425322362797,
      "learning_rate": 4.9722261485640584e-05,
      "loss": 0.2767,
      "step": 440
    },
    {
      "epoch": 0.23376623376623376,
      "grad_norm": 0.33775737625365165,
      "learning_rate": 4.9701028863059563e-05,
      "loss": 0.2753,
      "step": 450
    },
    {
      "epoch": 0.23896103896103896,
      "grad_norm": 0.38447368527387143,
      "learning_rate": 4.967901913046598e-05,
      "loss": 0.2805,
      "step": 460
    },
    {
      "epoch": 0.24415584415584415,
      "grad_norm": 0.36418471782114614,
      "learning_rate": 4.9656232980299976e-05,
      "loss": 0.2707,
      "step": 470
    },
    {
      "epoch": 0.24935064935064935,
      "grad_norm": 0.2826049309528878,
      "learning_rate": 4.963267112942826e-05,
      "loss": 0.2775,
      "step": 480
    },
    {
      "epoch": 0.2545454545454545,
      "grad_norm": 0.3501209948268494,
      "learning_rate": 4.9608334319121584e-05,
      "loss": 0.2731,
      "step": 490
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.3608819312798831,
      "learning_rate": 4.958322331503141e-05,
      "loss": 0.2707,
      "step": 500
    },
    {
      "epoch": 0.2649350649350649,
      "grad_norm": 0.299020142675044,
      "learning_rate": 4.9557338907165833e-05,
      "loss": 0.2732,
      "step": 510
    },
    {
      "epoch": 0.2701298701298701,
      "grad_norm": 0.2685575579055106,
      "learning_rate": 4.9530681909864724e-05,
      "loss": 0.2689,
      "step": 520
    },
    {
      "epoch": 0.2753246753246753,
      "grad_norm": 0.23478660501034596,
      "learning_rate": 4.950325316177409e-05,
      "loss": 0.2726,
      "step": 530
    },
    {
      "epoch": 0.2805194805194805,
      "grad_norm": 0.26267550311986976,
      "learning_rate": 4.947505352581974e-05,
      "loss": 0.2688,
      "step": 540
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.2804831055299143,
      "learning_rate": 4.944608388918005e-05,
      "loss": 0.2724,
      "step": 550
    },
    {
      "epoch": 0.2909090909090909,
      "grad_norm": 0.36597728037439853,
      "learning_rate": 4.941634516325816e-05,
      "loss": 0.2674,
      "step": 560
    },
    {
      "epoch": 0.2961038961038961,
      "grad_norm": 0.27054673370836463,
      "learning_rate": 4.9385838283653216e-05,
      "loss": 0.2649,
      "step": 570
    },
    {
      "epoch": 0.3012987012987013,
      "grad_norm": 0.3066834081735395,
      "learning_rate": 4.9354564210130976e-05,
      "loss": 0.2677,
      "step": 580
    },
    {
      "epoch": 0.3064935064935065,
      "grad_norm": 0.2950401672059928,
      "learning_rate": 4.93225239265936e-05,
      "loss": 0.2622,
      "step": 590
    },
    {
      "epoch": 0.3116883116883117,
      "grad_norm": 0.31772179112234966,
      "learning_rate": 4.928971844104868e-05,
      "loss": 0.2641,
      "step": 600
    },
    {
      "epoch": 0.3168831168831169,
      "grad_norm": 0.27924556453889027,
      "learning_rate": 4.9256148785577606e-05,
      "loss": 0.2647,
      "step": 610
    },
    {
      "epoch": 0.3220779220779221,
      "grad_norm": 0.28644663252200886,
      "learning_rate": 4.9221816016302966e-05,
      "loss": 0.2625,
      "step": 620
    },
    {
      "epoch": 0.32727272727272727,
      "grad_norm": 0.2606246292323375,
      "learning_rate": 4.9186721213355455e-05,
      "loss": 0.2636,
      "step": 630
    },
    {
      "epoch": 0.33246753246753247,
      "grad_norm": 0.3172496714001626,
      "learning_rate": 4.915086548083978e-05,
      "loss": 0.2683,
      "step": 640
    },
    {
      "epoch": 0.33766233766233766,
      "grad_norm": 0.24927905040341644,
      "learning_rate": 4.9114249946800003e-05,
      "loss": 0.2654,
      "step": 650
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 0.32250447729794757,
      "learning_rate": 4.907687576318401e-05,
      "loss": 0.2636,
      "step": 660
    },
    {
      "epoch": 0.34805194805194806,
      "grad_norm": 0.32565974721616914,
      "learning_rate": 4.903874410580731e-05,
      "loss": 0.2601,
      "step": 670
    },
    {
      "epoch": 0.35324675324675325,
      "grad_norm": 0.23517482221948124,
      "learning_rate": 4.899985617431597e-05,
      "loss": 0.2611,
      "step": 680
    },
    {
      "epoch": 0.35844155844155845,
      "grad_norm": 0.2438523561922534,
      "learning_rate": 4.896021319214895e-05,
      "loss": 0.2601,
      "step": 690
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.2668670929832916,
      "learning_rate": 4.8919816406499584e-05,
      "loss": 0.2696,
      "step": 700
    },
    {
      "epoch": 0.36883116883116884,
      "grad_norm": 0.24937135360115686,
      "learning_rate": 4.887866708827633e-05,
      "loss": 0.2602,
      "step": 710
    },
    {
      "epoch": 0.37402597402597404,
      "grad_norm": 0.23999629542679116,
      "learning_rate": 4.8836766532062804e-05,
      "loss": 0.2588,
      "step": 720
    },
    {
      "epoch": 0.37922077922077924,
      "grad_norm": 0.23334349462758497,
      "learning_rate": 4.879411605607704e-05,
      "loss": 0.2606,
      "step": 730
    },
    {
      "epoch": 0.38441558441558443,
      "grad_norm": 0.2215145938323352,
      "learning_rate": 4.8750717002130024e-05,
      "loss": 0.2567,
      "step": 740
    },
    {
      "epoch": 0.38961038961038963,
      "grad_norm": 0.3218548674660387,
      "learning_rate": 4.870657073558349e-05,
      "loss": 0.2627,
      "step": 750
    },
    {
      "epoch": 0.3948051948051948,
      "grad_norm": 0.29152150655446074,
      "learning_rate": 4.866167864530693e-05,
      "loss": 0.2561,
      "step": 760
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29823420885887736,
      "learning_rate": 4.8616042143633937e-05,
      "loss": 0.2594,
      "step": 770
    },
    {
      "epoch": 0.4051948051948052,
      "grad_norm": 0.256792339442467,
      "learning_rate": 4.856966266631777e-05,
      "loss": 0.2565,
      "step": 780
    },
    {
      "epoch": 0.4103896103896104,
      "grad_norm": 0.25804885977573755,
      "learning_rate": 4.8522541672486156e-05,
      "loss": 0.2577,
      "step": 790
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.2903609751193798,
      "learning_rate": 4.84746806445954e-05,
      "loss": 0.2534,
      "step": 800
    },
    {
      "epoch": 0.42077922077922075,
      "grad_norm": 0.26400859462593973,
      "learning_rate": 4.8426081088383756e-05,
      "loss": 0.2586,
      "step": 810
    },
    {
      "epoch": 0.42597402597402595,
      "grad_norm": 0.2960587838599708,
      "learning_rate": 4.837674453282404e-05,
      "loss": 0.261,
      "step": 820
    },
    {
      "epoch": 0.43116883116883115,
      "grad_norm": 0.23331561495605277,
      "learning_rate": 4.832667253007554e-05,
      "loss": 0.2536,
      "step": 830
    },
    {
      "epoch": 0.43636363636363634,
      "grad_norm": 0.24020811230350025,
      "learning_rate": 4.8275866655435175e-05,
      "loss": 0.2564,
      "step": 840
    },
    {
      "epoch": 0.44155844155844154,
      "grad_norm": 0.22023402453548904,
      "learning_rate": 4.8224328507287946e-05,
      "loss": 0.2562,
      "step": 850
    },
    {
      "epoch": 0.44675324675324674,
      "grad_norm": 0.293317498213313,
      "learning_rate": 4.8172059707056626e-05,
      "loss": 0.2565,
      "step": 860
    },
    {
      "epoch": 0.45194805194805193,
      "grad_norm": 0.2692215437341758,
      "learning_rate": 4.811906189915078e-05,
      "loss": 0.2506,
      "step": 870
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 0.23694698773474526,
      "learning_rate": 4.806533675091501e-05,
      "loss": 0.2518,
      "step": 880
    },
    {
      "epoch": 0.4623376623376623,
      "grad_norm": 0.22885916720084376,
      "learning_rate": 4.80108859525765e-05,
      "loss": 0.252,
      "step": 890
    },
    {
      "epoch": 0.4675324675324675,
      "grad_norm": 0.24916135308130166,
      "learning_rate": 4.795571121719187e-05,
      "loss": 0.253,
      "step": 900
    },
    {
      "epoch": 0.4727272727272727,
      "grad_norm": 0.24894984701102493,
      "learning_rate": 4.7899814280593226e-05,
      "loss": 0.2529,
      "step": 910
    },
    {
      "epoch": 0.4779220779220779,
      "grad_norm": 0.2723728137565129,
      "learning_rate": 4.78431969013336e-05,
      "loss": 0.2555,
      "step": 920
    },
    {
      "epoch": 0.4831168831168831,
      "grad_norm": 0.3183305552275493,
      "learning_rate": 4.778586086063159e-05,
      "loss": 0.2514,
      "step": 930
    },
    {
      "epoch": 0.4883116883116883,
      "grad_norm": 0.2414413013327865,
      "learning_rate": 4.772780796231537e-05,
      "loss": 0.2484,
      "step": 940
    },
    {
      "epoch": 0.4935064935064935,
      "grad_norm": 0.22563943539011178,
      "learning_rate": 4.766904003276589e-05,
      "loss": 0.2539,
      "step": 950
    },
    {
      "epoch": 0.4987012987012987,
      "grad_norm": 0.30205872999506944,
      "learning_rate": 4.760955892085942e-05,
      "loss": 0.2527,
      "step": 960
    },
    {
      "epoch": 0.5038961038961038,
      "grad_norm": 0.2860786126383834,
      "learning_rate": 4.754936649790942e-05,
      "loss": 0.2516,
      "step": 970
    },
    {
      "epoch": 0.509090909090909,
      "grad_norm": 0.27572406640999436,
      "learning_rate": 4.7488464657607635e-05,
      "loss": 0.2498,
      "step": 980
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 0.24673507061013106,
      "learning_rate": 4.7426855315964535e-05,
      "loss": 0.2531,
      "step": 990
    },
    {
      "epoch": 0.5194805194805194,
      "grad_norm": 0.21033978842271397,
      "learning_rate": 4.736454041124904e-05,
      "loss": 0.2504,
      "step": 1000
    },
    {
      "epoch": 0.5246753246753246,
      "grad_norm": 0.24381830272002009,
      "learning_rate": 4.7301521903927505e-05,
      "loss": 0.2428,
      "step": 1010
    },
    {
      "epoch": 0.5298701298701298,
      "grad_norm": 0.2334282560856222,
      "learning_rate": 4.723780177660209e-05,
      "loss": 0.2501,
      "step": 1020
    },
    {
      "epoch": 0.535064935064935,
      "grad_norm": 0.2751808654527514,
      "learning_rate": 4.717338203394836e-05,
      "loss": 0.2507,
      "step": 1030
    },
    {
      "epoch": 0.5402597402597402,
      "grad_norm": 0.2697637268340861,
      "learning_rate": 4.71082647026522e-05,
      "loss": 0.2503,
      "step": 1040
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 0.22403492548323756,
      "learning_rate": 4.7042451831346136e-05,
      "loss": 0.2495,
      "step": 1050
    },
    {
      "epoch": 0.5506493506493506,
      "grad_norm": 0.22425594786040917,
      "learning_rate": 4.697594549054474e-05,
      "loss": 0.2475,
      "step": 1060
    },
    {
      "epoch": 0.5558441558441558,
      "grad_norm": 0.24265650046282458,
      "learning_rate": 4.690874777257964e-05,
      "loss": 0.2491,
      "step": 1070
    },
    {
      "epoch": 0.561038961038961,
      "grad_norm": 0.22077426247794457,
      "learning_rate": 4.684086079153359e-05,
      "loss": 0.2449,
      "step": 1080
    },
    {
      "epoch": 0.5662337662337662,
      "grad_norm": 0.30484192484636535,
      "learning_rate": 4.6772286683174025e-05,
      "loss": 0.245,
      "step": 1090
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.22182045074526108,
      "learning_rate": 4.670302760488582e-05,
      "loss": 0.2477,
      "step": 1100
    },
    {
      "epoch": 0.5766233766233766,
      "grad_norm": 0.2200234503444333,
      "learning_rate": 4.663308573560343e-05,
      "loss": 0.2506,
      "step": 1110
    },
    {
      "epoch": 0.5818181818181818,
      "grad_norm": 0.21994287683589026,
      "learning_rate": 4.656246327574238e-05,
      "loss": 0.2421,
      "step": 1120
    },
    {
      "epoch": 0.587012987012987,
      "grad_norm": 0.272334351757034,
      "learning_rate": 4.649116244712998e-05,
      "loss": 0.2476,
      "step": 1130
    },
    {
      "epoch": 0.5922077922077922,
      "grad_norm": 0.2236806239622702,
      "learning_rate": 4.641918549293545e-05,
      "loss": 0.2454,
      "step": 1140
    },
    {
      "epoch": 0.5974025974025974,
      "grad_norm": 0.23546182291174625,
      "learning_rate": 4.634653467759936e-05,
      "loss": 0.2477,
      "step": 1150
    },
    {
      "epoch": 0.6025974025974026,
      "grad_norm": 0.2505748637769869,
      "learning_rate": 4.6273212286762376e-05,
      "loss": 0.2449,
      "step": 1160
    },
    {
      "epoch": 0.6077922077922078,
      "grad_norm": 0.18468403137918,
      "learning_rate": 4.619922062719335e-05,
      "loss": 0.2432,
      "step": 1170
    },
    {
      "epoch": 0.612987012987013,
      "grad_norm": 0.23235694779454488,
      "learning_rate": 4.6124562026716766e-05,
      "loss": 0.2457,
      "step": 1180
    },
    {
      "epoch": 0.6181818181818182,
      "grad_norm": 0.2587545121058708,
      "learning_rate": 4.604923883413946e-05,
      "loss": 0.2467,
      "step": 1190
    },
    {
      "epoch": 0.6233766233766234,
      "grad_norm": 0.23524218849591322,
      "learning_rate": 4.59732534191768e-05,
      "loss": 0.2425,
      "step": 1200
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 0.20482111091171828,
      "learning_rate": 4.589660817237805e-05,
      "loss": 0.2446,
      "step": 1210
    },
    {
      "epoch": 0.6337662337662338,
      "grad_norm": 0.23935891812153473,
      "learning_rate": 4.581930550505122e-05,
      "loss": 0.2359,
      "step": 1220
    },
    {
      "epoch": 0.638961038961039,
      "grad_norm": 0.2088944848253035,
      "learning_rate": 4.5741347849187186e-05,
      "loss": 0.2435,
      "step": 1230
    },
    {
      "epoch": 0.6441558441558441,
      "grad_norm": 0.19769218239953817,
      "learning_rate": 4.566273765738318e-05,
      "loss": 0.2429,
      "step": 1240
    },
    {
      "epoch": 0.6493506493506493,
      "grad_norm": 0.21104513975336958,
      "learning_rate": 4.558347740276562e-05,
      "loss": 0.2414,
      "step": 1250
    },
    {
      "epoch": 0.6545454545454545,
      "grad_norm": 0.22299730594468617,
      "learning_rate": 4.550356957891232e-05,
      "loss": 0.2405,
      "step": 1260
    },
    {
      "epoch": 0.6597402597402597,
      "grad_norm": 0.2553717349606562,
      "learning_rate": 4.5423016699774025e-05,
      "loss": 0.242,
      "step": 1270
    },
    {
      "epoch": 0.6649350649350649,
      "grad_norm": 0.184621184174687,
      "learning_rate": 4.5341821299595334e-05,
      "loss": 0.2377,
      "step": 1280
    },
    {
      "epoch": 0.6701298701298701,
      "grad_norm": 0.1931019421949112,
      "learning_rate": 4.525998593283496e-05,
      "loss": 0.2401,
      "step": 1290
    },
    {
      "epoch": 0.6753246753246753,
      "grad_norm": 0.2327800605150306,
      "learning_rate": 4.517751317408537e-05,
      "loss": 0.2405,
      "step": 1300
    },
    {
      "epoch": 0.6805194805194805,
      "grad_norm": 0.21060753158636902,
      "learning_rate": 4.5094405617991796e-05,
      "loss": 0.2363,
      "step": 1310
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.18780386461582757,
      "learning_rate": 4.501066587917058e-05,
      "loss": 0.2437,
      "step": 1320
    },
    {
      "epoch": 0.6909090909090909,
      "grad_norm": 0.2671736773226871,
      "learning_rate": 4.4926296592126946e-05,
      "loss": 0.2431,
      "step": 1330
    },
    {
      "epoch": 0.6961038961038961,
      "grad_norm": 0.23810395896846592,
      "learning_rate": 4.484130041117211e-05,
      "loss": 0.243,
      "step": 1340
    },
    {
      "epoch": 0.7012987012987013,
      "grad_norm": 0.18025186403432847,
      "learning_rate": 4.475568001033974e-05,
      "loss": 0.2457,
      "step": 1350
    },
    {
      "epoch": 0.7064935064935065,
      "grad_norm": 0.21115305940327297,
      "learning_rate": 4.466943808330189e-05,
      "loss": 0.2415,
      "step": 1360
    },
    {
      "epoch": 0.7116883116883117,
      "grad_norm": 0.25256979094205834,
      "learning_rate": 4.45825773432842e-05,
      "loss": 0.2407,
      "step": 1370
    },
    {
      "epoch": 0.7168831168831169,
      "grad_norm": 0.22014008453128092,
      "learning_rate": 4.449510052298056e-05,
      "loss": 0.2357,
      "step": 1380
    },
    {
      "epoch": 0.7220779220779221,
      "grad_norm": 0.20062628753000003,
      "learning_rate": 4.440701037446714e-05,
      "loss": 0.2396,
      "step": 1390
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.2083485579826855,
      "learning_rate": 4.431830966911582e-05,
      "loss": 0.2391,
      "step": 1400
    },
    {
      "epoch": 0.7324675324675325,
      "grad_norm": 0.18184305857175756,
      "learning_rate": 4.422900119750695e-05,
      "loss": 0.2355,
      "step": 1410
    },
    {
      "epoch": 0.7376623376623377,
      "grad_norm": 0.2048232593631597,
      "learning_rate": 4.4139087769341625e-05,
      "loss": 0.2332,
      "step": 1420
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 0.21132211989073768,
      "learning_rate": 4.4048572213353234e-05,
      "loss": 0.2422,
      "step": 1430
    },
    {
      "epoch": 0.7480519480519481,
      "grad_norm": 0.19792895604344352,
      "learning_rate": 4.39574573772185e-05,
      "loss": 0.2334,
      "step": 1440
    },
    {
      "epoch": 0.7532467532467533,
      "grad_norm": 0.22178836320367148,
      "learning_rate": 4.3865746127467876e-05,
      "loss": 0.2423,
      "step": 1450
    },
    {
      "epoch": 0.7584415584415585,
      "grad_norm": 0.21967650568135474,
      "learning_rate": 4.3773441349395374e-05,
      "loss": 0.2357,
      "step": 1460
    },
    {
      "epoch": 0.7636363636363637,
      "grad_norm": 0.1917556477695145,
      "learning_rate": 4.368054594696775e-05,
      "loss": 0.2443,
      "step": 1470
    },
    {
      "epoch": 0.7688311688311689,
      "grad_norm": 0.20969861600848638,
      "learning_rate": 4.3587062842733216e-05,
      "loss": 0.2341,
      "step": 1480
    },
    {
      "epoch": 0.7740259740259741,
      "grad_norm": 0.19871375239851857,
      "learning_rate": 4.349299497772945e-05,
      "loss": 0.2361,
      "step": 1490
    },
    {
      "epoch": 0.7792207792207793,
      "grad_norm": 0.19082750332598916,
      "learning_rate": 4.339834531139104e-05,
      "loss": 0.2316,
      "step": 1500
    },
    {
      "epoch": 0.7844155844155845,
      "grad_norm": 0.2177029161255871,
      "learning_rate": 4.330311682145645e-05,
      "loss": 0.2343,
      "step": 1510
    },
    {
      "epoch": 0.7896103896103897,
      "grad_norm": 0.20562958726540304,
      "learning_rate": 4.320731250387429e-05,
      "loss": 0.2401,
      "step": 1520
    },
    {
      "epoch": 0.7948051948051948,
      "grad_norm": 0.2070237852219627,
      "learning_rate": 4.311093537270905e-05,
      "loss": 0.2374,
      "step": 1530
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.18967395002327114,
      "learning_rate": 4.301398846004634e-05,
      "loss": 0.2363,
      "step": 1540
    },
    {
      "epoch": 0.8051948051948052,
      "grad_norm": 0.1970271386066234,
      "learning_rate": 4.291647481589742e-05,
      "loss": 0.2302,
      "step": 1550
    },
    {
      "epoch": 0.8103896103896104,
      "grad_norm": 0.18673676151020974,
      "learning_rate": 4.28183975081033e-05,
      "loss": 0.2416,
      "step": 1560
    },
    {
      "epoch": 0.8155844155844156,
      "grad_norm": 0.2183111540642943,
      "learning_rate": 4.271975962223821e-05,
      "loss": 0.2342,
      "step": 1570
    },
    {
      "epoch": 0.8207792207792208,
      "grad_norm": 0.1792298886397136,
      "learning_rate": 4.2620564261512496e-05,
      "loss": 0.2388,
      "step": 1580
    },
    {
      "epoch": 0.825974025974026,
      "grad_norm": 0.21429193275126804,
      "learning_rate": 4.2520814546675037e-05,
      "loss": 0.2323,
      "step": 1590
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 0.1923357673969473,
      "learning_rate": 4.242051361591505e-05,
      "loss": 0.2398,
      "step": 1600
    },
    {
      "epoch": 0.8363636363636363,
      "grad_norm": 0.1825902322292911,
      "learning_rate": 4.2319664624763325e-05,
      "loss": 0.2355,
      "step": 1610
    },
    {
      "epoch": 0.8415584415584415,
      "grad_norm": 0.1708452665847616,
      "learning_rate": 4.2218270745993016e-05,
      "loss": 0.2361,
      "step": 1620
    },
    {
      "epoch": 0.8467532467532467,
      "grad_norm": 0.2003983431936864,
      "learning_rate": 4.211633516951975e-05,
      "loss": 0.237,
      "step": 1630
    },
    {
      "epoch": 0.8519480519480519,
      "grad_norm": 0.1809948763155965,
      "learning_rate": 4.201386110230134e-05,
      "loss": 0.2291,
      "step": 1640
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.19621979591943875,
      "learning_rate": 4.1910851768236825e-05,
      "loss": 0.2284,
      "step": 1650
    },
    {
      "epoch": 0.8623376623376623,
      "grad_norm": 0.20784597945629102,
      "learning_rate": 4.180731040806511e-05,
      "loss": 0.2359,
      "step": 1660
    },
    {
      "epoch": 0.8675324675324675,
      "grad_norm": 0.22581034014160772,
      "learning_rate": 4.170324027926297e-05,
      "loss": 0.2329,
      "step": 1670
    },
    {
      "epoch": 0.8727272727272727,
      "grad_norm": 0.1789163119753752,
      "learning_rate": 4.159864465594255e-05,
      "loss": 0.2338,
      "step": 1680
    },
    {
      "epoch": 0.8779220779220779,
      "grad_norm": 0.1949206924337472,
      "learning_rate": 4.1493526828748416e-05,
      "loss": 0.2392,
      "step": 1690
    },
    {
      "epoch": 0.8831168831168831,
      "grad_norm": 0.20147429000086556,
      "learning_rate": 4.1387890104754004e-05,
      "loss": 0.233,
      "step": 1700
    },
    {
      "epoch": 0.8883116883116883,
      "grad_norm": 0.1537005161376695,
      "learning_rate": 4.128173780735753e-05,
      "loss": 0.2291,
      "step": 1710
    },
    {
      "epoch": 0.8935064935064935,
      "grad_norm": 0.17777763693741433,
      "learning_rate": 4.117507327617751e-05,
      "loss": 0.2291,
      "step": 1720
    },
    {
      "epoch": 0.8987012987012987,
      "grad_norm": 0.174198062693491,
      "learning_rate": 4.1067899866947665e-05,
      "loss": 0.2294,
      "step": 1730
    },
    {
      "epoch": 0.9038961038961039,
      "grad_norm": 0.1884364748511166,
      "learning_rate": 4.096022095141132e-05,
      "loss": 0.235,
      "step": 1740
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.1912652069094164,
      "learning_rate": 4.085203991721535e-05,
      "loss": 0.2318,
      "step": 1750
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.206558739242339,
      "learning_rate": 4.0743360167803614e-05,
      "loss": 0.2317,
      "step": 1760
    },
    {
      "epoch": 0.9194805194805195,
      "grad_norm": 0.18252166114267931,
      "learning_rate": 4.063418512230987e-05,
      "loss": 0.2346,
      "step": 1770
    },
    {
      "epoch": 0.9246753246753247,
      "grad_norm": 0.18463778266166328,
      "learning_rate": 4.0524518215450166e-05,
      "loss": 0.2306,
      "step": 1780
    },
    {
      "epoch": 0.9298701298701298,
      "grad_norm": 0.207923278938462,
      "learning_rate": 4.041436289741489e-05,
      "loss": 0.2301,
      "step": 1790
    },
    {
      "epoch": 0.935064935064935,
      "grad_norm": 0.25335695776490813,
      "learning_rate": 4.0303722633760085e-05,
      "loss": 0.2258,
      "step": 1800
    },
    {
      "epoch": 0.9402597402597402,
      "grad_norm": 0.16805426564943104,
      "learning_rate": 4.019260090529854e-05,
      "loss": 0.2284,
      "step": 1810
    },
    {
      "epoch": 0.9454545454545454,
      "grad_norm": 0.1884837989936669,
      "learning_rate": 4.008100120799019e-05,
      "loss": 0.2285,
      "step": 1820
    },
    {
      "epoch": 0.9506493506493506,
      "grad_norm": 0.19643081968195814,
      "learning_rate": 3.996892705283222e-05,
      "loss": 0.2354,
      "step": 1830
    },
    {
      "epoch": 0.9558441558441558,
      "grad_norm": 0.18949399067442121,
      "learning_rate": 3.9856381965748506e-05,
      "loss": 0.234,
      "step": 1840
    },
    {
      "epoch": 0.961038961038961,
      "grad_norm": 0.18422150470243814,
      "learning_rate": 3.974336948747879e-05,
      "loss": 0.228,
      "step": 1850
    },
    {
      "epoch": 0.9662337662337662,
      "grad_norm": 0.179837212988977,
      "learning_rate": 3.962989317346722e-05,
      "loss": 0.2283,
      "step": 1860
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 0.18190735457463206,
      "learning_rate": 3.951595659375048e-05,
      "loss": 0.2337,
      "step": 1870
    },
    {
      "epoch": 0.9766233766233766,
      "grad_norm": 0.17903757429753223,
      "learning_rate": 3.9401563332845545e-05,
      "loss": 0.2225,
      "step": 1880
    },
    {
      "epoch": 0.9818181818181818,
      "grad_norm": 0.18775008592274955,
      "learning_rate": 3.928671698963686e-05,
      "loss": 0.226,
      "step": 1890
    },
    {
      "epoch": 0.987012987012987,
      "grad_norm": 0.16101556678112905,
      "learning_rate": 3.917142117726312e-05,
      "loss": 0.2312,
      "step": 1900
    },
    {
      "epoch": 0.9922077922077922,
      "grad_norm": 0.18133929144854413,
      "learning_rate": 3.90556795230036e-05,
      "loss": 0.2264,
      "step": 1910
    },
    {
      "epoch": 0.9974025974025974,
      "grad_norm": 0.21238218367672568,
      "learning_rate": 3.893949566816404e-05,
      "loss": 0.2273,
      "step": 1920
    }
  ],
  "logging_steps": 10,
  "max_steps": 5775,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 6056617202155520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}