{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998489654130795,
  "eval_steps": 500,
  "global_step": 3310,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0030206917384080953,
      "grad_norm": 6.613835485331974,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 6.2562,
      "step": 10
    },
    {
      "epoch": 0.006041383476816191,
      "grad_norm": 4.759929469878327,
      "learning_rate": 6.505149978319905e-05,
      "loss": 3.3105,
      "step": 20
    },
    {
      "epoch": 0.009062075215224286,
      "grad_norm": 3.4302773742415433,
      "learning_rate": 7.385606273598311e-05,
      "loss": 1.6293,
      "step": 30
    },
    {
      "epoch": 0.012082766953632381,
      "grad_norm": 2.200390666433873,
      "learning_rate": 8.01029995663981e-05,
      "loss": 1.2723,
      "step": 40
    },
    {
      "epoch": 0.015103458692040477,
      "grad_norm": 1.9872314808738405,
      "learning_rate": 8.494850021680092e-05,
      "loss": 1.1221,
      "step": 50
    },
    {
      "epoch": 0.01812415043044857,
      "grad_norm": 2.0585565038414173,
      "learning_rate": 8.890756251918216e-05,
      "loss": 1.1238,
      "step": 60
    },
    {
      "epoch": 0.02114484216885667,
      "grad_norm": 1.5179740873933507,
      "learning_rate": 9.225490200071284e-05,
      "loss": 1.0437,
      "step": 70
    },
    {
      "epoch": 0.024165533907264763,
      "grad_norm": 1.7872492138745553,
      "learning_rate": 9.515449934959716e-05,
      "loss": 0.9537,
      "step": 80
    },
    {
      "epoch": 0.02718622564567286,
      "grad_norm": 1.59848853171592,
      "learning_rate": 9.771212547196623e-05,
      "loss": 0.9299,
      "step": 90
    },
    {
      "epoch": 0.030206917384080954,
      "grad_norm": 1.4459867462911404,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.8744,
      "step": 100
    },
    {
      "epoch": 0.03322760912248905,
      "grad_norm": 1.2666954985573535,
      "learning_rate": 9.97196261682243e-05,
      "loss": 0.9279,
      "step": 110
    },
    {
      "epoch": 0.03624830086089714,
      "grad_norm": 1.1375696912612132,
      "learning_rate": 9.940809968847352e-05,
      "loss": 0.943,
      "step": 120
    },
    {
      "epoch": 0.03926899259930524,
      "grad_norm": 1.5887317329510864,
      "learning_rate": 9.909657320872274e-05,
      "loss": 0.8703,
      "step": 130
    },
    {
      "epoch": 0.04228968433771334,
      "grad_norm": 1.5838324888073332,
      "learning_rate": 9.878504672897197e-05,
      "loss": 0.8943,
      "step": 140
    },
    {
      "epoch": 0.045310376076121435,
      "grad_norm": 1.247226983069568,
      "learning_rate": 9.847352024922119e-05,
      "loss": 0.8818,
      "step": 150
    },
    {
      "epoch": 0.048331067814529526,
      "grad_norm": 1.6083164873540423,
      "learning_rate": 9.816199376947041e-05,
      "loss": 0.8748,
      "step": 160
    },
    {
      "epoch": 0.05135175955293762,
      "grad_norm": 1.0686893898752028,
      "learning_rate": 9.785046728971962e-05,
      "loss": 0.875,
      "step": 170
    },
    {
      "epoch": 0.05437245129134572,
      "grad_norm": 1.3766933277373585,
      "learning_rate": 9.753894080996884e-05,
      "loss": 0.8324,
      "step": 180
    },
    {
      "epoch": 0.05739314302975381,
      "grad_norm": 1.0332571782362285,
      "learning_rate": 9.722741433021807e-05,
      "loss": 0.8414,
      "step": 190
    },
    {
      "epoch": 0.06041383476816191,
      "grad_norm": 1.2849086192566328,
      "learning_rate": 9.691588785046729e-05,
      "loss": 0.825,
      "step": 200
    },
    {
      "epoch": 0.06343452650657,
      "grad_norm": 1.125015534848599,
      "learning_rate": 9.660436137071651e-05,
      "loss": 0.827,
      "step": 210
    },
    {
      "epoch": 0.0664552182449781,
      "grad_norm": 1.0924381371211556,
      "learning_rate": 9.629283489096573e-05,
      "loss": 0.8381,
      "step": 220
    },
    {
      "epoch": 0.0694759099833862,
      "grad_norm": 1.1382918007143823,
      "learning_rate": 9.598130841121496e-05,
      "loss": 0.8047,
      "step": 230
    },
    {
      "epoch": 0.07249660172179428,
      "grad_norm": 1.263038309691145,
      "learning_rate": 9.566978193146418e-05,
      "loss": 0.7401,
      "step": 240
    },
    {
      "epoch": 0.07551729346020239,
      "grad_norm": 1.223191453286042,
      "learning_rate": 9.53582554517134e-05,
      "loss": 0.7566,
      "step": 250
    },
    {
      "epoch": 0.07853798519861048,
      "grad_norm": 1.4311949668852009,
      "learning_rate": 9.504672897196261e-05,
      "loss": 0.8324,
      "step": 260
    },
    {
      "epoch": 0.08155867693701857,
      "grad_norm": 1.1199798019887752,
      "learning_rate": 9.473520249221183e-05,
      "loss": 0.7918,
      "step": 270
    },
    {
      "epoch": 0.08457936867542667,
      "grad_norm": 1.1353541939369336,
      "learning_rate": 9.442367601246106e-05,
      "loss": 0.7982,
      "step": 280
    },
    {
      "epoch": 0.08760006041383477,
      "grad_norm": 1.0377240839273387,
      "learning_rate": 9.411214953271028e-05,
      "loss": 0.759,
      "step": 290
    },
    {
      "epoch": 0.09062075215224287,
      "grad_norm": 1.0430291081517695,
      "learning_rate": 9.38006230529595e-05,
      "loss": 0.7489,
      "step": 300
    },
    {
      "epoch": 0.09364144389065096,
      "grad_norm": 1.0556490523453508,
      "learning_rate": 9.348909657320873e-05,
      "loss": 0.7972,
      "step": 310
    },
    {
      "epoch": 0.09666213562905905,
      "grad_norm": 0.8377680272528175,
      "learning_rate": 9.317757009345795e-05,
      "loss": 0.7883,
      "step": 320
    },
    {
      "epoch": 0.09968282736746716,
      "grad_norm": 1.133534627320159,
      "learning_rate": 9.286604361370717e-05,
      "loss": 0.7963,
      "step": 330
    },
    {
      "epoch": 0.10270351910587525,
      "grad_norm": 1.115336306572395,
      "learning_rate": 9.255451713395639e-05,
      "loss": 0.7115,
      "step": 340
    },
    {
      "epoch": 0.10572421084428334,
      "grad_norm": 1.0771112865985937,
      "learning_rate": 9.22429906542056e-05,
      "loss": 0.6852,
      "step": 350
    },
    {
      "epoch": 0.10874490258269144,
      "grad_norm": 1.1279818268954833,
      "learning_rate": 9.193146417445482e-05,
      "loss": 0.752,
      "step": 360
    },
    {
      "epoch": 0.11176559432109953,
      "grad_norm": 1.1447909234403442,
      "learning_rate": 9.161993769470405e-05,
      "loss": 0.7813,
      "step": 370
    },
    {
      "epoch": 0.11478628605950762,
      "grad_norm": 0.9732050999087476,
      "learning_rate": 9.130841121495327e-05,
      "loss": 0.7529,
      "step": 380
    },
    {
      "epoch": 0.11780697779791573,
      "grad_norm": 0.9183505631954165,
      "learning_rate": 9.099688473520249e-05,
      "loss": 0.8184,
      "step": 390
    },
    {
      "epoch": 0.12082766953632382,
      "grad_norm": 1.3406866026704565,
      "learning_rate": 9.068535825545172e-05,
      "loss": 0.7546,
      "step": 400
    },
    {
      "epoch": 0.12384836127473191,
      "grad_norm": 1.080230905978024,
      "learning_rate": 9.037383177570094e-05,
      "loss": 0.7943,
      "step": 410
    },
    {
      "epoch": 0.12686905301314,
      "grad_norm": 0.981708271024814,
      "learning_rate": 9.006230529595016e-05,
      "loss": 0.7748,
      "step": 420
    },
    {
      "epoch": 0.1298897447515481,
      "grad_norm": 1.032944901370365,
      "learning_rate": 8.975077881619937e-05,
      "loss": 0.7367,
      "step": 430
    },
    {
      "epoch": 0.1329104364899562,
      "grad_norm": 0.9828039027398626,
      "learning_rate": 8.943925233644859e-05,
      "loss": 0.7562,
      "step": 440
    },
    {
      "epoch": 0.13593112822836428,
      "grad_norm": 1.025926244346921,
      "learning_rate": 8.912772585669782e-05,
      "loss": 0.7799,
      "step": 450
    },
    {
      "epoch": 0.1389518199667724,
      "grad_norm": 1.0500959648810009,
      "learning_rate": 8.881619937694704e-05,
      "loss": 0.7523,
      "step": 460
    },
    {
      "epoch": 0.1419725117051805,
      "grad_norm": 1.1701611318445844,
      "learning_rate": 8.850467289719626e-05,
      "loss": 0.7686,
      "step": 470
    },
    {
      "epoch": 0.14499320344358857,
      "grad_norm": 0.8990869023938298,
      "learning_rate": 8.819314641744548e-05,
      "loss": 0.7289,
      "step": 480
    },
    {
      "epoch": 0.14801389518199667,
      "grad_norm": 1.1095162409175816,
      "learning_rate": 8.78816199376947e-05,
      "loss": 0.7332,
      "step": 490
    },
    {
      "epoch": 0.15103458692040478,
      "grad_norm": 1.0108816660931794,
      "learning_rate": 8.757009345794393e-05,
      "loss": 0.7125,
      "step": 500
    },
    {
      "epoch": 0.15405527865881286,
      "grad_norm": 1.3324995099592942,
      "learning_rate": 8.725856697819315e-05,
      "loss": 0.7254,
      "step": 510
    },
    {
      "epoch": 0.15707597039722096,
      "grad_norm": 1.0845745603450858,
      "learning_rate": 8.694704049844236e-05,
      "loss": 0.7115,
      "step": 520
    },
    {
      "epoch": 0.16009666213562906,
      "grad_norm": 0.7967865715388287,
      "learning_rate": 8.663551401869158e-05,
      "loss": 0.7334,
      "step": 530
    },
    {
      "epoch": 0.16311735387403714,
      "grad_norm": 0.9490631327452955,
      "learning_rate": 8.63239875389408e-05,
      "loss": 0.7399,
      "step": 540
    },
    {
      "epoch": 0.16613804561244525,
      "grad_norm": 0.8962644592007071,
      "learning_rate": 8.601246105919003e-05,
      "loss": 0.7498,
      "step": 550
    },
    {
      "epoch": 0.16915873735085335,
      "grad_norm": 1.1678557286758378,
      "learning_rate": 8.570093457943925e-05,
      "loss": 0.7067,
      "step": 560
    },
    {
      "epoch": 0.17217942908926145,
      "grad_norm": 1.088949753627781,
      "learning_rate": 8.538940809968847e-05,
      "loss": 0.748,
      "step": 570
    },
    {
      "epoch": 0.17520012082766953,
      "grad_norm": 1.0893408136265033,
      "learning_rate": 8.50778816199377e-05,
      "loss": 0.7623,
      "step": 580
    },
    {
      "epoch": 0.17822081256607764,
      "grad_norm": 0.9840533089260339,
      "learning_rate": 8.476635514018692e-05,
      "loss": 0.6797,
      "step": 590
    },
    {
      "epoch": 0.18124150430448574,
      "grad_norm": 1.0812493862448191,
      "learning_rate": 8.445482866043614e-05,
      "loss": 0.7275,
      "step": 600
    },
    {
      "epoch": 0.18426219604289382,
      "grad_norm": 1.2195461231662528,
      "learning_rate": 8.414330218068535e-05,
      "loss": 0.6863,
      "step": 610
    },
    {
      "epoch": 0.18728288778130192,
      "grad_norm": 0.8428943981031285,
      "learning_rate": 8.383177570093457e-05,
      "loss": 0.7412,
      "step": 620
    },
    {
      "epoch": 0.19030357951971003,
      "grad_norm": 0.9484156439369784,
      "learning_rate": 8.35202492211838e-05,
      "loss": 0.7479,
      "step": 630
    },
    {
      "epoch": 0.1933242712581181,
      "grad_norm": 1.5268349287880287,
      "learning_rate": 8.320872274143302e-05,
      "loss": 0.7421,
      "step": 640
    },
    {
      "epoch": 0.1963449629965262,
      "grad_norm": 1.0210188261100828,
      "learning_rate": 8.289719626168224e-05,
      "loss": 0.735,
      "step": 650
    },
    {
      "epoch": 0.1993656547349343,
      "grad_norm": 1.0507811901351063,
      "learning_rate": 8.258566978193147e-05,
      "loss": 0.7174,
      "step": 660
    },
    {
      "epoch": 0.2023863464733424,
      "grad_norm": 0.977107559554224,
      "learning_rate": 8.227414330218069e-05,
      "loss": 0.6539,
      "step": 670
    },
    {
      "epoch": 0.2054070382117505,
      "grad_norm": 1.0747746808663372,
      "learning_rate": 8.196261682242991e-05,
      "loss": 0.6529,
      "step": 680
    },
    {
      "epoch": 0.2084277299501586,
      "grad_norm": 0.9082287831059276,
      "learning_rate": 8.165109034267912e-05,
      "loss": 0.732,
      "step": 690
    },
    {
      "epoch": 0.21144842168856667,
      "grad_norm": 0.9212172953657185,
      "learning_rate": 8.133956386292834e-05,
      "loss": 0.6734,
      "step": 700
    },
    {
      "epoch": 0.21446911342697478,
      "grad_norm": 0.7665370190006591,
      "learning_rate": 8.102803738317756e-05,
      "loss": 0.7048,
      "step": 710
    },
    {
      "epoch": 0.21748980516538288,
      "grad_norm": 0.8983656963975071,
      "learning_rate": 8.071651090342679e-05,
      "loss": 0.7166,
      "step": 720
    },
    {
      "epoch": 0.22051049690379096,
      "grad_norm": 0.996733266552071,
      "learning_rate": 8.040498442367601e-05,
      "loss": 0.7354,
      "step": 730
    },
    {
      "epoch": 0.22353118864219906,
      "grad_norm": 0.8782696913983529,
      "learning_rate": 8.009345794392523e-05,
      "loss": 0.6911,
      "step": 740
    },
    {
      "epoch": 0.22655188038060717,
      "grad_norm": 1.0464611268652604,
      "learning_rate": 7.978193146417447e-05,
      "loss": 0.6791,
      "step": 750
    },
    {
      "epoch": 0.22957257211901524,
      "grad_norm": 1.429521202085466,
      "learning_rate": 7.947040498442369e-05,
      "loss": 0.7254,
      "step": 760
    },
    {
      "epoch": 0.23259326385742335,
      "grad_norm": 0.8265481561359109,
      "learning_rate": 7.91588785046729e-05,
      "loss": 0.6933,
      "step": 770
    },
    {
      "epoch": 0.23561395559583145,
      "grad_norm": 1.1395351111445386,
      "learning_rate": 7.884735202492212e-05,
      "loss": 0.6927,
      "step": 780
    },
    {
      "epoch": 0.23863464733423953,
      "grad_norm": 0.8683132914875289,
      "learning_rate": 7.853582554517135e-05,
      "loss": 0.7115,
      "step": 790
    },
    {
      "epoch": 0.24165533907264763,
      "grad_norm": 1.0048357199050497,
      "learning_rate": 7.822429906542057e-05,
      "loss": 0.7258,
      "step": 800
    },
    {
      "epoch": 0.24467603081105574,
      "grad_norm": 0.9186551341926896,
      "learning_rate": 7.791277258566979e-05,
      "loss": 0.6937,
      "step": 810
    },
    {
      "epoch": 0.24769672254946382,
      "grad_norm": 0.90949443461049,
      "learning_rate": 7.760124610591901e-05,
      "loss": 0.717,
      "step": 820
    },
    {
      "epoch": 0.2507174142878719,
      "grad_norm": 1.2709292068990643,
      "learning_rate": 7.728971962616824e-05,
      "loss": 0.7313,
      "step": 830
    },
    {
      "epoch": 0.25373810602628,
      "grad_norm": 1.039112358866248,
      "learning_rate": 7.697819314641746e-05,
      "loss": 0.6961,
      "step": 840
    },
    {
      "epoch": 0.25675879776468813,
      "grad_norm": 0.9738088719115688,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.7418,
      "step": 850
    },
    {
      "epoch": 0.2597794895030962,
      "grad_norm": 1.0293122474854857,
      "learning_rate": 7.635514018691589e-05,
      "loss": 0.7208,
      "step": 860
    },
    {
      "epoch": 0.2628001812415043,
      "grad_norm": 0.9219605479548113,
      "learning_rate": 7.604361370716511e-05,
      "loss": 0.6573,
      "step": 870
    },
    {
      "epoch": 0.2658208729799124,
      "grad_norm": 0.986441148226433,
      "learning_rate": 7.573208722741434e-05,
      "loss": 0.6882,
      "step": 880
    },
    {
      "epoch": 0.2688415647183205,
      "grad_norm": 0.8721644047390303,
      "learning_rate": 7.542056074766356e-05,
      "loss": 0.699,
      "step": 890
    },
    {
      "epoch": 0.27186225645672857,
      "grad_norm": 0.8397538474783728,
      "learning_rate": 7.510903426791278e-05,
      "loss": 0.7473,
      "step": 900
    },
    {
      "epoch": 0.2748829481951367,
      "grad_norm": 0.9204492440191102,
      "learning_rate": 7.4797507788162e-05,
      "loss": 0.682,
      "step": 910
    },
    {
      "epoch": 0.2779036399335448,
      "grad_norm": 0.9916189555392334,
      "learning_rate": 7.448598130841123e-05,
      "loss": 0.6714,
      "step": 920
    },
    {
      "epoch": 0.28092433167195285,
      "grad_norm": 0.9959802073019116,
      "learning_rate": 7.417445482866045e-05,
      "loss": 0.7068,
      "step": 930
    },
    {
      "epoch": 0.283945023410361,
      "grad_norm": 0.822413918128088,
      "learning_rate": 7.386292834890966e-05,
      "loss": 0.6993,
      "step": 940
    },
    {
      "epoch": 0.28696571514876906,
      "grad_norm": 0.9857850966520738,
      "learning_rate": 7.355140186915888e-05,
      "loss": 0.708,
      "step": 950
    },
    {
      "epoch": 0.28998640688717714,
      "grad_norm": 0.7935815329490274,
      "learning_rate": 7.32398753894081e-05,
      "loss": 0.7262,
      "step": 960
    },
    {
      "epoch": 0.29300709862558527,
      "grad_norm": 0.9762914722861878,
      "learning_rate": 7.292834890965733e-05,
      "loss": 0.6663,
      "step": 970
    },
    {
      "epoch": 0.29602779036399335,
      "grad_norm": 0.9841123533490521,
      "learning_rate": 7.261682242990655e-05,
      "loss": 0.693,
      "step": 980
    },
    {
      "epoch": 0.2990484821024014,
      "grad_norm": 0.9136079490982437,
      "learning_rate": 7.230529595015577e-05,
      "loss": 0.7049,
      "step": 990
    },
    {
      "epoch": 0.30206917384080956,
      "grad_norm": 0.8902281683899582,
      "learning_rate": 7.1993769470405e-05,
      "loss": 0.6956,
      "step": 1000
    },
    {
      "epoch": 0.30508986557921763,
      "grad_norm": 0.9555339802943792,
      "learning_rate": 7.168224299065422e-05,
      "loss": 0.6923,
      "step": 1010
    },
    {
      "epoch": 0.3081105573176257,
      "grad_norm": 0.9226515554395557,
      "learning_rate": 7.137071651090344e-05,
      "loss": 0.7047,
      "step": 1020
    },
    {
      "epoch": 0.31113124905603384,
      "grad_norm": 1.0265101302708708,
      "learning_rate": 7.105919003115265e-05,
      "loss": 0.6342,
      "step": 1030
    },
    {
      "epoch": 0.3141519407944419,
      "grad_norm": 0.7354809813158237,
      "learning_rate": 7.074766355140187e-05,
      "loss": 0.7276,
      "step": 1040
    },
    {
      "epoch": 0.31717263253285,
      "grad_norm": 0.753423387786277,
      "learning_rate": 7.04361370716511e-05,
      "loss": 0.6743,
      "step": 1050
    },
    {
      "epoch": 0.32019332427125813,
      "grad_norm": 1.057091313600942,
      "learning_rate": 7.012461059190032e-05,
      "loss": 0.6715,
      "step": 1060
    },
    {
      "epoch": 0.3232140160096662,
      "grad_norm": 0.7838790412397071,
      "learning_rate": 6.981308411214954e-05,
      "loss": 0.6671,
      "step": 1070
    },
    {
      "epoch": 0.3262347077480743,
      "grad_norm": 0.8980932417690051,
      "learning_rate": 6.950155763239876e-05,
      "loss": 0.6774,
      "step": 1080
    },
    {
      "epoch": 0.3292553994864824,
      "grad_norm": 0.9220267366468554,
      "learning_rate": 6.919003115264799e-05,
      "loss": 0.6428,
      "step": 1090
    },
    {
      "epoch": 0.3322760912248905,
      "grad_norm": 0.9214463298045865,
      "learning_rate": 6.887850467289721e-05,
      "loss": 0.7271,
      "step": 1100
    },
    {
      "epoch": 0.3352967829632986,
      "grad_norm": 0.9321131041370911,
      "learning_rate": 6.856697819314643e-05,
      "loss": 0.7258,
      "step": 1110
    },
    {
      "epoch": 0.3383174747017067,
      "grad_norm": 1.279210720148468,
      "learning_rate": 6.825545171339564e-05,
      "loss": 0.6632,
      "step": 1120
    },
    {
      "epoch": 0.3413381664401148,
      "grad_norm": 1.0059923489692595,
      "learning_rate": 6.794392523364486e-05,
      "loss": 0.6364,
      "step": 1130
    },
    {
      "epoch": 0.3443588581785229,
      "grad_norm": 1.047129656921137,
      "learning_rate": 6.763239875389409e-05,
      "loss": 0.7291,
      "step": 1140
    },
    {
      "epoch": 0.347379549916931,
      "grad_norm": 0.9272803839751749,
      "learning_rate": 6.732087227414331e-05,
      "loss": 0.6395,
      "step": 1150
    },
    {
      "epoch": 0.35040024165533906,
      "grad_norm": 1.1151649879002739,
      "learning_rate": 6.700934579439253e-05,
      "loss": 0.6657,
      "step": 1160
    },
    {
      "epoch": 0.3534209333937472,
      "grad_norm": 0.9241781872831288,
      "learning_rate": 6.669781931464175e-05,
      "loss": 0.6653,
      "step": 1170
    },
    {
      "epoch": 0.35644162513215527,
      "grad_norm": 0.8358913806558893,
      "learning_rate": 6.638629283489098e-05,
      "loss": 0.7389,
      "step": 1180
    },
    {
      "epoch": 0.35946231687056335,
      "grad_norm": 1.0319211344473511,
      "learning_rate": 6.60747663551402e-05,
      "loss": 0.6954,
      "step": 1190
    },
    {
      "epoch": 0.3624830086089715,
      "grad_norm": 1.1505728351225282,
      "learning_rate": 6.576323987538941e-05,
      "loss": 0.7043,
      "step": 1200
    },
    {
      "epoch": 0.36550370034737956,
      "grad_norm": 0.7959153182677297,
      "learning_rate": 6.545171339563863e-05,
      "loss": 0.6695,
      "step": 1210
    },
    {
      "epoch": 0.36852439208578763,
      "grad_norm": 0.9337856655171398,
      "learning_rate": 6.514018691588785e-05,
      "loss": 0.7429,
      "step": 1220
    },
    {
      "epoch": 0.37154508382419577,
      "grad_norm": 0.8063590050110404,
      "learning_rate": 6.482866043613708e-05,
      "loss": 0.6899,
      "step": 1230
    },
    {
      "epoch": 0.37456577556260384,
      "grad_norm": 1.0599327700787213,
      "learning_rate": 6.45171339563863e-05,
      "loss": 0.6542,
      "step": 1240
    },
    {
      "epoch": 0.3775864673010119,
      "grad_norm": 0.9744322038123949,
      "learning_rate": 6.420560747663552e-05,
      "loss": 0.706,
      "step": 1250
    },
    {
      "epoch": 0.38060715903942005,
      "grad_norm": 0.8586307372164459,
      "learning_rate": 6.389408099688475e-05,
      "loss": 0.655,
      "step": 1260
    },
    {
      "epoch": 0.3836278507778281,
      "grad_norm": 0.9830477484209696,
      "learning_rate": 6.358255451713397e-05,
      "loss": 0.6742,
      "step": 1270
    },
    {
      "epoch": 0.3866485425162362,
      "grad_norm": 0.9902981913412452,
      "learning_rate": 6.327102803738319e-05,
      "loss": 0.6978,
      "step": 1280
    },
    {
      "epoch": 0.38966923425464434,
      "grad_norm": 1.1185182362842843,
      "learning_rate": 6.29595015576324e-05,
      "loss": 0.73,
      "step": 1290
    },
    {
      "epoch": 0.3926899259930524,
      "grad_norm": 0.877292130995857,
      "learning_rate": 6.264797507788162e-05,
      "loss": 0.7157,
      "step": 1300
    },
    {
      "epoch": 0.3957106177314605,
      "grad_norm": 0.9682140121699045,
      "learning_rate": 6.233644859813084e-05,
      "loss": 0.678,
      "step": 1310
    },
    {
      "epoch": 0.3987313094698686,
      "grad_norm": 1.0751962176923822,
      "learning_rate": 6.202492211838007e-05,
      "loss": 0.6393,
      "step": 1320
    },
    {
      "epoch": 0.4017520012082767,
      "grad_norm": 0.8236677739270527,
      "learning_rate": 6.171339563862929e-05,
      "loss": 0.6709,
      "step": 1330
    },
    {
      "epoch": 0.4047726929466848,
      "grad_norm": 0.9406647365344084,
      "learning_rate": 6.140186915887851e-05,
      "loss": 0.6588,
      "step": 1340
    },
    {
      "epoch": 0.4077933846850929,
      "grad_norm": 1.2046005814359064,
      "learning_rate": 6.109034267912774e-05,
      "loss": 0.6807,
      "step": 1350
    },
    {
      "epoch": 0.410814076423501,
      "grad_norm": 0.9066360882428902,
      "learning_rate": 6.077881619937695e-05,
      "loss": 0.7168,
      "step": 1360
    },
    {
      "epoch": 0.41383476816190906,
      "grad_norm": 0.9527150302995969,
      "learning_rate": 6.0467289719626174e-05,
      "loss": 0.6984,
      "step": 1370
    },
    {
      "epoch": 0.4168554599003172,
      "grad_norm": 0.8541783908332572,
      "learning_rate": 6.01557632398754e-05,
      "loss": 0.6596,
      "step": 1380
    },
    {
      "epoch": 0.41987615163872527,
      "grad_norm": 0.9929215209215169,
      "learning_rate": 5.984423676012462e-05,
      "loss": 0.6496,
      "step": 1390
    },
    {
      "epoch": 0.42289684337713335,
      "grad_norm": 0.8614209306432289,
      "learning_rate": 5.9532710280373836e-05,
      "loss": 0.6639,
      "step": 1400
    },
    {
      "epoch": 0.4259175351155415,
      "grad_norm": 0.9182204803735963,
      "learning_rate": 5.922118380062306e-05,
      "loss": 0.6599,
      "step": 1410
    },
    {
      "epoch": 0.42893822685394956,
      "grad_norm": 0.991639681929367,
      "learning_rate": 5.890965732087228e-05,
      "loss": 0.6997,
      "step": 1420
    },
    {
      "epoch": 0.43195891859235763,
      "grad_norm": 0.9192897320069349,
      "learning_rate": 5.8598130841121504e-05,
      "loss": 0.6911,
      "step": 1430
    },
    {
      "epoch": 0.43497961033076576,
      "grad_norm": 0.8791677727387652,
      "learning_rate": 5.828660436137072e-05,
      "loss": 0.6133,
      "step": 1440
    },
    {
      "epoch": 0.43800030206917384,
      "grad_norm": 0.9146654670599264,
      "learning_rate": 5.797507788161994e-05,
      "loss": 0.6671,
      "step": 1450
    },
    {
      "epoch": 0.4410209938075819,
      "grad_norm": 0.7904811312174689,
      "learning_rate": 5.7663551401869165e-05,
      "loss": 0.6664,
      "step": 1460
    },
    {
      "epoch": 0.44404168554599005,
      "grad_norm": 0.7986284407143894,
      "learning_rate": 5.735202492211839e-05,
      "loss": 0.7036,
      "step": 1470
    },
    {
      "epoch": 0.4470623772843981,
      "grad_norm": 1.1832322364502883,
      "learning_rate": 5.704049844236761e-05,
      "loss": 0.6819,
      "step": 1480
    },
    {
      "epoch": 0.4500830690228062,
      "grad_norm": 1.416546083504211,
      "learning_rate": 5.6728971962616826e-05,
      "loss": 0.6478,
      "step": 1490
    },
    {
      "epoch": 0.45310376076121434,
      "grad_norm": 0.9415627484625246,
      "learning_rate": 5.641744548286605e-05,
      "loss": 0.6146,
      "step": 1500
    },
    {
      "epoch": 0.4561244524996224,
      "grad_norm": 0.9223288773940925,
      "learning_rate": 5.610591900311527e-05,
      "loss": 0.7053,
      "step": 1510
    },
    {
      "epoch": 0.4591451442380305,
      "grad_norm": 0.970256765114157,
      "learning_rate": 5.5794392523364494e-05,
      "loss": 0.6443,
      "step": 1520
    },
    {
      "epoch": 0.4621658359764386,
      "grad_norm": 0.7403345850081653,
      "learning_rate": 5.548286604361371e-05,
      "loss": 0.6207,
      "step": 1530
    },
    {
      "epoch": 0.4651865277148467,
      "grad_norm": 1.1294910445688315,
      "learning_rate": 5.517133956386293e-05,
      "loss": 0.6837,
      "step": 1540
    },
    {
      "epoch": 0.4682072194532548,
      "grad_norm": 1.084685705504392,
      "learning_rate": 5.4859813084112156e-05,
      "loss": 0.6619,
      "step": 1550
    },
    {
      "epoch": 0.4712279111916629,
      "grad_norm": 0.9613159354858272,
      "learning_rate": 5.454828660436138e-05,
      "loss": 0.6515,
      "step": 1560
    },
    {
      "epoch": 0.474248602930071,
      "grad_norm": 0.9837692555022853,
      "learning_rate": 5.4236760124610594e-05,
      "loss": 0.5924,
      "step": 1570
    },
    {
      "epoch": 0.47726929466847906,
      "grad_norm": 0.9979615856918954,
      "learning_rate": 5.392523364485982e-05,
      "loss": 0.7001,
      "step": 1580
    },
    {
      "epoch": 0.4802899864068872,
      "grad_norm": 0.9728439737052923,
      "learning_rate": 5.361370716510904e-05,
      "loss": 0.6458,
      "step": 1590
    },
    {
      "epoch": 0.48331067814529527,
      "grad_norm": 1.0126438240737292,
      "learning_rate": 5.330218068535826e-05,
      "loss": 0.7399,
      "step": 1600
    },
    {
      "epoch": 0.48633136988370335,
      "grad_norm": 1.0282465120937962,
      "learning_rate": 5.2990654205607485e-05,
      "loss": 0.6591,
      "step": 1610
    },
    {
      "epoch": 0.4893520616221115,
      "grad_norm": 1.0608069664171718,
      "learning_rate": 5.26791277258567e-05,
      "loss": 0.6852,
      "step": 1620
    },
    {
      "epoch": 0.49237275336051955,
      "grad_norm": 1.06193177949756,
      "learning_rate": 5.2367601246105924e-05,
      "loss": 0.6256,
      "step": 1630
    },
    {
      "epoch": 0.49539344509892763,
      "grad_norm": 0.8994514739072711,
      "learning_rate": 5.2056074766355146e-05,
      "loss": 0.6157,
      "step": 1640
    },
    {
      "epoch": 0.49841413683733576,
      "grad_norm": 0.8831776925916133,
      "learning_rate": 5.174454828660437e-05,
      "loss": 0.6585,
      "step": 1650
    },
    {
      "epoch": 0.5014348285757438,
      "grad_norm": 0.9057465973326372,
      "learning_rate": 5.1433021806853585e-05,
      "loss": 0.6548,
      "step": 1660
    },
    {
      "epoch": 0.504455520314152,
      "grad_norm": 0.8470568048981005,
      "learning_rate": 5.112149532710281e-05,
      "loss": 0.5995,
      "step": 1670
    },
    {
      "epoch": 0.50747621205256,
      "grad_norm": 0.9274346159474457,
      "learning_rate": 5.080996884735203e-05,
      "loss": 0.6695,
      "step": 1680
    },
    {
      "epoch": 0.5104969037909681,
      "grad_norm": 0.8266785564731086,
      "learning_rate": 5.049844236760125e-05,
      "loss": 0.6749,
      "step": 1690
    },
    {
      "epoch": 0.5135175955293763,
      "grad_norm": 1.0808475616587816,
      "learning_rate": 5.018691588785047e-05,
      "loss": 0.6696,
      "step": 1700
    },
    {
      "epoch": 0.5165382872677843,
      "grad_norm": 0.938631750728695,
      "learning_rate": 4.987538940809969e-05,
      "loss": 0.6953,
      "step": 1710
    },
    {
      "epoch": 0.5195589790061924,
      "grad_norm": 1.0041642156156723,
      "learning_rate": 4.9563862928348914e-05,
      "loss": 0.6873,
      "step": 1720
    },
    {
      "epoch": 0.5225796707446005,
      "grad_norm": 0.8610395612409752,
      "learning_rate": 4.925233644859814e-05,
      "loss": 0.7225,
      "step": 1730
    },
    {
      "epoch": 0.5256003624830086,
      "grad_norm": 0.9496184351237903,
      "learning_rate": 4.894080996884736e-05,
      "loss": 0.6382,
      "step": 1740
    },
    {
      "epoch": 0.5286210542214167,
      "grad_norm": 1.0830488045922504,
      "learning_rate": 4.8629283489096575e-05,
      "loss": 0.6314,
      "step": 1750
    },
    {
      "epoch": 0.5316417459598248,
      "grad_norm": 1.3308812167096526,
      "learning_rate": 4.83177570093458e-05,
      "loss": 0.6279,
      "step": 1760
    },
    {
      "epoch": 0.5346624376982329,
      "grad_norm": 1.0384804569396457,
      "learning_rate": 4.800623052959502e-05,
      "loss": 0.6661,
      "step": 1770
    },
    {
      "epoch": 0.537683129436641,
      "grad_norm": 0.9268941184363317,
      "learning_rate": 4.7694704049844244e-05,
      "loss": 0.5982,
      "step": 1780
    },
    {
      "epoch": 0.5407038211750491,
      "grad_norm": 1.0041389636701843,
      "learning_rate": 4.738317757009346e-05,
      "loss": 0.5852,
      "step": 1790
    },
    {
      "epoch": 0.5437245129134571,
      "grad_norm": 1.0812438556748192,
      "learning_rate": 4.707165109034268e-05,
      "loss": 0.6445,
      "step": 1800
    },
    {
      "epoch": 0.5467452046518653,
      "grad_norm": 0.7998207518212354,
      "learning_rate": 4.6760124610591905e-05,
      "loss": 0.6212,
      "step": 1810
    },
    {
      "epoch": 0.5497658963902734,
      "grad_norm": 0.9860219965019006,
      "learning_rate": 4.644859813084113e-05,
      "loss": 0.6453,
      "step": 1820
    },
    {
      "epoch": 0.5527865881286814,
      "grad_norm": 1.0673930524321107,
      "learning_rate": 4.6137071651090343e-05,
      "loss": 0.5837,
      "step": 1830
    },
    {
      "epoch": 0.5558072798670896,
      "grad_norm": 1.143808636631285,
      "learning_rate": 4.5825545171339566e-05,
      "loss": 0.6996,
      "step": 1840
    },
    {
      "epoch": 0.5588279716054977,
      "grad_norm": 0.9805310678029436,
      "learning_rate": 4.551401869158879e-05,
      "loss": 0.6062,
      "step": 1850
    },
    {
      "epoch": 0.5618486633439057,
      "grad_norm": 1.0396445753595607,
      "learning_rate": 4.520249221183801e-05,
      "loss": 0.6249,
      "step": 1860
    },
    {
      "epoch": 0.5648693550823138,
      "grad_norm": 0.9626964932261893,
      "learning_rate": 4.4890965732087234e-05,
      "loss": 0.6571,
      "step": 1870
    },
    {
      "epoch": 0.567890046820722,
      "grad_norm": 0.9540429394126168,
      "learning_rate": 4.457943925233645e-05,
      "loss": 0.6324,
      "step": 1880
    },
    {
      "epoch": 0.57091073855913,
      "grad_norm": 0.9015174511995171,
      "learning_rate": 4.426791277258567e-05,
      "loss": 0.6863,
      "step": 1890
    },
    {
      "epoch": 0.5739314302975381,
      "grad_norm": 0.7989728677956948,
      "learning_rate": 4.3956386292834895e-05,
      "loss": 0.6107,
      "step": 1900
    },
    {
      "epoch": 0.5769521220359463,
      "grad_norm": 1.0283655646158905,
      "learning_rate": 4.364485981308412e-05,
      "loss": 0.5799,
      "step": 1910
    },
    {
      "epoch": 0.5799728137743543,
      "grad_norm": 0.8496110087014387,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.6114,
      "step": 1920
    },
    {
      "epoch": 0.5829935055127624,
      "grad_norm": 0.9378297749359396,
      "learning_rate": 4.302180685358256e-05,
      "loss": 0.6238,
      "step": 1930
    },
    {
      "epoch": 0.5860141972511705,
      "grad_norm": 0.7724041203791813,
      "learning_rate": 4.271028037383178e-05,
      "loss": 0.6684,
      "step": 1940
    },
    {
      "epoch": 0.5890348889895786,
      "grad_norm": 0.919952085375962,
      "learning_rate": 4.2398753894081e-05,
      "loss": 0.658,
      "step": 1950
    },
    {
      "epoch": 0.5920555807279867,
      "grad_norm": 1.0984535902148247,
      "learning_rate": 4.2087227414330225e-05,
      "loss": 0.6271,
      "step": 1960
    },
    {
      "epoch": 0.5950762724663948,
      "grad_norm": 1.089793595134447,
      "learning_rate": 4.177570093457944e-05,
      "loss": 0.6401,
      "step": 1970
    },
    {
      "epoch": 0.5980969642048029,
      "grad_norm": 0.9470634176090514,
      "learning_rate": 4.1464174454828663e-05,
      "loss": 0.6269,
      "step": 1980
    },
    {
      "epoch": 0.601117655943211,
      "grad_norm": 0.9586690317848461,
      "learning_rate": 4.1152647975077886e-05,
      "loss": 0.6455,
      "step": 1990
    },
    {
      "epoch": 0.6041383476816191,
      "grad_norm": 0.9355637416127348,
      "learning_rate": 4.084112149532711e-05,
      "loss": 0.6505,
      "step": 2000
    },
    {
      "epoch": 0.6071590394200271,
      "grad_norm": 0.9422028995953846,
      "learning_rate": 4.0529595015576325e-05,
      "loss": 0.5881,
      "step": 2010
    },
    {
      "epoch": 0.6101797311584353,
      "grad_norm": 1.0662379927757937,
      "learning_rate": 4.021806853582555e-05,
      "loss": 0.6965,
      "step": 2020
    },
    {
      "epoch": 0.6132004228968434,
      "grad_norm": 1.112340047069814,
      "learning_rate": 3.990654205607477e-05,
      "loss": 0.6696,
      "step": 2030
    },
    {
      "epoch": 0.6162211146352514,
      "grad_norm": 0.9811828345166662,
      "learning_rate": 3.959501557632399e-05,
      "loss": 0.6774,
      "step": 2040
    },
    {
      "epoch": 0.6192418063736596,
      "grad_norm": 0.8797632809116769,
      "learning_rate": 3.928348909657321e-05,
      "loss": 0.6814,
      "step": 2050
    },
    {
      "epoch": 0.6222624981120677,
      "grad_norm": 1.0012864041487008,
      "learning_rate": 3.897196261682243e-05,
      "loss": 0.6789,
      "step": 2060
    },
    {
      "epoch": 0.6252831898504757,
      "grad_norm": 0.9363857271307736,
      "learning_rate": 3.8660436137071654e-05,
      "loss": 0.6286,
      "step": 2070
    },
    {
      "epoch": 0.6283038815888838,
      "grad_norm": 0.8994398345785907,
      "learning_rate": 3.834890965732088e-05,
      "loss": 0.5957,
      "step": 2080
    },
    {
      "epoch": 0.631324573327292,
      "grad_norm": 0.9307639618859869,
      "learning_rate": 3.80373831775701e-05,
      "loss": 0.6365,
      "step": 2090
    },
    {
      "epoch": 0.6343452650657,
      "grad_norm": 0.8458149477874153,
      "learning_rate": 3.7725856697819315e-05,
      "loss": 0.682,
      "step": 2100
    },
    {
      "epoch": 0.6373659568041081,
      "grad_norm": 0.782298940529486,
      "learning_rate": 3.741433021806854e-05,
      "loss": 0.68,
      "step": 2110
    },
    {
      "epoch": 0.6403866485425163,
      "grad_norm": 0.9771834449277986,
      "learning_rate": 3.710280373831776e-05,
      "loss": 0.6732,
      "step": 2120
    },
    {
      "epoch": 0.6434073402809243,
      "grad_norm": 0.9295914583308844,
      "learning_rate": 3.6791277258566983e-05,
      "loss": 0.6173,
      "step": 2130
    },
    {
      "epoch": 0.6464280320193324,
      "grad_norm": 1.179909806414257,
      "learning_rate": 3.64797507788162e-05,
      "loss": 0.6193,
      "step": 2140
    },
    {
      "epoch": 0.6494487237577405,
      "grad_norm": 0.99878344996002,
      "learning_rate": 3.616822429906542e-05,
      "loss": 0.6116,
      "step": 2150
    },
    {
      "epoch": 0.6524694154961486,
      "grad_norm": 1.1005346639943379,
      "learning_rate": 3.5856697819314645e-05,
      "loss": 0.6794,
      "step": 2160
    },
    {
      "epoch": 0.6554901072345567,
      "grad_norm": 0.8754349383421103,
      "learning_rate": 3.554517133956387e-05,
      "loss": 0.627,
      "step": 2170
    },
    {
      "epoch": 0.6585107989729648,
      "grad_norm": 0.8266136534894406,
      "learning_rate": 3.523364485981308e-05,
      "loss": 0.5961,
      "step": 2180
    },
    {
      "epoch": 0.6615314907113728,
      "grad_norm": 1.0546584904026568,
      "learning_rate": 3.4922118380062306e-05,
      "loss": 0.6374,
      "step": 2190
    },
    {
      "epoch": 0.664552182449781,
      "grad_norm": 0.9197595680333482,
      "learning_rate": 3.461059190031153e-05,
      "loss": 0.6493,
      "step": 2200
    },
    {
      "epoch": 0.6675728741881891,
      "grad_norm": 0.8832736598142242,
      "learning_rate": 3.429906542056075e-05,
      "loss": 0.6243,
      "step": 2210
    },
    {
      "epoch": 0.6705935659265972,
      "grad_norm": 0.7953765164555499,
      "learning_rate": 3.3987538940809974e-05,
      "loss": 0.6413,
      "step": 2220
    },
    {
      "epoch": 0.6736142576650053,
      "grad_norm": 0.8616203793917606,
      "learning_rate": 3.367601246105919e-05,
      "loss": 0.6466,
      "step": 2230
    },
    {
      "epoch": 0.6766349494034134,
      "grad_norm": 0.9546691619897619,
      "learning_rate": 3.336448598130841e-05,
      "loss": 0.61,
      "step": 2240
    },
    {
      "epoch": 0.6796556411418215,
      "grad_norm": 0.9931567171367084,
      "learning_rate": 3.3052959501557635e-05,
      "loss": 0.6541,
      "step": 2250
    },
    {
      "epoch": 0.6826763328802296,
      "grad_norm": 0.910445823032414,
      "learning_rate": 3.274143302180686e-05,
      "loss": 0.6512,
      "step": 2260
    },
    {
      "epoch": 0.6856970246186377,
      "grad_norm": 0.9925620053257106,
      "learning_rate": 3.2429906542056074e-05,
      "loss": 0.6731,
      "step": 2270
    },
    {
      "epoch": 0.6887177163570458,
      "grad_norm": 0.9314964027183025,
      "learning_rate": 3.21183800623053e-05,
      "loss": 0.6349,
      "step": 2280
    },
    {
      "epoch": 0.6917384080954538,
      "grad_norm": 0.9924403199233388,
      "learning_rate": 3.180685358255452e-05,
      "loss": 0.6119,
      "step": 2290
    },
    {
      "epoch": 0.694759099833862,
      "grad_norm": 1.016617404918067,
      "learning_rate": 3.149532710280374e-05,
      "loss": 0.6364,
      "step": 2300
    },
    {
      "epoch": 0.6977797915722701,
      "grad_norm": 1.2100326550184481,
      "learning_rate": 3.118380062305296e-05,
      "loss": 0.6502,
      "step": 2310
    },
    {
      "epoch": 0.7008004833106781,
      "grad_norm": 0.8640417568500796,
      "learning_rate": 3.087227414330218e-05,
      "loss": 0.6129,
      "step": 2320
    },
    {
      "epoch": 0.7038211750490863,
      "grad_norm": 1.0875619618857508,
      "learning_rate": 3.05607476635514e-05,
      "loss": 0.6615,
      "step": 2330
    },
    {
      "epoch": 0.7068418667874944,
      "grad_norm": 0.8939128045261261,
      "learning_rate": 3.0249221183800626e-05,
      "loss": 0.6081,
      "step": 2340
    },
    {
      "epoch": 0.7098625585259024,
      "grad_norm": 0.8987589575526329,
      "learning_rate": 2.9937694704049845e-05,
      "loss": 0.6811,
      "step": 2350
    },
    {
      "epoch": 0.7128832502643105,
      "grad_norm": 0.9024007443688661,
      "learning_rate": 2.9626168224299068e-05,
      "loss": 0.6009,
      "step": 2360
    },
    {
      "epoch": 0.7159039420027187,
      "grad_norm": 0.963849254096249,
      "learning_rate": 2.9314641744548287e-05,
      "loss": 0.6509,
      "step": 2370
    },
    {
      "epoch": 0.7189246337411267,
      "grad_norm": 1.0800126230646785,
      "learning_rate": 2.900311526479751e-05,
      "loss": 0.6791,
      "step": 2380
    },
    {
      "epoch": 0.7219453254795348,
      "grad_norm": 0.8294159891121566,
      "learning_rate": 2.869158878504673e-05,
      "loss": 0.6334,
      "step": 2390
    },
    {
      "epoch": 0.724966017217943,
      "grad_norm": 1.0308203810536511,
      "learning_rate": 2.8380062305295952e-05,
      "loss": 0.627,
      "step": 2400
    },
    {
      "epoch": 0.727986708956351,
      "grad_norm": 0.9512924723408093,
      "learning_rate": 2.806853582554517e-05,
      "loss": 0.5754,
      "step": 2410
    },
    {
      "epoch": 0.7310074006947591,
      "grad_norm": 1.010781394411619,
      "learning_rate": 2.7757009345794394e-05,
      "loss": 0.6416,
      "step": 2420
    },
    {
      "epoch": 0.7340280924331672,
      "grad_norm": 1.0951399649684406,
      "learning_rate": 2.7445482866043613e-05,
      "loss": 0.6088,
      "step": 2430
    },
    {
      "epoch": 0.7370487841715753,
      "grad_norm": 1.0040444715395704,
      "learning_rate": 2.7133956386292836e-05,
      "loss": 0.6473,
      "step": 2440
    },
    {
      "epoch": 0.7400694759099834,
      "grad_norm": 0.9086640535011864,
      "learning_rate": 2.682242990654206e-05,
      "loss": 0.5991,
      "step": 2450
    },
    {
      "epoch": 0.7430901676483915,
      "grad_norm": 0.9979590888744413,
      "learning_rate": 2.6510903426791278e-05,
      "loss": 0.6554,
      "step": 2460
    },
    {
      "epoch": 0.7461108593867996,
      "grad_norm": 0.996844139395838,
      "learning_rate": 2.61993769470405e-05,
      "loss": 0.611,
      "step": 2470
    },
    {
      "epoch": 0.7491315511252077,
      "grad_norm": 1.0523632147482274,
      "learning_rate": 2.588785046728972e-05,
      "loss": 0.6327,
      "step": 2480
    },
    {
      "epoch": 0.7521522428636158,
      "grad_norm": 1.1077614436376035,
      "learning_rate": 2.5576323987538943e-05,
      "loss": 0.6846,
      "step": 2490
    },
    {
      "epoch": 0.7551729346020238,
      "grad_norm": 1.0062846906925333,
      "learning_rate": 2.5264797507788162e-05,
      "loss": 0.632,
      "step": 2500
    },
    {
      "epoch": 0.758193626340432,
      "grad_norm": 0.8044057354860985,
      "learning_rate": 2.4953271028037385e-05,
      "loss": 0.6423,
      "step": 2510
    },
    {
      "epoch": 0.7612143180788401,
      "grad_norm": 1.0776815216184104,
      "learning_rate": 2.4641744548286604e-05,
      "loss": 0.5959,
      "step": 2520
    },
    {
      "epoch": 0.7642350098172481,
      "grad_norm": 1.3229091185592443,
      "learning_rate": 2.4330218068535827e-05,
      "loss": 0.6308,
      "step": 2530
    },
    {
      "epoch": 0.7672557015556563,
      "grad_norm": 1.0134005011914144,
      "learning_rate": 2.4018691588785046e-05,
      "loss": 0.6082,
      "step": 2540
    },
    {
      "epoch": 0.7702763932940644,
      "grad_norm": 1.0710215536966488,
      "learning_rate": 2.370716510903427e-05,
      "loss": 0.5964,
      "step": 2550
    },
    {
      "epoch": 0.7732970850324724,
      "grad_norm": 1.0857761269794204,
      "learning_rate": 2.3395638629283488e-05,
      "loss": 0.5891,
      "step": 2560
    },
    {
      "epoch": 0.7763177767708805,
      "grad_norm": 0.9369701363776228,
      "learning_rate": 2.308411214953271e-05,
      "loss": 0.6104,
      "step": 2570
    },
    {
      "epoch": 0.7793384685092887,
      "grad_norm": 0.8529120199979476,
      "learning_rate": 2.2772585669781933e-05,
      "loss": 0.6451,
      "step": 2580
    },
    {
      "epoch": 0.7823591602476967,
      "grad_norm": 0.9061359038046752,
      "learning_rate": 2.2461059190031153e-05,
      "loss": 0.6287,
      "step": 2590
    },
    {
      "epoch": 0.7853798519861048,
      "grad_norm": 1.022842846924079,
      "learning_rate": 2.2149532710280375e-05,
      "loss": 0.6424,
      "step": 2600
    },
    {
      "epoch": 0.788400543724513,
      "grad_norm": 1.1644483297932762,
      "learning_rate": 2.1838006230529595e-05,
      "loss": 0.6165,
      "step": 2610
    },
    {
      "epoch": 0.791421235462921,
      "grad_norm": 0.901213276967422,
      "learning_rate": 2.1526479750778817e-05,
      "loss": 0.6631,
      "step": 2620
    },
    {
      "epoch": 0.7944419272013291,
      "grad_norm": 0.939968751090103,
      "learning_rate": 2.1214953271028037e-05,
      "loss": 0.6562,
      "step": 2630
    },
    {
      "epoch": 0.7974626189397372,
      "grad_norm": 0.9433346580971289,
      "learning_rate": 2.090342679127726e-05,
      "loss": 0.6181,
      "step": 2640
    },
    {
      "epoch": 0.8004833106781453,
      "grad_norm": 1.0194887755025552,
      "learning_rate": 2.059190031152648e-05,
      "loss": 0.6399,
      "step": 2650
    },
    {
      "epoch": 0.8035040024165534,
      "grad_norm": 1.1638844007159146,
      "learning_rate": 2.02803738317757e-05,
      "loss": 0.6491,
      "step": 2660
    },
    {
      "epoch": 0.8065246941549615,
      "grad_norm": 1.0276398667542848,
      "learning_rate": 1.996884735202492e-05,
      "loss": 0.6145,
      "step": 2670
    },
    {
      "epoch": 0.8095453858933696,
      "grad_norm": 0.8210616041641313,
      "learning_rate": 1.9657320872274143e-05,
      "loss": 0.6466,
      "step": 2680
    },
    {
      "epoch": 0.8125660776317777,
      "grad_norm": 1.093871978484254,
      "learning_rate": 1.9345794392523366e-05,
      "loss": 0.6194,
      "step": 2690
    },
    {
      "epoch": 0.8155867693701858,
      "grad_norm": 1.2620240032817829,
      "learning_rate": 1.9034267912772585e-05,
      "loss": 0.6183,
      "step": 2700
    },
    {
      "epoch": 0.8186074611085938,
      "grad_norm": 0.8779223909750274,
      "learning_rate": 1.8722741433021808e-05,
      "loss": 0.6586,
      "step": 2710
    },
    {
      "epoch": 0.821628152847002,
      "grad_norm": 0.7699492596332658,
      "learning_rate": 1.8411214953271027e-05,
      "loss": 0.5942,
      "step": 2720
    },
    {
      "epoch": 0.8246488445854101,
      "grad_norm": 1.1685873990863083,
      "learning_rate": 1.809968847352025e-05,
      "loss": 0.6443,
      "step": 2730
    },
    {
      "epoch": 0.8276695363238181,
      "grad_norm": 1.083337396850288,
      "learning_rate": 1.778816199376947e-05,
      "loss": 0.6482,
      "step": 2740
    },
    {
      "epoch": 0.8306902280622263,
      "grad_norm": 0.8063684272668464,
      "learning_rate": 1.7476635514018695e-05,
      "loss": 0.606,
      "step": 2750
    },
    {
      "epoch": 0.8337109198006344,
      "grad_norm": 0.824663714107508,
      "learning_rate": 1.7165109034267915e-05,
      "loss": 0.586,
      "step": 2760
    },
    {
      "epoch": 0.8367316115390424,
      "grad_norm": 0.9760183523742957,
      "learning_rate": 1.6853582554517137e-05,
      "loss": 0.5791,
      "step": 2770
    },
    {
      "epoch": 0.8397523032774505,
      "grad_norm": 1.0330407004022142,
      "learning_rate": 1.6542056074766357e-05,
      "loss": 0.6037,
      "step": 2780
    },
    {
      "epoch": 0.8427729950158587,
      "grad_norm": 0.9234529390674108,
      "learning_rate": 1.623052959501558e-05,
      "loss": 0.61,
      "step": 2790
    },
    {
      "epoch": 0.8457936867542667,
      "grad_norm": 1.0821057854566793,
      "learning_rate": 1.59190031152648e-05,
      "loss": 0.6463,
      "step": 2800
    },
    {
      "epoch": 0.8488143784926748,
      "grad_norm": 1.0890558497862515,
      "learning_rate": 1.560747663551402e-05,
      "loss": 0.6438,
      "step": 2810
    },
    {
      "epoch": 0.851835070231083,
      "grad_norm": 0.7374750576790412,
      "learning_rate": 1.529595015576324e-05,
      "loss": 0.6205,
      "step": 2820
    },
    {
      "epoch": 0.854855761969491,
      "grad_norm": 0.8883192786185565,
      "learning_rate": 1.4984423676012463e-05,
      "loss": 0.6249,
      "step": 2830
    },
    {
      "epoch": 0.8578764537078991,
      "grad_norm": 1.1008000896235184,
      "learning_rate": 1.4672897196261684e-05,
      "loss": 0.5964,
      "step": 2840
    },
    {
      "epoch": 0.8608971454463072,
      "grad_norm": 0.9588924680911977,
      "learning_rate": 1.4361370716510905e-05,
      "loss": 0.6246,
      "step": 2850
    },
    {
      "epoch": 0.8639178371847153,
      "grad_norm": 1.224312320705181,
      "learning_rate": 1.4049844236760126e-05,
      "loss": 0.6571,
      "step": 2860
    },
    {
      "epoch": 0.8669385289231234,
      "grad_norm": 0.9905124714238223,
      "learning_rate": 1.3738317757009347e-05,
      "loss": 0.6687,
      "step": 2870
    },
    {
      "epoch": 0.8699592206615315,
      "grad_norm": 0.8353778159477359,
      "learning_rate": 1.3426791277258568e-05,
      "loss": 0.6,
      "step": 2880
    },
    {
      "epoch": 0.8729799123999396,
      "grad_norm": 0.905312921332729,
      "learning_rate": 1.311526479750779e-05,
      "loss": 0.606,
      "step": 2890
    },
    {
      "epoch": 0.8760006041383477,
      "grad_norm": 0.8649231871193698,
      "learning_rate": 1.280373831775701e-05,
      "loss": 0.5743,
      "step": 2900
    },
    {
      "epoch": 0.8790212958767558,
      "grad_norm": 1.0254699594493124,
      "learning_rate": 1.2492211838006231e-05,
      "loss": 0.6204,
      "step": 2910
    },
    {
      "epoch": 0.8820419876151638,
      "grad_norm": 1.0278406308415082,
      "learning_rate": 1.2180685358255452e-05,
      "loss": 0.5853,
      "step": 2920
    },
    {
      "epoch": 0.885062679353572,
      "grad_norm": 0.9990676256188065,
      "learning_rate": 1.1869158878504673e-05,
      "loss": 0.6224,
      "step": 2930
    },
    {
      "epoch": 0.8880833710919801,
      "grad_norm": 1.0949270827311683,
      "learning_rate": 1.1557632398753894e-05,
      "loss": 0.6125,
      "step": 2940
    },
    {
      "epoch": 0.8911040628303881,
      "grad_norm": 1.2006202540115667,
      "learning_rate": 1.1246105919003117e-05,
      "loss": 0.6392,
      "step": 2950
    },
    {
      "epoch": 0.8941247545687963,
      "grad_norm": 1.1187121516645784,
      "learning_rate": 1.0934579439252338e-05,
      "loss": 0.6357,
      "step": 2960
    },
    {
      "epoch": 0.8971454463072044,
      "grad_norm": 0.8638598049799623,
      "learning_rate": 1.0623052959501559e-05,
      "loss": 0.5989,
      "step": 2970
    },
    {
      "epoch": 0.9001661380456124,
      "grad_norm": 0.9297895550418135,
      "learning_rate": 1.031152647975078e-05,
      "loss": 0.6528,
      "step": 2980
    },
    {
      "epoch": 0.9031868297840205,
      "grad_norm": 0.9615985258637098,
      "learning_rate": 1e-05,
      "loss": 0.6123,
      "step": 2990
    },
    {
      "epoch": 0.9062075215224287,
      "grad_norm": 0.8309736923021925,
      "learning_rate": 9.688473520249222e-06,
      "loss": 0.6459,
      "step": 3000
    },
    {
      "epoch": 0.9092282132608367,
      "grad_norm": 0.9815142795709194,
      "learning_rate": 9.376947040498443e-06,
      "loss": 0.571,
      "step": 3010
    },
    {
      "epoch": 0.9122489049992448,
      "grad_norm": 0.8008160564268554,
      "learning_rate": 9.065420560747664e-06,
      "loss": 0.6037,
      "step": 3020
    },
    {
      "epoch": 0.915269596737653,
      "grad_norm": 1.0819593992572683,
      "learning_rate": 8.753894080996885e-06,
      "loss": 0.5869,
      "step": 3030
    },
    {
      "epoch": 0.918290288476061,
      "grad_norm": 0.8665948995270025,
      "learning_rate": 8.442367601246107e-06,
      "loss": 0.611,
      "step": 3040
    },
    {
      "epoch": 0.9213109802144691,
      "grad_norm": 0.8583950697207722,
      "learning_rate": 8.130841121495328e-06,
      "loss": 0.579,
      "step": 3050
    },
    {
      "epoch": 0.9243316719528772,
      "grad_norm": 0.921012636719109,
      "learning_rate": 7.81931464174455e-06,
      "loss": 0.6101,
      "step": 3060
    },
    {
      "epoch": 0.9273523636912853,
      "grad_norm": 0.8604105683200605,
      "learning_rate": 7.5077881619937705e-06,
      "loss": 0.585,
      "step": 3070
    },
    {
      "epoch": 0.9303730554296934,
      "grad_norm": 0.9924409498073824,
      "learning_rate": 7.1962616822429915e-06,
      "loss": 0.5966,
      "step": 3080
    },
    {
      "epoch": 0.9333937471681015,
      "grad_norm": 0.8554914123213949,
      "learning_rate": 6.8847352024922125e-06,
      "loss": 0.6001,
      "step": 3090
    },
    {
      "epoch": 0.9364144389065095,
      "grad_norm": 1.0189344509024982,
      "learning_rate": 6.5732087227414335e-06,
      "loss": 0.6072,
      "step": 3100
    },
    {
      "epoch": 0.9394351306449177,
      "grad_norm": 0.7551552875967538,
      "learning_rate": 6.2616822429906544e-06,
      "loss": 0.6195,
      "step": 3110
    },
    {
      "epoch": 0.9424558223833258,
      "grad_norm": 0.7876647280151428,
      "learning_rate": 5.9501557632398754e-06,
      "loss": 0.5982,
      "step": 3120
    },
    {
      "epoch": 0.9454765141217338,
      "grad_norm": 1.0198335362611242,
      "learning_rate": 5.638629283489097e-06,
      "loss": 0.6261,
      "step": 3130
    },
    {
      "epoch": 0.948497205860142,
      "grad_norm": 1.0604464411076657,
      "learning_rate": 5.327102803738318e-06,
      "loss": 0.6302,
      "step": 3140
    },
    {
      "epoch": 0.9515178975985501,
      "grad_norm": 0.951744438195523,
      "learning_rate": 5.015576323987539e-06,
      "loss": 0.6093,
      "step": 3150
    },
    {
      "epoch": 0.9545385893369581,
      "grad_norm": 0.8133396952083934,
      "learning_rate": 4.70404984423676e-06,
      "loss": 0.5976,
      "step": 3160
    },
    {
      "epoch": 0.9575592810753663,
      "grad_norm": 0.9048179249325315,
      "learning_rate": 4.392523364485981e-06,
      "loss": 0.5917,
      "step": 3170
    },
    {
      "epoch": 0.9605799728137744,
      "grad_norm": 0.9211845494853066,
      "learning_rate": 4.080996884735202e-06,
      "loss": 0.6598,
      "step": 3180
    },
    {
      "epoch": 0.9636006645521824,
      "grad_norm": 1.1245992131314346,
      "learning_rate": 3.7694704049844237e-06,
      "loss": 0.6366,
      "step": 3190
    },
    {
      "epoch": 0.9666213562905905,
      "grad_norm": 0.8622643092177033,
      "learning_rate": 3.4579439252336447e-06,
      "loss": 0.6032,
      "step": 3200
    },
    {
      "epoch": 0.9696420480289987,
      "grad_norm": 0.9757245422407422,
      "learning_rate": 3.146417445482866e-06,
      "loss": 0.6109,
      "step": 3210
    },
    {
      "epoch": 0.9726627397674067,
      "grad_norm": 0.9177066421934855,
      "learning_rate": 2.8348909657320875e-06,
      "loss": 0.5852,
      "step": 3220
    },
    {
      "epoch": 0.9756834315058148,
      "grad_norm": 1.0589749886784343,
      "learning_rate": 2.5233644859813085e-06,
      "loss": 0.6162,
      "step": 3230
    },
    {
      "epoch": 0.978704123244223,
      "grad_norm": 0.8439326188110577,
      "learning_rate": 2.21183800623053e-06,
      "loss": 0.5679,
      "step": 3240
    },
    {
      "epoch": 0.981724814982631,
      "grad_norm": 0.8843365815693212,
      "learning_rate": 1.900311526479751e-06,
      "loss": 0.5856,
      "step": 3250
    },
    {
      "epoch": 0.9847455067210391,
      "grad_norm": 0.91811081864421,
      "learning_rate": 1.588785046728972e-06,
      "loss": 0.5229,
      "step": 3260
    },
    {
      "epoch": 0.9877661984594472,
      "grad_norm": 0.8575199450142503,
      "learning_rate": 1.2772585669781931e-06,
      "loss": 0.6195,
      "step": 3270
    },
    {
      "epoch": 0.9907868901978553,
      "grad_norm": 0.9360300893347252,
      "learning_rate": 9.657320872274145e-07,
      "loss": 0.6176,
      "step": 3280
    },
    {
      "epoch": 0.9938075819362634,
      "grad_norm": 0.9000581153467687,
      "learning_rate": 6.542056074766356e-07,
      "loss": 0.5858,
      "step": 3290
    },
    {
      "epoch": 0.9968282736746715,
      "grad_norm": 0.9999683040058044,
      "learning_rate": 3.4267912772585674e-07,
      "loss": 0.5731,
      "step": 3300
    },
    {
      "epoch": 0.9998489654130795,
      "grad_norm": 0.9739409873164033,
      "learning_rate": 3.1152647975077883e-08,
      "loss": 0.6497,
      "step": 3310
    }
  ],
  "logging_steps": 10,
  "max_steps": 3310,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}