{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1941,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015455950540958269,
      "grad_norm": 9.875,
      "learning_rate": 7.627118644067798e-06,
      "loss": 3.7349,
      "step": 10
    },
    {
      "epoch": 0.030911901081916538,
      "grad_norm": 8.625,
      "learning_rate": 1.6101694915254237e-05,
      "loss": 3.5453,
      "step": 20
    },
    {
      "epoch": 0.04636785162287481,
      "grad_norm": 7.40625,
      "learning_rate": 2.457627118644068e-05,
      "loss": 3.441,
      "step": 30
    },
    {
      "epoch": 0.061823802163833076,
      "grad_norm": 5.71875,
      "learning_rate": 3.305084745762712e-05,
      "loss": 3.4128,
      "step": 40
    },
    {
      "epoch": 0.07727975270479134,
      "grad_norm": 6.21875,
      "learning_rate": 4.152542372881356e-05,
      "loss": 3.1485,
      "step": 50
    },
    {
      "epoch": 0.09273570324574962,
      "grad_norm": 5.0,
      "learning_rate": 5e-05,
      "loss": 3.0359,
      "step": 60
    },
    {
      "epoch": 0.10819165378670788,
      "grad_norm": 4.1875,
      "learning_rate": 4.999651694417285e-05,
      "loss": 3.2478,
      "step": 70
    },
    {
      "epoch": 0.12364760432766615,
      "grad_norm": 3.5,
      "learning_rate": 4.9986068747225644e-05,
      "loss": 3.1223,
      "step": 80
    },
    {
      "epoch": 0.1391035548686244,
      "grad_norm": 3.890625,
      "learning_rate": 4.9968658320490636e-05,
      "loss": 3.0477,
      "step": 90
    },
    {
      "epoch": 0.1545595054095827,
      "grad_norm": 3.859375,
      "learning_rate": 4.994429051528689e-05,
      "loss": 3.0829,
      "step": 100
    },
    {
      "epoch": 0.17001545595054096,
      "grad_norm": 3.734375,
      "learning_rate": 4.991297212156848e-05,
      "loss": 2.9565,
      "step": 110
    },
    {
      "epoch": 0.18547140649149924,
      "grad_norm": 3.671875,
      "learning_rate": 4.9874711866032495e-05,
      "loss": 3.142,
      "step": 120
    },
    {
      "epoch": 0.2009273570324575,
      "grad_norm": 5.15625,
      "learning_rate": 4.982952040968744e-05,
      "loss": 3.0807,
      "step": 130
    },
    {
      "epoch": 0.21638330757341576,
      "grad_norm": 4.625,
      "learning_rate": 4.977741034488251e-05,
      "loss": 2.9314,
      "step": 140
    },
    {
      "epoch": 0.23183925811437403,
      "grad_norm": 4.125,
      "learning_rate": 4.971839619179892e-05,
      "loss": 3.1325,
      "step": 150
    },
    {
      "epoch": 0.2472952086553323,
      "grad_norm": 4.65625,
      "learning_rate": 4.965249439440384e-05,
      "loss": 2.8936,
      "step": 160
    },
    {
      "epoch": 0.26275115919629055,
      "grad_norm": 5.0625,
      "learning_rate": 4.957972331586843e-05,
      "loss": 2.9205,
      "step": 170
    },
    {
      "epoch": 0.2782071097372488,
      "grad_norm": 4.34375,
      "learning_rate": 4.950010323345103e-05,
      "loss": 2.9776,
      "step": 180
    },
    {
      "epoch": 0.2936630602782071,
      "grad_norm": 4.53125,
      "learning_rate": 4.941365633284699e-05,
      "loss": 2.9432,
      "step": 190
    },
    {
      "epoch": 0.3091190108191654,
      "grad_norm": 4.25,
      "learning_rate": 4.932040670200677e-05,
      "loss": 2.9593,
      "step": 200
    },
    {
      "epoch": 0.32457496136012365,
      "grad_norm": 5.03125,
      "learning_rate": 4.9220380324424006e-05,
      "loss": 2.9826,
      "step": 210
    },
    {
      "epoch": 0.3400309119010819,
      "grad_norm": 4.28125,
      "learning_rate": 4.911360507189526e-05,
      "loss": 2.8009,
      "step": 220
    },
    {
      "epoch": 0.3554868624420402,
      "grad_norm": 4.5625,
      "learning_rate": 4.900011069675378e-05,
      "loss": 2.7733,
      "step": 230
    },
    {
      "epoch": 0.37094281298299847,
      "grad_norm": 3.609375,
      "learning_rate": 4.8879928823579136e-05,
      "loss": 2.8954,
      "step": 240
    },
    {
      "epoch": 0.38639876352395675,
      "grad_norm": 3.1875,
      "learning_rate": 4.875309294038523e-05,
      "loss": 2.8972,
      "step": 250
    },
    {
      "epoch": 0.401854714064915,
      "grad_norm": 4.5,
      "learning_rate": 4.8619638389289026e-05,
      "loss": 2.9377,
      "step": 260
    },
    {
      "epoch": 0.41731066460587324,
      "grad_norm": 3.796875,
      "learning_rate": 4.8479602356662665e-05,
      "loss": 2.8918,
      "step": 270
    },
    {
      "epoch": 0.4327666151468315,
      "grad_norm": 5.0625,
      "learning_rate": 4.833302386277171e-05,
      "loss": 2.8245,
      "step": 280
    },
    {
      "epoch": 0.4482225656877898,
      "grad_norm": 3.03125,
      "learning_rate": 4.817994375090233e-05,
      "loss": 2.8384,
      "step": 290
    },
    {
      "epoch": 0.46367851622874806,
      "grad_norm": 3.390625,
      "learning_rate": 4.80204046759806e-05,
      "loss": 2.9315,
      "step": 300
    },
    {
      "epoch": 0.47913446676970634,
      "grad_norm": 3.75,
      "learning_rate": 4.785445109268687e-05,
      "loss": 2.8416,
      "step": 310
    },
    {
      "epoch": 0.4945904173106646,
      "grad_norm": 3.734375,
      "learning_rate": 4.768212924306877e-05,
      "loss": 2.8615,
      "step": 320
    },
    {
      "epoch": 0.5100463678516228,
      "grad_norm": 3.28125,
      "learning_rate": 4.75034871436561e-05,
      "loss": 2.8885,
      "step": 330
    },
    {
      "epoch": 0.5255023183925811,
      "grad_norm": 3.96875,
      "learning_rate": 4.7318574572081275e-05,
      "loss": 2.8238,
      "step": 340
    },
    {
      "epoch": 0.5409582689335394,
      "grad_norm": 4.53125,
      "learning_rate": 4.712744305320911e-05,
      "loss": 2.816,
      "step": 350
    },
    {
      "epoch": 0.5564142194744977,
      "grad_norm": 2.9375,
      "learning_rate": 4.6930145844779626e-05,
      "loss": 2.6657,
      "step": 360
    },
    {
      "epoch": 0.5718701700154559,
      "grad_norm": 3.78125,
      "learning_rate": 4.672673792256816e-05,
      "loss": 2.8454,
      "step": 370
    },
    {
      "epoch": 0.5873261205564142,
      "grad_norm": 3.375,
      "learning_rate": 4.651727596506661e-05,
      "loss": 2.8263,
      "step": 380
    },
    {
      "epoch": 0.6027820710973725,
      "grad_norm": 4.03125,
      "learning_rate": 4.63018183376903e-05,
      "loss": 2.9417,
      "step": 390
    },
    {
      "epoch": 0.6182380216383307,
      "grad_norm": 3.453125,
      "learning_rate": 4.60804250765148e-05,
      "loss": 2.8449,
      "step": 400
    },
    {
      "epoch": 0.633693972179289,
      "grad_norm": 3.703125,
      "learning_rate": 4.5853157871547184e-05,
      "loss": 2.8171,
      "step": 410
    },
    {
      "epoch": 0.6491499227202473,
      "grad_norm": 3.25,
      "learning_rate": 4.562008004953644e-05,
      "loss": 2.9253,
      "step": 420
    },
    {
      "epoch": 0.6646058732612056,
      "grad_norm": 4.5,
      "learning_rate": 4.538125655632789e-05,
      "loss": 2.8203,
      "step": 430
    },
    {
      "epoch": 0.6800618238021638,
      "grad_norm": 3.71875,
      "learning_rate": 4.5136753938766286e-05,
      "loss": 2.8078,
      "step": 440
    },
    {
      "epoch": 0.6955177743431221,
      "grad_norm": 3.265625,
      "learning_rate": 4.4886640326152985e-05,
      "loss": 2.7962,
      "step": 450
    },
    {
      "epoch": 0.7109737248840804,
      "grad_norm": 4.1875,
      "learning_rate": 4.4630985411262054e-05,
      "loss": 2.7565,
      "step": 460
    },
    {
      "epoch": 0.7264296754250387,
      "grad_norm": 4.0,
      "learning_rate": 4.4369860430920776e-05,
      "loss": 2.7677,
      "step": 470
    },
    {
      "epoch": 0.7418856259659969,
      "grad_norm": 3.0,
      "learning_rate": 4.41033381461599e-05,
      "loss": 2.7879,
      "step": 480
    },
    {
      "epoch": 0.7573415765069552,
      "grad_norm": 4.71875,
      "learning_rate": 4.383149282193919e-05,
      "loss": 2.6332,
      "step": 490
    },
    {
      "epoch": 0.7727975270479135,
      "grad_norm": 4.96875,
      "learning_rate": 4.35544002064539e-05,
      "loss": 2.7762,
      "step": 500
    },
    {
      "epoch": 0.7882534775888718,
      "grad_norm": 4.375,
      "learning_rate": 4.327213751002794e-05,
      "loss": 2.8478,
      "step": 510
    },
    {
      "epoch": 0.80370942812983,
      "grad_norm": 4.0625,
      "learning_rate": 4.298478338359968e-05,
      "loss": 2.7518,
      "step": 520
    },
    {
      "epoch": 0.8191653786707882,
      "grad_norm": 4.40625,
      "learning_rate": 4.269241789680629e-05,
      "loss": 2.747,
      "step": 530
    },
    {
      "epoch": 0.8346213292117465,
      "grad_norm": 3.4375,
      "learning_rate": 4.239512251567275e-05,
      "loss": 2.7868,
      "step": 540
    },
    {
      "epoch": 0.8500772797527048,
      "grad_norm": 3.859375,
      "learning_rate": 4.2092980079911836e-05,
      "loss": 2.694,
      "step": 550
    },
    {
      "epoch": 0.865533230293663,
      "grad_norm": 5.34375,
      "learning_rate": 4.178607477984128e-05,
      "loss": 2.693,
      "step": 560
    },
    {
      "epoch": 0.8809891808346213,
      "grad_norm": 6.53125,
      "learning_rate": 4.147449213292457e-05,
      "loss": 2.7596,
      "step": 570
    },
    {
      "epoch": 0.8964451313755796,
      "grad_norm": 3.875,
      "learning_rate": 4.115831895994203e-05,
      "loss": 2.7299,
      "step": 580
    },
    {
      "epoch": 0.9119010819165378,
      "grad_norm": 3.796875,
      "learning_rate": 4.0837643360798666e-05,
      "loss": 2.8138,
      "step": 590
    },
    {
      "epoch": 0.9273570324574961,
      "grad_norm": 4.21875,
      "learning_rate": 4.051255468997562e-05,
      "loss": 2.7069,
      "step": 600
    },
    {
      "epoch": 0.9428129829984544,
      "grad_norm": 4.125,
      "learning_rate": 4.018314353163202e-05,
      "loss": 2.6954,
      "step": 610
    },
    {
      "epoch": 0.9582689335394127,
      "grad_norm": 3.78125,
      "learning_rate": 3.984950167436423e-05,
      "loss": 2.5214,
      "step": 620
    },
    {
      "epoch": 0.973724884080371,
      "grad_norm": 3.859375,
      "learning_rate": 3.9511722085629485e-05,
      "loss": 2.7825,
      "step": 630
    },
    {
      "epoch": 0.9891808346213292,
      "grad_norm": 3.234375,
      "learning_rate": 3.916989888584095e-05,
      "loss": 2.7346,
      "step": 640
    },
    {
      "epoch": 1.0046367851622875,
      "grad_norm": 3.9375,
      "learning_rate": 3.882412732214167e-05,
      "loss": 2.5301,
      "step": 650
    },
    {
      "epoch": 1.0200927357032457,
      "grad_norm": 3.65625,
      "learning_rate": 3.847450374186441e-05,
      "loss": 2.1356,
      "step": 660
    },
    {
      "epoch": 1.035548686244204,
      "grad_norm": 3.78125,
      "learning_rate": 3.812112556568508e-05,
      "loss": 2.0337,
      "step": 670
    },
    {
      "epoch": 1.0510046367851622,
      "grad_norm": 4.96875,
      "learning_rate": 3.776409126047692e-05,
      "loss": 2.1394,
      "step": 680
    },
    {
      "epoch": 1.0664605873261206,
      "grad_norm": 5.15625,
      "learning_rate": 3.740350031187332e-05,
      "loss": 1.9865,
      "step": 690
    },
    {
      "epoch": 1.0819165378670788,
      "grad_norm": 3.671875,
      "learning_rate": 3.7039453196546645e-05,
      "loss": 2.0067,
      "step": 700
    },
    {
      "epoch": 1.0973724884080371,
      "grad_norm": 3.65625,
      "learning_rate": 3.6672051354211035e-05,
      "loss": 2.2291,
      "step": 710
    },
    {
      "epoch": 1.1128284389489953,
      "grad_norm": 3.421875,
      "learning_rate": 3.6301397159356696e-05,
      "loss": 2.0544,
      "step": 720
    },
    {
      "epoch": 1.1282843894899537,
      "grad_norm": 4.3125,
      "learning_rate": 3.592759389272389e-05,
      "loss": 1.9978,
      "step": 730
    },
    {
      "epoch": 1.1437403400309119,
      "grad_norm": 4.1875,
      "learning_rate": 3.555074571252431e-05,
      "loss": 1.9575,
      "step": 740
    },
    {
      "epoch": 1.1591962905718702,
      "grad_norm": 4.34375,
      "learning_rate": 3.517095762541795e-05,
      "loss": 2.0236,
      "step": 750
    },
    {
      "epoch": 1.1746522411128284,
      "grad_norm": 4.34375,
      "learning_rate": 3.4788335457253595e-05,
      "loss": 1.9811,
      "step": 760
    },
    {
      "epoch": 1.1901081916537868,
      "grad_norm": 5.25,
      "learning_rate": 3.4402985823581046e-05,
      "loss": 2.0005,
      "step": 770
    },
    {
      "epoch": 1.205564142194745,
      "grad_norm": 4.875,
      "learning_rate": 3.401501609994326e-05,
      "loss": 1.9412,
      "step": 780
    },
    {
      "epoch": 1.2210200927357033,
      "grad_norm": 4.5625,
      "learning_rate": 3.362453439195679e-05,
      "loss": 2.0814,
      "step": 790
    },
    {
      "epoch": 1.2364760432766615,
      "grad_norm": 4.03125,
      "learning_rate": 3.323164950518868e-05,
      "loss": 1.9947,
      "step": 800
    },
    {
      "epoch": 1.2519319938176197,
      "grad_norm": 5.71875,
      "learning_rate": 3.283647091483849e-05,
      "loss": 1.9282,
      "step": 810
    },
    {
      "epoch": 1.267387944358578,
      "grad_norm": 4.40625,
      "learning_rate": 3.243910873523356e-05,
      "loss": 1.9283,
      "step": 820
    },
    {
      "epoch": 1.2828438948995364,
      "grad_norm": 6.84375,
      "learning_rate": 3.203967368914631e-05,
      "loss": 1.9607,
      "step": 830
    },
    {
      "epoch": 1.2982998454404946,
      "grad_norm": 6.15625,
      "learning_rate": 3.163827707694193e-05,
      "loss": 2.0587,
      "step": 840
    },
    {
      "epoch": 1.3137557959814528,
      "grad_norm": 4.84375,
      "learning_rate": 3.123503074556513e-05,
      "loss": 2.0544,
      "step": 850
    },
    {
      "epoch": 1.3292117465224111,
      "grad_norm": 4.40625,
      "learning_rate": 3.083004705737468e-05,
      "loss": 2.02,
      "step": 860
    },
    {
      "epoch": 1.3446676970633695,
      "grad_norm": 4.4375,
      "learning_rate": 3.0423438858834174e-05,
      "loss": 2.0169,
      "step": 870
    },
    {
      "epoch": 1.3601236476043277,
      "grad_norm": 5.15625,
      "learning_rate": 3.0015319449068025e-05,
      "loss": 2.0551,
      "step": 880
    },
    {
      "epoch": 1.3755795981452859,
      "grad_norm": 5.40625,
      "learning_rate": 2.960580254829131e-05,
      "loss": 1.8902,
      "step": 890
    },
    {
      "epoch": 1.3910355486862442,
      "grad_norm": 5.28125,
      "learning_rate": 2.919500226612224e-05,
      "loss": 2.1326,
      "step": 900
    },
    {
      "epoch": 1.4064914992272024,
      "grad_norm": 5.5625,
      "learning_rate": 2.8783033069786132e-05,
      "loss": 1.7767,
      "step": 910
    },
    {
      "epoch": 1.4219474497681608,
      "grad_norm": 4.40625,
      "learning_rate": 2.8370009752219788e-05,
      "loss": 1.9215,
      "step": 920
    },
    {
      "epoch": 1.437403400309119,
      "grad_norm": 5.78125,
      "learning_rate": 2.795604740008504e-05,
      "loss": 1.9508,
      "step": 930
    },
    {
      "epoch": 1.4528593508500773,
      "grad_norm": 4.59375,
      "learning_rate": 2.7541261361700514e-05,
      "loss": 2.0011,
      "step": 940
    },
    {
      "epoch": 1.4683153013910355,
      "grad_norm": 4.34375,
      "learning_rate": 2.7125767214900455e-05,
      "loss": 2.0049,
      "step": 950
    },
    {
      "epoch": 1.4837712519319939,
      "grad_norm": 4.875,
      "learning_rate": 2.67096807348296e-05,
      "loss": 2.0328,
      "step": 960
    },
    {
      "epoch": 1.499227202472952,
      "grad_norm": 5.375,
      "learning_rate": 2.629311786168306e-05,
      "loss": 2.0937,
      "step": 970
    },
    {
      "epoch": 1.5146831530139102,
      "grad_norm": 3.984375,
      "learning_rate": 2.5876194668400256e-05,
      "loss": 2.0844,
      "step": 980
    },
    {
      "epoch": 1.5301391035548686,
      "grad_norm": 4.46875,
      "learning_rate": 2.545902732832181e-05,
      "loss": 1.9383,
      "step": 990
    },
    {
      "epoch": 1.545595054095827,
      "grad_norm": 4.90625,
      "learning_rate": 2.5041732082818503e-05,
      "loss": 1.963,
      "step": 1000
    },
    {
      "epoch": 1.5610510046367851,
      "grad_norm": 4.09375,
      "learning_rate": 2.462442520890127e-05,
      "loss": 2.0792,
      "step": 1010
    },
    {
      "epoch": 1.5765069551777433,
      "grad_norm": 5.09375,
      "learning_rate": 2.4207222986821194e-05,
      "loss": 2.0617,
      "step": 1020
    },
    {
      "epoch": 1.5919629057187017,
      "grad_norm": 5.3125,
      "learning_rate": 2.379024166766876e-05,
      "loss": 1.9914,
      "step": 1030
    },
    {
      "epoch": 1.60741885625966,
      "grad_norm": 5.0,
      "learning_rate": 2.3373597440981027e-05,
      "loss": 1.9685,
      "step": 1040
    },
    {
      "epoch": 1.6228748068006182,
      "grad_norm": 5.96875,
      "learning_rate": 2.295740640236614e-05,
      "loss": 1.9599,
      "step": 1050
    },
    {
      "epoch": 1.6383307573415764,
      "grad_norm": 6.4375,
      "learning_rate": 2.2541784521153875e-05,
      "loss": 2.0456,
      "step": 1060
    },
    {
      "epoch": 1.6537867078825348,
      "grad_norm": 5.625,
      "learning_rate": 2.2126847608081434e-05,
      "loss": 1.9414,
      "step": 1070
    },
    {
      "epoch": 1.6692426584234932,
      "grad_norm": 5.25,
      "learning_rate": 2.1712711283023482e-05,
      "loss": 2.0061,
      "step": 1080
    },
    {
      "epoch": 1.6846986089644513,
      "grad_norm": 4.96875,
      "learning_rate": 2.1299490942775213e-05,
      "loss": 2.0977,
      "step": 1090
    },
    {
      "epoch": 1.7001545595054095,
      "grad_norm": 5.125,
      "learning_rate": 2.088730172889777e-05,
      "loss": 2.0847,
      "step": 1100
    },
    {
      "epoch": 1.7156105100463679,
      "grad_norm": 5.65625,
      "learning_rate": 2.0476258495634594e-05,
      "loss": 2.1084,
      "step": 1110
    },
    {
      "epoch": 1.7310664605873263,
      "grad_norm": 4.125,
      "learning_rate": 2.0066475777908005e-05,
      "loss": 1.9382,
      "step": 1120
    },
    {
      "epoch": 1.7465224111282844,
      "grad_norm": 6.40625,
      "learning_rate": 1.9658067759404625e-05,
      "loss": 2.0278,
      "step": 1130
    },
    {
      "epoch": 1.7619783616692426,
      "grad_norm": 5.5625,
      "learning_rate": 1.925114824075876e-05,
      "loss": 2.0224,
      "step": 1140
    },
    {
      "epoch": 1.7774343122102008,
      "grad_norm": 4.8125,
      "learning_rate": 1.8845830607842438e-05,
      "loss": 2.0613,
      "step": 1150
    },
    {
      "epoch": 1.7928902627511591,
      "grad_norm": 5.375,
      "learning_rate": 1.8442227800171125e-05,
      "loss": 1.9505,
      "step": 1160
    },
    {
      "epoch": 1.8083462132921175,
      "grad_norm": 5.46875,
      "learning_rate": 1.8040452279433716e-05,
      "loss": 1.9729,
      "step": 1170
    },
    {
      "epoch": 1.8238021638330757,
      "grad_norm": 4.59375,
      "learning_rate": 1.7640615998155693e-05,
      "loss": 1.9394,
      "step": 1180
    },
    {
      "epoch": 1.8392581143740339,
      "grad_norm": 4.875,
      "learning_rate": 1.7242830368504227e-05,
      "loss": 1.8939,
      "step": 1190
    },
    {
      "epoch": 1.8547140649149922,
      "grad_norm": 3.984375,
      "learning_rate": 1.6847206231243718e-05,
      "loss": 1.8111,
      "step": 1200
    },
    {
      "epoch": 1.8701700154559506,
      "grad_norm": 4.9375,
      "learning_rate": 1.6453853824850726e-05,
      "loss": 1.9524,
      "step": 1210
    },
    {
      "epoch": 1.8856259659969088,
      "grad_norm": 3.953125,
      "learning_rate": 1.606288275479652e-05,
      "loss": 2.0018,
      "step": 1220
    },
    {
      "epoch": 1.901081916537867,
      "grad_norm": 5.15625,
      "learning_rate": 1.5674401963006235e-05,
      "loss": 2.0096,
      "step": 1230
    },
    {
      "epoch": 1.9165378670788253,
      "grad_norm": 5.5625,
      "learning_rate": 1.5288519697502697e-05,
      "loss": 1.9857,
      "step": 1240
    },
    {
      "epoch": 1.9319938176197837,
      "grad_norm": 6.625,
      "learning_rate": 1.4905343482243775e-05,
      "loss": 1.9265,
      "step": 1250
    },
    {
      "epoch": 1.947449768160742,
      "grad_norm": 4.875,
      "learning_rate": 1.4524980087161438e-05,
      "loss": 2.0875,
      "step": 1260
    },
    {
      "epoch": 1.9629057187017,
      "grad_norm": 3.484375,
      "learning_rate": 1.4147535498410853e-05,
      "loss": 2.0144,
      "step": 1270
    },
    {
      "epoch": 1.9783616692426584,
      "grad_norm": 4.25,
      "learning_rate": 1.3773114888837957e-05,
      "loss": 1.9803,
      "step": 1280
    },
    {
      "epoch": 1.9938176197836168,
      "grad_norm": 4.90625,
      "learning_rate": 1.3401822588673636e-05,
      "loss": 2.0252,
      "step": 1290
    },
    {
      "epoch": 2.009273570324575,
      "grad_norm": 4.4375,
      "learning_rate": 1.3033762056462654e-05,
      "loss": 1.7441,
      "step": 1300
    },
    {
      "epoch": 2.024729520865533,
      "grad_norm": 3.84375,
      "learning_rate": 1.2669035850235536e-05,
      "loss": 1.58,
      "step": 1310
    },
    {
      "epoch": 2.0401854714064913,
      "grad_norm": 4.6875,
      "learning_rate": 1.230774559893131e-05,
      "loss": 1.5253,
      "step": 1320
    },
    {
      "epoch": 2.05564142194745,
      "grad_norm": 5.46875,
      "learning_rate": 1.1949991974079197e-05,
      "loss": 1.5205,
      "step": 1330
    },
    {
      "epoch": 2.071097372488408,
      "grad_norm": 5.1875,
      "learning_rate": 1.1595874661746986e-05,
      "loss": 1.6223,
      "step": 1340
    },
    {
      "epoch": 2.0865533230293662,
      "grad_norm": 4.90625,
      "learning_rate": 1.1245492334764169e-05,
      "loss": 1.5583,
      "step": 1350
    },
    {
      "epoch": 2.1020092735703244,
      "grad_norm": 4.90625,
      "learning_rate": 1.0898942625227168e-05,
      "loss": 1.5896,
      "step": 1360
    },
    {
      "epoch": 2.117465224111283,
      "grad_norm": 6.78125,
      "learning_rate": 1.0556322097294835e-05,
      "loss": 1.6192,
      "step": 1370
    },
    {
      "epoch": 2.132921174652241,
      "grad_norm": 5.0,
      "learning_rate": 1.0217726220281243e-05,
      "loss": 1.6102,
      "step": 1380
    },
    {
      "epoch": 2.1483771251931993,
      "grad_norm": 5.8125,
      "learning_rate": 9.883249342053821e-06,
      "loss": 1.5518,
      "step": 1390
    },
    {
      "epoch": 2.1638330757341575,
      "grad_norm": 5.6875,
      "learning_rate": 9.552984662743716e-06,
      "loss": 1.4824,
      "step": 1400
    },
    {
      "epoch": 2.179289026275116,
      "grad_norm": 5.09375,
      "learning_rate": 9.227024208776224e-06,
      "loss": 1.4147,
      "step": 1410
    },
    {
      "epoch": 2.1947449768160743,
      "grad_norm": 5.5,
      "learning_rate": 8.905458807228013e-06,
      "loss": 1.5841,
      "step": 1420
    },
    {
      "epoch": 2.2102009273570324,
      "grad_norm": 5.09375,
      "learning_rate": 8.588378060518738e-06,
      "loss": 1.6099,
      "step": 1430
    },
    {
      "epoch": 2.2256568778979906,
      "grad_norm": 6.6875,
      "learning_rate": 8.275870321443802e-06,
      "loss": 1.5302,
      "step": 1440
    },
    {
      "epoch": 2.2411128284389488,
      "grad_norm": 5.21875,
      "learning_rate": 7.968022668555334e-06,
      "loss": 1.5991,
      "step": 1450
    },
    {
      "epoch": 2.2565687789799074,
      "grad_norm": 6.8125,
      "learning_rate": 7.664920881898236e-06,
      "loss": 1.6388,
      "step": 1460
    },
    {
      "epoch": 2.2720247295208655,
      "grad_norm": 6.78125,
      "learning_rate": 7.3666494191080455e-06,
      "loss": 1.5273,
      "step": 1470
    },
    {
      "epoch": 2.2874806800618237,
      "grad_norm": 5.5625,
      "learning_rate": 7.073291391877288e-06,
      "loss": 1.502,
      "step": 1480
    },
    {
      "epoch": 2.3029366306027823,
      "grad_norm": 5.9375,
      "learning_rate": 6.7849285427968586e-06,
      "loss": 1.4714,
      "step": 1490
    },
    {
      "epoch": 2.3183925811437405,
      "grad_norm": 6.4375,
      "learning_rate": 6.5016412225789e-06,
      "loss": 1.4667,
      "step": 1500
    },
    {
      "epoch": 2.3338485316846986,
      "grad_norm": 6.625,
      "learning_rate": 6.223508367667538e-06,
      "loss": 1.604,
      "step": 1510
    },
    {
      "epoch": 2.349304482225657,
      "grad_norm": 5.4375,
      "learning_rate": 5.9506074782436275e-06,
      "loss": 1.6265,
      "step": 1520
    },
    {
      "epoch": 2.364760432766615,
      "grad_norm": 5.1875,
      "learning_rate": 5.683014596629857e-06,
      "loss": 1.4042,
      "step": 1530
    },
    {
      "epoch": 2.3802163833075736,
      "grad_norm": 4.40625,
      "learning_rate": 5.420804286101838e-06,
      "loss": 1.6431,
      "step": 1540
    },
    {
      "epoch": 2.3956723338485317,
      "grad_norm": 4.84375,
      "learning_rate": 5.164049610111607e-06,
      "loss": 1.6615,
      "step": 1550
    },
    {
      "epoch": 2.41112828438949,
      "grad_norm": 4.15625,
      "learning_rate": 4.912822111928767e-06,
      "loss": 1.6007,
      "step": 1560
    },
    {
      "epoch": 2.426584234930448,
      "grad_norm": 6.8125,
      "learning_rate": 4.667191794705444e-06,
      "loss": 1.5768,
      "step": 1570
    },
    {
      "epoch": 2.4420401854714067,
      "grad_norm": 7.09375,
      "learning_rate": 4.427227101970261e-06,
      "loss": 1.6504,
      "step": 1580
    },
    {
      "epoch": 2.457496136012365,
      "grad_norm": 4.125,
      "learning_rate": 4.192994898556921e-06,
      "loss": 1.5617,
      "step": 1590
    },
    {
      "epoch": 2.472952086553323,
      "grad_norm": 4.9375,
      "learning_rate": 3.964560451972705e-06,
      "loss": 1.4937,
      "step": 1600
    },
    {
      "epoch": 2.488408037094281,
      "grad_norm": 4.34375,
      "learning_rate": 3.7419874142120363e-06,
      "loss": 1.5398,
      "step": 1610
    },
    {
      "epoch": 2.5038639876352393,
      "grad_norm": 7.125,
      "learning_rate": 3.525337804020212e-06,
      "loss": 1.6331,
      "step": 1620
    },
    {
      "epoch": 2.519319938176198,
      "grad_norm": 4.65625,
      "learning_rate": 3.314671989612195e-06,
      "loss": 1.5092,
      "step": 1630
    },
    {
      "epoch": 2.534775888717156,
      "grad_norm": 6.0,
      "learning_rate": 3.110048671851404e-06,
      "loss": 1.6244,
      "step": 1640
    },
    {
      "epoch": 2.5502318392581143,
      "grad_norm": 5.875,
      "learning_rate": 2.911524867892956e-06,
      "loss": 1.5014,
      "step": 1650
    },
    {
      "epoch": 2.565687789799073,
      "grad_norm": 5.71875,
      "learning_rate": 2.719155895296255e-06,
      "loss": 1.5084,
      "step": 1660
    },
    {
      "epoch": 2.581143740340031,
      "grad_norm": 5.40625,
      "learning_rate": 2.5329953566109537e-06,
      "loss": 1.6218,
      "step": 1670
    },
    {
      "epoch": 2.596599690880989,
      "grad_norm": 6.28125,
      "learning_rate": 2.353095124440999e-06,
      "loss": 1.5937,
      "step": 1680
    },
    {
      "epoch": 2.6120556414219473,
      "grad_norm": 5.78125,
      "learning_rate": 2.17950532699053e-06,
      "loss": 1.5456,
      "step": 1690
    },
    {
      "epoch": 2.6275115919629055,
      "grad_norm": 4.15625,
      "learning_rate": 2.012274334095998e-06,
      "loss": 1.6362,
      "step": 1700
    },
    {
      "epoch": 2.642967542503864,
      "grad_norm": 5.28125,
      "learning_rate": 1.8514487437481436e-06,
      "loss": 1.5103,
      "step": 1710
    },
    {
      "epoch": 2.6584234930448223,
      "grad_norm": 5.53125,
      "learning_rate": 1.6970733691077378e-06,
      "loss": 1.5271,
      "step": 1720
    },
    {
      "epoch": 2.6738794435857804,
      "grad_norm": 5.34375,
      "learning_rate": 1.549191226018637e-06,
      "loss": 1.5694,
      "step": 1730
    },
    {
      "epoch": 2.689335394126739,
      "grad_norm": 5.21875,
      "learning_rate": 1.4078435210216568e-06,
      "loss": 1.564,
      "step": 1740
    },
    {
      "epoch": 2.704791344667697,
      "grad_norm": 4.8125,
      "learning_rate": 1.273069639872601e-06,
      "loss": 1.6323,
      "step": 1750
    },
    {
      "epoch": 2.7202472952086554,
      "grad_norm": 4.75,
      "learning_rate": 1.1449071365676356e-06,
      "loss": 1.6231,
      "step": 1760
    },
    {
      "epoch": 2.7357032457496135,
      "grad_norm": 6.125,
      "learning_rate": 1.0233917228790828e-06,
      "loss": 1.5671,
      "step": 1770
    },
    {
      "epoch": 2.7511591962905717,
      "grad_norm": 5.53125,
      "learning_rate": 9.08557258404516e-07,
      "loss": 1.5081,
      "step": 1780
    },
    {
      "epoch": 2.76661514683153,
      "grad_norm": 4.9375,
      "learning_rate": 8.004357411319857e-07,
      "loss": 1.5764,
      "step": 1790
    },
    {
      "epoch": 2.7820710973724885,
      "grad_norm": 6.46875,
      "learning_rate": 6.990572985239612e-07,
      "loss": 1.5529,
      "step": 1800
    },
    {
      "epoch": 2.7975270479134466,
      "grad_norm": 5.6875,
      "learning_rate": 6.044501791224539e-07,
      "loss": 1.5893,
      "step": 1810
    },
    {
      "epoch": 2.812982998454405,
      "grad_norm": 7.125,
      "learning_rate": 5.166407446777572e-07,
      "loss": 1.6085,
      "step": 1820
    },
    {
      "epoch": 2.8284389489953634,
      "grad_norm": 5.25,
      "learning_rate": 4.3565346280284305e-07,
      "loss": 1.4917,
      "step": 1830
    },
    {
      "epoch": 2.8438948995363216,
      "grad_norm": 5.3125,
      "learning_rate": 3.6151090015565103e-07,
      "loss": 1.5961,
      "step": 1840
    },
    {
      "epoch": 2.8593508500772797,
      "grad_norm": 5.0,
      "learning_rate": 2.9423371615096005e-07,
      "loss": 1.621,
      "step": 1850
    },
    {
      "epoch": 2.874806800618238,
      "grad_norm": 4.375,
      "learning_rate": 2.3384065720379878e-07,
      "loss": 1.5848,
      "step": 1860
    },
    {
      "epoch": 2.890262751159196,
      "grad_norm": 4.90625,
      "learning_rate": 1.803485515058323e-07,
      "loss": 1.47,
      "step": 1870
    },
    {
      "epoch": 2.9057187017001547,
      "grad_norm": 7.21875,
      "learning_rate": 1.3377230433630205e-07,
      "loss": 1.5546,
      "step": 1880
    },
    {
      "epoch": 2.921174652241113,
      "grad_norm": 5.6875,
      "learning_rate": 9.412489390873414e-08,
      "loss": 1.6426,
      "step": 1890
    },
    {
      "epoch": 2.936630602782071,
      "grad_norm": 5.78125,
      "learning_rate": 6.141736775464313e-08,
      "loss": 1.5036,
      "step": 1900
    },
    {
      "epoch": 2.9520865533230296,
      "grad_norm": 7.125,
      "learning_rate": 3.5658839645194565e-08,
      "loss": 1.5698,
      "step": 1910
    },
    {
      "epoch": 2.9675425038639878,
      "grad_norm": 5.75,
      "learning_rate": 1.6856487051700177e-08,
      "loss": 1.5624,
      "step": 1920
    },
    {
      "epoch": 2.982998454404946,
      "grad_norm": 5.6875,
      "learning_rate": 5.015549145664933e-09,
      "loss": 1.7195,
      "step": 1930
    },
    {
      "epoch": 2.998454404945904,
      "grad_norm": 6.15625,
      "learning_rate": 1.393253389103677e-10,
      "loss": 1.6643,
      "step": 1940
    },
    {
      "epoch": 3.0,
      "step": 1941,
      "total_flos": 7.627263225499238e+16,
      "train_loss": 2.1592409425884838,
      "train_runtime": 1942.4867,
      "train_samples_per_second": 15.985,
      "train_steps_per_second": 0.999
    }
  ],
  "logging_steps": 10,
  "max_steps": 1941,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.627263225499238e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}