{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.21279671332927572,
  "eval_steps": 500,
  "global_step": 2797,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007608034083992696,
      "grad_norm": 766.052521836789,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 2.887490081787109,
      "step": 10
    },
    {
      "epoch": 0.0015216068167985392,
      "grad_norm": 219.21834016884335,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.9272476196289063,
      "step": 20
    },
    {
      "epoch": 0.002282410225197809,
      "grad_norm": 697.9334513096925,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.0680903434753417,
      "step": 30
    },
    {
      "epoch": 0.0030432136335970784,
      "grad_norm": 84.53394094934715,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.8754222869873047,
      "step": 40
    },
    {
      "epoch": 0.003804017041996348,
      "grad_norm": 69.64463705089611,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.8158416748046875,
      "step": 50
    },
    {
      "epoch": 0.004564820450395618,
      "grad_norm": 28.95668284276755,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.7302873134613037,
      "step": 60
    },
    {
      "epoch": 0.0053256238587948874,
      "grad_norm": 54.00260757854784,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.7970684051513672,
      "step": 70
    },
    {
      "epoch": 0.006086427267194157,
      "grad_norm": 292.96231799204514,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.7267954826354981,
      "step": 80
    },
    {
      "epoch": 0.006847230675593427,
      "grad_norm": 726.0873753124116,
      "learning_rate": 1.005e-05,
      "loss": 0.7285655975341797,
      "step": 90
    },
    {
      "epoch": 0.007608034083992696,
      "grad_norm": 27.829181455179807,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.8418128013610839,
      "step": 100
    },
    {
      "epoch": 0.008368837492391966,
      "grad_norm": 38.50424143410114,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.7732617378234863,
      "step": 110
    },
    {
      "epoch": 0.009129640900791236,
      "grad_norm": 38.005830230959454,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.7485276222229004,
      "step": 120
    },
    {
      "epoch": 0.009890444309190505,
      "grad_norm": 239.32716231811452,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.7527177333831787,
      "step": 130
    },
    {
      "epoch": 0.010651247717589775,
      "grad_norm": 6.455966289941813,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.805815601348877,
      "step": 140
    },
    {
      "epoch": 0.011412051125989045,
      "grad_norm": 18.693317281807552,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.7609236717224122,
      "step": 150
    },
    {
      "epoch": 0.012172854534388313,
      "grad_norm": 11.963698194011737,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.7418763637542725,
      "step": 160
    },
    {
      "epoch": 0.012933657942787584,
      "grad_norm": 6.958042928688908,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.7683084011077881,
      "step": 170
    },
    {
      "epoch": 0.013694461351186854,
      "grad_norm": 11.571845696402532,
      "learning_rate": 2e-05,
      "loss": 0.7141913890838623,
      "step": 180
    },
    {
      "epoch": 0.014455264759586122,
      "grad_norm": 8.273402278038633,
      "learning_rate": 2e-05,
      "loss": 0.7291435241699219,
      "step": 190
    },
    {
      "epoch": 0.015216068167985392,
      "grad_norm": 1.8068788190209832,
      "learning_rate": 2e-05,
      "loss": 0.7061244010925293,
      "step": 200
    },
    {
      "epoch": 0.015976871576384662,
      "grad_norm": 3.1806280744048605,
      "learning_rate": 2e-05,
      "loss": 0.7077610969543457,
      "step": 210
    },
    {
      "epoch": 0.016737674984783932,
      "grad_norm": 280.189044907266,
      "learning_rate": 2e-05,
      "loss": 0.6421300411224365,
      "step": 220
    },
    {
      "epoch": 0.017498478393183203,
      "grad_norm": 2.55416598370708,
      "learning_rate": 2e-05,
      "loss": 0.6898053169250489,
      "step": 230
    },
    {
      "epoch": 0.018259281801582473,
      "grad_norm": 7.553555757256519,
      "learning_rate": 2e-05,
      "loss": 0.6969669342041016,
      "step": 240
    },
    {
      "epoch": 0.01902008520998174,
      "grad_norm": 477.54050832124796,
      "learning_rate": 2e-05,
      "loss": 0.764107608795166,
      "step": 250
    },
    {
      "epoch": 0.01978088861838101,
      "grad_norm": 6.159675954810047,
      "learning_rate": 2e-05,
      "loss": 0.9087591171264648,
      "step": 260
    },
    {
      "epoch": 0.02054169202678028,
      "grad_norm": 14.248123008291211,
      "learning_rate": 2e-05,
      "loss": 0.7264697551727295,
      "step": 270
    },
    {
      "epoch": 0.02130249543517955,
      "grad_norm": 15.77402781933041,
      "learning_rate": 2e-05,
      "loss": 0.7285306453704834,
      "step": 280
    },
    {
      "epoch": 0.02206329884357882,
      "grad_norm": 24.137047826837957,
      "learning_rate": 2e-05,
      "loss": 0.6912951469421387,
      "step": 290
    },
    {
      "epoch": 0.02282410225197809,
      "grad_norm": 2.3148760857715445,
      "learning_rate": 2e-05,
      "loss": 0.6992810726165771,
      "step": 300
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 126.46425865282905,
      "learning_rate": 2e-05,
      "loss": 0.7007718563079834,
      "step": 310
    },
    {
      "epoch": 0.024345709068776627,
      "grad_norm": 2.6039447021239726,
      "learning_rate": 2e-05,
      "loss": 0.6786778926849365,
      "step": 320
    },
    {
      "epoch": 0.025106512477175897,
      "grad_norm": 2.392207141310159,
      "learning_rate": 2e-05,
      "loss": 0.669903564453125,
      "step": 330
    },
    {
      "epoch": 0.025867315885575167,
      "grad_norm": 9.247678789298908,
      "learning_rate": 2e-05,
      "loss": 0.6987538814544678,
      "step": 340
    },
    {
      "epoch": 0.026628119293974437,
      "grad_norm": 6.464711837428858,
      "learning_rate": 2e-05,
      "loss": 0.6893903255462647,
      "step": 350
    },
    {
      "epoch": 0.027388922702373707,
      "grad_norm": 5.1880207521662705,
      "learning_rate": 2e-05,
      "loss": 0.6841831207275391,
      "step": 360
    },
    {
      "epoch": 0.028149726110772978,
      "grad_norm": 50.12816854251406,
      "learning_rate": 2e-05,
      "loss": 0.7154038906097412,
      "step": 370
    },
    {
      "epoch": 0.028910529519172244,
      "grad_norm": 3.0117477081232598,
      "learning_rate": 2e-05,
      "loss": 0.7240267753601074,
      "step": 380
    },
    {
      "epoch": 0.029671332927571514,
      "grad_norm": 13.836877796796628,
      "learning_rate": 2e-05,
      "loss": 0.7010954856872559,
      "step": 390
    },
    {
      "epoch": 0.030432136335970784,
      "grad_norm": 2.006973973891417,
      "learning_rate": 2e-05,
      "loss": 0.7020359039306641,
      "step": 400
    },
    {
      "epoch": 0.031192939744370055,
      "grad_norm": 27.591581159248626,
      "learning_rate": 2e-05,
      "loss": 0.6682147979736328,
      "step": 410
    },
    {
      "epoch": 0.031953743152769325,
      "grad_norm": 19.791706144588765,
      "learning_rate": 2e-05,
      "loss": 0.6723065853118897,
      "step": 420
    },
    {
      "epoch": 0.032714546561168595,
      "grad_norm": 2.0177081819744553,
      "learning_rate": 2e-05,
      "loss": 0.7360958099365235,
      "step": 430
    },
    {
      "epoch": 0.033475349969567865,
      "grad_norm": 2.188210526838186,
      "learning_rate": 2e-05,
      "loss": 0.6793450832366943,
      "step": 440
    },
    {
      "epoch": 0.034236153377967135,
      "grad_norm": 2.5384186698997313,
      "learning_rate": 2e-05,
      "loss": 0.6992023468017579,
      "step": 450
    },
    {
      "epoch": 0.034996956786366405,
      "grad_norm": 1.5352126431908675,
      "learning_rate": 2e-05,
      "loss": 0.6813825130462646,
      "step": 460
    },
    {
      "epoch": 0.035757760194765675,
      "grad_norm": 5.422494483821066,
      "learning_rate": 2e-05,
      "loss": 0.6625070095062255,
      "step": 470
    },
    {
      "epoch": 0.036518563603164945,
      "grad_norm": 2.164648016182954,
      "learning_rate": 2e-05,
      "loss": 0.6671664237976074,
      "step": 480
    },
    {
      "epoch": 0.03727936701156421,
      "grad_norm": 5.128707982424324,
      "learning_rate": 2e-05,
      "loss": 0.6608222007751465,
      "step": 490
    },
    {
      "epoch": 0.03804017041996348,
      "grad_norm": 2.6188100469545605,
      "learning_rate": 2e-05,
      "loss": 0.6627635002136231,
      "step": 500
    },
    {
      "epoch": 0.03880097382836275,
      "grad_norm": 2.2690007433780646,
      "learning_rate": 2e-05,
      "loss": 0.7100722312927246,
      "step": 510
    },
    {
      "epoch": 0.03956177723676202,
      "grad_norm": 1.6955425896534853,
      "learning_rate": 2e-05,
      "loss": 0.7041080474853516,
      "step": 520
    },
    {
      "epoch": 0.04032258064516129,
      "grad_norm": 1.7481547195068348,
      "learning_rate": 2e-05,
      "loss": 0.6757216453552246,
      "step": 530
    },
    {
      "epoch": 0.04108338405356056,
      "grad_norm": 1.6344561599660914,
      "learning_rate": 2e-05,
      "loss": 0.6713863372802734,
      "step": 540
    },
    {
      "epoch": 0.04184418746195983,
      "grad_norm": 3.2078736781026254,
      "learning_rate": 2e-05,
      "loss": 0.6618888854980469,
      "step": 550
    },
    {
      "epoch": 0.0426049908703591,
      "grad_norm": 1.8379942627724124,
      "learning_rate": 2e-05,
      "loss": 0.7351431846618652,
      "step": 560
    },
    {
      "epoch": 0.04336579427875837,
      "grad_norm": 1.5888838232838909,
      "learning_rate": 2e-05,
      "loss": 0.6826735496520996,
      "step": 570
    },
    {
      "epoch": 0.04412659768715764,
      "grad_norm": 2.0735189856680214,
      "learning_rate": 2e-05,
      "loss": 0.6519746780395508,
      "step": 580
    },
    {
      "epoch": 0.04488740109555691,
      "grad_norm": 1.6672247701461744,
      "learning_rate": 2e-05,
      "loss": 0.6651469707489014,
      "step": 590
    },
    {
      "epoch": 0.04564820450395618,
      "grad_norm": 2.331882937366215,
      "learning_rate": 2e-05,
      "loss": 0.6785001277923584,
      "step": 600
    },
    {
      "epoch": 0.04640900791235545,
      "grad_norm": 1.5170691943378647,
      "learning_rate": 2e-05,
      "loss": 0.6592602729797363,
      "step": 610
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 1.3648878825695145,
      "learning_rate": 2e-05,
      "loss": 0.6565546989440918,
      "step": 620
    },
    {
      "epoch": 0.047930614729153984,
      "grad_norm": 1.6864881230555355,
      "learning_rate": 2e-05,
      "loss": 0.6754720211029053,
      "step": 630
    },
    {
      "epoch": 0.048691418137553254,
      "grad_norm": 1.6332391825289212,
      "learning_rate": 2e-05,
      "loss": 0.6423514366149903,
      "step": 640
    },
    {
      "epoch": 0.049452221545952524,
      "grad_norm": 2.7059106535186572,
      "learning_rate": 2e-05,
      "loss": 0.661496639251709,
      "step": 650
    },
    {
      "epoch": 0.050213024954351794,
      "grad_norm": 2.896510703741966,
      "learning_rate": 2e-05,
      "loss": 0.6534607887268067,
      "step": 660
    },
    {
      "epoch": 0.050973828362751064,
      "grad_norm": 2.2363116679018598,
      "learning_rate": 2e-05,
      "loss": 0.6489164352416992,
      "step": 670
    },
    {
      "epoch": 0.051734631771150334,
      "grad_norm": 4.461890971477879,
      "learning_rate": 2e-05,
      "loss": 0.6854209899902344,
      "step": 680
    },
    {
      "epoch": 0.052495435179549604,
      "grad_norm": 2.916869559890112,
      "learning_rate": 2e-05,
      "loss": 0.6794282913208007,
      "step": 690
    },
    {
      "epoch": 0.053256238587948874,
      "grad_norm": 6.088998851699956,
      "learning_rate": 2e-05,
      "loss": 0.7020540237426758,
      "step": 700
    },
    {
      "epoch": 0.054017041996348145,
      "grad_norm": 2.4051352056906787,
      "learning_rate": 2e-05,
      "loss": 0.7063278198242188,
      "step": 710
    },
    {
      "epoch": 0.054777845404747415,
      "grad_norm": 3.094845534842579,
      "learning_rate": 2e-05,
      "loss": 0.6743041515350342,
      "step": 720
    },
    {
      "epoch": 0.055538648813146685,
      "grad_norm": 1.300597375588879,
      "learning_rate": 2e-05,
      "loss": 0.6632575035095215,
      "step": 730
    },
    {
      "epoch": 0.056299452221545955,
      "grad_norm": 3.279701114868982,
      "learning_rate": 2e-05,
      "loss": 0.6596969127655029,
      "step": 740
    },
    {
      "epoch": 0.057060255629945225,
      "grad_norm": 1.6596836234521832,
      "learning_rate": 2e-05,
      "loss": 0.6648564815521241,
      "step": 750
    },
    {
      "epoch": 0.05782105903834449,
      "grad_norm": 3.3066247075012956,
      "learning_rate": 2e-05,
      "loss": 0.6650107383728028,
      "step": 760
    },
    {
      "epoch": 0.05858186244674376,
      "grad_norm": 3.929474692350062,
      "learning_rate": 2e-05,
      "loss": 0.6553778648376465,
      "step": 770
    },
    {
      "epoch": 0.05934266585514303,
      "grad_norm": 2.518127583230988,
      "learning_rate": 2e-05,
      "loss": 0.690962553024292,
      "step": 780
    },
    {
      "epoch": 0.0601034692635423,
      "grad_norm": 1.2367931748447925,
      "learning_rate": 2e-05,
      "loss": 0.6664523601531982,
      "step": 790
    },
    {
      "epoch": 0.06086427267194157,
      "grad_norm": 4.142969089978007,
      "learning_rate": 2e-05,
      "loss": 0.6557377815246582,
      "step": 800
    },
    {
      "epoch": 0.06162507608034084,
      "grad_norm": 1.4239753399947923,
      "learning_rate": 2e-05,
      "loss": 0.655963659286499,
      "step": 810
    },
    {
      "epoch": 0.06238587948874011,
      "grad_norm": 1.100863203530169,
      "learning_rate": 2e-05,
      "loss": 0.6471473693847656,
      "step": 820
    },
    {
      "epoch": 0.06314668289713937,
      "grad_norm": 1.232804036614322,
      "learning_rate": 2e-05,
      "loss": 0.6807149410247803,
      "step": 830
    },
    {
      "epoch": 0.06390748630553865,
      "grad_norm": 1.2630476712708762,
      "learning_rate": 2e-05,
      "loss": 0.6472922325134277,
      "step": 840
    },
    {
      "epoch": 0.06466828971393791,
      "grad_norm": 3.1776755109142627,
      "learning_rate": 2e-05,
      "loss": 0.6446826934814454,
      "step": 850
    },
    {
      "epoch": 0.06542909312233719,
      "grad_norm": 75.45896872532501,
      "learning_rate": 2e-05,
      "loss": 0.6587884902954102,
      "step": 860
    },
    {
      "epoch": 0.06618989653073645,
      "grad_norm": 1.9742341717753686,
      "learning_rate": 2e-05,
      "loss": 0.6437560081481933,
      "step": 870
    },
    {
      "epoch": 0.06695069993913573,
      "grad_norm": 2.964541109902538,
      "learning_rate": 2e-05,
      "loss": 0.6286462783813477,
      "step": 880
    },
    {
      "epoch": 0.067711503347535,
      "grad_norm": 3.7026446921551486,
      "learning_rate": 2e-05,
      "loss": 0.6568585872650147,
      "step": 890
    },
    {
      "epoch": 0.06847230675593427,
      "grad_norm": 9.317395483241429,
      "learning_rate": 2e-05,
      "loss": 0.6981778144836426,
      "step": 900
    },
    {
      "epoch": 0.06923311016433353,
      "grad_norm": 1.0948327360661168,
      "learning_rate": 2e-05,
      "loss": 0.6652321815490723,
      "step": 910
    },
    {
      "epoch": 0.06999391357273281,
      "grad_norm": 1.0354718206717286,
      "learning_rate": 2e-05,
      "loss": 0.6647524833679199,
      "step": 920
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 1.0304901518571212,
      "learning_rate": 2e-05,
      "loss": 0.6425517559051513,
      "step": 930
    },
    {
      "epoch": 0.07151552038953135,
      "grad_norm": 1.5686213831558329,
      "learning_rate": 2e-05,
      "loss": 0.6413394927978515,
      "step": 940
    },
    {
      "epoch": 0.07227632379793061,
      "grad_norm": 1.7897054437110276,
      "learning_rate": 2e-05,
      "loss": 0.6535541534423828,
      "step": 950
    },
    {
      "epoch": 0.07303712720632989,
      "grad_norm": 0.9359855078368583,
      "learning_rate": 2e-05,
      "loss": 0.639809513092041,
      "step": 960
    },
    {
      "epoch": 0.07379793061472915,
      "grad_norm": 1.512951285597644,
      "learning_rate": 2e-05,
      "loss": 0.6573833465576172,
      "step": 970
    },
    {
      "epoch": 0.07455873402312842,
      "grad_norm": 2.359700069484156,
      "learning_rate": 2e-05,
      "loss": 0.6440255165100097,
      "step": 980
    },
    {
      "epoch": 0.0753195374315277,
      "grad_norm": 1.7130709676637312,
      "learning_rate": 2e-05,
      "loss": 0.6478070259094239,
      "step": 990
    },
    {
      "epoch": 0.07608034083992696,
      "grad_norm": 1.3177759770170199,
      "learning_rate": 2e-05,
      "loss": 0.6684154510498047,
      "step": 1000
    },
    {
      "epoch": 0.07684114424832623,
      "grad_norm": 4.135754705114454,
      "learning_rate": 2e-05,
      "loss": 0.6643803596496582,
      "step": 1010
    },
    {
      "epoch": 0.0776019476567255,
      "grad_norm": 1.1349353267975015,
      "learning_rate": 2e-05,
      "loss": 0.6595885276794433,
      "step": 1020
    },
    {
      "epoch": 0.07836275106512477,
      "grad_norm": 1.0289702425560623,
      "learning_rate": 2e-05,
      "loss": 0.6449689388275146,
      "step": 1030
    },
    {
      "epoch": 0.07912355447352404,
      "grad_norm": 1.1123711190266128,
      "learning_rate": 2e-05,
      "loss": 0.6514264583587647,
      "step": 1040
    },
    {
      "epoch": 0.07988435788192332,
      "grad_norm": 1.1479004223710685,
      "learning_rate": 2e-05,
      "loss": 0.6790945053100585,
      "step": 1050
    },
    {
      "epoch": 0.08064516129032258,
      "grad_norm": 1.3082960682160845,
      "learning_rate": 2e-05,
      "loss": 0.6506200313568116,
      "step": 1060
    },
    {
      "epoch": 0.08140596469872186,
      "grad_norm": 2.106064887485788,
      "learning_rate": 2e-05,
      "loss": 0.6225467681884765,
      "step": 1070
    },
    {
      "epoch": 0.08216676810712112,
      "grad_norm": 3.4175935958057004,
      "learning_rate": 2e-05,
      "loss": 0.6518382549285888,
      "step": 1080
    },
    {
      "epoch": 0.0829275715155204,
      "grad_norm": 1.1113744031391726,
      "learning_rate": 2e-05,
      "loss": 0.6582504272460937,
      "step": 1090
    },
    {
      "epoch": 0.08368837492391966,
      "grad_norm": 0.9903911423489237,
      "learning_rate": 2e-05,
      "loss": 0.6358994960784912,
      "step": 1100
    },
    {
      "epoch": 0.08444917833231892,
      "grad_norm": 0.8851123136905361,
      "learning_rate": 2e-05,
      "loss": 0.6215919494628906,
      "step": 1110
    },
    {
      "epoch": 0.0852099817407182,
      "grad_norm": 0.8853327646542226,
      "learning_rate": 2e-05,
      "loss": 0.6268770217895507,
      "step": 1120
    },
    {
      "epoch": 0.08597078514911746,
      "grad_norm": 0.902800886984458,
      "learning_rate": 2e-05,
      "loss": 0.6309802055358886,
      "step": 1130
    },
    {
      "epoch": 0.08673158855751674,
      "grad_norm": 10.357475980832238,
      "learning_rate": 2e-05,
      "loss": 0.6268141746520997,
      "step": 1140
    },
    {
      "epoch": 0.087492391965916,
      "grad_norm": 4.536094472750066,
      "learning_rate": 2e-05,
      "loss": 0.6583127975463867,
      "step": 1150
    },
    {
      "epoch": 0.08825319537431528,
      "grad_norm": 3.161273337598724,
      "learning_rate": 2e-05,
      "loss": 0.6397143363952636,
      "step": 1160
    },
    {
      "epoch": 0.08901399878271454,
      "grad_norm": 1.9628805920204075,
      "learning_rate": 2e-05,
      "loss": 0.6532254695892334,
      "step": 1170
    },
    {
      "epoch": 0.08977480219111382,
      "grad_norm": 1.3058782788410006,
      "learning_rate": 2e-05,
      "loss": 0.6381869316101074,
      "step": 1180
    },
    {
      "epoch": 0.09053560559951308,
      "grad_norm": 10.319238456330314,
      "learning_rate": 2e-05,
      "loss": 0.6643379211425782,
      "step": 1190
    },
    {
      "epoch": 0.09129640900791236,
      "grad_norm": 8.807866849572596,
      "learning_rate": 2e-05,
      "loss": 0.6664113521575927,
      "step": 1200
    },
    {
      "epoch": 0.09205721241631162,
      "grad_norm": 1.3436238512439613,
      "learning_rate": 2e-05,
      "loss": 0.6490108013153076,
      "step": 1210
    },
    {
      "epoch": 0.0928180158247109,
      "grad_norm": 4.3500091772799,
      "learning_rate": 2e-05,
      "loss": 0.6590385437011719,
      "step": 1220
    },
    {
      "epoch": 0.09357881923311016,
      "grad_norm": 1.0603429372503828,
      "learning_rate": 2e-05,
      "loss": 0.6429415702819824,
      "step": 1230
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.610503564476208,
      "learning_rate": 2e-05,
      "loss": 0.6506116390228271,
      "step": 1240
    },
    {
      "epoch": 0.0951004260499087,
      "grad_norm": 2.125651897332447,
      "learning_rate": 2e-05,
      "loss": 0.6554283142089844,
      "step": 1250
    },
    {
      "epoch": 0.09586122945830797,
      "grad_norm": 1.5217172296899981,
      "learning_rate": 2e-05,
      "loss": 0.6519949913024903,
      "step": 1260
    },
    {
      "epoch": 0.09662203286670724,
      "grad_norm": 1.430228287443949,
      "learning_rate": 2e-05,
      "loss": 0.6444644451141357,
      "step": 1270
    },
    {
      "epoch": 0.09738283627510651,
      "grad_norm": 1.0951036531124672,
      "learning_rate": 2e-05,
      "loss": 0.6609076499938965,
      "step": 1280
    },
    {
      "epoch": 0.09814363968350578,
      "grad_norm": 1.897316036200272,
      "learning_rate": 2e-05,
      "loss": 0.6357399940490722,
      "step": 1290
    },
    {
      "epoch": 0.09890444309190505,
      "grad_norm": 9.680273322175012,
      "learning_rate": 2e-05,
      "loss": 0.6786548614501953,
      "step": 1300
    },
    {
      "epoch": 0.09966524650030432,
      "grad_norm": 3.280416382171216,
      "learning_rate": 2e-05,
      "loss": 0.6191361904144287,
      "step": 1310
    },
    {
      "epoch": 0.10042604990870359,
      "grad_norm": 5.004121996330226,
      "learning_rate": 2e-05,
      "loss": 0.6557157516479493,
      "step": 1320
    },
    {
      "epoch": 0.10118685331710287,
      "grad_norm": 2.4488618272689857,
      "learning_rate": 2e-05,
      "loss": 0.6534897804260253,
      "step": 1330
    },
    {
      "epoch": 0.10194765672550213,
      "grad_norm": 3.949040164888651,
      "learning_rate": 2e-05,
      "loss": 0.6422089576721192,
      "step": 1340
    },
    {
      "epoch": 0.1027084601339014,
      "grad_norm": 1.4977923859383218,
      "learning_rate": 2e-05,
      "loss": 0.6565611839294434,
      "step": 1350
    },
    {
      "epoch": 0.10346926354230067,
      "grad_norm": 3.835777101408433,
      "learning_rate": 2e-05,
      "loss": 0.6588279724121093,
      "step": 1360
    },
    {
      "epoch": 0.10423006695069995,
      "grad_norm": 2.0073417233896174,
      "learning_rate": 2e-05,
      "loss": 0.6190773010253906,
      "step": 1370
    },
    {
      "epoch": 0.10499087035909921,
      "grad_norm": 2.180412133473381,
      "learning_rate": 2e-05,
      "loss": 0.6264517784118653,
      "step": 1380
    },
    {
      "epoch": 0.10575167376749847,
      "grad_norm": 1.0608691574541762,
      "learning_rate": 2e-05,
      "loss": 0.6519461631774902,
      "step": 1390
    },
    {
      "epoch": 0.10651247717589775,
      "grad_norm": 2.8077241127945483,
      "learning_rate": 2e-05,
      "loss": 0.6312610626220703,
      "step": 1400
    },
    {
      "epoch": 0.10727328058429701,
      "grad_norm": 2.4692034985049167,
      "learning_rate": 2e-05,
      "loss": 0.6593670845031738,
      "step": 1410
    },
    {
      "epoch": 0.10803408399269629,
      "grad_norm": 1.1688082965213138,
      "learning_rate": 2e-05,
      "loss": 0.6476199626922607,
      "step": 1420
    },
    {
      "epoch": 0.10879488740109555,
      "grad_norm": 2.1444655878822156,
      "learning_rate": 2e-05,
      "loss": 0.6519205093383789,
      "step": 1430
    },
    {
      "epoch": 0.10955569080949483,
      "grad_norm": 2.330806151670356,
      "learning_rate": 2e-05,
      "loss": 0.673212718963623,
      "step": 1440
    },
    {
      "epoch": 0.11031649421789409,
      "grad_norm": 1.1797329434550838,
      "learning_rate": 2e-05,
      "loss": 0.629892110824585,
      "step": 1450
    },
    {
      "epoch": 0.11107729762629337,
      "grad_norm": 1.370178616571048,
      "learning_rate": 2e-05,
      "loss": 0.6480928421020508,
      "step": 1460
    },
    {
      "epoch": 0.11183810103469263,
      "grad_norm": 1.701804101440024,
      "learning_rate": 2e-05,
      "loss": 0.6533725261688232,
      "step": 1470
    },
    {
      "epoch": 0.11259890444309191,
      "grad_norm": 1.5918247074640068,
      "learning_rate": 2e-05,
      "loss": 0.6213892936706543,
      "step": 1480
    },
    {
      "epoch": 0.11335970785149117,
      "grad_norm": 4.479701036046454,
      "learning_rate": 2e-05,
      "loss": 0.6552947998046875,
      "step": 1490
    },
    {
      "epoch": 0.11412051125989045,
      "grad_norm": 1.1821034084350328,
      "learning_rate": 2e-05,
      "loss": 0.6669442176818847,
      "step": 1500
    },
    {
      "epoch": 0.11488131466828971,
      "grad_norm": 2.56260643167876,
      "learning_rate": 2e-05,
      "loss": 0.6492503643035888,
      "step": 1510
    },
    {
      "epoch": 0.11564211807668898,
      "grad_norm": 1.1572098215301176,
      "learning_rate": 2e-05,
      "loss": 0.6260011196136475,
      "step": 1520
    },
    {
      "epoch": 0.11640292148508825,
      "grad_norm": 1.2381901297643017,
      "learning_rate": 2e-05,
      "loss": 0.6485986232757568,
      "step": 1530
    },
    {
      "epoch": 0.11716372489348752,
      "grad_norm": 0.8407626307915014,
      "learning_rate": 2e-05,
      "loss": 0.6335528373718262,
      "step": 1540
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 1.1522494208742629,
      "learning_rate": 2e-05,
      "loss": 0.6174094676971436,
      "step": 1550
    },
    {
      "epoch": 0.11868533171028606,
      "grad_norm": 1.4358247956839025,
      "learning_rate": 2e-05,
      "loss": 0.6372797012329101,
      "step": 1560
    },
    {
      "epoch": 0.11944613511868533,
      "grad_norm": 1.309941398010788,
      "learning_rate": 2e-05,
      "loss": 0.6266646385192871,
      "step": 1570
    },
    {
      "epoch": 0.1202069385270846,
      "grad_norm": 1.3808985287266267,
      "learning_rate": 2e-05,
      "loss": 0.6343531608581543,
      "step": 1580
    },
    {
      "epoch": 0.12096774193548387,
      "grad_norm": 1.3528927662199868,
      "learning_rate": 2e-05,
      "loss": 0.6284969329833985,
      "step": 1590
    },
    {
      "epoch": 0.12172854534388314,
      "grad_norm": 3.0519694955006105,
      "learning_rate": 2e-05,
      "loss": 0.6424075126647949,
      "step": 1600
    },
    {
      "epoch": 0.12248934875228241,
      "grad_norm": 2.174281639710108,
      "learning_rate": 2e-05,
      "loss": 0.6337644577026367,
      "step": 1610
    },
    {
      "epoch": 0.12325015216068168,
      "grad_norm": 1.561704499585276,
      "learning_rate": 2e-05,
      "loss": 0.6298858165740967,
      "step": 1620
    },
    {
      "epoch": 0.12401095556908096,
      "grad_norm": 1.224072880782582,
      "learning_rate": 2e-05,
      "loss": 0.6224400997161865,
      "step": 1630
    },
    {
      "epoch": 0.12477175897748022,
      "grad_norm": 1.0462139926853118,
      "learning_rate": 2e-05,
      "loss": 0.6482504844665528,
      "step": 1640
    },
    {
      "epoch": 0.12553256238587948,
      "grad_norm": 9.21967648588717,
      "learning_rate": 2e-05,
      "loss": 0.6525485992431641,
      "step": 1650
    },
    {
      "epoch": 0.12629336579427874,
      "grad_norm": 2.007629694651067,
      "learning_rate": 2e-05,
      "loss": 0.6432745933532715,
      "step": 1660
    },
    {
      "epoch": 0.12705416920267804,
      "grad_norm": 0.9720933358173028,
      "learning_rate": 2e-05,
      "loss": 0.6356534004211426,
      "step": 1670
    },
    {
      "epoch": 0.1278149726110773,
      "grad_norm": 1.4468055933940205,
      "learning_rate": 2e-05,
      "loss": 0.6624721050262451,
      "step": 1680
    },
    {
      "epoch": 0.12857577601947656,
      "grad_norm": 1.3207000999132357,
      "learning_rate": 2e-05,
      "loss": 0.6150018692016601,
      "step": 1690
    },
    {
      "epoch": 0.12933657942787583,
      "grad_norm": 1.1806546194990943,
      "learning_rate": 2e-05,
      "loss": 0.6312203407287598,
      "step": 1700
    },
    {
      "epoch": 0.13009738283627512,
      "grad_norm": 1.5786931266899833,
      "learning_rate": 2e-05,
      "loss": 0.6413074493408203,
      "step": 1710
    },
    {
      "epoch": 0.13085818624467438,
      "grad_norm": 1.0702890272477146,
      "learning_rate": 2e-05,
      "loss": 0.642481517791748,
      "step": 1720
    },
    {
      "epoch": 0.13161898965307364,
      "grad_norm": 1.4900158944450277,
      "learning_rate": 2e-05,
      "loss": 0.6344526290893555,
      "step": 1730
    },
    {
      "epoch": 0.1323797930614729,
      "grad_norm": 1.4116190947678835,
      "learning_rate": 2e-05,
      "loss": 0.6302139282226562,
      "step": 1740
    },
    {
      "epoch": 0.1331405964698722,
      "grad_norm": 1.7448570810060797,
      "learning_rate": 2e-05,
      "loss": 0.6293466567993165,
      "step": 1750
    },
    {
      "epoch": 0.13390139987827146,
      "grad_norm": 3.3383049060870595,
      "learning_rate": 2e-05,
      "loss": 0.6328795909881592,
      "step": 1760
    },
    {
      "epoch": 0.13466220328667072,
      "grad_norm": 3.631205476735395,
      "learning_rate": 2e-05,
      "loss": 0.615032958984375,
      "step": 1770
    },
    {
      "epoch": 0.13542300669507,
      "grad_norm": 1.5518542674647309,
      "learning_rate": 2e-05,
      "loss": 0.6316960334777832,
      "step": 1780
    },
    {
      "epoch": 0.13618381010346928,
      "grad_norm": 1.7492350745877743,
      "learning_rate": 2e-05,
      "loss": 0.6442355155944824,
      "step": 1790
    },
    {
      "epoch": 0.13694461351186854,
      "grad_norm": 1.9258665395327041,
      "learning_rate": 2e-05,
      "loss": 0.6106966018676758,
      "step": 1800
    },
    {
      "epoch": 0.1377054169202678,
      "grad_norm": 6.827483693001483,
      "learning_rate": 2e-05,
      "loss": 0.642854642868042,
      "step": 1810
    },
    {
      "epoch": 0.13846622032866707,
      "grad_norm": 1.1009026570397635,
      "learning_rate": 2e-05,
      "loss": 0.6260102748870849,
      "step": 1820
    },
    {
      "epoch": 0.13922702373706633,
      "grad_norm": 1.049291470879374,
      "learning_rate": 2e-05,
      "loss": 0.6057558059692383,
      "step": 1830
    },
    {
      "epoch": 0.13998782714546562,
      "grad_norm": 1.8353318571336608,
      "learning_rate": 2e-05,
      "loss": 0.6460296154022217,
      "step": 1840
    },
    {
      "epoch": 0.14074863055386488,
      "grad_norm": 11.056378441909427,
      "learning_rate": 2e-05,
      "loss": 0.6394613265991211,
      "step": 1850
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 39.436296571275825,
      "learning_rate": 2e-05,
      "loss": 0.6526387691497803,
      "step": 1860
    },
    {
      "epoch": 0.1422702373706634,
      "grad_norm": 1.8312409048902327,
      "learning_rate": 2e-05,
      "loss": 0.6392415523529053,
      "step": 1870
    },
    {
      "epoch": 0.1430310407790627,
      "grad_norm": 1.071587830023617,
      "learning_rate": 2e-05,
      "loss": 0.6196507453918457,
      "step": 1880
    },
    {
      "epoch": 0.14379184418746196,
      "grad_norm": 1.4789771369042652,
      "learning_rate": 2e-05,
      "loss": 0.6441464900970459,
      "step": 1890
    },
    {
      "epoch": 0.14455264759586123,
      "grad_norm": 1.8376505782240828,
      "learning_rate": 2e-05,
      "loss": 0.6638864040374756,
      "step": 1900
    },
    {
      "epoch": 0.1453134510042605,
      "grad_norm": 1.3745479307166024,
      "learning_rate": 2e-05,
      "loss": 0.6252675533294678,
      "step": 1910
    },
    {
      "epoch": 0.14607425441265978,
      "grad_norm": 1.1668435873091556,
      "learning_rate": 2e-05,
      "loss": 0.6090609550476074,
      "step": 1920
    },
    {
      "epoch": 0.14683505782105905,
      "grad_norm": 1.23594711006607,
      "learning_rate": 2e-05,
      "loss": 0.6123690605163574,
      "step": 1930
    },
    {
      "epoch": 0.1475958612294583,
      "grad_norm": 1.2418726773487727,
      "learning_rate": 2e-05,
      "loss": 0.6570170879364013,
      "step": 1940
    },
    {
      "epoch": 0.14835666463785757,
      "grad_norm": 1.0504866082911768,
      "learning_rate": 2e-05,
      "loss": 0.6538400173187255,
      "step": 1950
    },
    {
      "epoch": 0.14911746804625683,
      "grad_norm": 2.014671731533198,
      "learning_rate": 2e-05,
      "loss": 0.6170327186584472,
      "step": 1960
    },
    {
      "epoch": 0.14987827145465613,
      "grad_norm": 2.1058379351030108,
      "learning_rate": 2e-05,
      "loss": 0.6381420135498047,
      "step": 1970
    },
    {
      "epoch": 0.1506390748630554,
      "grad_norm": 1.3150633416083708,
      "learning_rate": 2e-05,
      "loss": 0.6363181114196778,
      "step": 1980
    },
    {
      "epoch": 0.15139987827145465,
      "grad_norm": 0.8932530375434491,
      "learning_rate": 2e-05,
      "loss": 0.6291966915130616,
      "step": 1990
    },
    {
      "epoch": 0.15216068167985392,
      "grad_norm": 2.957084421448231,
      "learning_rate": 2e-05,
      "loss": 0.6210861206054688,
      "step": 2000
    },
    {
      "epoch": 0.1529214850882532,
      "grad_norm": 0.8890058850680974,
      "learning_rate": 2e-05,
      "loss": 0.6021919250488281,
      "step": 2010
    },
    {
      "epoch": 0.15368228849665247,
      "grad_norm": 0.9932774221445535,
      "learning_rate": 2e-05,
      "loss": 0.6515641689300538,
      "step": 2020
    },
    {
      "epoch": 0.15444309190505173,
      "grad_norm": 1.0023975736435211,
      "learning_rate": 2e-05,
      "loss": 0.6417403221130371,
      "step": 2030
    },
    {
      "epoch": 0.155203895313451,
      "grad_norm": 0.8554859168766407,
      "learning_rate": 2e-05,
      "loss": 0.6303350448608398,
      "step": 2040
    },
    {
      "epoch": 0.1559646987218503,
      "grad_norm": 1.5251333977211128,
      "learning_rate": 2e-05,
      "loss": 0.6416450500488281,
      "step": 2050
    },
    {
      "epoch": 0.15672550213024955,
      "grad_norm": 1.0986596009489755,
      "learning_rate": 2e-05,
      "loss": 0.6401923179626465,
      "step": 2060
    },
    {
      "epoch": 0.1574863055386488,
      "grad_norm": 1.1539013637626214,
      "learning_rate": 2e-05,
      "loss": 0.6259047508239746,
      "step": 2070
    },
    {
      "epoch": 0.15824710894704808,
      "grad_norm": 1.4348698553263404,
      "learning_rate": 2e-05,
      "loss": 0.6216594696044921,
      "step": 2080
    },
    {
      "epoch": 0.15900791235544734,
      "grad_norm": 0.9526520497956816,
      "learning_rate": 2e-05,
      "loss": 0.6363525390625,
      "step": 2090
    },
    {
      "epoch": 0.15976871576384663,
      "grad_norm": 3.45947923470152,
      "learning_rate": 2e-05,
      "loss": 0.6570816040039062,
      "step": 2100
    },
    {
      "epoch": 0.1605295191722459,
      "grad_norm": 2.0280020107564853,
      "learning_rate": 2e-05,
      "loss": 0.6443814754486084,
      "step": 2110
    },
    {
      "epoch": 0.16129032258064516,
      "grad_norm": 1.2746739038674486,
      "learning_rate": 2e-05,
      "loss": 0.6273581981658936,
      "step": 2120
    },
    {
      "epoch": 0.16205112598904442,
      "grad_norm": 1.1054316756226754,
      "learning_rate": 2e-05,
      "loss": 0.6392014980316162,
      "step": 2130
    },
    {
      "epoch": 0.1628119293974437,
      "grad_norm": 0.7393965027847225,
      "learning_rate": 2e-05,
      "loss": 0.6142825603485107,
      "step": 2140
    },
    {
      "epoch": 0.16357273280584297,
      "grad_norm": 0.8045230869463124,
      "learning_rate": 2e-05,
      "loss": 0.6058244228363037,
      "step": 2150
    },
    {
      "epoch": 0.16433353621424224,
      "grad_norm": 0.8285093876406089,
      "learning_rate": 2e-05,
      "loss": 0.6443311214447022,
      "step": 2160
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 0.8596670201373968,
      "learning_rate": 2e-05,
      "loss": 0.6266818046569824,
      "step": 2170
    },
    {
      "epoch": 0.1658551430310408,
      "grad_norm": 0.9253480501904824,
      "learning_rate": 2e-05,
      "loss": 0.6671144008636475,
      "step": 2180
    },
    {
      "epoch": 0.16661594643944005,
      "grad_norm": 0.8320885887267414,
      "learning_rate": 2e-05,
      "loss": 0.62307767868042,
      "step": 2190
    },
    {
      "epoch": 0.16737674984783932,
      "grad_norm": 2.6787131895939384,
      "learning_rate": 2e-05,
      "loss": 0.6703015804290772,
      "step": 2200
    },
    {
      "epoch": 0.16813755325623858,
      "grad_norm": 0.8902868417037517,
      "learning_rate": 2e-05,
      "loss": 0.6234861373901367,
      "step": 2210
    },
    {
      "epoch": 0.16889835666463784,
      "grad_norm": 0.8847567666362576,
      "learning_rate": 2e-05,
      "loss": 0.6103476524353028,
      "step": 2220
    },
    {
      "epoch": 0.16965916007303714,
      "grad_norm": 0.8408811071144497,
      "learning_rate": 2e-05,
      "loss": 0.6184403419494628,
      "step": 2230
    },
    {
      "epoch": 0.1704199634814364,
      "grad_norm": 1.6892917345046299,
      "learning_rate": 2e-05,
      "loss": 0.6205358505249023,
      "step": 2240
    },
    {
      "epoch": 0.17118076688983566,
      "grad_norm": 0.7405946865412475,
      "learning_rate": 2e-05,
      "loss": 0.61587553024292,
      "step": 2250
    },
    {
      "epoch": 0.17194157029823492,
      "grad_norm": 0.7506110237045676,
      "learning_rate": 2e-05,
      "loss": 0.6244790554046631,
      "step": 2260
    },
    {
      "epoch": 0.17270237370663422,
      "grad_norm": 2.153478514759513,
      "learning_rate": 2e-05,
      "loss": 0.6198970794677734,
      "step": 2270
    },
    {
      "epoch": 0.17346317711503348,
      "grad_norm": 10.500995119599999,
      "learning_rate": 2e-05,
      "loss": 0.6225154876708985,
      "step": 2280
    },
    {
      "epoch": 0.17422398052343274,
      "grad_norm": 1.213251555227088,
      "learning_rate": 2e-05,
      "loss": 0.6163989067077636,
      "step": 2290
    },
    {
      "epoch": 0.174984783931832,
      "grad_norm": 0.9729698503142967,
      "learning_rate": 2e-05,
      "loss": 0.610928726196289,
      "step": 2300
    },
    {
      "epoch": 0.1757455873402313,
      "grad_norm": 1.0702561000731816,
      "learning_rate": 2e-05,
      "loss": 0.6329403877258301,
      "step": 2310
    },
    {
      "epoch": 0.17650639074863056,
      "grad_norm": 0.9932357555439623,
      "learning_rate": 2e-05,
      "loss": 0.6477541923522949,
      "step": 2320
    },
    {
      "epoch": 0.17726719415702982,
      "grad_norm": 0.8974295172792705,
      "learning_rate": 2e-05,
      "loss": 0.6245576858520507,
      "step": 2330
    },
    {
      "epoch": 0.17802799756542909,
      "grad_norm": 0.8948226638723901,
      "learning_rate": 2e-05,
      "loss": 0.6292768478393554,
      "step": 2340
    },
    {
      "epoch": 0.17878880097382835,
      "grad_norm": 1.0441356789609721,
      "learning_rate": 2e-05,
      "loss": 0.6134624481201172,
      "step": 2350
    },
    {
      "epoch": 0.17954960438222764,
      "grad_norm": 1.0579592934994417,
      "learning_rate": 2e-05,
      "loss": 0.6160748958587646,
      "step": 2360
    },
    {
      "epoch": 0.1803104077906269,
      "grad_norm": 1.0688656877042852,
      "learning_rate": 2e-05,
      "loss": 0.6353797912597656,
      "step": 2370
    },
    {
      "epoch": 0.18107121119902617,
      "grad_norm": 0.9630873249684109,
      "learning_rate": 2e-05,
      "loss": 0.6212389945983887,
      "step": 2380
    },
    {
      "epoch": 0.18183201460742543,
      "grad_norm": 1.299642947772091,
      "learning_rate": 2e-05,
      "loss": 0.6184343338012696,
      "step": 2390
    },
    {
      "epoch": 0.18259281801582472,
      "grad_norm": 1.9371331971789227,
      "learning_rate": 2e-05,
      "loss": 0.6408385276794434,
      "step": 2400
    },
    {
      "epoch": 0.18335362142422398,
      "grad_norm": 2.3484601024255074,
      "learning_rate": 2e-05,
      "loss": 0.6273296356201172,
      "step": 2410
    },
    {
      "epoch": 0.18411442483262325,
      "grad_norm": 2.421637945440785,
      "learning_rate": 2e-05,
      "loss": 0.6258901119232178,
      "step": 2420
    },
    {
      "epoch": 0.1848752282410225,
      "grad_norm": 7.087299118182203,
      "learning_rate": 2e-05,
      "loss": 0.6362738609313965,
      "step": 2430
    },
    {
      "epoch": 0.1856360316494218,
      "grad_norm": 1.60152608667479,
      "learning_rate": 2e-05,
      "loss": 0.6333796977996826,
      "step": 2440
    },
    {
      "epoch": 0.18639683505782106,
      "grad_norm": 1.5388500323071719,
      "learning_rate": 2e-05,
      "loss": 0.6134243011474609,
      "step": 2450
    },
    {
      "epoch": 0.18715763846622033,
      "grad_norm": 1.6406255566044918,
      "learning_rate": 2e-05,
      "loss": 0.6321418285369873,
      "step": 2460
    },
    {
      "epoch": 0.1879184418746196,
      "grad_norm": 1.751258694616863,
      "learning_rate": 2e-05,
      "loss": 0.6141924858093262,
      "step": 2470
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 2.2372643207413128,
      "learning_rate": 2e-05,
      "loss": 0.6251022338867187,
      "step": 2480
    },
    {
      "epoch": 0.18944004869141814,
      "grad_norm": 3.178260160359772,
      "learning_rate": 2e-05,
      "loss": 0.6332900047302246,
      "step": 2490
    },
    {
      "epoch": 0.1902008520998174,
      "grad_norm": 2.5115652770876618,
      "learning_rate": 2e-05,
      "loss": 0.6328558921813965,
      "step": 2500
    },
    {
      "epoch": 0.19096165550821667,
      "grad_norm": 3.551106098546657,
      "learning_rate": 2e-05,
      "loss": 0.6492326736450196,
      "step": 2510
    },
    {
      "epoch": 0.19172245891661593,
      "grad_norm": 1.1091484806788126,
      "learning_rate": 2e-05,
      "loss": 0.6217524528503418,
      "step": 2520
    },
    {
      "epoch": 0.19248326232501523,
      "grad_norm": 0.7761699186852694,
      "learning_rate": 2e-05,
      "loss": 0.630795669555664,
      "step": 2530
    },
    {
      "epoch": 0.1932440657334145,
      "grad_norm": 2.8016530566761406,
      "learning_rate": 2e-05,
      "loss": 0.6455306529998779,
      "step": 2540
    },
    {
      "epoch": 0.19400486914181375,
      "grad_norm": 0.9122456764667124,
      "learning_rate": 1.997261400182901e-05,
      "loss": 0.5890992709568569,
      "step": 2550
    },
    {
      "epoch": 0.19476567255021301,
      "grad_norm": 1.4671668897736663,
      "learning_rate": 1.9805800746810423e-05,
      "loss": 0.6393177032470703,
      "step": 2560
    },
    {
      "epoch": 0.1955264759586123,
      "grad_norm": 1.4520027598679992,
      "learning_rate": 1.9489934220110274e-05,
      "loss": 0.6147416591644287,
      "step": 2570
    },
    {
      "epoch": 0.19628727936701157,
      "grad_norm": 1.6176004030706503,
      "learning_rate": 1.9029840364820392e-05,
      "loss": 0.6248357772827149,
      "step": 2580
    },
    {
      "epoch": 0.19704808277541083,
      "grad_norm": 2.313171666822779,
      "learning_rate": 1.8432548690479908e-05,
      "loss": 0.6295966148376465,
      "step": 2590
    },
    {
      "epoch": 0.1978088861838101,
      "grad_norm": 1.328636140347456,
      "learning_rate": 1.7707184873239772e-05,
      "loss": 0.6219944953918457,
      "step": 2600
    },
    {
      "epoch": 0.1985696895922094,
      "grad_norm": 0.9388828161454343,
      "learning_rate": 1.6864831329902565e-05,
      "loss": 0.6297315120697021,
      "step": 2610
    },
    {
      "epoch": 0.19933049300060865,
      "grad_norm": 0.8093123205260678,
      "learning_rate": 1.591835789604667e-05,
      "loss": 0.616849708557129,
      "step": 2620
    },
    {
      "epoch": 0.2000912964090079,
      "grad_norm": 9.388227295921142,
      "learning_rate": 1.4882225195206459e-05,
      "loss": 0.6245549201965332,
      "step": 2630
    },
    {
      "epoch": 0.20085209981740718,
      "grad_norm": 2.702077235649169,
      "learning_rate": 1.3772263703317695e-05,
      "loss": 0.5944196224212647,
      "step": 2640
    },
    {
      "epoch": 0.20161290322580644,
      "grad_norm": 1.302064895055697,
      "learning_rate": 1.2605431883975547e-05,
      "loss": 0.6241308212280273,
      "step": 2650
    },
    {
      "epoch": 0.20237370663420573,
      "grad_norm": 1.6738062163144525,
      "learning_rate": 1.1399557089817574e-05,
      "loss": 0.6194950103759765,
      "step": 2660
    },
    {
      "epoch": 0.203134510042605,
      "grad_norm": 1.3712463140149325,
      "learning_rate": 1.0173063188650463e-05,
      "loss": 0.6252943992614746,
      "step": 2670
    },
    {
      "epoch": 0.20389531345100426,
      "grad_norm": 1.7099387917926658,
      "learning_rate": 8.944689075764396e-06,
      "loss": 0.6121802806854248,
      "step": 2680
    },
    {
      "epoch": 0.20465611685940352,
      "grad_norm": 0.9605147921202484,
      "learning_rate": 7.73320237312362e-06,
      "loss": 0.6088183403015137,
      "step": 2690
    },
    {
      "epoch": 0.2054169202678028,
      "grad_norm": 1.7875449645111638,
      "learning_rate": 6.557112689658629e-06,
      "loss": 0.6046363830566406,
      "step": 2700
    },
    {
      "epoch": 0.20617772367620207,
      "grad_norm": 1.0533119000224143,
      "learning_rate": 5.434388823591213e-06,
      "loss": 0.605999755859375,
      "step": 2710
    },
    {
      "epoch": 0.20693852708460134,
      "grad_norm": 0.8273808331393275,
      "learning_rate": 4.382184227495424e-06,
      "loss": 0.616620397567749,
      "step": 2720
    },
    {
      "epoch": 0.2076993304930006,
      "grad_norm": 0.7909298332499916,
      "learning_rate": 3.4165749305560203e-06,
      "loss": 0.5991597652435303,
      "step": 2730
    },
    {
      "epoch": 0.2084601339013999,
      "grad_norm": 2.6532718694190747,
      "learning_rate": 2.5523139221595556e-06,
      "loss": 0.598800802230835,
      "step": 2740
    },
    {
      "epoch": 0.20922093730979915,
      "grad_norm": 0.778594658362869,
      "learning_rate": 1.8026057494499821e-06,
      "loss": 0.5857003688812256,
      "step": 2750
    },
    {
      "epoch": 0.20998174071819842,
      "grad_norm": 0.8347927148811665,
      "learning_rate": 1.178904772643223e-06,
      "loss": 0.5856118679046631,
      "step": 2760
    },
    {
      "epoch": 0.21074254412659768,
      "grad_norm": 1.259833364466379,
      "learning_rate": 6.90740160441754e-07,
      "loss": 0.6020267486572266,
      "step": 2770
    },
    {
      "epoch": 0.21150334753499694,
      "grad_norm": 3.164007567728581,
      "learning_rate": 3.45570299343632e-07,
      "loss": 0.6088290214538574,
      "step": 2780
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 2.1073570615099513,
      "learning_rate": 1.486688412425152e-07,
      "loss": 0.585694408416748,
      "step": 2790
    },
    {
      "epoch": 0.21279671332927572,
      "step": 2797,
      "total_flos": 1324435843842048.0,
      "train_loss": 0.055548980059604965,
      "train_runtime": 3706.492,
      "train_samples_per_second": 9.055,
      "train_steps_per_second": 0.755
    }
  ],
  "logging_steps": 10,
  "max_steps": 2797,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2543,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1324435843842048.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}