{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 7.051180674009678,
      "learning_rate": 1.437699680511182e-07,
      "loss": 0.45998425483703614,
      "step": 10
    },
    {
      "epoch": 0.0064,
      "grad_norm": 6.386709802443142,
      "learning_rate": 3.0351437699680514e-07,
      "loss": 0.44952831268310545,
      "step": 20
    },
    {
      "epoch": 0.0096,
      "grad_norm": 5.3631010908380015,
      "learning_rate": 4.6325878594249205e-07,
      "loss": 0.3993690013885498,
      "step": 30
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.852155255839625,
      "learning_rate": 6.230031948881789e-07,
      "loss": 0.3118258237838745,
      "step": 40
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.4705571475990448,
      "learning_rate": 7.82747603833866e-07,
      "loss": 0.2786674976348877,
      "step": 50
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.39720799855122535,
      "learning_rate": 9.424920127795528e-07,
      "loss": 0.2685645580291748,
      "step": 60
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.34144681090493506,
      "learning_rate": 1.1022364217252397e-06,
      "loss": 0.27388153076171873,
      "step": 70
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.29670665469044527,
      "learning_rate": 1.2619808306709266e-06,
      "loss": 0.2562382221221924,
      "step": 80
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.2721949763566226,
      "learning_rate": 1.4217252396166134e-06,
      "loss": 0.2521932125091553,
      "step": 90
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.30509418891876505,
      "learning_rate": 1.5814696485623005e-06,
      "loss": 0.2553669214248657,
      "step": 100
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.2710599378904947,
      "learning_rate": 1.7412140575079875e-06,
      "loss": 0.2548961162567139,
      "step": 110
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.3180117403374185,
      "learning_rate": 1.9009584664536742e-06,
      "loss": 0.2442842960357666,
      "step": 120
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.2695352733592907,
      "learning_rate": 2.060702875399361e-06,
      "loss": 0.24766459465026855,
      "step": 130
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.3064535363854503,
      "learning_rate": 2.220447284345048e-06,
      "loss": 0.23845260143280028,
      "step": 140
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.29068646435586043,
      "learning_rate": 2.380191693290735e-06,
      "loss": 0.23559024333953857,
      "step": 150
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.3186203915842237,
      "learning_rate": 2.539936102236422e-06,
      "loss": 0.23029537200927735,
      "step": 160
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.32754075011707046,
      "learning_rate": 2.699680511182109e-06,
      "loss": 0.2385089635848999,
      "step": 170
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.38484368515577855,
      "learning_rate": 2.8594249201277955e-06,
      "loss": 0.23111426830291748,
      "step": 180
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.299502856060473,
      "learning_rate": 3.0191693290734825e-06,
      "loss": 0.23530282974243164,
      "step": 190
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.3078123289936782,
      "learning_rate": 3.17891373801917e-06,
      "loss": 0.23611860275268554,
      "step": 200
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.30717572422426626,
      "learning_rate": 3.3386581469648564e-06,
      "loss": 0.23241891860961914,
      "step": 210
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.30949630760689323,
      "learning_rate": 3.4984025559105434e-06,
      "loss": 0.2257370948791504,
      "step": 220
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.2734080768093611,
      "learning_rate": 3.6581469648562303e-06,
      "loss": 0.22820501327514647,
      "step": 230
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.31986420438753294,
      "learning_rate": 3.817891373801918e-06,
      "loss": 0.22324295043945314,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3271935835910018,
      "learning_rate": 3.977635782747604e-06,
      "loss": 0.22092509269714355,
      "step": 250
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.28164810489138675,
      "learning_rate": 4.137380191693291e-06,
      "loss": 0.22088565826416015,
      "step": 260
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.2806581165168549,
      "learning_rate": 4.297124600638978e-06,
      "loss": 0.2235860824584961,
      "step": 270
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.2818314341404028,
      "learning_rate": 4.456869009584665e-06,
      "loss": 0.21951718330383302,
      "step": 280
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.2755068214230404,
      "learning_rate": 4.616613418530352e-06,
      "loss": 0.22480430603027343,
      "step": 290
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.2991295097090295,
      "learning_rate": 4.776357827476039e-06,
      "loss": 0.22600164413452148,
      "step": 300
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.3239664056294863,
      "learning_rate": 4.936102236421725e-06,
      "loss": 0.21372499465942382,
      "step": 310
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.2881723034602484,
      "learning_rate": 4.999943833158769e-06,
      "loss": 0.21513206958770753,
      "step": 320
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.27877645475403023,
      "learning_rate": 4.999600600490783e-06,
      "loss": 0.22072982788085938,
      "step": 330
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.28224550070191395,
      "learning_rate": 4.9989453817439345e-06,
      "loss": 0.2146312713623047,
      "step": 340
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.26853026150431764,
      "learning_rate": 4.997978258698942e-06,
      "loss": 0.21449072360992433,
      "step": 350
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.2706003676564934,
      "learning_rate": 4.996699352066659e-06,
      "loss": 0.2151791572570801,
      "step": 360
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.28539700359373177,
      "learning_rate": 4.995108821473014e-06,
      "loss": 0.21470160484313966,
      "step": 370
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.29207494833659137,
      "learning_rate": 4.993206865439084e-06,
      "loss": 0.21086468696594238,
      "step": 380
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.263064572322246,
      "learning_rate": 4.990993721356317e-06,
      "loss": 0.20984139442443847,
      "step": 390
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.2865097413347111,
      "learning_rate": 4.988469665456901e-06,
      "loss": 0.21040558815002441,
      "step": 400
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.2738592744136949,
      "learning_rate": 4.985635012779288e-06,
      "loss": 0.21828360557556153,
      "step": 410
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.24970618963972283,
      "learning_rate": 4.98249011712887e-06,
      "loss": 0.2106489658355713,
      "step": 420
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.2629431248486553,
      "learning_rate": 4.979035371033824e-06,
      "loss": 0.20979018211364747,
      "step": 430
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.2725505982701801,
      "learning_rate": 4.975271205696115e-06,
      "loss": 0.20948367118835448,
      "step": 440
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.2704053444924022,
      "learning_rate": 4.971198090937671e-06,
      "loss": 0.2033768653869629,
      "step": 450
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.31765035973786815,
      "learning_rate": 4.966816535141756e-06,
      "loss": 0.20044360160827637,
      "step": 460
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.2502347867419884,
      "learning_rate": 4.9621270851895035e-06,
      "loss": 0.2100567102432251,
      "step": 470
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.2934932151321077,
      "learning_rate": 4.957130326391662e-06,
      "loss": 0.21090621948242189,
      "step": 480
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.26660410583968774,
      "learning_rate": 4.951826882415544e-06,
      "loss": 0.20775444507598878,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.28519626596006936,
      "learning_rate": 4.946217415207177e-06,
      "loss": 0.20256528854370118,
      "step": 500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.2798675045050625,
      "learning_rate": 4.940302624908689e-06,
      "loss": 0.20623595714569093,
      "step": 510
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.28222884808809434,
      "learning_rate": 4.934083249770912e-06,
      "loss": 0.20097856521606444,
      "step": 520
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.2788085638053828,
      "learning_rate": 4.927560066061251e-06,
      "loss": 0.20387496948242187,
      "step": 530
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.27554368272722524,
      "learning_rate": 4.920733887966783e-06,
      "loss": 0.21524934768676757,
      "step": 540
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.26559833530971816,
      "learning_rate": 4.913605567492636e-06,
      "loss": 0.20402135848999023,
      "step": 550
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.26554772115650926,
      "learning_rate": 4.906175994355656e-06,
      "loss": 0.20598478317260743,
      "step": 560
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.3223429392292309,
      "learning_rate": 4.898446095873345e-06,
      "loss": 0.20747475624084472,
      "step": 570
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.24355730693567182,
      "learning_rate": 4.890416836848128e-06,
      "loss": 0.20512137413024903,
      "step": 580
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.3001767473938059,
      "learning_rate": 4.882089219446925e-06,
      "loss": 0.19992779493331908,
      "step": 590
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.2835389086432711,
      "learning_rate": 4.873464283076074e-06,
      "loss": 0.20495295524597168,
      "step": 600
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.26019712508927473,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.2035728931427002,
      "step": 610
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.2657949563176517,
      "learning_rate": 4.855326796464798e-06,
      "loss": 0.20619282722473145,
      "step": 620
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.28295912792439204,
      "learning_rate": 4.8458165100433725e-06,
      "loss": 0.2016925811767578,
      "step": 630
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.2902924299127114,
      "learning_rate": 4.836013432007738e-06,
      "loss": 0.20164456367492675,
      "step": 640
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.24256417788990398,
      "learning_rate": 4.825918785922921e-06,
      "loss": 0.20648303031921386,
      "step": 650
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.27122351891055063,
      "learning_rate": 4.8155338317458315e-06,
      "loss": 0.20356349945068358,
      "step": 660
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.2600569122055766,
      "learning_rate": 4.804859865668002e-06,
      "loss": 0.19959055185317992,
      "step": 670
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.25345624369635567,
      "learning_rate": 4.793898219953804e-06,
      "loss": 0.2007960557937622,
      "step": 680
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.2544929334444299,
      "learning_rate": 4.782650262774164e-06,
      "loss": 0.20300769805908203,
      "step": 690
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.2897145189307127,
      "learning_rate": 4.7711173980357886e-06,
      "loss": 0.19880002737045288,
      "step": 700
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.2560542526589546,
      "learning_rate": 4.759301065205947e-06,
      "loss": 0.19960763454437255,
      "step": 710
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.3097914904022575,
      "learning_rate": 4.7472027391328e-06,
      "loss": 0.2003918170928955,
      "step": 720
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.2887809607696432,
      "learning_rate": 4.734823929861317e-06,
      "loss": 0.20292911529541016,
      "step": 730
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.28917619670340877,
      "learning_rate": 4.722166182444801e-06,
      "loss": 0.20004558563232422,
      "step": 740
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.27043264841658887,
      "learning_rate": 4.709231076752045e-06,
      "loss": 0.19843683242797852,
      "step": 750
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.2645651727770741,
      "learning_rate": 4.696020227270142e-06,
      "loss": 0.20258240699768065,
      "step": 760
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.2777282429222742,
      "learning_rate": 4.6825352829029705e-06,
      "loss": 0.1994302749633789,
      "step": 770
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.28182340391383837,
      "learning_rate": 4.668777926765392e-06,
      "loss": 0.197939932346344,
      "step": 780
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.2390403179508666,
      "learning_rate": 4.6547498759731725e-06,
      "loss": 0.19328031539916993,
      "step": 790
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.30761446053746666,
      "learning_rate": 4.6404528814286575e-06,
      "loss": 0.1962287425994873,
      "step": 800
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.26058296777263723,
      "learning_rate": 4.6258887276022425e-06,
      "loss": 0.20304152965545655,
      "step": 810
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.3023946784650888,
      "learning_rate": 4.611059232309639e-06,
      "loss": 0.19789116382598876,
      "step": 820
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.28736962727648746,
      "learning_rate": 4.595966246484986e-06,
      "loss": 0.19968997240066527,
      "step": 830
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.28571881200537336,
      "learning_rate": 4.580611653949829e-06,
      "loss": 0.20007586479187012,
      "step": 840
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.295019179491335,
      "learning_rate": 4.564997371177992e-06,
      "loss": 0.19763822555541993,
      "step": 850
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.29653404936460237,
      "learning_rate": 4.54912534705637e-06,
      "loss": 0.19755616188049316,
      "step": 860
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.2642449071502374,
      "learning_rate": 4.532997562641683e-06,
      "loss": 0.19439829587936402,
      "step": 870
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.25657475126133233,
      "learning_rate": 4.516616030913214e-06,
      "loss": 0.1987127423286438,
      "step": 880
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.28458590654874555,
      "learning_rate": 4.499982796521556e-06,
      "loss": 0.19352295398712158,
      "step": 890
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.2793448530701338,
      "learning_rate": 4.48309993553341e-06,
      "loss": 0.1959349274635315,
      "step": 900
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.3163250873932861,
      "learning_rate": 4.465969555172468e-06,
      "loss": 0.1957021713256836,
      "step": 910
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.2933329400631374,
      "learning_rate": 4.448593793556391e-06,
      "loss": 0.20156097412109375,
      "step": 920
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.2688085579058971,
      "learning_rate": 4.430974819429954e-06,
      "loss": 0.1948945164680481,
      "step": 930
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.28553708068341715,
      "learning_rate": 4.413114831894344e-06,
      "loss": 0.18995710611343383,
      "step": 940
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.26518275753825254,
      "learning_rate": 4.3950160601326865e-06,
      "loss": 0.18871839046478273,
      "step": 950
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.28692003913342795,
      "learning_rate": 4.376680763131811e-06,
      "loss": 0.19533849954605104,
      "step": 960
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.27227233815166896,
      "learning_rate": 4.358111229400296e-06,
      "loss": 0.19751427173614503,
      "step": 970
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.27245831220598377,
      "learning_rate": 4.33930977668283e-06,
      "loss": 0.20111453533172607,
      "step": 980
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.2482632152661181,
      "learning_rate": 4.320278751670922e-06,
      "loss": 0.19406617879867555,
      "step": 990
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.2892442073812178,
      "learning_rate": 4.301020529710009e-06,
      "loss": 0.19525597095489503,
      "step": 1000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.26392559431034407,
      "learning_rate": 4.281537514502962e-06,
      "loss": 0.19918107986450195,
      "step": 1010
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.27003912401002855,
      "learning_rate": 4.261832137810093e-06,
      "loss": 0.1964997172355652,
      "step": 1020
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.2664017566726753,
      "learning_rate": 4.241906859145611e-06,
      "loss": 0.19660145044326782,
      "step": 1030
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.2744161118643581,
      "learning_rate": 4.221764165470661e-06,
      "loss": 0.1935626745223999,
      "step": 1040
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.2717693030089869,
      "learning_rate": 4.201406570882898e-06,
      "loss": 0.19286205768585205,
      "step": 1050
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.259292524653773,
      "learning_rate": 4.180836616302704e-06,
      "loss": 0.1922353148460388,
      "step": 1060
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.2739674960468982,
      "learning_rate": 4.160056869156041e-06,
      "loss": 0.19553282260894775,
      "step": 1070
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.272965837223612,
      "learning_rate": 4.139069923053995e-06,
      "loss": 0.19367674589157105,
      "step": 1080
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.2463436566122966,
      "learning_rate": 4.117878397469062e-06,
      "loss": 0.19772920608520508,
      "step": 1090
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.24672019869428047,
      "learning_rate": 4.096484937408195e-06,
      "loss": 0.1892393112182617,
      "step": 1100
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.2673060417093708,
      "learning_rate": 4.074892213082676e-06,
      "loss": 0.1892371416091919,
      "step": 1110
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.26767314750680543,
      "learning_rate": 4.0531029195748265e-06,
      "loss": 0.19717614650726317,
      "step": 1120
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.2796524343786416,
      "learning_rate": 4.03111977650163e-06,
      "loss": 0.19503848552703856,
      "step": 1130
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.2816284710404393,
      "learning_rate": 4.008945527675281e-06,
      "loss": 0.19529366493225098,
      "step": 1140
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.31949481569871324,
      "learning_rate": 3.986582940760717e-06,
      "loss": 0.18451136350631714,
      "step": 1150
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.2723449306170863,
      "learning_rate": 3.9640348069301785e-06,
      "loss": 0.191510009765625,
      "step": 1160
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.27747112521567696,
      "learning_rate": 3.941303940514826e-06,
      "loss": 0.19263410568237305,
      "step": 1170
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.2719099807762723,
      "learning_rate": 3.918393178653472e-06,
      "loss": 0.19341590404510497,
      "step": 1180
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.29074805846664115,
      "learning_rate": 3.895305380938468e-06,
      "loss": 0.19099385738372804,
      "step": 1190
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.2517462589595264,
      "learning_rate": 3.872043429058783e-06,
      "loss": 0.18874506950378417,
      "step": 1200
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.2591827841853763,
      "learning_rate": 3.84861022644033e-06,
      "loss": 0.19069148302078248,
      "step": 1210
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.2702770742629986,
      "learning_rate": 3.825008697883574e-06,
      "loss": 0.19928838014602662,
      "step": 1220
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.27788866885326635,
      "learning_rate": 3.8012417891984776e-06,
      "loss": 0.19237933158874512,
      "step": 1230
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.2656255469668472,
      "learning_rate": 3.777312466836819e-06,
      "loss": 0.19055767059326173,
      "step": 1240
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.28446496354107703,
      "learning_rate": 3.7532237175219378e-06,
      "loss": 0.18940582275390624,
      "step": 1250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.4152862546777316,
      "learning_rate": 3.728978547875948e-06,
      "loss": 0.19362914562225342,
      "step": 1260
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.28537432061728957,
      "learning_rate": 3.7045799840444712e-06,
      "loss": 0.1886904716491699,
      "step": 1270
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.29038310854731697,
      "learning_rate": 3.6800310713189258e-06,
      "loss": 0.18923617601394654,
      "step": 1280
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.32132086585692904,
      "learning_rate": 3.6553348737564328e-06,
      "loss": 0.19005811214447021,
      "step": 1290
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.2669423345384319,
      "learning_rate": 3.6304944737973794e-06,
      "loss": 0.19575085639953613,
      "step": 1300
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.28931030301965927,
      "learning_rate": 3.6055129718806836e-06,
      "loss": 0.18975239992141724,
      "step": 1310
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.28948269391746034,
      "learning_rate": 3.5803934860568134e-06,
      "loss": 0.18510067462921143,
      "step": 1320
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.2821484963772758,
      "learning_rate": 3.5551391515986163e-06,
      "loss": 0.1907583475112915,
      "step": 1330
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.27423888046510925,
      "learning_rate": 3.529753120609982e-06,
      "loss": 0.18690071105957032,
      "step": 1340
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.30811658453814883,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.19000139236450195,
      "step": 1350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.24402420223179272,
      "learning_rate": 3.4785986592495934e-06,
      "loss": 0.18803791999816893,
      "step": 1360
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.24576039119812526,
      "learning_rate": 3.452836613689803e-06,
      "loss": 0.1866163969039917,
      "step": 1370
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.2949022587874532,
      "learning_rate": 3.426955640426584e-06,
      "loss": 0.1890486001968384,
      "step": 1380
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.2582182081996982,
      "learning_rate": 3.4009589697773605e-06,
      "loss": 0.18851635456085206,
      "step": 1390
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.2722482128131903,
      "learning_rate": 3.3748498465002475e-06,
      "loss": 0.18554195165634155,
      "step": 1400
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.27484686642107964,
      "learning_rate": 3.3486315293890693e-06,
      "loss": 0.19425587654113768,
      "step": 1410
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.28258316073925427,
      "learning_rate": 3.3223072908666053e-06,
      "loss": 0.1843653440475464,
      "step": 1420
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.28555979247115143,
      "learning_rate": 3.295880416576153e-06,
      "loss": 0.1941524863243103,
      "step": 1430
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.2969010932820601,
      "learning_rate": 3.269354204971427e-06,
      "loss": 0.18759560585021973,
      "step": 1440
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.30795851200957197,
      "learning_rate": 3.242731966904865e-06,
      "loss": 0.18544803857803344,
      "step": 1450
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.28527072571260903,
      "learning_rate": 3.2160170252143913e-06,
      "loss": 0.18547136783599855,
      "step": 1460
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.2533866816866613,
      "learning_rate": 3.1892127143086716e-06,
      "loss": 0.19228132963180541,
      "step": 1470
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.2776942873045479,
      "learning_rate": 3.1623223797509347e-06,
      "loss": 0.1812342882156372,
      "step": 1480
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.2744584915099732,
      "learning_rate": 3.135349377841396e-06,
      "loss": 0.1853887915611267,
      "step": 1490
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2866639604297882,
      "learning_rate": 3.1082970751983497e-06,
      "loss": 0.1918737769126892,
      "step": 1500
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.26310322890713356,
      "learning_rate": 3.0811688483379546e-06,
      "loss": 0.18995790481567382,
      "step": 1510
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.28320054398109096,
      "learning_rate": 3.0539680832528074e-06,
      "loss": 0.18962399959564208,
      "step": 1520
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.2654570443815982,
      "learning_rate": 3.026698174989316e-06,
      "loss": 0.18734774589538575,
      "step": 1530
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.2658181920127404,
      "learning_rate": 2.999362527223952e-06,
      "loss": 0.1873406410217285,
      "step": 1540
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.29250213703445505,
      "learning_rate": 2.9719645518384194e-06,
      "loss": 0.1892526626586914,
      "step": 1550
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.3090995402302473,
      "learning_rate": 2.944507668493807e-06,
      "loss": 0.19257349967956544,
      "step": 1560
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.28272052629438726,
      "learning_rate": 2.9169953042037623e-06,
      "loss": 0.18868753910064698,
      "step": 1570
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.3954198531333443,
      "learning_rate": 2.889430892906754e-06,
      "loss": 0.18459179401397705,
      "step": 1580
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.2563261821009193,
      "learning_rate": 2.861817875037462e-06,
      "loss": 0.18163517713546753,
      "step": 1590
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.28115388072993086,
      "learning_rate": 2.8341596970973683e-06,
      "loss": 0.19087796211242675,
      "step": 1600
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.27079102831839946,
      "learning_rate": 2.80645981122458e-06,
      "loss": 0.1863863706588745,
      "step": 1610
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.27596423249252744,
      "learning_rate": 2.7787216747629508e-06,
      "loss": 0.19303735494613647,
      "step": 1620
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.2682301223547138,
      "learning_rate": 2.7509487498305615e-06,
      "loss": 0.18045294284820557,
      "step": 1630
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.27817197846381203,
      "learning_rate": 2.7231445028875924e-06,
      "loss": 0.18653267621994019,
      "step": 1640
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.25176165708531945,
      "learning_rate": 2.6953124043036604e-06,
      "loss": 0.18530716896057128,
      "step": 1650
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.272299195118528,
      "learning_rate": 2.667455927924667e-06,
      "loss": 0.18495219945907593,
      "step": 1660
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.26513870922757315,
      "learning_rate": 2.6395785506392164e-06,
      "loss": 0.18016864061355592,
      "step": 1670
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.26899577641448663,
      "learning_rate": 2.6116837519446407e-06,
      "loss": 0.18437364101409912,
      "step": 1680
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.29589553270345376,
      "learning_rate": 2.5837750135127192e-06,
      "loss": 0.18141529560089112,
      "step": 1690
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.28180995392351926,
      "learning_rate": 2.555855818755108e-06,
      "loss": 0.18680166006088256,
      "step": 1700
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.29608650413456306,
      "learning_rate": 2.5279296523885636e-06,
      "loss": 0.18486298322677613,
      "step": 1710
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.28475957723655715,
      "learning_rate": 2.5e-06,
      "loss": 0.1850725531578064,
      "step": 1720
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.27856833997611247,
      "learning_rate": 2.472070347611437e-06,
      "loss": 0.1791991949081421,
      "step": 1730
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.30516489860119894,
      "learning_rate": 2.444144181244893e-06,
      "loss": 0.18483606576919556,
      "step": 1740
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.29804656625996045,
      "learning_rate": 2.416224986487282e-06,
      "loss": 0.18195321559906005,
      "step": 1750
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.30740179095263215,
      "learning_rate": 2.3883162480553605e-06,
      "loss": 0.17964634895324708,
      "step": 1760
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.29672245353605753,
      "learning_rate": 2.3604214493607844e-06,
      "loss": 0.18308933973312377,
      "step": 1770
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.2837212145176832,
      "learning_rate": 2.332544072075333e-06,
      "loss": 0.18688681125640869,
      "step": 1780
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.28451872958084823,
      "learning_rate": 2.30468759569634e-06,
      "loss": 0.18532857894897461,
      "step": 1790
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.29734825652467917,
      "learning_rate": 2.276855497112408e-06,
      "loss": 0.18262310028076173,
      "step": 1800
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.3012944650683003,
      "learning_rate": 2.2490512501694394e-06,
      "loss": 0.17781586647033693,
      "step": 1810
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.2692920477116042,
      "learning_rate": 2.2212783252370496e-06,
      "loss": 0.18318163156509398,
      "step": 1820
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.2700619255739624,
      "learning_rate": 2.1935401887754213e-06,
      "loss": 0.18857367038726808,
      "step": 1830
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.2868516489290536,
      "learning_rate": 2.165840302902632e-06,
      "loss": 0.18190672397613525,
      "step": 1840
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.28726300225812107,
      "learning_rate": 2.1381821249625383e-06,
      "loss": 0.1867521286010742,
      "step": 1850
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.2995145996099388,
      "learning_rate": 2.1105691070932465e-06,
      "loss": 0.17851842641830445,
      "step": 1860
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.28575212768410063,
      "learning_rate": 2.083004695796238e-06,
      "loss": 0.17741835117340088,
      "step": 1870
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.31284763297048707,
      "learning_rate": 2.055492331506194e-06,
      "loss": 0.1843113422393799,
      "step": 1880
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.3170666816206652,
      "learning_rate": 2.0280354481615814e-06,
      "loss": 0.18248820304870605,
      "step": 1890
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.30950907311465886,
      "learning_rate": 2.000637472776049e-06,
      "loss": 0.1839754819869995,
      "step": 1900
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.2536972696685391,
      "learning_rate": 1.973301825010685e-06,
      "loss": 0.17841637134552002,
      "step": 1910
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.291862692607901,
      "learning_rate": 1.9460319167471934e-06,
      "loss": 0.18339977264404297,
      "step": 1920
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.2848109477155621,
      "learning_rate": 1.9188311516620466e-06,
      "loss": 0.17915148735046388,
      "step": 1930
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.3060077712638729,
      "learning_rate": 1.891702924801651e-06,
      "loss": 0.1848907709121704,
      "step": 1940
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.27297816434517674,
      "learning_rate": 1.864650622158604e-06,
      "loss": 0.18888840675354004,
      "step": 1950
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.2781302448691454,
      "learning_rate": 1.8376776202490666e-06,
      "loss": 0.1847243309020996,
      "step": 1960
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.31527749144779466,
      "learning_rate": 1.8107872856913293e-06,
      "loss": 0.17888798713684081,
      "step": 1970
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.2981389294211551,
      "learning_rate": 1.7839829747856096e-06,
      "loss": 0.18081605434417725,
      "step": 1980
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.29438595992497246,
      "learning_rate": 1.7572680330951359e-06,
      "loss": 0.17975808382034303,
      "step": 1990
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2777422843592099,
      "learning_rate": 1.7306457950285747e-06,
      "loss": 0.1812159538269043,
      "step": 2000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.3068388373590525,
      "learning_rate": 1.704119583423848e-06,
      "loss": 0.17536230087280275,
      "step": 2010
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.272885194568128,
      "learning_rate": 1.677692709133396e-06,
      "loss": 0.18365554809570311,
      "step": 2020
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.3023336412584975,
      "learning_rate": 1.6513684706109311e-06,
      "loss": 0.18368566036224365,
      "step": 2030
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.28988866387653284,
      "learning_rate": 1.6251501534997529e-06,
      "loss": 0.18175660371780394,
      "step": 2040
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.28123365590903454,
      "learning_rate": 1.5990410302226405e-06,
      "loss": 0.17483808994293212,
      "step": 2050
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.28187049939921544,
      "learning_rate": 1.5730443595734162e-06,
      "loss": 0.18124582767486572,
      "step": 2060
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.31643189708694724,
      "learning_rate": 1.5471633863101982e-06,
      "loss": 0.18188211917877198,
      "step": 2070
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.3071146379480691,
      "learning_rate": 1.521401340750407e-06,
      "loss": 0.18458983898162842,
      "step": 2080
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.30923765962914507,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.18291953802108765,
      "step": 2090
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.31506268222239586,
      "learning_rate": 1.4702468793900187e-06,
      "loss": 0.18112607002258302,
      "step": 2100
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.2991031913192095,
      "learning_rate": 1.444860848401384e-06,
      "loss": 0.18132129907608033,
      "step": 2110
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.306957825954438,
      "learning_rate": 1.4196065139431866e-06,
      "loss": 0.18091821670532227,
      "step": 2120
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.30984784981623864,
      "learning_rate": 1.3944870281193178e-06,
      "loss": 0.17975277900695802,
      "step": 2130
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.33685631116321924,
      "learning_rate": 1.3695055262026208e-06,
      "loss": 0.18606040477752686,
      "step": 2140
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.28362188085343176,
      "learning_rate": 1.3446651262435679e-06,
      "loss": 0.17845985889434815,
      "step": 2150
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.28046286761312267,
      "learning_rate": 1.3199689286810746e-06,
      "loss": 0.18048195838928222,
      "step": 2160
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.29900090645940436,
      "learning_rate": 1.2954200159555294e-06,
      "loss": 0.17538446187973022,
      "step": 2170
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.32576508972663926,
      "learning_rate": 1.2710214521240527e-06,
      "loss": 0.18001599311828614,
      "step": 2180
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.30869890145158635,
      "learning_rate": 1.246776282478063e-06,
      "loss": 0.18135268688201905,
      "step": 2190
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.28612747319198,
      "learning_rate": 1.222687533163181e-06,
      "loss": 0.18038851022720337,
      "step": 2200
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.32303440375726766,
      "learning_rate": 1.1987582108015228e-06,
      "loss": 0.18109045028686524,
      "step": 2210
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.3093047688685527,
      "learning_rate": 1.1749913021164255e-06,
      "loss": 0.18254566192626953,
      "step": 2220
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.2882548432858515,
      "learning_rate": 1.1513897735596702e-06,
      "loss": 0.17732615470886232,
      "step": 2230
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.29445166285798274,
      "learning_rate": 1.127956570941218e-06,
      "loss": 0.17425966262817383,
      "step": 2240
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3514589237334647,
      "learning_rate": 1.104694619061533e-06,
      "loss": 0.18296418190002442,
      "step": 2250
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.32323021290499837,
      "learning_rate": 1.0816068213465295e-06,
      "loss": 0.1851881265640259,
      "step": 2260
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.30421571681673176,
      "learning_rate": 1.0586960594851762e-06,
      "loss": 0.180436372756958,
      "step": 2270
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.31911631321578676,
      "learning_rate": 1.0359651930698217e-06,
      "loss": 0.17929892539978026,
      "step": 2280
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.30015899620754233,
      "learning_rate": 1.0134170592392837e-06,
      "loss": 0.18022915124893188,
      "step": 2290
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.31786084969492157,
      "learning_rate": 9.910544723247204e-07,
      "loss": 0.17959039211273192,
      "step": 2300
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.31599364626026827,
      "learning_rate": 9.688802234983706e-07,
      "loss": 0.17806137800216676,
      "step": 2310
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.3303243768736776,
      "learning_rate": 9.468970804251742e-07,
      "loss": 0.1811964988708496,
      "step": 2320
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.3312986961423255,
      "learning_rate": 9.251077869173244e-07,
      "loss": 0.17583439350128174,
      "step": 2330
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.30030412592967864,
      "learning_rate": 9.035150625918054e-07,
      "loss": 0.17623555660247803,
      "step": 2340
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.3177646626866783,
      "learning_rate": 8.821216025309395e-07,
      "loss": 0.18003884553909302,
      "step": 2350
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.3012142976429357,
      "learning_rate": 8.609300769460055e-07,
      "loss": 0.17543296813964843,
      "step": 2360
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.3177168816443014,
      "learning_rate": 8.399431308439592e-07,
      "loss": 0.18021781444549562,
      "step": 2370
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.34248252589513506,
      "learning_rate": 8.191633836972962e-07,
      "loss": 0.18125417232513427,
      "step": 2380
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.29292480325152365,
      "learning_rate": 7.985934291171024e-07,
      "loss": 0.17757056951522826,
      "step": 2390
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.3257764859746147,
      "learning_rate": 7.7823583452934e-07,
      "loss": 0.18096057176589966,
      "step": 2400
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.28892062916284306,
      "learning_rate": 7.58093140854389e-07,
      "loss": 0.18015010356903077,
      "step": 2410
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.32360358107292697,
      "learning_rate": 7.381678621899077e-07,
      "loss": 0.18067935705184937,
      "step": 2420
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.3139428787829718,
      "learning_rate": 7.184624854970379e-07,
      "loss": 0.1768512487411499,
      "step": 2430
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.3182311104789415,
      "learning_rate": 6.989794702899932e-07,
      "loss": 0.17589566707611085,
      "step": 2440
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.3112954733861784,
      "learning_rate": 6.797212483290777e-07,
      "loss": 0.177903413772583,
      "step": 2450
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.31026727362843554,
      "learning_rate": 6.60690223317171e-07,
      "loss": 0.17535500526428222,
      "step": 2460
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.2855504901999764,
      "learning_rate": 6.418887705997046e-07,
      "loss": 0.1787285327911377,
      "step": 2470
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.33581031525319194,
      "learning_rate": 6.23319236868189e-07,
      "loss": 0.181508469581604,
      "step": 2480
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.30084134655605693,
      "learning_rate": 6.049839398673141e-07,
      "loss": 0.18244649171829225,
      "step": 2490
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3207759323449182,
      "learning_rate": 5.868851681056567e-07,
      "loss": 0.18296375274658203,
      "step": 2500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.3103299858846911,
      "learning_rate": 5.690251805700467e-07,
      "loss": 0.18089601993560792,
      "step": 2510
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.3310470653200237,
      "learning_rate": 5.514062064436096e-07,
      "loss": 0.1829407334327698,
      "step": 2520
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.31783823046596615,
      "learning_rate": 5.34030444827533e-07,
      "loss": 0.17886234521865846,
      "step": 2530
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.3279151171862584,
      "learning_rate": 5.169000644665895e-07,
      "loss": 0.17618993520736695,
      "step": 2540
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.3006249030100123,
      "learning_rate": 5.000172034784442e-07,
      "loss": 0.17779455184936524,
      "step": 2550
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.3164261324675526,
      "learning_rate": 4.833839690867853e-07,
      "loss": 0.18002912998199463,
      "step": 2560
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.31374931318878396,
      "learning_rate": 4.6700243735831705e-07,
      "loss": 0.173567795753479,
      "step": 2570
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.31170459979916293,
      "learning_rate": 4.508746529436311e-07,
      "loss": 0.1724323034286499,
      "step": 2580
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.3080863565290302,
      "learning_rate": 4.350026288220083e-07,
      "loss": 0.1794981598854065,
      "step": 2590
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.30618951989415283,
      "learning_rate": 4.1938834605017133e-07,
      "loss": 0.1761255979537964,
      "step": 2600
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.3029510706797137,
      "learning_rate": 4.0403375351501515e-07,
      "loss": 0.17623082399368287,
      "step": 2610
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.336336912959277,
      "learning_rate": 3.88940767690362e-07,
      "loss": 0.1757615327835083,
      "step": 2620
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.32859024308656015,
      "learning_rate": 3.7411127239775774e-07,
      "loss": 0.17455869913101196,
      "step": 2630
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.3174124959768476,
      "learning_rate": 3.595471185713431e-07,
      "loss": 0.17312180995941162,
      "step": 2640
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.3247217043719523,
      "learning_rate": 3.4525012402682826e-07,
      "loss": 0.17421470880508422,
      "step": 2650
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.3290462164412991,
      "learning_rate": 3.3122207323460804e-07,
      "loss": 0.17708632946014405,
      "step": 2660
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.3024938333869805,
      "learning_rate": 3.1746471709702963e-07,
      "loss": 0.17333836555480958,
      "step": 2670
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.32678703604131465,
      "learning_rate": 3.039797727298585e-07,
      "loss": 0.1801586151123047,
      "step": 2680
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.32985764106850785,
      "learning_rate": 2.9076892324795546e-07,
      "loss": 0.17783432006835936,
      "step": 2690
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.31242585953952057,
      "learning_rate": 2.778338175551995e-07,
      "loss": 0.17357670068740844,
      "step": 2700
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.3220012856306909,
      "learning_rate": 2.6517607013868326e-07,
      "loss": 0.18131563663482667,
      "step": 2710
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.33350326064348024,
      "learning_rate": 2.527972608672002e-07,
      "loss": 0.17757024765014648,
      "step": 2720
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.335919926946263,
      "learning_rate": 2.40698934794053e-07,
      "loss": 0.17683808803558348,
      "step": 2730
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.3209912976041497,
      "learning_rate": 2.2888260196421237e-07,
      "loss": 0.17635661363601685,
      "step": 2740
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3165955269677658,
      "learning_rate": 2.1734973722583735e-07,
      "loss": 0.17913974523544313,
      "step": 2750
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.31474674596852353,
      "learning_rate": 2.0610178004619564e-07,
      "loss": 0.17095563411712647,
      "step": 2760
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.305115903859637,
      "learning_rate": 1.9514013433199834e-07,
      "loss": 0.18293533325195313,
      "step": 2770
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.3164297745100823,
      "learning_rate": 1.8446616825416958e-07,
      "loss": 0.18138229846954346,
      "step": 2780
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.3526140625065779,
      "learning_rate": 1.7408121407708007e-07,
      "loss": 0.18163397312164306,
      "step": 2790
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.3224933819196559,
      "learning_rate": 1.6398656799226253e-07,
      "loss": 0.1705089807510376,
      "step": 2800
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.31764589677400257,
      "learning_rate": 1.5418348995662773e-07,
      "loss": 0.17652597427368164,
      "step": 2810
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.3414067132784035,
      "learning_rate": 1.4467320353520275e-07,
      "loss": 0.17487871646881104,
      "step": 2820
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.3138098972679996,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.17592911720275878,
      "step": 2830
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.31560573280288073,
      "learning_rate": 1.26535716923927e-07,
      "loss": 0.18197228908538818,
      "step": 2840
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.3188962184685744,
      "learning_rate": 1.1791078055307493e-07,
      "loss": 0.1777464509010315,
      "step": 2850
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.31575220367713525,
      "learning_rate": 1.0958316315187289e-07,
      "loss": 0.17706483602523804,
      "step": 2860
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.3131837624055497,
      "learning_rate": 1.0155390412665528e-07,
      "loss": 0.17496002912521363,
      "step": 2870
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.32248583567737266,
      "learning_rate": 9.38240056443443e-08,
      "loss": 0.17229046821594238,
      "step": 2880
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.3101253584845484,
      "learning_rate": 8.639443250736402e-08,
      "loss": 0.17552309036254882,
      "step": 2890
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.33217431742972764,
      "learning_rate": 7.926611203321777e-08,
      "loss": 0.17659810781478882,
      "step": 2900
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.33918124282098266,
      "learning_rate": 7.243993393874882e-08,
      "loss": 0.17737939357757568,
      "step": 2910
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.31351790893613213,
      "learning_rate": 6.591675022908805e-08,
      "loss": 0.1745692253112793,
      "step": 2920
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.33783778867129854,
      "learning_rate": 5.969737509131241e-08,
      "loss": 0.1722058415412903,
      "step": 2930
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.308776655874055,
      "learning_rate": 5.3782584792823334e-08,
      "loss": 0.17710112333297728,
      "step": 2940
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.3142338038371378,
      "learning_rate": 4.817311758445686e-08,
      "loss": 0.178252911567688,
      "step": 2950
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.33048986218580767,
      "learning_rate": 4.286967360833866e-08,
      "loss": 0.1782402753829956,
      "step": 2960
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.3110909627270251,
      "learning_rate": 3.787291481049754e-08,
      "loss": 0.17829475402832032,
      "step": 2970
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.33726065147122686,
      "learning_rate": 3.3183464858244364e-08,
      "loss": 0.18406097888946532,
      "step": 2980
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.3326393750086487,
      "learning_rate": 2.8801909062328992e-08,
      "loss": 0.17060396671295167,
      "step": 2990
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.32948960265922206,
      "learning_rate": 2.4728794303886248e-08,
      "loss": 0.16899311542510986,
      "step": 3000
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.33211982053439487,
      "learning_rate": 2.0964628966175794e-08,
      "loss": 0.17517964839935302,
      "step": 3010
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.30613498697830943,
      "learning_rate": 1.750988287113009e-08,
      "loss": 0.17458994388580323,
      "step": 3020
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.3027770955918648,
      "learning_rate": 1.4364987220713278e-08,
      "loss": 0.18178436756134034,
      "step": 3030
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.3292318037983906,
      "learning_rate": 1.1530334543099763e-08,
      "loss": 0.1790144681930542,
      "step": 3040
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.33300755292787143,
      "learning_rate": 9.006278643683697e-09,
      "loss": 0.1808505654335022,
      "step": 3050
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.32631723332989787,
      "learning_rate": 6.793134560916514e-09,
      "loss": 0.18275127410888672,
      "step": 3060
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.3082787331662993,
      "learning_rate": 4.891178526986451e-09,
      "loss": 0.1783647656440735,
      "step": 3070
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.32341550392390483,
      "learning_rate": 3.3006479333413943e-09,
      "loss": 0.18126009702682494,
      "step": 3080
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.30931371888762194,
      "learning_rate": 2.021741301058422e-09,
      "loss": 0.17681236267089845,
      "step": 3090
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.3419672311636941,
      "learning_rate": 1.0546182560652872e-09,
      "loss": 0.17989683151245117,
      "step": 3100
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.3111951639834393,
      "learning_rate": 3.9939950921774607e-10,
      "loss": 0.17482796907424927,
      "step": 3110
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.3230413672933209,
      "learning_rate": 5.616684123160854e-11,
      "loss": 0.17436976432800294,
      "step": 3120
    },
    {
      "epoch": 1.0,
      "step": 3125,
      "total_flos": 1955525886476288.0,
      "train_loss": 0.1948647116279602,
      "train_runtime": 28413.61,
      "train_samples_per_second": 7.039,
      "train_steps_per_second": 0.11
    }
  ],
  "logging_steps": 10,
  "max_steps": 3125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1955525886476288.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}