{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5535,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005420054200542005,
      "grad_norm": 0.0,
      "learning_rate": 1.8050541516245488e-06,
      "loss": 1.3502,
      "step": 10
    },
    {
      "epoch": 0.01084010840108401,
      "grad_norm": 0.0,
      "learning_rate": 3.6101083032490977e-06,
      "loss": 1.3193,
      "step": 20
    },
    {
      "epoch": 0.016260162601626018,
      "grad_norm": 0.0,
      "learning_rate": 5.4151624548736465e-06,
      "loss": 1.3226,
      "step": 30
    },
    {
      "epoch": 0.02168021680216802,
      "grad_norm": 0.0,
      "learning_rate": 7.220216606498195e-06,
      "loss": 1.2298,
      "step": 40
    },
    {
      "epoch": 0.02710027100271003,
      "grad_norm": 0.0,
      "learning_rate": 9.025270758122744e-06,
      "loss": 1.1955,
      "step": 50
    },
    {
      "epoch": 0.032520325203252036,
      "grad_norm": 0.0,
      "learning_rate": 1.0830324909747293e-05,
      "loss": 1.3366,
      "step": 60
    },
    {
      "epoch": 0.037940379403794036,
      "grad_norm": 0.0,
      "learning_rate": 1.263537906137184e-05,
      "loss": 1.2038,
      "step": 70
    },
    {
      "epoch": 0.04336043360433604,
      "grad_norm": 0.0,
      "learning_rate": 1.444043321299639e-05,
      "loss": 1.2876,
      "step": 80
    },
    {
      "epoch": 0.04878048780487805,
      "grad_norm": 0.0,
      "learning_rate": 1.624548736462094e-05,
      "loss": 1.0663,
      "step": 90
    },
    {
      "epoch": 0.05420054200542006,
      "grad_norm": 0.0,
      "learning_rate": 1.805054151624549e-05,
      "loss": 1.0746,
      "step": 100
    },
    {
      "epoch": 0.05962059620596206,
      "grad_norm": 0.0,
      "learning_rate": 1.9855595667870036e-05,
      "loss": 1.1507,
      "step": 110
    },
    {
      "epoch": 0.06504065040650407,
      "grad_norm": 0.0,
      "learning_rate": 2.1660649819494586e-05,
      "loss": 1.108,
      "step": 120
    },
    {
      "epoch": 0.07046070460704607,
      "grad_norm": 0.0,
      "learning_rate": 2.3465703971119137e-05,
      "loss": 1.0929,
      "step": 130
    },
    {
      "epoch": 0.07588075880758807,
      "grad_norm": 0.0,
      "learning_rate": 2.527075812274368e-05,
      "loss": 1.0056,
      "step": 140
    },
    {
      "epoch": 0.08130081300813008,
      "grad_norm": 0.0,
      "learning_rate": 2.7075812274368234e-05,
      "loss": 1.2323,
      "step": 150
    },
    {
      "epoch": 0.08672086720867209,
      "grad_norm": 0.0,
      "learning_rate": 2.888086642599278e-05,
      "loss": 1.1624,
      "step": 160
    },
    {
      "epoch": 0.0921409214092141,
      "grad_norm": 0.0,
      "learning_rate": 3.0685920577617325e-05,
      "loss": 1.1477,
      "step": 170
    },
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 0.0,
      "learning_rate": 3.249097472924188e-05,
      "loss": 1.1886,
      "step": 180
    },
    {
      "epoch": 0.10298102981029811,
      "grad_norm": 0.0,
      "learning_rate": 3.4296028880866426e-05,
      "loss": 1.0738,
      "step": 190
    },
    {
      "epoch": 0.10840108401084012,
      "grad_norm": 0.0,
      "learning_rate": 3.610108303249098e-05,
      "loss": 1.103,
      "step": 200
    },
    {
      "epoch": 0.11382113821138211,
      "grad_norm": 0.0,
      "learning_rate": 3.790613718411553e-05,
      "loss": 1.25,
      "step": 210
    },
    {
      "epoch": 0.11924119241192412,
      "grad_norm": 0.0,
      "learning_rate": 3.971119133574007e-05,
      "loss": 1.188,
      "step": 220
    },
    {
      "epoch": 0.12466124661246612,
      "grad_norm": 0.0,
      "learning_rate": 4.151624548736462e-05,
      "loss": 1.11,
      "step": 230
    },
    {
      "epoch": 0.13008130081300814,
      "grad_norm": 0.0,
      "learning_rate": 4.332129963898917e-05,
      "loss": 1.0991,
      "step": 240
    },
    {
      "epoch": 0.13550135501355012,
      "grad_norm": 0.0,
      "learning_rate": 4.5126353790613716e-05,
      "loss": 1.07,
      "step": 250
    },
    {
      "epoch": 0.14092140921409213,
      "grad_norm": 0.0,
      "learning_rate": 4.693140794223827e-05,
      "loss": 1.2724,
      "step": 260
    },
    {
      "epoch": 0.14634146341463414,
      "grad_norm": 0.0,
      "learning_rate": 4.873646209386282e-05,
      "loss": 1.1382,
      "step": 270
    },
    {
      "epoch": 0.15176151761517614,
      "grad_norm": 0.0,
      "learning_rate": 5.054151624548736e-05,
      "loss": 1.1774,
      "step": 280
    },
    {
      "epoch": 0.15718157181571815,
      "grad_norm": 0.0,
      "learning_rate": 5.234657039711192e-05,
      "loss": 1.1395,
      "step": 290
    },
    {
      "epoch": 0.16260162601626016,
      "grad_norm": 0.0,
      "learning_rate": 5.415162454873647e-05,
      "loss": 1.192,
      "step": 300
    },
    {
      "epoch": 0.16802168021680217,
      "grad_norm": 0.0,
      "learning_rate": 5.595667870036101e-05,
      "loss": 1.1848,
      "step": 310
    },
    {
      "epoch": 0.17344173441734417,
      "grad_norm": 0.0,
      "learning_rate": 5.776173285198556e-05,
      "loss": 1.1632,
      "step": 320
    },
    {
      "epoch": 0.17886178861788618,
      "grad_norm": 0.0,
      "learning_rate": 5.9566787003610113e-05,
      "loss": 1.1659,
      "step": 330
    },
    {
      "epoch": 0.1842818428184282,
      "grad_norm": 0.0,
      "learning_rate": 6.137184115523465e-05,
      "loss": 1.0097,
      "step": 340
    },
    {
      "epoch": 0.1897018970189702,
      "grad_norm": 0.0,
      "learning_rate": 6.31768953068592e-05,
      "loss": 1.068,
      "step": 350
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.0,
      "learning_rate": 6.498194945848377e-05,
      "loss": 1.1652,
      "step": 360
    },
    {
      "epoch": 0.2005420054200542,
      "grad_norm": 0.0,
      "learning_rate": 6.678700361010832e-05,
      "loss": 1.0985,
      "step": 370
    },
    {
      "epoch": 0.20596205962059622,
      "grad_norm": 0.0,
      "learning_rate": 6.859205776173285e-05,
      "loss": 1.0488,
      "step": 380
    },
    {
      "epoch": 0.21138211382113822,
      "grad_norm": 0.0,
      "learning_rate": 7.03971119133574e-05,
      "loss": 1.1854,
      "step": 390
    },
    {
      "epoch": 0.21680216802168023,
      "grad_norm": 0.0,
      "learning_rate": 7.220216606498195e-05,
      "loss": 1.1238,
      "step": 400
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.0,
      "learning_rate": 7.40072202166065e-05,
      "loss": 1.145,
      "step": 410
    },
    {
      "epoch": 0.22764227642276422,
      "grad_norm": 0.0,
      "learning_rate": 7.581227436823105e-05,
      "loss": 1.2215,
      "step": 420
    },
    {
      "epoch": 0.23306233062330622,
      "grad_norm": 0.0,
      "learning_rate": 7.76173285198556e-05,
      "loss": 1.143,
      "step": 430
    },
    {
      "epoch": 0.23848238482384823,
      "grad_norm": 0.0,
      "learning_rate": 7.942238267148014e-05,
      "loss": 1.1652,
      "step": 440
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.0,
      "learning_rate": 8.122743682310469e-05,
      "loss": 1.2227,
      "step": 450
    },
    {
      "epoch": 0.24932249322493225,
      "grad_norm": 0.0,
      "learning_rate": 8.303249097472924e-05,
      "loss": 1.1247,
      "step": 460
    },
    {
      "epoch": 0.25474254742547425,
      "grad_norm": 0.0,
      "learning_rate": 8.48375451263538e-05,
      "loss": 1.1026,
      "step": 470
    },
    {
      "epoch": 0.2601626016260163,
      "grad_norm": 0.0,
      "learning_rate": 8.664259927797834e-05,
      "loss": 1.0836,
      "step": 480
    },
    {
      "epoch": 0.26558265582655827,
      "grad_norm": 0.0,
      "learning_rate": 8.84476534296029e-05,
      "loss": 1.0682,
      "step": 490
    },
    {
      "epoch": 0.27100271002710025,
      "grad_norm": 0.0,
      "learning_rate": 9.025270758122743e-05,
      "loss": 1.0829,
      "step": 500
    },
    {
      "epoch": 0.2764227642276423,
      "grad_norm": 0.0,
      "learning_rate": 9.205776173285198e-05,
      "loss": 1.0599,
      "step": 510
    },
    {
      "epoch": 0.28184281842818426,
      "grad_norm": 0.0,
      "learning_rate": 9.386281588447655e-05,
      "loss": 1.1322,
      "step": 520
    },
    {
      "epoch": 0.2872628726287263,
      "grad_norm": 0.0,
      "learning_rate": 9.56678700361011e-05,
      "loss": 1.1264,
      "step": 530
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 0.0,
      "learning_rate": 9.747292418772563e-05,
      "loss": 1.1008,
      "step": 540
    },
    {
      "epoch": 0.2981029810298103,
      "grad_norm": 0.0,
      "learning_rate": 9.927797833935018e-05,
      "loss": 1.1783,
      "step": 550
    },
    {
      "epoch": 0.3035230352303523,
      "grad_norm": 0.0,
      "learning_rate": 9.999964197887486e-05,
      "loss": 1.1658,
      "step": 560
    },
    {
      "epoch": 0.3089430894308943,
      "grad_norm": 0.0,
      "learning_rate": 9.999745409056648e-05,
      "loss": 1.1876,
      "step": 570
    },
    {
      "epoch": 0.3143631436314363,
      "grad_norm": 0.0,
      "learning_rate": 9.999327730150358e-05,
      "loss": 1.0611,
      "step": 580
    },
    {
      "epoch": 0.31978319783197834,
      "grad_norm": 0.0,
      "learning_rate": 9.998711177783901e-05,
      "loss": 1.0863,
      "step": 590
    },
    {
      "epoch": 0.3252032520325203,
      "grad_norm": 0.0,
      "learning_rate": 9.997895776483754e-05,
      "loss": 1.114,
      "step": 600
    },
    {
      "epoch": 0.33062330623306235,
      "grad_norm": 0.0,
      "learning_rate": 9.996881558686615e-05,
      "loss": 1.0857,
      "step": 610
    },
    {
      "epoch": 0.33604336043360433,
      "grad_norm": 0.0,
      "learning_rate": 9.995668564738108e-05,
      "loss": 1.0991,
      "step": 620
    },
    {
      "epoch": 0.34146341463414637,
      "grad_norm": 0.0,
      "learning_rate": 9.994256842891181e-05,
      "loss": 1.0993,
      "step": 630
    },
    {
      "epoch": 0.34688346883468835,
      "grad_norm": 0.0,
      "learning_rate": 9.992646449304188e-05,
      "loss": 1.0537,
      "step": 640
    },
    {
      "epoch": 0.3523035230352303,
      "grad_norm": 0.0,
      "learning_rate": 9.990837448038653e-05,
      "loss": 1.0829,
      "step": 650
    },
    {
      "epoch": 0.35772357723577236,
      "grad_norm": 0.0,
      "learning_rate": 9.988829911056716e-05,
      "loss": 1.0441,
      "step": 660
    },
    {
      "epoch": 0.36314363143631434,
      "grad_norm": 0.0,
      "learning_rate": 9.98662391821828e-05,
      "loss": 1.0902,
      "step": 670
    },
    {
      "epoch": 0.3685636856368564,
      "grad_norm": 0.0,
      "learning_rate": 9.984219557277832e-05,
      "loss": 1.1888,
      "step": 680
    },
    {
      "epoch": 0.37398373983739835,
      "grad_norm": 0.0,
      "learning_rate": 9.981616923880947e-05,
      "loss": 1.0265,
      "step": 690
    },
    {
      "epoch": 0.3794037940379404,
      "grad_norm": 0.0,
      "learning_rate": 9.978816121560487e-05,
      "loss": 1.1118,
      "step": 700
    },
    {
      "epoch": 0.38482384823848237,
      "grad_norm": 0.0,
      "learning_rate": 9.975817261732483e-05,
      "loss": 1.0043,
      "step": 710
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 0.0,
      "learning_rate": 9.972620463691699e-05,
      "loss": 1.0356,
      "step": 720
    },
    {
      "epoch": 0.3956639566395664,
      "grad_norm": 0.0,
      "learning_rate": 9.969225854606892e-05,
      "loss": 1.1946,
      "step": 730
    },
    {
      "epoch": 0.4010840108401084,
      "grad_norm": 0.0,
      "learning_rate": 9.965633569515749e-05,
      "loss": 1.0684,
      "step": 740
    },
    {
      "epoch": 0.4065040650406504,
      "grad_norm": 0.0,
      "learning_rate": 9.961843751319517e-05,
      "loss": 1.1165,
      "step": 750
    },
    {
      "epoch": 0.41192411924119243,
      "grad_norm": 0.0,
      "learning_rate": 9.957856550777318e-05,
      "loss": 1.106,
      "step": 760
    },
    {
      "epoch": 0.4173441734417344,
      "grad_norm": 0.0,
      "learning_rate": 9.95367212650015e-05,
      "loss": 1.0912,
      "step": 770
    },
    {
      "epoch": 0.42276422764227645,
      "grad_norm": 0.0,
      "learning_rate": 9.949290644944585e-05,
      "loss": 1.0344,
      "step": 780
    },
    {
      "epoch": 0.4281842818428184,
      "grad_norm": 0.0,
      "learning_rate": 9.944712280406132e-05,
      "loss": 1.148,
      "step": 790
    },
    {
      "epoch": 0.43360433604336046,
      "grad_norm": 0.0,
      "learning_rate": 9.93993721501232e-05,
      "loss": 1.0802,
      "step": 800
    },
    {
      "epoch": 0.43902439024390244,
      "grad_norm": 0.0,
      "learning_rate": 9.934965638715446e-05,
      "loss": 1.0527,
      "step": 810
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.0,
      "learning_rate": 9.929797749285015e-05,
      "loss": 1.0383,
      "step": 820
    },
    {
      "epoch": 0.44986449864498645,
      "grad_norm": 0.0,
      "learning_rate": 9.92443375229988e-05,
      "loss": 1.0799,
      "step": 830
    },
    {
      "epoch": 0.45528455284552843,
      "grad_norm": 0.0,
      "learning_rate": 9.918873861140058e-05,
      "loss": 1.0581,
      "step": 840
    },
    {
      "epoch": 0.46070460704607047,
      "grad_norm": 0.0,
      "learning_rate": 9.913118296978246e-05,
      "loss": 1.0526,
      "step": 850
    },
    {
      "epoch": 0.46612466124661245,
      "grad_norm": 0.0,
      "learning_rate": 9.907167288771019e-05,
      "loss": 1.1422,
      "step": 860
    },
    {
      "epoch": 0.4715447154471545,
      "grad_norm": 0.0,
      "learning_rate": 9.901021073249725e-05,
      "loss": 1.0433,
      "step": 870
    },
    {
      "epoch": 0.47696476964769646,
      "grad_norm": 0.0,
      "learning_rate": 9.89467989491107e-05,
      "loss": 1.1253,
      "step": 880
    },
    {
      "epoch": 0.4823848238482385,
      "grad_norm": 0.0,
      "learning_rate": 9.888144006007382e-05,
      "loss": 1.0904,
      "step": 890
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.0,
      "learning_rate": 9.881413666536589e-05,
      "loss": 1.0581,
      "step": 900
    },
    {
      "epoch": 0.4932249322493225,
      "grad_norm": 0.0,
      "learning_rate": 9.874489144231869e-05,
      "loss": 1.0418,
      "step": 910
    },
    {
      "epoch": 0.4986449864498645,
      "grad_norm": 0.0,
      "learning_rate": 9.867370714551e-05,
      "loss": 1.191,
      "step": 920
    },
    {
      "epoch": 0.5040650406504065,
      "grad_norm": 0.0,
      "learning_rate": 9.860058660665404e-05,
      "loss": 1.0701,
      "step": 930
    },
    {
      "epoch": 0.5094850948509485,
      "grad_norm": 0.0,
      "learning_rate": 9.852553273448879e-05,
      "loss": 1.1752,
      "step": 940
    },
    {
      "epoch": 0.5149051490514905,
      "grad_norm": 0.0,
      "learning_rate": 9.844854851466037e-05,
      "loss": 1.0127,
      "step": 950
    },
    {
      "epoch": 0.5203252032520326,
      "grad_norm": 0.0,
      "learning_rate": 9.836963700960416e-05,
      "loss": 1.1023,
      "step": 960
    },
    {
      "epoch": 0.5257452574525745,
      "grad_norm": 0.0,
      "learning_rate": 9.828880135842302e-05,
      "loss": 1.1257,
      "step": 970
    },
    {
      "epoch": 0.5311653116531165,
      "grad_norm": 0.0,
      "learning_rate": 9.820604477676247e-05,
      "loss": 1.0576,
      "step": 980
    },
    {
      "epoch": 0.5365853658536586,
      "grad_norm": 0.0,
      "learning_rate": 9.812137055668267e-05,
      "loss": 1.0565,
      "step": 990
    },
    {
      "epoch": 0.5420054200542005,
      "grad_norm": 0.0,
      "learning_rate": 9.803478206652752e-05,
      "loss": 1.0173,
      "step": 1000
    },
    {
      "epoch": 0.5474254742547425,
      "grad_norm": 0.0,
      "learning_rate": 9.794628275079069e-05,
      "loss": 1.0592,
      "step": 1010
    },
    {
      "epoch": 0.5528455284552846,
      "grad_norm": 0.0,
      "learning_rate": 9.785587612997857e-05,
      "loss": 0.9735,
      "step": 1020
    },
    {
      "epoch": 0.5582655826558266,
      "grad_norm": 0.0,
      "learning_rate": 9.776356580047019e-05,
      "loss": 0.9913,
      "step": 1030
    },
    {
      "epoch": 0.5636856368563685,
      "grad_norm": 0.0,
      "learning_rate": 9.766935543437421e-05,
      "loss": 1.0708,
      "step": 1040
    },
    {
      "epoch": 0.5691056910569106,
      "grad_norm": 0.0,
      "learning_rate": 9.757324877938283e-05,
      "loss": 1.1463,
      "step": 1050
    },
    {
      "epoch": 0.5745257452574526,
      "grad_norm": 0.0,
      "learning_rate": 9.747524965862267e-05,
      "loss": 1.0462,
      "step": 1060
    },
    {
      "epoch": 0.5799457994579946,
      "grad_norm": 0.0,
      "learning_rate": 9.737536197050275e-05,
      "loss": 1.0178,
      "step": 1070
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 0.0,
      "learning_rate": 9.727358968855932e-05,
      "loss": 0.8617,
      "step": 1080
    },
    {
      "epoch": 0.5907859078590786,
      "grad_norm": 0.0,
      "learning_rate": 9.716993686129796e-05,
      "loss": 0.9766,
      "step": 1090
    },
    {
      "epoch": 0.5962059620596206,
      "grad_norm": 0.0,
      "learning_rate": 9.706440761203232e-05,
      "loss": 1.0065,
      "step": 1100
    },
    {
      "epoch": 0.6016260162601627,
      "grad_norm": 0.0,
      "learning_rate": 9.695700613872019e-05,
      "loss": 0.8501,
      "step": 1110
    },
    {
      "epoch": 0.6070460704607046,
      "grad_norm": 0.0,
      "learning_rate": 9.684773671379658e-05,
      "loss": 1.1225,
      "step": 1120
    },
    {
      "epoch": 0.6124661246612466,
      "grad_norm": 0.0,
      "learning_rate": 9.673660368400364e-05,
      "loss": 1.0852,
      "step": 1130
    },
    {
      "epoch": 0.6178861788617886,
      "grad_norm": 0.0,
      "learning_rate": 9.662361147021779e-05,
      "loss": 0.9858,
      "step": 1140
    },
    {
      "epoch": 0.6233062330623306,
      "grad_norm": 0.0,
      "learning_rate": 9.650876456727392e-05,
      "loss": 1.0293,
      "step": 1150
    },
    {
      "epoch": 0.6287262872628726,
      "grad_norm": 0.0,
      "learning_rate": 9.639206754378646e-05,
      "loss": 1.06,
      "step": 1160
    },
    {
      "epoch": 0.6341463414634146,
      "grad_norm": 0.0,
      "learning_rate": 9.627352504196777e-05,
      "loss": 1.1035,
      "step": 1170
    },
    {
      "epoch": 0.6395663956639567,
      "grad_norm": 0.0,
      "learning_rate": 9.615314177744337e-05,
      "loss": 1.0686,
      "step": 1180
    },
    {
      "epoch": 0.6449864498644986,
      "grad_norm": 0.0,
      "learning_rate": 9.603092253906441e-05,
      "loss": 0.9771,
      "step": 1190
    },
    {
      "epoch": 0.6504065040650406,
      "grad_norm": 0.0,
      "learning_rate": 9.590687218871719e-05,
      "loss": 1.0134,
      "step": 1200
    },
    {
      "epoch": 0.6558265582655827,
      "grad_norm": 0.0,
      "learning_rate": 9.578099566112964e-05,
      "loss": 1.013,
      "step": 1210
    },
    {
      "epoch": 0.6612466124661247,
      "grad_norm": 0.0,
      "learning_rate": 9.565329796367516e-05,
      "loss": 1.012,
      "step": 1220
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.0,
      "learning_rate": 9.552378417617336e-05,
      "loss": 0.9936,
      "step": 1230
    },
    {
      "epoch": 0.6720867208672087,
      "grad_norm": 0.0,
      "learning_rate": 9.539245945068794e-05,
      "loss": 0.9983,
      "step": 1240
    },
    {
      "epoch": 0.6775067750677507,
      "grad_norm": 0.0,
      "learning_rate": 9.525932901132183e-05,
      "loss": 0.989,
      "step": 1250
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 0.0,
      "learning_rate": 9.512439815400931e-05,
      "loss": 1.0803,
      "step": 1260
    },
    {
      "epoch": 0.6883468834688347,
      "grad_norm": 0.0,
      "learning_rate": 9.498767224630537e-05,
      "loss": 1.0493,
      "step": 1270
    },
    {
      "epoch": 0.6937669376693767,
      "grad_norm": 0.0,
      "learning_rate": 9.484915672717213e-05,
      "loss": 1.0222,
      "step": 1280
    },
    {
      "epoch": 0.6991869918699187,
      "grad_norm": 0.0,
      "learning_rate": 9.470885710676265e-05,
      "loss": 0.9564,
      "step": 1290
    },
    {
      "epoch": 0.7046070460704607,
      "grad_norm": 0.0,
      "learning_rate": 9.456677896620145e-05,
      "loss": 0.9883,
      "step": 1300
    },
    {
      "epoch": 0.7100271002710027,
      "grad_norm": 0.0,
      "learning_rate": 9.442292795736278e-05,
      "loss": 1.0013,
      "step": 1310
    },
    {
      "epoch": 0.7154471544715447,
      "grad_norm": 0.0,
      "learning_rate": 9.427730980264559e-05,
      "loss": 0.9472,
      "step": 1320
    },
    {
      "epoch": 0.7208672086720868,
      "grad_norm": 0.0,
      "learning_rate": 9.412993029474602e-05,
      "loss": 1.0772,
      "step": 1330
    },
    {
      "epoch": 0.7262872628726287,
      "grad_norm": 0.0,
      "learning_rate": 9.398079529642685e-05,
      "loss": 1.0644,
      "step": 1340
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.0,
      "learning_rate": 9.382991074028436e-05,
      "loss": 1.1249,
      "step": 1350
    },
    {
      "epoch": 0.7371273712737128,
      "grad_norm": 0.0,
      "learning_rate": 9.367728262851237e-05,
      "loss": 0.9928,
      "step": 1360
    },
    {
      "epoch": 0.7425474254742548,
      "grad_norm": 0.0,
      "learning_rate": 9.352291703266331e-05,
      "loss": 1.1657,
      "step": 1370
    },
    {
      "epoch": 0.7479674796747967,
      "grad_norm": 0.0,
      "learning_rate": 9.33668200934069e-05,
      "loss": 1.0173,
      "step": 1380
    },
    {
      "epoch": 0.7533875338753387,
      "grad_norm": 0.0,
      "learning_rate": 9.32089980202857e-05,
      "loss": 0.9503,
      "step": 1390
    },
    {
      "epoch": 0.7588075880758808,
      "grad_norm": 0.0,
      "learning_rate": 9.304945709146819e-05,
      "loss": 1.1083,
      "step": 1400
    },
    {
      "epoch": 0.7642276422764228,
      "grad_norm": 0.0,
      "learning_rate": 9.2888203653499e-05,
      "loss": 0.9929,
      "step": 1410
    },
    {
      "epoch": 0.7696476964769647,
      "grad_norm": 0.0,
      "learning_rate": 9.272524412104645e-05,
      "loss": 1.1123,
      "step": 1420
    },
    {
      "epoch": 0.7750677506775068,
      "grad_norm": 0.0,
      "learning_rate": 9.256058497664734e-05,
      "loss": 1.0397,
      "step": 1430
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 0.0,
      "learning_rate": 9.239423277044917e-05,
      "loss": 1.0311,
      "step": 1440
    },
    {
      "epoch": 0.7859078590785907,
      "grad_norm": 0.0,
      "learning_rate": 9.22261941199494e-05,
      "loss": 0.9421,
      "step": 1450
    },
    {
      "epoch": 0.7913279132791328,
      "grad_norm": 0.0,
      "learning_rate": 9.205647570973239e-05,
      "loss": 0.9313,
      "step": 1460
    },
    {
      "epoch": 0.7967479674796748,
      "grad_norm": 0.0,
      "learning_rate": 9.188508429120342e-05,
      "loss": 1.0034,
      "step": 1470
    },
    {
      "epoch": 0.8021680216802168,
      "grad_norm": 0.0,
      "learning_rate": 9.171202668232007e-05,
      "loss": 1.036,
      "step": 1480
    },
    {
      "epoch": 0.8075880758807588,
      "grad_norm": 0.0,
      "learning_rate": 9.153730976732101e-05,
      "loss": 1.0326,
      "step": 1490
    },
    {
      "epoch": 0.8130081300813008,
      "grad_norm": 0.0,
      "learning_rate": 9.136094049645226e-05,
      "loss": 0.8734,
      "step": 1500
    },
    {
      "epoch": 0.8184281842818428,
      "grad_norm": 0.0,
      "learning_rate": 9.118292588569057e-05,
      "loss": 0.8989,
      "step": 1510
    },
    {
      "epoch": 0.8238482384823849,
      "grad_norm": 0.0,
      "learning_rate": 9.100327301646439e-05,
      "loss": 1.0225,
      "step": 1520
    },
    {
      "epoch": 0.8292682926829268,
      "grad_norm": 0.0,
      "learning_rate": 9.082198903537213e-05,
      "loss": 1.0594,
      "step": 1530
    },
    {
      "epoch": 0.8346883468834688,
      "grad_norm": 0.0,
      "learning_rate": 9.063908115389794e-05,
      "loss": 0.9244,
      "step": 1540
    },
    {
      "epoch": 0.8401084010840109,
      "grad_norm": 0.0,
      "learning_rate": 9.045455664812478e-05,
      "loss": 1.0654,
      "step": 1550
    },
    {
      "epoch": 0.8455284552845529,
      "grad_norm": 0.0,
      "learning_rate": 9.026842285844499e-05,
      "loss": 1.0126,
      "step": 1560
    },
    {
      "epoch": 0.8509485094850948,
      "grad_norm": 0.0,
      "learning_rate": 9.008068718926825e-05,
      "loss": 1.0821,
      "step": 1570
    },
    {
      "epoch": 0.8563685636856369,
      "grad_norm": 0.0,
      "learning_rate": 8.989135710872713e-05,
      "loss": 1.0105,
      "step": 1580
    },
    {
      "epoch": 0.8617886178861789,
      "grad_norm": 0.0,
      "learning_rate": 8.970044014837994e-05,
      "loss": 0.9684,
      "step": 1590
    },
    {
      "epoch": 0.8672086720867209,
      "grad_norm": 0.0,
      "learning_rate": 8.950794390291108e-05,
      "loss": 1.0457,
      "step": 1600
    },
    {
      "epoch": 0.8726287262872628,
      "grad_norm": 0.0,
      "learning_rate": 8.931387602982903e-05,
      "loss": 1.0847,
      "step": 1610
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 0.0,
      "learning_rate": 8.911824424916169e-05,
      "loss": 1.1106,
      "step": 1620
    },
    {
      "epoch": 0.8834688346883469,
      "grad_norm": 0.0,
      "learning_rate": 8.89210563431492e-05,
      "loss": 1.6764,
      "step": 1630
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.0,
      "learning_rate": 8.872232015593447e-05,
      "loss": 0.9737,
      "step": 1640
    },
    {
      "epoch": 0.8943089430894309,
      "grad_norm": 0.0,
      "learning_rate": 8.852204359325115e-05,
      "loss": 1.0723,
      "step": 1650
    },
    {
      "epoch": 0.8997289972899729,
      "grad_norm": 0.0,
      "learning_rate": 8.8320234622109e-05,
      "loss": 1.026,
      "step": 1660
    },
    {
      "epoch": 0.9051490514905149,
      "grad_norm": 0.0,
      "learning_rate": 8.811690127047707e-05,
      "loss": 0.9524,
      "step": 1670
    },
    {
      "epoch": 0.9105691056910569,
      "grad_norm": 0.0,
      "learning_rate": 8.791205162696439e-05,
      "loss": 0.9443,
      "step": 1680
    },
    {
      "epoch": 0.9159891598915989,
      "grad_norm": 0.0,
      "learning_rate": 8.770569384049806e-05,
      "loss": 1.0877,
      "step": 1690
    },
    {
      "epoch": 0.9214092140921409,
      "grad_norm": 0.0,
      "learning_rate": 8.749783611999924e-05,
      "loss": 1.0068,
      "step": 1700
    },
    {
      "epoch": 0.926829268292683,
      "grad_norm": 0.0,
      "learning_rate": 8.72884867340565e-05,
      "loss": 1.0288,
      "step": 1710
    },
    {
      "epoch": 0.9322493224932249,
      "grad_norm": 0.0,
      "learning_rate": 8.707765401059689e-05,
      "loss": 1.0115,
      "step": 1720
    },
    {
      "epoch": 0.9376693766937669,
      "grad_norm": 0.0,
      "learning_rate": 8.68653463365547e-05,
      "loss": 0.9857,
      "step": 1730
    },
    {
      "epoch": 0.943089430894309,
      "grad_norm": 0.0,
      "learning_rate": 8.665157215753788e-05,
      "loss": 0.9941,
      "step": 1740
    },
    {
      "epoch": 0.948509485094851,
      "grad_norm": 0.0,
      "learning_rate": 8.64363399774919e-05,
      "loss": 0.9,
      "step": 1750
    },
    {
      "epoch": 0.9539295392953929,
      "grad_norm": 0.0,
      "learning_rate": 8.621965835836165e-05,
      "loss": 0.8777,
      "step": 1760
    },
    {
      "epoch": 0.959349593495935,
      "grad_norm": 0.0,
      "learning_rate": 8.600153591975073e-05,
      "loss": 0.952,
      "step": 1770
    },
    {
      "epoch": 0.964769647696477,
      "grad_norm": 0.0,
      "learning_rate": 8.578198133857863e-05,
      "loss": 0.9876,
      "step": 1780
    },
    {
      "epoch": 0.9701897018970189,
      "grad_norm": 0.0,
      "learning_rate": 8.556100334873549e-05,
      "loss": 0.9553,
      "step": 1790
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.0,
      "learning_rate": 8.533861074073472e-05,
      "loss": 0.9153,
      "step": 1800
    },
    {
      "epoch": 0.981029810298103,
      "grad_norm": 0.0,
      "learning_rate": 8.511481236136329e-05,
      "loss": 0.9828,
      "step": 1810
    },
    {
      "epoch": 0.986449864498645,
      "grad_norm": 0.0,
      "learning_rate": 8.488961711332986e-05,
      "loss": 0.98,
      "step": 1820
    },
    {
      "epoch": 0.991869918699187,
      "grad_norm": 0.0,
      "learning_rate": 8.466303395491051e-05,
      "loss": 0.9471,
      "step": 1830
    },
    {
      "epoch": 0.997289972899729,
      "grad_norm": 0.0,
      "learning_rate": 8.443507189959248e-05,
      "loss": 0.9981,
      "step": 1840
    },
    {
      "epoch": 1.002710027100271,
      "grad_norm": 0.0,
      "learning_rate": 8.420574001571561e-05,
      "loss": 0.9234,
      "step": 1850
    },
    {
      "epoch": 1.008130081300813,
      "grad_norm": 0.0,
      "learning_rate": 8.397504742611155e-05,
      "loss": 0.8865,
      "step": 1860
    },
    {
      "epoch": 1.013550135501355,
      "grad_norm": 0.0,
      "learning_rate": 8.37430033077409e-05,
      "loss": 0.8455,
      "step": 1870
    },
    {
      "epoch": 1.018970189701897,
      "grad_norm": 0.0,
      "learning_rate": 8.350961689132808e-05,
      "loss": 0.8807,
      "step": 1880
    },
    {
      "epoch": 1.024390243902439,
      "grad_norm": 0.0,
      "learning_rate": 8.327489746099422e-05,
      "loss": 0.8296,
      "step": 1890
    },
    {
      "epoch": 1.029810298102981,
      "grad_norm": 0.0,
      "learning_rate": 8.303885435388783e-05,
      "loss": 0.8678,
      "step": 1900
    },
    {
      "epoch": 1.0352303523035231,
      "grad_norm": 0.0,
      "learning_rate": 8.280149695981325e-05,
      "loss": 0.7886,
      "step": 1910
    },
    {
      "epoch": 1.040650406504065,
      "grad_norm": 0.0,
      "learning_rate": 8.25628347208573e-05,
      "loss": 0.8248,
      "step": 1920
    },
    {
      "epoch": 1.046070460704607,
      "grad_norm": 0.0,
      "learning_rate": 8.232287713101351e-05,
      "loss": 0.8579,
      "step": 1930
    },
    {
      "epoch": 1.051490514905149,
      "grad_norm": 0.0,
      "learning_rate": 8.208163373580456e-05,
      "loss": 0.7809,
      "step": 1940
    },
    {
      "epoch": 1.056910569105691,
      "grad_norm": 0.0,
      "learning_rate": 8.183911413190251e-05,
      "loss": 0.7519,
      "step": 1950
    },
    {
      "epoch": 1.062330623306233,
      "grad_norm": 0.0,
      "learning_rate": 8.159532796674706e-05,
      "loss": 0.8322,
      "step": 1960
    },
    {
      "epoch": 1.067750677506775,
      "grad_norm": 0.0,
      "learning_rate": 8.135028493816174e-05,
      "loss": 0.8875,
      "step": 1970
    },
    {
      "epoch": 1.0731707317073171,
      "grad_norm": 0.0,
      "learning_rate": 8.110399479396822e-05,
      "loss": 0.6955,
      "step": 1980
    },
    {
      "epoch": 1.0785907859078592,
      "grad_norm": 0.0,
      "learning_rate": 8.085646733159841e-05,
      "loss": 0.8317,
      "step": 1990
    },
    {
      "epoch": 1.084010840108401,
      "grad_norm": 0.0,
      "learning_rate": 8.060771239770482e-05,
      "loss": 0.9015,
      "step": 2000
    },
    {
      "epoch": 1.089430894308943,
      "grad_norm": 0.0,
      "learning_rate": 8.035773988776883e-05,
      "loss": 0.9467,
      "step": 2010
    },
    {
      "epoch": 1.094850948509485,
      "grad_norm": 0.0,
      "learning_rate": 8.010655974570703e-05,
      "loss": 0.9237,
      "step": 2020
    },
    {
      "epoch": 1.100271002710027,
      "grad_norm": 0.0,
      "learning_rate": 7.985418196347568e-05,
      "loss": 0.8042,
      "step": 2030
    },
    {
      "epoch": 1.1056910569105691,
      "grad_norm": 0.0,
      "learning_rate": 7.960061658067322e-05,
      "loss": 0.7585,
      "step": 2040
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.0,
      "learning_rate": 7.934587368414084e-05,
      "loss": 0.8525,
      "step": 2050
    },
    {
      "epoch": 1.1165311653116532,
      "grad_norm": 0.0,
      "learning_rate": 7.908996340756136e-05,
      "loss": 0.8898,
      "step": 2060
    },
    {
      "epoch": 1.1219512195121952,
      "grad_norm": 0.0,
      "learning_rate": 7.883289593105597e-05,
      "loss": 0.851,
      "step": 2070
    },
    {
      "epoch": 1.127371273712737,
      "grad_norm": 0.0,
      "learning_rate": 7.857468148077926e-05,
      "loss": 0.7651,
      "step": 2080
    },
    {
      "epoch": 1.132791327913279,
      "grad_norm": 0.0,
      "learning_rate": 7.831533032851264e-05,
      "loss": 0.8294,
      "step": 2090
    },
    {
      "epoch": 1.1382113821138211,
      "grad_norm": 0.0,
      "learning_rate": 7.805485279125543e-05,
      "loss": 0.7916,
      "step": 2100
    },
    {
      "epoch": 1.1436314363143631,
      "grad_norm": 0.0,
      "learning_rate": 7.779325923081465e-05,
      "loss": 0.9365,
      "step": 2110
    },
    {
      "epoch": 1.1490514905149052,
      "grad_norm": 0.0,
      "learning_rate": 7.753056005339277e-05,
      "loss": 0.8456,
      "step": 2120
    },
    {
      "epoch": 1.1544715447154472,
      "grad_norm": 0.0,
      "learning_rate": 7.726676570917372e-05,
      "loss": 0.8459,
      "step": 2130
    },
    {
      "epoch": 1.1598915989159893,
      "grad_norm": 0.0,
      "learning_rate": 7.700188669190727e-05,
      "loss": 0.8237,
      "step": 2140
    },
    {
      "epoch": 1.165311653116531,
      "grad_norm": 0.0,
      "learning_rate": 7.673593353849143e-05,
      "loss": 0.8186,
      "step": 2150
    },
    {
      "epoch": 1.170731707317073,
      "grad_norm": 0.0,
      "learning_rate": 7.646891682855347e-05,
      "loss": 0.8047,
      "step": 2160
    },
    {
      "epoch": 1.1761517615176151,
      "grad_norm": 0.0,
      "learning_rate": 7.620084718402895e-05,
      "loss": 0.8191,
      "step": 2170
    },
    {
      "epoch": 1.1815718157181572,
      "grad_norm": 0.0,
      "learning_rate": 7.593173526873925e-05,
      "loss": 0.8245,
      "step": 2180
    },
    {
      "epoch": 1.1869918699186992,
      "grad_norm": 0.0,
      "learning_rate": 7.566159178796726e-05,
      "loss": 0.9024,
      "step": 2190
    },
    {
      "epoch": 1.1924119241192412,
      "grad_norm": 0.0,
      "learning_rate": 7.539042748803162e-05,
      "loss": 0.7802,
      "step": 2200
    },
    {
      "epoch": 1.1978319783197833,
      "grad_norm": 0.0,
      "learning_rate": 7.511825315585921e-05,
      "loss": 0.7251,
      "step": 2210
    },
    {
      "epoch": 1.203252032520325,
      "grad_norm": 0.0,
      "learning_rate": 7.484507961855598e-05,
      "loss": 0.8132,
      "step": 2220
    },
    {
      "epoch": 1.2086720867208671,
      "grad_norm": 0.0,
      "learning_rate": 7.45709177429764e-05,
      "loss": 0.7704,
      "step": 2230
    },
    {
      "epoch": 1.2140921409214092,
      "grad_norm": 0.0,
      "learning_rate": 7.429577843529096e-05,
      "loss": 0.7359,
      "step": 2240
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.0,
      "learning_rate": 7.401967264055253e-05,
      "loss": 0.8472,
      "step": 2250
    },
    {
      "epoch": 1.2249322493224932,
      "grad_norm": 0.0,
      "learning_rate": 7.374261134226083e-05,
      "loss": 0.7575,
      "step": 2260
    },
    {
      "epoch": 1.2303523035230353,
      "grad_norm": 0.0,
      "learning_rate": 7.346460556192557e-05,
      "loss": 0.8207,
      "step": 2270
    },
    {
      "epoch": 1.2357723577235773,
      "grad_norm": 0.0,
      "learning_rate": 7.318566635862797e-05,
      "loss": 0.863,
      "step": 2280
    },
    {
      "epoch": 1.2411924119241193,
      "grad_norm": 0.0,
      "learning_rate": 7.290580482858088e-05,
      "loss": 0.9107,
      "step": 2290
    },
    {
      "epoch": 1.2466124661246614,
      "grad_norm": 0.0,
      "learning_rate": 7.262503210468728e-05,
      "loss": 0.7615,
      "step": 2300
    },
    {
      "epoch": 1.2520325203252032,
      "grad_norm": 0.0,
      "learning_rate": 7.234335935609761e-05,
      "loss": 0.8431,
      "step": 2310
    },
    {
      "epoch": 1.2574525745257452,
      "grad_norm": 0.0,
      "learning_rate": 7.20607977877652e-05,
      "loss": 0.8447,
      "step": 2320
    },
    {
      "epoch": 1.2628726287262872,
      "grad_norm": 0.0,
      "learning_rate": 7.177735864000073e-05,
      "loss": 0.8853,
      "step": 2330
    },
    {
      "epoch": 1.2682926829268293,
      "grad_norm": 0.0,
      "learning_rate": 7.149305318802502e-05,
      "loss": 0.9468,
      "step": 2340
    },
    {
      "epoch": 1.2737127371273713,
      "grad_norm": 0.0,
      "learning_rate": 7.120789274152049e-05,
      "loss": 0.7821,
      "step": 2350
    },
    {
      "epoch": 1.2791327913279134,
      "grad_norm": 0.0,
      "learning_rate": 7.092188864418127e-05,
      "loss": 0.6905,
      "step": 2360
    },
    {
      "epoch": 1.2845528455284554,
      "grad_norm": 0.0,
      "learning_rate": 7.063505227326198e-05,
      "loss": 0.7971,
      "step": 2370
    },
    {
      "epoch": 1.2899728997289972,
      "grad_norm": 0.0,
      "learning_rate": 7.03473950391251e-05,
      "loss": 0.6959,
      "step": 2380
    },
    {
      "epoch": 1.2953929539295392,
      "grad_norm": 0.0,
      "learning_rate": 7.005892838478711e-05,
      "loss": 0.8096,
      "step": 2390
    },
    {
      "epoch": 1.3008130081300813,
      "grad_norm": 0.0,
      "learning_rate": 6.976966378546322e-05,
      "loss": 0.93,
      "step": 2400
    },
    {
      "epoch": 1.3062330623306233,
      "grad_norm": 0.0,
      "learning_rate": 6.947961274811093e-05,
      "loss": 0.7973,
      "step": 2410
    },
    {
      "epoch": 1.3116531165311653,
      "grad_norm": 0.0,
      "learning_rate": 6.918878681097232e-05,
      "loss": 0.85,
      "step": 2420
    },
    {
      "epoch": 1.3170731707317074,
      "grad_norm": 0.0,
      "learning_rate": 6.889719754311495e-05,
      "loss": 0.7707,
      "step": 2430
    },
    {
      "epoch": 1.3224932249322494,
      "grad_norm": 0.0,
      "learning_rate": 6.860485654397173e-05,
      "loss": 0.7808,
      "step": 2440
    },
    {
      "epoch": 1.3279132791327912,
      "grad_norm": 0.0,
      "learning_rate": 6.831177544287952e-05,
      "loss": 0.8128,
      "step": 2450
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.0,
      "learning_rate": 6.80179658986164e-05,
      "loss": 0.7977,
      "step": 2460
    },
    {
      "epoch": 1.3387533875338753,
      "grad_norm": 0.0,
      "learning_rate": 6.7723439598938e-05,
      "loss": 0.7492,
      "step": 2470
    },
    {
      "epoch": 1.3441734417344173,
      "grad_norm": 0.0,
      "learning_rate": 6.742820826011245e-05,
      "loss": 0.8064,
      "step": 2480
    },
    {
      "epoch": 1.3495934959349594,
      "grad_norm": 0.0,
      "learning_rate": 6.71322836264544e-05,
      "loss": 0.807,
      "step": 2490
    },
    {
      "epoch": 1.3550135501355014,
      "grad_norm": 0.0,
      "learning_rate": 6.683567746985783e-05,
      "loss": 0.8208,
      "step": 2500
    },
    {
      "epoch": 1.3604336043360434,
      "grad_norm": 0.0,
      "learning_rate": 6.653840158932769e-05,
      "loss": 0.7258,
      "step": 2510
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 0.0,
      "learning_rate": 6.624046781051055e-05,
      "loss": 0.7142,
      "step": 2520
    },
    {
      "epoch": 1.3712737127371275,
      "grad_norm": 0.0,
      "learning_rate": 6.594188798522428e-05,
      "loss": 0.7837,
      "step": 2530
    },
    {
      "epoch": 1.3766937669376693,
      "grad_norm": 0.0,
      "learning_rate": 6.56426739909864e-05,
      "loss": 0.7829,
      "step": 2540
    },
    {
      "epoch": 1.3821138211382114,
      "grad_norm": 0.0,
      "learning_rate": 6.53428377305418e-05,
      "loss": 0.7371,
      "step": 2550
    },
    {
      "epoch": 1.3875338753387534,
      "grad_norm": 0.0,
      "learning_rate": 6.504239113138908e-05,
      "loss": 0.9048,
      "step": 2560
    },
    {
      "epoch": 1.3929539295392954,
      "grad_norm": 0.0,
      "learning_rate": 6.474134614530617e-05,
      "loss": 0.8665,
      "step": 2570
    },
    {
      "epoch": 1.3983739837398375,
      "grad_norm": 0.0,
      "learning_rate": 6.443971474787483e-05,
      "loss": 0.8496,
      "step": 2580
    },
    {
      "epoch": 1.4037940379403793,
      "grad_norm": 0.0,
      "learning_rate": 6.413750893800432e-05,
      "loss": 0.7842,
      "step": 2590
    },
    {
      "epoch": 1.4092140921409215,
      "grad_norm": 0.0,
      "learning_rate": 6.3834740737454e-05,
      "loss": 0.8222,
      "step": 2600
    },
    {
      "epoch": 1.4146341463414633,
      "grad_norm": 0.0,
      "learning_rate": 6.353142219035521e-05,
      "loss": 0.7915,
      "step": 2610
    },
    {
      "epoch": 1.4200542005420054,
      "grad_norm": 0.0,
      "learning_rate": 6.322756536273208e-05,
      "loss": 0.8833,
      "step": 2620
    },
    {
      "epoch": 1.4254742547425474,
      "grad_norm": 0.0,
      "learning_rate": 6.292318234202155e-05,
      "loss": 0.8277,
      "step": 2630
    },
    {
      "epoch": 1.4308943089430894,
      "grad_norm": 0.0,
      "learning_rate": 6.261828523659253e-05,
      "loss": 0.743,
      "step": 2640
    },
    {
      "epoch": 1.4363143631436315,
      "grad_norm": 0.0,
      "learning_rate": 6.231288617526427e-05,
      "loss": 0.6826,
      "step": 2650
    },
    {
      "epoch": 1.4417344173441735,
      "grad_norm": 0.0,
      "learning_rate": 6.200699730682384e-05,
      "loss": 0.7916,
      "step": 2660
    },
    {
      "epoch": 1.4471544715447155,
      "grad_norm": 0.0,
      "learning_rate": 6.170063079954283e-05,
      "loss": 0.7353,
      "step": 2670
    },
    {
      "epoch": 1.4525745257452574,
      "grad_norm": 0.0,
      "learning_rate": 6.139379884069337e-05,
      "loss": 0.8918,
      "step": 2680
    },
    {
      "epoch": 1.4579945799457994,
      "grad_norm": 0.0,
      "learning_rate": 6.108651363606324e-05,
      "loss": 0.7956,
      "step": 2690
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 0.0,
      "learning_rate": 6.077878740947035e-05,
      "loss": 0.808,
      "step": 2700
    },
    {
      "epoch": 1.4688346883468835,
      "grad_norm": 0.0,
      "learning_rate": 6.04706324022765e-05,
      "loss": 0.7646,
      "step": 2710
    },
    {
      "epoch": 1.4742547425474255,
      "grad_norm": 0.0,
      "learning_rate": 6.016206087290038e-05,
      "loss": 0.8579,
      "step": 2720
    },
    {
      "epoch": 1.4796747967479675,
      "grad_norm": 0.0,
      "learning_rate": 5.985308509633e-05,
      "loss": 0.887,
      "step": 2730
    },
    {
      "epoch": 1.4850948509485096,
      "grad_norm": 0.0,
      "learning_rate": 5.9543717363634263e-05,
      "loss": 0.8292,
      "step": 2740
    },
    {
      "epoch": 1.4905149051490514,
      "grad_norm": 0.0,
      "learning_rate": 5.9233969981474156e-05,
      "loss": 0.8304,
      "step": 2750
    },
    {
      "epoch": 1.4959349593495934,
      "grad_norm": 0.0,
      "learning_rate": 5.892385527161311e-05,
      "loss": 0.7978,
      "step": 2760
    },
    {
      "epoch": 1.5013550135501355,
      "grad_norm": 0.0,
      "learning_rate": 5.861338557042691e-05,
      "loss": 0.7651,
      "step": 2770
    },
    {
      "epoch": 1.5067750677506775,
      "grad_norm": 0.0,
      "learning_rate": 5.830257322841285e-05,
      "loss": 0.9219,
      "step": 2780
    },
    {
      "epoch": 1.5121951219512195,
      "grad_norm": 0.0,
      "learning_rate": 5.799143060969854e-05,
      "loss": 0.7874,
      "step": 2790
    },
    {
      "epoch": 1.5176151761517616,
      "grad_norm": 0.0,
      "learning_rate": 5.767997009154997e-05,
      "loss": 0.7405,
      "step": 2800
    },
    {
      "epoch": 1.5230352303523036,
      "grad_norm": 0.0,
      "learning_rate": 5.736820406387923e-05,
      "loss": 0.845,
      "step": 2810
    },
    {
      "epoch": 1.5284552845528454,
      "grad_norm": 0.0,
      "learning_rate": 5.705614492875153e-05,
      "loss": 0.8464,
      "step": 2820
    },
    {
      "epoch": 1.5338753387533877,
      "grad_norm": 0.0,
      "learning_rate": 5.6743805099891955e-05,
      "loss": 0.9143,
      "step": 2830
    },
    {
      "epoch": 1.5392953929539295,
      "grad_norm": 0.0,
      "learning_rate": 5.6431197002191594e-05,
      "loss": 0.804,
      "step": 2840
    },
    {
      "epoch": 1.5447154471544715,
      "grad_norm": 0.0,
      "learning_rate": 5.611833307121327e-05,
      "loss": 0.7703,
      "step": 2850
    },
    {
      "epoch": 1.5501355013550135,
      "grad_norm": 0.0,
      "learning_rate": 5.580522575269684e-05,
      "loss": 0.8818,
      "step": 2860
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.0,
      "learning_rate": 5.54918875020642e-05,
      "loss": 0.7634,
      "step": 2870
    },
    {
      "epoch": 1.5609756097560976,
      "grad_norm": 0.0,
      "learning_rate": 5.517833078392365e-05,
      "loss": 0.7817,
      "step": 2880
    },
    {
      "epoch": 1.5663956639566394,
      "grad_norm": 0.0,
      "learning_rate": 5.486456807157417e-05,
      "loss": 0.7829,
      "step": 2890
    },
    {
      "epoch": 1.5718157181571817,
      "grad_norm": 0.0,
      "learning_rate": 5.455061184650921e-05,
      "loss": 0.7587,
      "step": 2900
    },
    {
      "epoch": 1.5772357723577235,
      "grad_norm": 0.0,
      "learning_rate": 5.423647459792018e-05,
      "loss": 0.6822,
      "step": 2910
    },
    {
      "epoch": 1.5826558265582655,
      "grad_norm": 0.0,
      "learning_rate": 5.392216882219956e-05,
      "loss": 0.8737,
      "step": 2920
    },
    {
      "epoch": 1.5880758807588076,
      "grad_norm": 0.0,
      "learning_rate": 5.360770702244388e-05,
      "loss": 0.8793,
      "step": 2930
    },
    {
      "epoch": 1.5934959349593496,
      "grad_norm": 0.0,
      "learning_rate": 5.3293101707956306e-05,
      "loss": 0.8142,
      "step": 2940
    },
    {
      "epoch": 1.5989159891598916,
      "grad_norm": 0.0,
      "learning_rate": 5.2978365393749044e-05,
      "loss": 0.718,
      "step": 2950
    },
    {
      "epoch": 1.6043360433604335,
      "grad_norm": 0.0,
      "learning_rate": 5.266351060004544e-05,
      "loss": 0.8058,
      "step": 2960
    },
    {
      "epoch": 1.6097560975609757,
      "grad_norm": 0.0,
      "learning_rate": 5.234854985178196e-05,
      "loss": 0.8089,
      "step": 2970
    },
    {
      "epoch": 1.6151761517615175,
      "grad_norm": 0.0,
      "learning_rate": 5.203349567811e-05,
      "loss": 0.7888,
      "step": 2980
    },
    {
      "epoch": 1.6205962059620598,
      "grad_norm": 0.0,
      "learning_rate": 5.1718360611897374e-05,
      "loss": 0.761,
      "step": 2990
    },
    {
      "epoch": 1.6260162601626016,
      "grad_norm": 0.0,
      "learning_rate": 5.140315718922979e-05,
      "loss": 0.7682,
      "step": 3000
    },
    {
      "epoch": 1.6314363143631436,
      "grad_norm": 0.0,
      "learning_rate": 5.108789794891224e-05,
      "loss": 0.8024,
      "step": 3010
    },
    {
      "epoch": 1.6368563685636857,
      "grad_norm": 0.0,
      "learning_rate": 5.0772595431970095e-05,
      "loss": 0.8088,
      "step": 3020
    },
    {
      "epoch": 1.6422764227642277,
      "grad_norm": 0.0,
      "learning_rate": 5.0457262181150286e-05,
      "loss": 0.7911,
      "step": 3030
    },
    {
      "epoch": 1.6476964769647697,
      "grad_norm": 0.0,
      "learning_rate": 5.014191074042233e-05,
      "loss": 0.7997,
      "step": 3040
    },
    {
      "epoch": 1.6531165311653115,
      "grad_norm": 0.0,
      "learning_rate": 4.9826553654479366e-05,
      "loss": 0.7467,
      "step": 3050
    },
    {
      "epoch": 1.6585365853658538,
      "grad_norm": 0.0,
      "learning_rate": 4.9511203468239064e-05,
      "loss": 0.8015,
      "step": 3060
    },
    {
      "epoch": 1.6639566395663956,
      "grad_norm": 0.0,
      "learning_rate": 4.9195872726344655e-05,
      "loss": 0.8341,
      "step": 3070
    },
    {
      "epoch": 1.6693766937669376,
      "grad_norm": 0.0,
      "learning_rate": 4.888057397266585e-05,
      "loss": 0.7597,
      "step": 3080
    },
    {
      "epoch": 1.6747967479674797,
      "grad_norm": 0.0,
      "learning_rate": 4.856531974979987e-05,
      "loss": 0.7855,
      "step": 3090
    },
    {
      "epoch": 1.6802168021680217,
      "grad_norm": 0.0,
      "learning_rate": 4.8250122598572537e-05,
      "loss": 0.7871,
      "step": 3100
    },
    {
      "epoch": 1.6856368563685638,
      "grad_norm": 0.0,
      "learning_rate": 4.79349950575393e-05,
      "loss": 0.8974,
      "step": 3110
    },
    {
      "epoch": 1.6910569105691056,
      "grad_norm": 0.0,
      "learning_rate": 4.7619949662486554e-05,
      "loss": 0.8021,
      "step": 3120
    },
    {
      "epoch": 1.6964769647696478,
      "grad_norm": 0.0,
      "learning_rate": 4.7304998945932904e-05,
      "loss": 0.7876,
      "step": 3130
    },
    {
      "epoch": 1.7018970189701896,
      "grad_norm": 0.0,
      "learning_rate": 4.699015543663066e-05,
      "loss": 0.901,
      "step": 3140
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.0,
      "learning_rate": 4.667543165906739e-05,
      "loss": 0.7647,
      "step": 3150
    },
    {
      "epoch": 1.7127371273712737,
      "grad_norm": 0.0,
      "learning_rate": 4.6360840132967735e-05,
      "loss": 0.9019,
      "step": 3160
    },
    {
      "epoch": 1.7181571815718157,
      "grad_norm": 0.0,
      "learning_rate": 4.604639337279543e-05,
      "loss": 0.7046,
      "step": 3170
    },
    {
      "epoch": 1.7235772357723578,
      "grad_norm": 0.0,
      "learning_rate": 4.5732103887255354e-05,
      "loss": 0.7644,
      "step": 3180
    },
    {
      "epoch": 1.7289972899728996,
      "grad_norm": 0.0,
      "learning_rate": 4.541798417879598e-05,
      "loss": 0.8395,
      "step": 3190
    },
    {
      "epoch": 1.7344173441734418,
      "grad_norm": 0.0,
      "learning_rate": 4.5104046743112094e-05,
      "loss": 0.7812,
      "step": 3200
    },
    {
      "epoch": 1.7398373983739837,
      "grad_norm": 0.0,
      "learning_rate": 4.479030406864763e-05,
      "loss": 0.7638,
      "step": 3210
    },
    {
      "epoch": 1.7452574525745257,
      "grad_norm": 0.0,
      "learning_rate": 4.447676863609892e-05,
      "loss": 0.7856,
      "step": 3220
    },
    {
      "epoch": 1.7506775067750677,
      "grad_norm": 0.0,
      "learning_rate": 4.4163452917918194e-05,
      "loss": 0.8102,
      "step": 3230
    },
    {
      "epoch": 1.7560975609756098,
      "grad_norm": 0.0,
      "learning_rate": 4.3850369377817465e-05,
      "loss": 0.9204,
      "step": 3240
    },
    {
      "epoch": 1.7615176151761518,
      "grad_norm": 0.0,
      "learning_rate": 4.353753047027269e-05,
      "loss": 0.8147,
      "step": 3250
    },
    {
      "epoch": 1.7669376693766936,
      "grad_norm": 0.0,
      "learning_rate": 4.32249486400283e-05,
      "loss": 0.7163,
      "step": 3260
    },
    {
      "epoch": 1.7723577235772359,
      "grad_norm": 0.0,
      "learning_rate": 4.2912636321602186e-05,
      "loss": 0.7531,
      "step": 3270
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.0,
      "learning_rate": 4.260060593879108e-05,
      "loss": 0.8307,
      "step": 3280
    },
    {
      "epoch": 1.78319783197832,
      "grad_norm": 0.0,
      "learning_rate": 4.228886990417627e-05,
      "loss": 1.0805,
      "step": 3290
    },
    {
      "epoch": 1.7886178861788617,
      "grad_norm": 0.0,
      "learning_rate": 4.197744061862987e-05,
      "loss": 0.7975,
      "step": 3300
    },
    {
      "epoch": 1.7940379403794038,
      "grad_norm": 0.0,
      "learning_rate": 4.1666330470821536e-05,
      "loss": 0.6771,
      "step": 3310
    },
    {
      "epoch": 1.7994579945799458,
      "grad_norm": 0.0,
      "learning_rate": 4.135555183672557e-05,
      "loss": 0.7911,
      "step": 3320
    },
    {
      "epoch": 1.8048780487804879,
      "grad_norm": 0.0,
      "learning_rate": 4.104511707912866e-05,
      "loss": 0.9151,
      "step": 3330
    },
    {
      "epoch": 1.8102981029810299,
      "grad_norm": 0.0,
      "learning_rate": 4.0735038547138065e-05,
      "loss": 0.7033,
      "step": 3340
    },
    {
      "epoch": 1.8157181571815717,
      "grad_norm": 0.0,
      "learning_rate": 4.042532857569039e-05,
      "loss": 0.725,
      "step": 3350
    },
    {
      "epoch": 1.821138211382114,
      "grad_norm": 0.0,
      "learning_rate": 4.011599948506088e-05,
      "loss": 0.7157,
      "step": 3360
    },
    {
      "epoch": 1.8265582655826558,
      "grad_norm": 0.0,
      "learning_rate": 3.9807063580373314e-05,
      "loss": 0.8203,
      "step": 3370
    },
    {
      "epoch": 1.8319783197831978,
      "grad_norm": 0.0,
      "learning_rate": 3.949853315111055e-05,
      "loss": 0.7513,
      "step": 3380
    },
    {
      "epoch": 1.8373983739837398,
      "grad_norm": 0.0,
      "learning_rate": 3.919042047062558e-05,
      "loss": 0.7372,
      "step": 3390
    },
    {
      "epoch": 1.8428184281842819,
      "grad_norm": 0.0,
      "learning_rate": 3.8882737795653346e-05,
      "loss": 0.7178,
      "step": 3400
    },
    {
      "epoch": 1.848238482384824,
      "grad_norm": 0.0,
      "learning_rate": 3.857549736582316e-05,
      "loss": 0.7819,
      "step": 3410
    },
    {
      "epoch": 1.8536585365853657,
      "grad_norm": 0.0,
      "learning_rate": 3.8268711403171805e-05,
      "loss": 0.7438,
      "step": 3420
    },
    {
      "epoch": 1.859078590785908,
      "grad_norm": 0.0,
      "learning_rate": 3.796239211165733e-05,
      "loss": 0.7724,
      "step": 3430
    },
    {
      "epoch": 1.8644986449864498,
      "grad_norm": 0.0,
      "learning_rate": 3.765655167667358e-05,
      "loss": 0.8343,
      "step": 3440
    },
    {
      "epoch": 1.8699186991869918,
      "grad_norm": 0.0,
      "learning_rate": 3.735120226456551e-05,
      "loss": 0.7478,
      "step": 3450
    },
    {
      "epoch": 1.8753387533875339,
      "grad_norm": 0.0,
      "learning_rate": 3.704635602214513e-05,
      "loss": 0.7864,
      "step": 3460
    },
    {
      "epoch": 1.880758807588076,
      "grad_norm": 0.0,
      "learning_rate": 3.6742025076208344e-05,
      "loss": 0.7534,
      "step": 3470
    },
    {
      "epoch": 1.886178861788618,
      "grad_norm": 0.0,
      "learning_rate": 3.643822153305252e-05,
      "loss": 0.7924,
      "step": 3480
    },
    {
      "epoch": 1.8915989159891597,
      "grad_norm": 0.0,
      "learning_rate": 3.613495747799498e-05,
      "loss": 0.8663,
      "step": 3490
    },
    {
      "epoch": 1.897018970189702,
      "grad_norm": 0.0,
      "learning_rate": 3.583224497489213e-05,
      "loss": 0.8439,
      "step": 3500
    },
    {
      "epoch": 1.9024390243902438,
      "grad_norm": 0.0,
      "learning_rate": 3.553009606565962e-05,
      "loss": 0.7788,
      "step": 3510
    },
    {
      "epoch": 1.907859078590786,
      "grad_norm": 0.0,
      "learning_rate": 3.522852276979337e-05,
      "loss": 0.8515,
      "step": 3520
    },
    {
      "epoch": 1.9132791327913279,
      "grad_norm": 0.0,
      "learning_rate": 3.492753708389134e-05,
      "loss": 0.7187,
      "step": 3530
    },
    {
      "epoch": 1.91869918699187,
      "grad_norm": 0.0,
      "learning_rate": 3.462715098117632e-05,
      "loss": 0.7172,
      "step": 3540
    },
    {
      "epoch": 1.924119241192412,
      "grad_norm": 0.0,
      "learning_rate": 3.4327376411019675e-05,
      "loss": 0.7059,
      "step": 3550
    },
    {
      "epoch": 1.9295392953929538,
      "grad_norm": 0.0,
      "learning_rate": 3.4028225298465987e-05,
      "loss": 0.9021,
      "step": 3560
    },
    {
      "epoch": 1.934959349593496,
      "grad_norm": 0.0,
      "learning_rate": 3.372970954375867e-05,
      "loss": 0.7364,
      "step": 3570
    },
    {
      "epoch": 1.9403794037940378,
      "grad_norm": 0.0,
      "learning_rate": 3.343184102186655e-05,
      "loss": 0.8083,
      "step": 3580
    },
    {
      "epoch": 1.94579945799458,
      "grad_norm": 0.0,
      "learning_rate": 3.313463158201157e-05,
      "loss": 0.7949,
      "step": 3590
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.0,
      "learning_rate": 3.283809304719728e-05,
      "loss": 0.7554,
      "step": 3600
    },
    {
      "epoch": 1.956639566395664,
      "grad_norm": 0.0,
      "learning_rate": 3.254223721373866e-05,
      "loss": 0.7353,
      "step": 3610
    },
    {
      "epoch": 1.962059620596206,
      "grad_norm": 0.0,
      "learning_rate": 3.2247075850792774e-05,
      "loss": 0.6905,
      "step": 3620
    },
    {
      "epoch": 1.967479674796748,
      "grad_norm": 0.0,
      "learning_rate": 3.195262069989064e-05,
      "loss": 0.7416,
      "step": 3630
    },
    {
      "epoch": 1.97289972899729,
      "grad_norm": 0.0,
      "learning_rate": 3.1658883474470096e-05,
      "loss": 0.8705,
      "step": 3640
    },
    {
      "epoch": 1.9783197831978319,
      "grad_norm": 0.0,
      "learning_rate": 3.136587585940991e-05,
      "loss": 0.8878,
      "step": 3650
    },
    {
      "epoch": 1.9837398373983741,
      "grad_norm": 0.0,
      "learning_rate": 3.107360951056493e-05,
      "loss": 0.6716,
      "step": 3660
    },
    {
      "epoch": 1.989159891598916,
      "grad_norm": 0.0,
      "learning_rate": 3.0782096054302375e-05,
      "loss": 0.7695,
      "step": 3670
    },
    {
      "epoch": 1.994579945799458,
      "grad_norm": 0.0,
      "learning_rate": 3.0491347087039363e-05,
      "loss": 0.8582,
      "step": 3680
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.0,
      "learning_rate": 3.020137417478163e-05,
      "loss": 0.7794,
      "step": 3690
    },
    {
      "epoch": 2.005420054200542,
      "grad_norm": 0.0,
      "learning_rate": 2.9912188852663404e-05,
      "loss": 0.6789,
      "step": 3700
    },
    {
      "epoch": 2.010840108401084,
      "grad_norm": 0.0,
      "learning_rate": 2.962380262448855e-05,
      "loss": 0.6068,
      "step": 3710
    },
    {
      "epoch": 2.016260162601626,
      "grad_norm": 0.0,
      "learning_rate": 2.9336226962272904e-05,
      "loss": 0.5462,
      "step": 3720
    },
    {
      "epoch": 2.021680216802168,
      "grad_norm": 0.0,
      "learning_rate": 2.9049473305788045e-05,
      "loss": 0.5795,
      "step": 3730
    },
    {
      "epoch": 2.02710027100271,
      "grad_norm": 0.0,
      "learning_rate": 2.876355306210604e-05,
      "loss": 0.5308,
      "step": 3740
    },
    {
      "epoch": 2.032520325203252,
      "grad_norm": 0.0,
      "learning_rate": 2.8478477605145815e-05,
      "loss": 0.5021,
      "step": 3750
    },
    {
      "epoch": 2.037940379403794,
      "grad_norm": 0.0,
      "learning_rate": 2.819425827522061e-05,
      "loss": 0.5725,
      "step": 3760
    },
    {
      "epoch": 2.043360433604336,
      "grad_norm": 0.0,
      "learning_rate": 2.791090637858692e-05,
      "loss": 0.5371,
      "step": 3770
    },
    {
      "epoch": 2.048780487804878,
      "grad_norm": 0.0,
      "learning_rate": 2.7628433186994705e-05,
      "loss": 0.4408,
      "step": 3780
    },
    {
      "epoch": 2.05420054200542,
      "grad_norm": 0.0,
      "learning_rate": 2.7346849937238983e-05,
      "loss": 0.5324,
      "step": 3790
    },
    {
      "epoch": 2.059620596205962,
      "grad_norm": 0.0,
      "learning_rate": 2.7066167830712897e-05,
      "loss": 0.537,
      "step": 3800
    },
    {
      "epoch": 2.065040650406504,
      "grad_norm": 0.0,
      "learning_rate": 2.678639803296203e-05,
      "loss": 0.6198,
      "step": 3810
    },
    {
      "epoch": 2.0704607046070462,
      "grad_norm": 0.0,
      "learning_rate": 2.6507551673240295e-05,
      "loss": 0.6005,
      "step": 3820
    },
    {
      "epoch": 2.075880758807588,
      "grad_norm": 0.0,
      "learning_rate": 2.622963984406722e-05,
      "loss": 0.5428,
      "step": 3830
    },
    {
      "epoch": 2.08130081300813,
      "grad_norm": 0.0,
      "learning_rate": 2.595267360078665e-05,
      "loss": 0.5816,
      "step": 3840
    },
    {
      "epoch": 2.086720867208672,
      "grad_norm": 0.0,
      "learning_rate": 2.567666396112699e-05,
      "loss": 0.5656,
      "step": 3850
    },
    {
      "epoch": 2.092140921409214,
      "grad_norm": 0.0,
      "learning_rate": 2.5401621904762896e-05,
      "loss": 0.5424,
      "step": 3860
    },
    {
      "epoch": 2.097560975609756,
      "grad_norm": 0.0,
      "learning_rate": 2.512755837287859e-05,
      "loss": 0.5393,
      "step": 3870
    },
    {
      "epoch": 2.102981029810298,
      "grad_norm": 0.0,
      "learning_rate": 2.485448426773248e-05,
      "loss": 0.4829,
      "step": 3880
    },
    {
      "epoch": 2.1084010840108403,
      "grad_norm": 0.0,
      "learning_rate": 2.458241045222356e-05,
      "loss": 0.5376,
      "step": 3890
    },
    {
      "epoch": 2.113821138211382,
      "grad_norm": 0.0,
      "learning_rate": 2.4311347749459252e-05,
      "loss": 0.5107,
      "step": 3900
    },
    {
      "epoch": 2.1192411924119243,
      "grad_norm": 0.0,
      "learning_rate": 2.4041306942324892e-05,
      "loss": 0.5824,
      "step": 3910
    },
    {
      "epoch": 2.124661246612466,
      "grad_norm": 0.0,
      "learning_rate": 2.3772298773054757e-05,
      "loss": 0.5408,
      "step": 3920
    },
    {
      "epoch": 2.130081300813008,
      "grad_norm": 0.0,
      "learning_rate": 2.3504333942804728e-05,
      "loss": 0.4924,
      "step": 3930
    },
    {
      "epoch": 2.13550135501355,
      "grad_norm": 0.0,
      "learning_rate": 2.323742311122667e-05,
      "loss": 0.4882,
      "step": 3940
    },
    {
      "epoch": 2.140921409214092,
      "grad_norm": 0.0,
      "learning_rate": 2.2971576896044283e-05,
      "loss": 0.5444,
      "step": 3950
    },
    {
      "epoch": 2.1463414634146343,
      "grad_norm": 0.0,
      "learning_rate": 2.2706805872630816e-05,
      "loss": 0.4971,
      "step": 3960
    },
    {
      "epoch": 2.151761517615176,
      "grad_norm": 0.0,
      "learning_rate": 2.2443120573588323e-05,
      "loss": 0.6078,
      "step": 3970
    },
    {
      "epoch": 2.1571815718157183,
      "grad_norm": 0.0,
      "learning_rate": 2.218053148832872e-05,
      "loss": 0.5503,
      "step": 3980
    },
    {
      "epoch": 2.16260162601626,
      "grad_norm": 0.0,
      "learning_rate": 2.1919049062656476e-05,
      "loss": 0.4805,
      "step": 3990
    },
    {
      "epoch": 2.168021680216802,
      "grad_norm": 0.0,
      "learning_rate": 2.1658683698353095e-05,
      "loss": 0.5476,
      "step": 4000
    },
    {
      "epoch": 2.1734417344173442,
      "grad_norm": 0.0,
      "learning_rate": 2.1399445752763374e-05,
      "loss": 0.4429,
      "step": 4010
    },
    {
      "epoch": 2.178861788617886,
      "grad_norm": 0.0,
      "learning_rate": 2.1141345538383306e-05,
      "loss": 0.5418,
      "step": 4020
    },
    {
      "epoch": 2.1842818428184283,
      "grad_norm": 0.0,
      "learning_rate": 2.0884393322449895e-05,
      "loss": 0.55,
      "step": 4030
    },
    {
      "epoch": 2.18970189701897,
      "grad_norm": 0.0,
      "learning_rate": 2.0628599326532738e-05,
      "loss": 0.4787,
      "step": 4040
    },
    {
      "epoch": 2.1951219512195124,
      "grad_norm": 0.0,
      "learning_rate": 2.0373973726127378e-05,
      "loss": 0.5278,
      "step": 4050
    },
    {
      "epoch": 2.200542005420054,
      "grad_norm": 0.0,
      "learning_rate": 2.0120526650250533e-05,
      "loss": 0.5121,
      "step": 4060
    },
    {
      "epoch": 2.205962059620596,
      "grad_norm": 0.0,
      "learning_rate": 1.9868268181037185e-05,
      "loss": 0.5864,
      "step": 4070
    },
    {
      "epoch": 2.2113821138211383,
      "grad_norm": 0.0,
      "learning_rate": 1.961720835333949e-05,
      "loss": 0.6388,
      "step": 4080
    },
    {
      "epoch": 2.21680216802168,
      "grad_norm": 0.0,
      "learning_rate": 1.9367357154327576e-05,
      "loss": 0.576,
      "step": 4090
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.0,
      "learning_rate": 1.91187245230923e-05,
      "loss": 0.56,
      "step": 4100
    },
    {
      "epoch": 2.227642276422764,
      "grad_norm": 0.0,
      "learning_rate": 1.8871320350249833e-05,
      "loss": 0.5696,
      "step": 4110
    },
    {
      "epoch": 2.2330623306233064,
      "grad_norm": 0.0,
      "learning_rate": 1.862515447754819e-05,
      "loss": 0.5115,
      "step": 4120
    },
    {
      "epoch": 2.238482384823848,
      "grad_norm": 0.0,
      "learning_rate": 1.8380236697475832e-05,
      "loss": 0.5578,
      "step": 4130
    },
    {
      "epoch": 2.2439024390243905,
      "grad_norm": 0.0,
      "learning_rate": 1.813657675287198e-05,
      "loss": 0.5048,
      "step": 4140
    },
    {
      "epoch": 2.2493224932249323,
      "grad_norm": 0.0,
      "learning_rate": 1.789418433653912e-05,
      "loss": 0.5528,
      "step": 4150
    },
    {
      "epoch": 2.254742547425474,
      "grad_norm": 0.0,
      "learning_rate": 1.7653069090857434e-05,
      "loss": 0.4945,
      "step": 4160
    },
    {
      "epoch": 2.2601626016260163,
      "grad_norm": 0.0,
      "learning_rate": 1.7413240607401164e-05,
      "loss": 0.4937,
      "step": 4170
    },
    {
      "epoch": 2.265582655826558,
      "grad_norm": 0.0,
      "learning_rate": 1.717470842655715e-05,
      "loss": 0.5175,
      "step": 4180
    },
    {
      "epoch": 2.2710027100271004,
      "grad_norm": 0.0,
      "learning_rate": 1.693748203714522e-05,
      "loss": 0.5842,
      "step": 4190
    },
    {
      "epoch": 2.2764227642276422,
      "grad_norm": 0.0,
      "learning_rate": 1.6701570876040777e-05,
      "loss": 0.5089,
      "step": 4200
    },
    {
      "epoch": 2.281842818428184,
      "grad_norm": 0.0,
      "learning_rate": 1.6466984327799385e-05,
      "loss": 0.5788,
      "step": 4210
    },
    {
      "epoch": 2.2872628726287263,
      "grad_norm": 0.0,
      "learning_rate": 1.6233731724283492e-05,
      "loss": 0.5136,
      "step": 4220
    },
    {
      "epoch": 2.292682926829268,
      "grad_norm": 0.0,
      "learning_rate": 1.6001822344291124e-05,
      "loss": 0.4981,
      "step": 4230
    },
    {
      "epoch": 2.2981029810298104,
      "grad_norm": 0.0,
      "learning_rate": 1.5771265413186832e-05,
      "loss": 0.5172,
      "step": 4240
    },
    {
      "epoch": 2.303523035230352,
      "grad_norm": 0.0,
      "learning_rate": 1.5542070102534706e-05,
      "loss": 0.5126,
      "step": 4250
    },
    {
      "epoch": 2.3089430894308944,
      "grad_norm": 0.0,
      "learning_rate": 1.5314245529733507e-05,
      "loss": 0.5028,
      "step": 4260
    },
    {
      "epoch": 2.3143631436314362,
      "grad_norm": 0.0,
      "learning_rate": 1.5087800757653997e-05,
      "loss": 0.4974,
      "step": 4270
    },
    {
      "epoch": 2.3197831978319785,
      "grad_norm": 0.0,
      "learning_rate": 1.48627447942784e-05,
      "loss": 0.4955,
      "step": 4280
    },
    {
      "epoch": 2.3252032520325203,
      "grad_norm": 0.0,
      "learning_rate": 1.4639086592342099e-05,
      "loss": 0.5425,
      "step": 4290
    },
    {
      "epoch": 2.330623306233062,
      "grad_norm": 0.0,
      "learning_rate": 1.4416835048977445e-05,
      "loss": 0.5073,
      "step": 4300
    },
    {
      "epoch": 2.3360433604336044,
      "grad_norm": 0.0,
      "learning_rate": 1.419599900535985e-05,
      "loss": 0.6215,
      "step": 4310
    },
    {
      "epoch": 2.341463414634146,
      "grad_norm": 0.0,
      "learning_rate": 1.3976587246356116e-05,
      "loss": 0.5859,
      "step": 4320
    },
    {
      "epoch": 2.3468834688346885,
      "grad_norm": 0.0,
      "learning_rate": 1.3758608500174908e-05,
      "loss": 0.497,
      "step": 4330
    },
    {
      "epoch": 2.3523035230352303,
      "grad_norm": 0.0,
      "learning_rate": 1.3542071438019622e-05,
      "loss": 0.5852,
      "step": 4340
    },
    {
      "epoch": 2.3577235772357725,
      "grad_norm": 0.0,
      "learning_rate": 1.3326984673743353e-05,
      "loss": 0.4882,
      "step": 4350
    },
    {
      "epoch": 2.3631436314363143,
      "grad_norm": 0.0,
      "learning_rate": 1.3113356763506363e-05,
      "loss": 0.5861,
      "step": 4360
    },
    {
      "epoch": 2.3685636856368566,
      "grad_norm": 0.0,
      "learning_rate": 1.2901196205435578e-05,
      "loss": 0.5987,
      "step": 4370
    },
    {
      "epoch": 2.3739837398373984,
      "grad_norm": 0.0,
      "learning_rate": 1.2690511439286623e-05,
      "loss": 0.5111,
      "step": 4380
    },
    {
      "epoch": 2.3794037940379402,
      "grad_norm": 0.0,
      "learning_rate": 1.248131084610804e-05,
      "loss": 0.555,
      "step": 4390
    },
    {
      "epoch": 2.3848238482384825,
      "grad_norm": 0.0,
      "learning_rate": 1.227360274790793e-05,
      "loss": 0.5821,
      "step": 4400
    },
    {
      "epoch": 2.3902439024390243,
      "grad_norm": 0.0,
      "learning_rate": 1.206739540732288e-05,
      "loss": 0.5988,
      "step": 4410
    },
    {
      "epoch": 2.3956639566395665,
      "grad_norm": 0.0,
      "learning_rate": 1.1862697027289254e-05,
      "loss": 0.5745,
      "step": 4420
    },
    {
      "epoch": 2.4010840108401084,
      "grad_norm": 0.0,
      "learning_rate": 1.1659515750716955e-05,
      "loss": 0.481,
      "step": 4430
    },
    {
      "epoch": 2.40650406504065,
      "grad_norm": 0.0,
      "learning_rate": 1.1457859660165404e-05,
      "loss": 0.5362,
      "step": 4440
    },
    {
      "epoch": 2.4119241192411924,
      "grad_norm": 0.0,
      "learning_rate": 1.1257736777522072e-05,
      "loss": 0.6335,
      "step": 4450
    },
    {
      "epoch": 2.4173441734417342,
      "grad_norm": 0.0,
      "learning_rate": 1.1059155063683357e-05,
      "loss": 0.5322,
      "step": 4460
    },
    {
      "epoch": 2.4227642276422765,
      "grad_norm": 0.0,
      "learning_rate": 1.0862122418237907e-05,
      "loss": 0.5453,
      "step": 4470
    },
    {
      "epoch": 2.4281842818428183,
      "grad_norm": 0.0,
      "learning_rate": 1.0666646679152365e-05,
      "loss": 0.5394,
      "step": 4480
    },
    {
      "epoch": 2.4336043360433606,
      "grad_norm": 0.0,
      "learning_rate": 1.0472735622459545e-05,
      "loss": 0.6681,
      "step": 4490
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.0,
      "learning_rate": 1.028039696194919e-05,
      "loss": 0.5178,
      "step": 4500
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.0,
      "learning_rate": 1.0089638348861024e-05,
      "loss": 0.4984,
      "step": 4510
    },
    {
      "epoch": 2.4498644986449865,
      "grad_norm": 0.0,
      "learning_rate": 9.900467371580407e-06,
      "loss": 0.5896,
      "step": 4520
    },
    {
      "epoch": 2.4552845528455283,
      "grad_norm": 0.0,
      "learning_rate": 9.712891555336517e-06,
      "loss": 0.4909,
      "step": 4530
    },
    {
      "epoch": 2.4607046070460705,
      "grad_norm": 0.0,
      "learning_rate": 9.52691836190293e-06,
      "loss": 0.6343,
      "step": 4540
    },
    {
      "epoch": 2.4661246612466123,
      "grad_norm": 0.0,
      "learning_rate": 9.342555189300838e-06,
      "loss": 0.5115,
      "step": 4550
    },
    {
      "epoch": 2.4715447154471546,
      "grad_norm": 0.0,
      "learning_rate": 9.159809371504723e-06,
      "loss": 0.5469,
      "step": 4560
    },
    {
      "epoch": 2.4769647696476964,
      "grad_norm": 0.0,
      "learning_rate": 8.978688178150668e-06,
      "loss": 0.5606,
      "step": 4570
    },
    {
      "epoch": 2.4823848238482387,
      "grad_norm": 0.0,
      "learning_rate": 8.79919881424709e-06,
      "loss": 0.5203,
      "step": 4580
    },
    {
      "epoch": 2.4878048780487805,
      "grad_norm": 0.0,
      "learning_rate": 8.621348419888165e-06,
      "loss": 0.5111,
      "step": 4590
    },
    {
      "epoch": 2.4932249322493227,
      "grad_norm": 0.0,
      "learning_rate": 8.445144069969812e-06,
      "loss": 0.5115,
      "step": 4600
    },
    {
      "epoch": 2.4986449864498645,
      "grad_norm": 0.0,
      "learning_rate": 8.270592773908232e-06,
      "loss": 0.4814,
      "step": 4610
    },
    {
      "epoch": 2.5040650406504064,
      "grad_norm": 0.0,
      "learning_rate": 8.09770147536107e-06,
      "loss": 0.426,
      "step": 4620
    },
    {
      "epoch": 2.5094850948509486,
      "grad_norm": 0.0,
      "learning_rate": 7.926477051951192e-06,
      "loss": 0.4801,
      "step": 4630
    },
    {
      "epoch": 2.5149051490514904,
      "grad_norm": 0.0,
      "learning_rate": 7.756926314993135e-06,
      "loss": 0.6409,
      "step": 4640
    },
    {
      "epoch": 2.5203252032520327,
      "grad_norm": 0.0,
      "learning_rate": 7.589056009222089e-06,
      "loss": 0.5287,
      "step": 4650
    },
    {
      "epoch": 2.5257452574525745,
      "grad_norm": 0.0,
      "learning_rate": 7.422872812525633e-06,
      "loss": 0.5553,
      "step": 4660
    },
    {
      "epoch": 2.5311653116531163,
      "grad_norm": 0.0,
      "learning_rate": 7.258383335678098e-06,
      "loss": 0.5598,
      "step": 4670
    },
    {
      "epoch": 2.5365853658536586,
      "grad_norm": 0.0,
      "learning_rate": 7.095594122077542e-06,
      "loss": 0.5477,
      "step": 4680
    },
    {
      "epoch": 2.5420054200542004,
      "grad_norm": 0.0,
      "learning_rate": 6.934511647485503e-06,
      "loss": 0.5337,
      "step": 4690
    },
    {
      "epoch": 2.5474254742547426,
      "grad_norm": 0.0,
      "learning_rate": 6.775142319769351e-06,
      "loss": 0.4678,
      "step": 4700
    },
    {
      "epoch": 2.5528455284552845,
      "grad_norm": 0.0,
      "learning_rate": 6.61749247864743e-06,
      "loss": 0.5446,
      "step": 4710
    },
    {
      "epoch": 2.5582655826558267,
      "grad_norm": 0.0,
      "learning_rate": 6.461568395436812e-06,
      "loss": 0.5185,
      "step": 4720
    },
    {
      "epoch": 2.5636856368563685,
      "grad_norm": 0.0,
      "learning_rate": 6.307376272803861e-06,
      "loss": 0.5106,
      "step": 4730
    },
    {
      "epoch": 2.569105691056911,
      "grad_norm": 0.0,
      "learning_rate": 6.154922244517486e-06,
      "loss": 0.535,
      "step": 4740
    },
    {
      "epoch": 2.5745257452574526,
      "grad_norm": 0.0,
      "learning_rate": 6.004212375205115e-06,
      "loss": 0.6075,
      "step": 4750
    },
    {
      "epoch": 2.5799457994579944,
      "grad_norm": 0.0,
      "learning_rate": 5.855252660111482e-06,
      "loss": 0.4473,
      "step": 4760
    },
    {
      "epoch": 2.5853658536585367,
      "grad_norm": 0.0,
      "learning_rate": 5.708049024860085e-06,
      "loss": 0.5698,
      "step": 4770
    },
    {
      "epoch": 2.5907859078590785,
      "grad_norm": 0.0,
      "learning_rate": 5.562607325217533e-06,
      "loss": 0.534,
      "step": 4780
    },
    {
      "epoch": 2.5962059620596207,
      "grad_norm": 0.0,
      "learning_rate": 5.418933346860538e-06,
      "loss": 0.5458,
      "step": 4790
    },
    {
      "epoch": 2.6016260162601625,
      "grad_norm": 0.0,
      "learning_rate": 5.277032805145799e-06,
      "loss": 0.5138,
      "step": 4800
    },
    {
      "epoch": 2.6070460704607044,
      "grad_norm": 0.0,
      "learning_rate": 5.136911344882622e-06,
      "loss": 0.5076,
      "step": 4810
    },
    {
      "epoch": 2.6124661246612466,
      "grad_norm": 0.0,
      "learning_rate": 4.998574540108392e-06,
      "loss": 0.5299,
      "step": 4820
    },
    {
      "epoch": 2.617886178861789,
      "grad_norm": 0.0,
      "learning_rate": 4.862027893866822e-06,
      "loss": 0.4938,
      "step": 4830
    },
    {
      "epoch": 2.6233062330623307,
      "grad_norm": 0.0,
      "learning_rate": 4.7272768379890365e-06,
      "loss": 0.5444,
      "step": 4840
    },
    {
      "epoch": 2.6287262872628725,
      "grad_norm": 0.0,
      "learning_rate": 4.594326732877524e-06,
      "loss": 0.5202,
      "step": 4850
    },
    {
      "epoch": 2.6341463414634148,
      "grad_norm": 0.0,
      "learning_rate": 4.463182867292865e-06,
      "loss": 0.5111,
      "step": 4860
    },
    {
      "epoch": 2.6395663956639566,
      "grad_norm": 0.0,
      "learning_rate": 4.333850458143357e-06,
      "loss": 0.5141,
      "step": 4870
    },
    {
      "epoch": 2.644986449864499,
      "grad_norm": 0.0,
      "learning_rate": 4.206334650277483e-06,
      "loss": 0.4874,
      "step": 4880
    },
    {
      "epoch": 2.6504065040650406,
      "grad_norm": 0.0,
      "learning_rate": 4.080640516279266e-06,
      "loss": 0.4867,
      "step": 4890
    },
    {
      "epoch": 2.6558265582655824,
      "grad_norm": 0.0,
      "learning_rate": 3.956773056266466e-06,
      "loss": 0.5755,
      "step": 4900
    },
    {
      "epoch": 2.6612466124661247,
      "grad_norm": 0.0,
      "learning_rate": 3.8347371976916516e-06,
      "loss": 0.5755,
      "step": 4910
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.0,
      "learning_rate": 3.714537795146261e-06,
      "loss": 0.523,
      "step": 4920
    },
    {
      "epoch": 2.6720867208672088,
      "grad_norm": 0.0,
      "learning_rate": 3.5961796301674012e-06,
      "loss": 0.4475,
      "step": 4930
    },
    {
      "epoch": 2.6775067750677506,
      "grad_norm": 0.0,
      "learning_rate": 3.479667411047677e-06,
      "loss": 0.4947,
      "step": 4940
    },
    {
      "epoch": 2.682926829268293,
      "grad_norm": 0.0,
      "learning_rate": 3.3650057726479e-06,
      "loss": 0.5437,
      "step": 4950
    },
    {
      "epoch": 2.6883468834688347,
      "grad_norm": 0.0,
      "learning_rate": 3.2521992762127085e-06,
      "loss": 0.6523,
      "step": 4960
    },
    {
      "epoch": 2.693766937669377,
      "grad_norm": 0.0,
      "learning_rate": 3.141252409189099e-06,
      "loss": 0.5232,
      "step": 4970
    },
    {
      "epoch": 2.6991869918699187,
      "grad_norm": 0.0,
      "learning_rate": 3.032169585047956e-06,
      "loss": 0.5144,
      "step": 4980
    },
    {
      "epoch": 2.7046070460704605,
      "grad_norm": 0.0,
      "learning_rate": 2.92495514310846e-06,
      "loss": 0.5177,
      "step": 4990
    },
    {
      "epoch": 2.710027100271003,
      "grad_norm": 0.0,
      "learning_rate": 2.819613348365463e-06,
      "loss": 0.4604,
      "step": 5000
    },
    {
      "epoch": 2.7154471544715446,
      "grad_norm": 0.0,
      "learning_rate": 2.7161483913198392e-06,
      "loss": 0.5745,
      "step": 5010
    },
    {
      "epoch": 2.720867208672087,
      "grad_norm": 0.0,
      "learning_rate": 2.6145643878117744e-06,
      "loss": 0.4852,
      "step": 5020
    },
    {
      "epoch": 2.7262872628726287,
      "grad_norm": 0.0,
      "learning_rate": 2.5148653788570676e-06,
      "loss": 0.6759,
      "step": 5030
    },
    {
      "epoch": 2.7317073170731705,
      "grad_norm": 0.0,
      "learning_rate": 2.4170553304863287e-06,
      "loss": 0.5064,
      "step": 5040
    },
    {
      "epoch": 2.7371273712737128,
      "grad_norm": 0.0,
      "learning_rate": 2.321138133587264e-06,
      "loss": 0.4882,
      "step": 5050
    },
    {
      "epoch": 2.742547425474255,
      "grad_norm": 0.0,
      "learning_rate": 2.227117603749862e-06,
      "loss": 0.5716,
      "step": 5060
    },
    {
      "epoch": 2.747967479674797,
      "grad_norm": 0.0,
      "learning_rate": 2.13499748111462e-06,
      "loss": 0.5078,
      "step": 5070
    },
    {
      "epoch": 2.7533875338753386,
      "grad_norm": 0.0,
      "learning_rate": 2.044781430223758e-06,
      "loss": 0.4927,
      "step": 5080
    },
    {
      "epoch": 2.758807588075881,
      "grad_norm": 0.0,
      "learning_rate": 1.956473039875445e-06,
      "loss": 0.6156,
      "step": 5090
    },
    {
      "epoch": 2.7642276422764227,
      "grad_norm": 0.0,
      "learning_rate": 1.8700758229810266e-06,
      "loss": 0.6286,
      "step": 5100
    },
    {
      "epoch": 2.769647696476965,
      "grad_norm": 0.0,
      "learning_rate": 1.785593216425313e-06,
      "loss": 0.5391,
      "step": 5110
    },
    {
      "epoch": 2.7750677506775068,
      "grad_norm": 0.0,
      "learning_rate": 1.7030285809298107e-06,
      "loss": 0.5114,
      "step": 5120
    },
    {
      "epoch": 2.7804878048780486,
      "grad_norm": 0.0,
      "learning_rate": 1.6223852009190854e-06,
      "loss": 0.4548,
      "step": 5130
    },
    {
      "epoch": 2.785907859078591,
      "grad_norm": 0.0,
      "learning_rate": 1.5436662843900718e-06,
      "loss": 0.5117,
      "step": 5140
    },
    {
      "epoch": 2.7913279132791327,
      "grad_norm": 0.0,
      "learning_rate": 1.4668749627844536e-06,
      "loss": 0.5998,
      "step": 5150
    },
    {
      "epoch": 2.796747967479675,
      "grad_norm": 0.0,
      "learning_rate": 1.3920142908641353e-06,
      "loss": 0.5348,
      "step": 5160
    },
    {
      "epoch": 2.8021680216802167,
      "grad_norm": 0.0,
      "learning_rate": 1.3190872465896786e-06,
      "loss": 0.5632,
      "step": 5170
    },
    {
      "epoch": 2.8075880758807585,
      "grad_norm": 0.0,
      "learning_rate": 1.2480967310018743e-06,
      "loss": 0.5922,
      "step": 5180
    },
    {
      "epoch": 2.813008130081301,
      "grad_norm": 0.0,
      "learning_rate": 1.1790455681063018e-06,
      "loss": 0.4896,
      "step": 5190
    },
    {
      "epoch": 2.818428184281843,
      "grad_norm": 0.0,
      "learning_rate": 1.1119365047610409e-06,
      "loss": 0.5172,
      "step": 5200
    },
    {
      "epoch": 2.823848238482385,
      "grad_norm": 0.0,
      "learning_rate": 1.046772210567354e-06,
      "loss": 0.5374,
      "step": 5210
    },
    {
      "epoch": 2.8292682926829267,
      "grad_norm": 0.0,
      "learning_rate": 9.835552777635092e-07,
      "loss": 0.5543,
      "step": 5220
    },
    {
      "epoch": 2.834688346883469,
      "grad_norm": 0.0,
      "learning_rate": 9.222882211216588e-07,
      "loss": 0.5452,
      "step": 5230
    },
    {
      "epoch": 2.8401084010840107,
      "grad_norm": 0.0,
      "learning_rate": 8.62973477847806e-07,
      "loss": 0.5049,
      "step": 5240
    },
    {
      "epoch": 2.845528455284553,
      "grad_norm": 0.0,
      "learning_rate": 8.05613407484851e-07,
      "loss": 0.5682,
      "step": 5250
    },
    {
      "epoch": 2.850948509485095,
      "grad_norm": 0.0,
      "learning_rate": 7.502102918187203e-07,
      "loss": 0.448,
      "step": 5260
    },
    {
      "epoch": 2.8563685636856366,
      "grad_norm": 0.0,
      "learning_rate": 6.967663347876019e-07,
      "loss": 0.4991,
      "step": 5270
    },
    {
      "epoch": 2.861788617886179,
      "grad_norm": 0.0,
      "learning_rate": 6.452836623942859e-07,
      "loss": 0.5162,
      "step": 5280
    },
    {
      "epoch": 2.867208672086721,
      "grad_norm": 0.0,
      "learning_rate": 5.957643226215615e-07,
      "loss": 0.4567,
      "step": 5290
    },
    {
      "epoch": 2.872628726287263,
      "grad_norm": 0.0,
      "learning_rate": 5.482102853507809e-07,
      "loss": 0.5041,
      "step": 5300
    },
    {
      "epoch": 2.8780487804878048,
      "grad_norm": 0.0,
      "learning_rate": 5.02623442283473e-07,
      "loss": 0.4718,
      "step": 5310
    },
    {
      "epoch": 2.883468834688347,
      "grad_norm": 0.0,
      "learning_rate": 4.590056068661197e-07,
      "loss": 0.5498,
      "step": 5320
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.0,
      "learning_rate": 4.1735851421796925e-07,
      "loss": 0.5377,
      "step": 5330
    },
    {
      "epoch": 2.894308943089431,
      "grad_norm": 0.0,
      "learning_rate": 3.776838210620692e-07,
      "loss": 0.5376,
      "step": 5340
    },
    {
      "epoch": 2.899728997289973,
      "grad_norm": 0.0,
      "learning_rate": 3.39983105659325e-07,
      "loss": 0.5345,
      "step": 5350
    },
    {
      "epoch": 2.9051490514905147,
      "grad_norm": 0.0,
      "learning_rate": 3.0425786774572753e-07,
      "loss": 0.5777,
      "step": 5360
    },
    {
      "epoch": 2.910569105691057,
      "grad_norm": 0.0,
      "learning_rate": 2.705095284726844e-07,
      "loss": 0.5314,
      "step": 5370
    },
    {
      "epoch": 2.915989159891599,
      "grad_norm": 0.0,
      "learning_rate": 2.3873943035050417e-07,
      "loss": 0.5036,
      "step": 5380
    },
    {
      "epoch": 2.921409214092141,
      "grad_norm": 0.0,
      "learning_rate": 2.089488371949888e-07,
      "loss": 0.5007,
      "step": 5390
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.0,
      "learning_rate": 1.8113893407713523e-07,
      "loss": 0.5986,
      "step": 5400
    },
    {
      "epoch": 2.9322493224932247,
      "grad_norm": 0.0,
      "learning_rate": 1.5531082727602285e-07,
      "loss": 0.653,
      "step": 5410
    },
    {
      "epoch": 2.937669376693767,
      "grad_norm": 0.0,
      "learning_rate": 1.3146554423478785e-07,
      "loss": 0.523,
      "step": 5420
    },
    {
      "epoch": 2.943089430894309,
      "grad_norm": 0.0,
      "learning_rate": 1.0960403351976123e-07,
      "loss": 0.4574,
      "step": 5430
    },
    {
      "epoch": 2.948509485094851,
      "grad_norm": 0.0,
      "learning_rate": 8.972716478272691e-08,
      "loss": 0.6004,
      "step": 5440
    },
    {
      "epoch": 2.953929539295393,
      "grad_norm": 0.0,
      "learning_rate": 7.183572872632715e-08,
      "loss": 0.5666,
      "step": 5450
    },
    {
      "epoch": 2.959349593495935,
      "grad_norm": 0.0,
      "learning_rate": 5.593043707262102e-08,
      "loss": 0.5306,
      "step": 5460
    },
    {
      "epoch": 2.964769647696477,
      "grad_norm": 0.0,
      "learning_rate": 4.201192253474595e-08,
      "loss": 0.4908,
      "step": 5470
    },
    {
      "epoch": 2.970189701897019,
      "grad_norm": 0.0,
      "learning_rate": 3.0080738791782304e-08,
      "loss": 0.4865,
      "step": 5480
    },
    {
      "epoch": 2.975609756097561,
      "grad_norm": 0.0,
      "learning_rate": 2.0137360466704337e-08,
      "loss": 0.4843,
      "step": 5490
    },
    {
      "epoch": 2.9810298102981028,
      "grad_norm": 0.0,
      "learning_rate": 1.2182183107506406e-08,
      "loss": 0.5347,
      "step": 5500
    },
    {
      "epoch": 2.986449864498645,
      "grad_norm": 0.0,
      "learning_rate": 6.215523171465565e-09,
      "loss": 0.567,
      "step": 5510
    },
    {
      "epoch": 2.991869918699187,
      "grad_norm": 0.0,
      "learning_rate": 2.2376180125571743e-09,
      "loss": 0.5512,
      "step": 5520
    },
    {
      "epoch": 2.997289972899729,
      "grad_norm": 0.0,
      "learning_rate": 2.4862587201246457e-10,
      "loss": 0.5884,
      "step": 5530
    },
    {
      "epoch": 3.0,
      "step": 5535,
      "total_flos": 3.0626432822122906e+18,
      "train_loss": 0.8061501924692238,
      "train_runtime": 22473.7337,
      "train_samples_per_second": 0.246,
      "train_steps_per_second": 0.246
    }
  ],
  "logging_steps": 10,
  "max_steps": 5535,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 2500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.0626432822122906e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}