{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4258216676810712,
  "eval_steps": 500,
  "global_step": 5597,
  "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": 0.9817069331556703,
      "learning_rate": 2e-05,
      "loss": 0.5892092841012138,
      "step": 2550
    },
    {
      "epoch": 0.19476567255021301,
      "grad_norm": 1.1506642297473906,
      "learning_rate": 2e-05,
      "loss": 0.6403939247131347,
      "step": 2560
    },
    {
      "epoch": 0.1955264759586123,
      "grad_norm": 1.023859977623109,
      "learning_rate": 2e-05,
      "loss": 0.6122564315795899,
      "step": 2570
    },
    {
      "epoch": 0.19628727936701157,
      "grad_norm": 1.5469625525063975,
      "learning_rate": 2e-05,
      "loss": 0.6223512172698975,
      "step": 2580
    },
    {
      "epoch": 0.19704808277541083,
      "grad_norm": 1.0680850837639781,
      "learning_rate": 2e-05,
      "loss": 0.6277861595153809,
      "step": 2590
    },
    {
      "epoch": 0.1978088861838101,
      "grad_norm": 0.9216390380997925,
      "learning_rate": 2e-05,
      "loss": 0.6205940246582031,
      "step": 2600
    },
    {
      "epoch": 0.1985696895922094,
      "grad_norm": 0.9069803099730847,
      "learning_rate": 2e-05,
      "loss": 0.629159688949585,
      "step": 2610
    },
    {
      "epoch": 0.19933049300060865,
      "grad_norm": 0.9065994563342014,
      "learning_rate": 2e-05,
      "loss": 0.618154239654541,
      "step": 2620
    },
    {
      "epoch": 0.2000912964090079,
      "grad_norm": 2.1363162353810567,
      "learning_rate": 2e-05,
      "loss": 0.6302952766418457,
      "step": 2630
    },
    {
      "epoch": 0.20085209981740718,
      "grad_norm": 1.3495231395011489,
      "learning_rate": 2e-05,
      "loss": 0.602275037765503,
      "step": 2640
    },
    {
      "epoch": 0.20161290322580644,
      "grad_norm": 1.268006641731718,
      "learning_rate": 2e-05,
      "loss": 0.633419418334961,
      "step": 2650
    },
    {
      "epoch": 0.20237370663420573,
      "grad_norm": 1.1198642302804405,
      "learning_rate": 2e-05,
      "loss": 0.628083324432373,
      "step": 2660
    },
    {
      "epoch": 0.203134510042605,
      "grad_norm": 0.921063027206477,
      "learning_rate": 2e-05,
      "loss": 0.6306491374969483,
      "step": 2670
    },
    {
      "epoch": 0.20389531345100426,
      "grad_norm": 0.9131822079624049,
      "learning_rate": 2e-05,
      "loss": 0.6213677883148193,
      "step": 2680
    },
    {
      "epoch": 0.20465611685940352,
      "grad_norm": 0.9626227664170652,
      "learning_rate": 2e-05,
      "loss": 0.6190157890319824,
      "step": 2690
    },
    {
      "epoch": 0.2054169202678028,
      "grad_norm": 1.5952273952274294,
      "learning_rate": 2e-05,
      "loss": 0.6162997245788574,
      "step": 2700
    },
    {
      "epoch": 0.20617772367620207,
      "grad_norm": 1.6032390147900086,
      "learning_rate": 2e-05,
      "loss": 0.6197282791137695,
      "step": 2710
    },
    {
      "epoch": 0.20693852708460134,
      "grad_norm": 1.5025225130869249,
      "learning_rate": 2e-05,
      "loss": 0.6324640274047851,
      "step": 2720
    },
    {
      "epoch": 0.2076993304930006,
      "grad_norm": 1.15091373470265,
      "learning_rate": 2e-05,
      "loss": 0.6154145240783692,
      "step": 2730
    },
    {
      "epoch": 0.2084601339013999,
      "grad_norm": 2.503142503929874,
      "learning_rate": 2e-05,
      "loss": 0.6177960395812988,
      "step": 2740
    },
    {
      "epoch": 0.20922093730979915,
      "grad_norm": 3.225448811792671,
      "learning_rate": 2e-05,
      "loss": 0.6053067207336426,
      "step": 2750
    },
    {
      "epoch": 0.20998174071819842,
      "grad_norm": 0.9340037286792306,
      "learning_rate": 2e-05,
      "loss": 0.6067095756530761,
      "step": 2760
    },
    {
      "epoch": 0.21074254412659768,
      "grad_norm": 0.7839092401023494,
      "learning_rate": 2e-05,
      "loss": 0.6238336086273193,
      "step": 2770
    },
    {
      "epoch": 0.21150334753499694,
      "grad_norm": 0.9703683681711317,
      "learning_rate": 2e-05,
      "loss": 0.630821943283081,
      "step": 2780
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 1.1956854322866988,
      "learning_rate": 2e-05,
      "loss": 0.6082069396972656,
      "step": 2790
    },
    {
      "epoch": 0.2130249543517955,
      "grad_norm": 2.5970281442565692,
      "learning_rate": 2e-05,
      "loss": 0.6003427505493164,
      "step": 2800
    },
    {
      "epoch": 0.21378575776019476,
      "grad_norm": 1.181050183007216,
      "learning_rate": 2e-05,
      "loss": 0.6277856349945068,
      "step": 2810
    },
    {
      "epoch": 0.21454656116859402,
      "grad_norm": 1.0961834404857531,
      "learning_rate": 2e-05,
      "loss": 0.6375556468963623,
      "step": 2820
    },
    {
      "epoch": 0.21530736457699332,
      "grad_norm": 1.107277032190709,
      "learning_rate": 2e-05,
      "loss": 0.6170720100402832,
      "step": 2830
    },
    {
      "epoch": 0.21606816798539258,
      "grad_norm": 0.9876856488355245,
      "learning_rate": 2e-05,
      "loss": 0.6377505302429199,
      "step": 2840
    },
    {
      "epoch": 0.21682897139379184,
      "grad_norm": 2.8138017098561487,
      "learning_rate": 2e-05,
      "loss": 0.6230205535888672,
      "step": 2850
    },
    {
      "epoch": 0.2175897748021911,
      "grad_norm": 0.9569008919457225,
      "learning_rate": 2e-05,
      "loss": 0.6139896392822266,
      "step": 2860
    },
    {
      "epoch": 0.2183505782105904,
      "grad_norm": 1.2166946479977514,
      "learning_rate": 2e-05,
      "loss": 0.6102657318115234,
      "step": 2870
    },
    {
      "epoch": 0.21911138161898966,
      "grad_norm": 0.9439320682429234,
      "learning_rate": 2e-05,
      "loss": 0.6210070610046386,
      "step": 2880
    },
    {
      "epoch": 0.21987218502738892,
      "grad_norm": 1.9444963486987974,
      "learning_rate": 2e-05,
      "loss": 0.6143630504608154,
      "step": 2890
    },
    {
      "epoch": 0.22063298843578819,
      "grad_norm": 1.0731632672697757,
      "learning_rate": 2e-05,
      "loss": 0.6242578029632568,
      "step": 2900
    },
    {
      "epoch": 0.22139379184418745,
      "grad_norm": 1.4416846334633917,
      "learning_rate": 2e-05,
      "loss": 0.6301780700683594,
      "step": 2910
    },
    {
      "epoch": 0.22215459525258674,
      "grad_norm": 1.0398958563814207,
      "learning_rate": 2e-05,
      "loss": 0.6249309539794922,
      "step": 2920
    },
    {
      "epoch": 0.222915398660986,
      "grad_norm": 1.5915685165282702,
      "learning_rate": 2e-05,
      "loss": 0.6256701946258545,
      "step": 2930
    },
    {
      "epoch": 0.22367620206938527,
      "grad_norm": 1.97421945040368,
      "learning_rate": 2e-05,
      "loss": 0.6315007686614991,
      "step": 2940
    },
    {
      "epoch": 0.22443700547778453,
      "grad_norm": 0.9439951164923264,
      "learning_rate": 2e-05,
      "loss": 0.6190637111663818,
      "step": 2950
    },
    {
      "epoch": 0.22519780888618382,
      "grad_norm": 1.9216827306482598,
      "learning_rate": 2e-05,
      "loss": 0.6048786163330078,
      "step": 2960
    },
    {
      "epoch": 0.22595861229458308,
      "grad_norm": 1.3758385768580352,
      "learning_rate": 2e-05,
      "loss": 0.6186373710632325,
      "step": 2970
    },
    {
      "epoch": 0.22671941570298235,
      "grad_norm": 3.0227729807206907,
      "learning_rate": 2e-05,
      "loss": 0.6130160808563232,
      "step": 2980
    },
    {
      "epoch": 0.2274802191113816,
      "grad_norm": 1.0965618740158993,
      "learning_rate": 2e-05,
      "loss": 0.6068836212158203,
      "step": 2990
    },
    {
      "epoch": 0.2282410225197809,
      "grad_norm": 1.3276497304928805,
      "learning_rate": 2e-05,
      "loss": 0.6168122291564941,
      "step": 3000
    },
    {
      "epoch": 0.22900182592818016,
      "grad_norm": 2.3525701651750768,
      "learning_rate": 2e-05,
      "loss": 0.6179019451141358,
      "step": 3010
    },
    {
      "epoch": 0.22976262933657943,
      "grad_norm": 0.7100746650791652,
      "learning_rate": 2e-05,
      "loss": 0.5935450553894043,
      "step": 3020
    },
    {
      "epoch": 0.2305234327449787,
      "grad_norm": 2.967110575052808,
      "learning_rate": 2e-05,
      "loss": 0.6108531475067138,
      "step": 3030
    },
    {
      "epoch": 0.23128423615337795,
      "grad_norm": 1.0742537412854092,
      "learning_rate": 2e-05,
      "loss": 0.5901834011077881,
      "step": 3040
    },
    {
      "epoch": 0.23204503956177724,
      "grad_norm": 1.0149513822294771,
      "learning_rate": 2e-05,
      "loss": 0.6123186111450195,
      "step": 3050
    },
    {
      "epoch": 0.2328058429701765,
      "grad_norm": 1.129503907318443,
      "learning_rate": 2e-05,
      "loss": 0.6262495994567872,
      "step": 3060
    },
    {
      "epoch": 0.23356664637857577,
      "grad_norm": 1.5795916128532546,
      "learning_rate": 2e-05,
      "loss": 0.6286283016204834,
      "step": 3070
    },
    {
      "epoch": 0.23432744978697503,
      "grad_norm": 1.5737449340106107,
      "learning_rate": 2e-05,
      "loss": 0.6367819786071778,
      "step": 3080
    },
    {
      "epoch": 0.23508825319537432,
      "grad_norm": 0.837301889940201,
      "learning_rate": 2e-05,
      "loss": 0.5960247039794921,
      "step": 3090
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 1.7129589642475265,
      "learning_rate": 2e-05,
      "loss": 0.6188911437988281,
      "step": 3100
    },
    {
      "epoch": 0.23660986001217285,
      "grad_norm": 1.0978797863870255,
      "learning_rate": 2e-05,
      "loss": 0.6011454582214355,
      "step": 3110
    },
    {
      "epoch": 0.23737066342057211,
      "grad_norm": 1.3392547341281262,
      "learning_rate": 2e-05,
      "loss": 0.6084640502929688,
      "step": 3120
    },
    {
      "epoch": 0.2381314668289714,
      "grad_norm": 1.1896487886307119,
      "learning_rate": 2e-05,
      "loss": 0.6010078430175781,
      "step": 3130
    },
    {
      "epoch": 0.23889227023737067,
      "grad_norm": 0.9724148059510948,
      "learning_rate": 2e-05,
      "loss": 0.625861120223999,
      "step": 3140
    },
    {
      "epoch": 0.23965307364576993,
      "grad_norm": 0.8546534339092822,
      "learning_rate": 2e-05,
      "loss": 0.5815396785736084,
      "step": 3150
    },
    {
      "epoch": 0.2404138770541692,
      "grad_norm": 1.0358608032712895,
      "learning_rate": 2e-05,
      "loss": 0.5957612037658692,
      "step": 3160
    },
    {
      "epoch": 0.24117468046256849,
      "grad_norm": 0.9532257041483436,
      "learning_rate": 2e-05,
      "loss": 0.6230683326721191,
      "step": 3170
    },
    {
      "epoch": 0.24193548387096775,
      "grad_norm": 0.856584963672178,
      "learning_rate": 2e-05,
      "loss": 0.6225377559661865,
      "step": 3180
    },
    {
      "epoch": 0.242696287279367,
      "grad_norm": 0.9701495366122038,
      "learning_rate": 2e-05,
      "loss": 0.5935500144958497,
      "step": 3190
    },
    {
      "epoch": 0.24345709068776628,
      "grad_norm": 0.8719862670241907,
      "learning_rate": 2e-05,
      "loss": 0.6125042915344239,
      "step": 3200
    },
    {
      "epoch": 0.24421789409616554,
      "grad_norm": 0.8876554659779671,
      "learning_rate": 2e-05,
      "loss": 0.6382464408874512,
      "step": 3210
    },
    {
      "epoch": 0.24497869750456483,
      "grad_norm": 0.663293931503932,
      "learning_rate": 2e-05,
      "loss": 0.613660192489624,
      "step": 3220
    },
    {
      "epoch": 0.2457395009129641,
      "grad_norm": 0.9157732445075698,
      "learning_rate": 2e-05,
      "loss": 0.6227169036865234,
      "step": 3230
    },
    {
      "epoch": 0.24650030432136336,
      "grad_norm": 0.9169230676438861,
      "learning_rate": 2e-05,
      "loss": 0.6264325618743897,
      "step": 3240
    },
    {
      "epoch": 0.24726110772976262,
      "grad_norm": 0.9267464760873556,
      "learning_rate": 2e-05,
      "loss": 0.6031881332397461,
      "step": 3250
    },
    {
      "epoch": 0.2480219111381619,
      "grad_norm": 1.6276096678949572,
      "learning_rate": 2e-05,
      "loss": 0.6197366714477539,
      "step": 3260
    },
    {
      "epoch": 0.24878271454656117,
      "grad_norm": 0.8661493898082891,
      "learning_rate": 2e-05,
      "loss": 0.6157424449920654,
      "step": 3270
    },
    {
      "epoch": 0.24954351795496044,
      "grad_norm": 0.8642949072778816,
      "learning_rate": 2e-05,
      "loss": 0.6269526481628418,
      "step": 3280
    },
    {
      "epoch": 0.2503043213633597,
      "grad_norm": 1.1643303657998856,
      "learning_rate": 2e-05,
      "loss": 0.6230345726013183,
      "step": 3290
    },
    {
      "epoch": 0.25106512477175896,
      "grad_norm": 1.1115822075885138,
      "learning_rate": 2e-05,
      "loss": 0.6161396026611328,
      "step": 3300
    },
    {
      "epoch": 0.25182592818015825,
      "grad_norm": 1.3716226645242158,
      "learning_rate": 2e-05,
      "loss": 0.6055039882659912,
      "step": 3310
    },
    {
      "epoch": 0.2525867315885575,
      "grad_norm": 1.4005747784917617,
      "learning_rate": 2e-05,
      "loss": 0.6360630035400391,
      "step": 3320
    },
    {
      "epoch": 0.2533475349969568,
      "grad_norm": 1.0243200200378946,
      "learning_rate": 2e-05,
      "loss": 0.6756389617919922,
      "step": 3330
    },
    {
      "epoch": 0.25410833840535607,
      "grad_norm": 1.4886693526090016,
      "learning_rate": 2e-05,
      "loss": 0.6030360698699951,
      "step": 3340
    },
    {
      "epoch": 0.2548691418137553,
      "grad_norm": 1.294277490662169,
      "learning_rate": 2e-05,
      "loss": 0.6214393615722656,
      "step": 3350
    },
    {
      "epoch": 0.2556299452221546,
      "grad_norm": 0.8814030358404723,
      "learning_rate": 2e-05,
      "loss": 0.6296616554260254,
      "step": 3360
    },
    {
      "epoch": 0.2563907486305539,
      "grad_norm": 1.797589540426131,
      "learning_rate": 2e-05,
      "loss": 0.6054345607757569,
      "step": 3370
    },
    {
      "epoch": 0.2571515520389531,
      "grad_norm": 1.1534914208205673,
      "learning_rate": 2e-05,
      "loss": 0.5965017318725586,
      "step": 3380
    },
    {
      "epoch": 0.2579123554473524,
      "grad_norm": 1.094261609741512,
      "learning_rate": 2e-05,
      "loss": 0.6489345073699951,
      "step": 3390
    },
    {
      "epoch": 0.25867315885575165,
      "grad_norm": 1.085371831991245,
      "learning_rate": 2e-05,
      "loss": 0.615788745880127,
      "step": 3400
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 2.914563733064073,
      "learning_rate": 2e-05,
      "loss": 0.611840009689331,
      "step": 3410
    },
    {
      "epoch": 0.26019476567255023,
      "grad_norm": 0.745256249202797,
      "learning_rate": 2e-05,
      "loss": 0.5985809326171875,
      "step": 3420
    },
    {
      "epoch": 0.26095556908094947,
      "grad_norm": 1.0176582915284635,
      "learning_rate": 2e-05,
      "loss": 0.6228419303894043,
      "step": 3430
    },
    {
      "epoch": 0.26171637248934876,
      "grad_norm": 1.2161548866693805,
      "learning_rate": 2e-05,
      "loss": 0.627016830444336,
      "step": 3440
    },
    {
      "epoch": 0.26247717589774805,
      "grad_norm": 1.0364686040834363,
      "learning_rate": 2e-05,
      "loss": 0.6175382137298584,
      "step": 3450
    },
    {
      "epoch": 0.2632379793061473,
      "grad_norm": 1.1636472172805075,
      "learning_rate": 2e-05,
      "loss": 0.6075883388519288,
      "step": 3460
    },
    {
      "epoch": 0.2639987827145466,
      "grad_norm": 1.7614077254816805,
      "learning_rate": 2e-05,
      "loss": 0.6019673824310303,
      "step": 3470
    },
    {
      "epoch": 0.2647595861229458,
      "grad_norm": 0.863434633984219,
      "learning_rate": 2e-05,
      "loss": 0.601746654510498,
      "step": 3480
    },
    {
      "epoch": 0.2655203895313451,
      "grad_norm": 1.304041982690625,
      "learning_rate": 2e-05,
      "loss": 0.6177839279174805,
      "step": 3490
    },
    {
      "epoch": 0.2662811929397444,
      "grad_norm": 1.883012196044417,
      "learning_rate": 2e-05,
      "loss": 0.6234899044036866,
      "step": 3500
    },
    {
      "epoch": 0.26704199634814363,
      "grad_norm": 0.8964938484441732,
      "learning_rate": 2e-05,
      "loss": 0.6390236854553223,
      "step": 3510
    },
    {
      "epoch": 0.2678027997565429,
      "grad_norm": 1.4284801033824128,
      "learning_rate": 2e-05,
      "loss": 0.6097182273864746,
      "step": 3520
    },
    {
      "epoch": 0.26856360316494216,
      "grad_norm": 0.8210642254657347,
      "learning_rate": 2e-05,
      "loss": 0.6138957500457763,
      "step": 3530
    },
    {
      "epoch": 0.26932440657334145,
      "grad_norm": 1.0521207645749941,
      "learning_rate": 2e-05,
      "loss": 0.6080638885498046,
      "step": 3540
    },
    {
      "epoch": 0.27008520998174074,
      "grad_norm": 6.735721511812288,
      "learning_rate": 2e-05,
      "loss": 0.6256732940673828,
      "step": 3550
    },
    {
      "epoch": 0.27084601339014,
      "grad_norm": 2.799485569675146,
      "learning_rate": 2e-05,
      "loss": 0.5936986923217773,
      "step": 3560
    },
    {
      "epoch": 0.27160681679853926,
      "grad_norm": 0.9111946254259832,
      "learning_rate": 2e-05,
      "loss": 0.6204559326171875,
      "step": 3570
    },
    {
      "epoch": 0.27236762020693855,
      "grad_norm": 1.8582172885915664,
      "learning_rate": 2e-05,
      "loss": 0.615900707244873,
      "step": 3580
    },
    {
      "epoch": 0.2731284236153378,
      "grad_norm": 0.7739816201039059,
      "learning_rate": 2e-05,
      "loss": 0.5894556999206543,
      "step": 3590
    },
    {
      "epoch": 0.2738892270237371,
      "grad_norm": 1.441208937322419,
      "learning_rate": 2e-05,
      "loss": 0.6097617149353027,
      "step": 3600
    },
    {
      "epoch": 0.2746500304321363,
      "grad_norm": 0.9465929067889907,
      "learning_rate": 2e-05,
      "loss": 0.6167520046234131,
      "step": 3610
    },
    {
      "epoch": 0.2754108338405356,
      "grad_norm": 0.9840465275588058,
      "learning_rate": 2e-05,
      "loss": 0.6279431819915772,
      "step": 3620
    },
    {
      "epoch": 0.2761716372489349,
      "grad_norm": 1.6678784102631918,
      "learning_rate": 2e-05,
      "loss": 0.6081015586853027,
      "step": 3630
    },
    {
      "epoch": 0.27693244065733413,
      "grad_norm": 0.8555695729438972,
      "learning_rate": 2e-05,
      "loss": 0.6070639610290527,
      "step": 3640
    },
    {
      "epoch": 0.2776932440657334,
      "grad_norm": 1.756659504956261,
      "learning_rate": 2e-05,
      "loss": 0.6191421508789062,
      "step": 3650
    },
    {
      "epoch": 0.27845404747413266,
      "grad_norm": 1.2982748593593505,
      "learning_rate": 2e-05,
      "loss": 0.6194043636322022,
      "step": 3660
    },
    {
      "epoch": 0.27921485088253195,
      "grad_norm": 0.7922690486284613,
      "learning_rate": 2e-05,
      "loss": 0.5996121883392334,
      "step": 3670
    },
    {
      "epoch": 0.27997565429093124,
      "grad_norm": 0.8341378633449698,
      "learning_rate": 2e-05,
      "loss": 0.6137380123138427,
      "step": 3680
    },
    {
      "epoch": 0.2807364576993305,
      "grad_norm": 1.2791128722007485,
      "learning_rate": 2e-05,
      "loss": 0.6054240703582764,
      "step": 3690
    },
    {
      "epoch": 0.28149726110772977,
      "grad_norm": 0.8939937791977226,
      "learning_rate": 2e-05,
      "loss": 0.6308682918548584,
      "step": 3700
    },
    {
      "epoch": 0.28225806451612906,
      "grad_norm": 3.628759918703947,
      "learning_rate": 2e-05,
      "loss": 0.6237603664398194,
      "step": 3710
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 4.577204658525299,
      "learning_rate": 2e-05,
      "loss": 0.6311593532562256,
      "step": 3720
    },
    {
      "epoch": 0.2837796713329276,
      "grad_norm": 0.8948798691151667,
      "learning_rate": 2e-05,
      "loss": 0.608569860458374,
      "step": 3730
    },
    {
      "epoch": 0.2845404747413268,
      "grad_norm": 0.9194449656139297,
      "learning_rate": 2e-05,
      "loss": 0.6357901573181153,
      "step": 3740
    },
    {
      "epoch": 0.2853012781497261,
      "grad_norm": 0.8171348243537644,
      "learning_rate": 2e-05,
      "loss": 0.613353681564331,
      "step": 3750
    },
    {
      "epoch": 0.2860620815581254,
      "grad_norm": 1.013927897669978,
      "learning_rate": 2e-05,
      "loss": 0.6140854835510254,
      "step": 3760
    },
    {
      "epoch": 0.28682288496652464,
      "grad_norm": 0.7210071956307897,
      "learning_rate": 2e-05,
      "loss": 0.6217946529388427,
      "step": 3770
    },
    {
      "epoch": 0.28758368837492393,
      "grad_norm": 0.8011534818888779,
      "learning_rate": 2e-05,
      "loss": 0.5974589347839355,
      "step": 3780
    },
    {
      "epoch": 0.28834449178332316,
      "grad_norm": 0.9011845655705722,
      "learning_rate": 2e-05,
      "loss": 0.6087424278259277,
      "step": 3790
    },
    {
      "epoch": 0.28910529519172246,
      "grad_norm": 0.8814420360391584,
      "learning_rate": 2e-05,
      "loss": 0.6011880874633789,
      "step": 3800
    },
    {
      "epoch": 0.28986609860012175,
      "grad_norm": 0.8315296955618288,
      "learning_rate": 2e-05,
      "loss": 0.5989766120910645,
      "step": 3810
    },
    {
      "epoch": 0.290626902008521,
      "grad_norm": 0.8120554977139738,
      "learning_rate": 2e-05,
      "loss": 0.6353487968444824,
      "step": 3820
    },
    {
      "epoch": 0.2913877054169203,
      "grad_norm": 2.682256590029768,
      "learning_rate": 2e-05,
      "loss": 0.5951233863830566,
      "step": 3830
    },
    {
      "epoch": 0.29214850882531956,
      "grad_norm": 1.5678136476459417,
      "learning_rate": 2e-05,
      "loss": 0.6109321594238282,
      "step": 3840
    },
    {
      "epoch": 0.2929093122337188,
      "grad_norm": 1.6518908694551735,
      "learning_rate": 2e-05,
      "loss": 0.6007917404174805,
      "step": 3850
    },
    {
      "epoch": 0.2936701156421181,
      "grad_norm": 0.8390332769455142,
      "learning_rate": 2e-05,
      "loss": 0.5966996192932129,
      "step": 3860
    },
    {
      "epoch": 0.2944309190505173,
      "grad_norm": 0.9939817097512031,
      "learning_rate": 2e-05,
      "loss": 0.6230123519897461,
      "step": 3870
    },
    {
      "epoch": 0.2951917224589166,
      "grad_norm": 0.8449526081587111,
      "learning_rate": 2e-05,
      "loss": 0.6368628978729248,
      "step": 3880
    },
    {
      "epoch": 0.2959525258673159,
      "grad_norm": 1.3118101691079025,
      "learning_rate": 2e-05,
      "loss": 0.6220353603363037,
      "step": 3890
    },
    {
      "epoch": 0.29671332927571514,
      "grad_norm": 0.978940167656836,
      "learning_rate": 2e-05,
      "loss": 0.6155681610107422,
      "step": 3900
    },
    {
      "epoch": 0.29747413268411443,
      "grad_norm": 1.7975965748511273,
      "learning_rate": 2e-05,
      "loss": 0.5918474197387695,
      "step": 3910
    },
    {
      "epoch": 0.29823493609251367,
      "grad_norm": 1.6706236390444285,
      "learning_rate": 2e-05,
      "loss": 0.6126898765563965,
      "step": 3920
    },
    {
      "epoch": 0.29899573950091296,
      "grad_norm": 0.8955060435507843,
      "learning_rate": 2e-05,
      "loss": 0.5828713417053223,
      "step": 3930
    },
    {
      "epoch": 0.29975654290931225,
      "grad_norm": 0.9057540496762859,
      "learning_rate": 2e-05,
      "loss": 0.6285027980804443,
      "step": 3940
    },
    {
      "epoch": 0.3005173463177115,
      "grad_norm": 0.8513770306376237,
      "learning_rate": 2e-05,
      "loss": 0.5995490074157714,
      "step": 3950
    },
    {
      "epoch": 0.3012781497261108,
      "grad_norm": 0.9018811929887697,
      "learning_rate": 2e-05,
      "loss": 0.6095849514007569,
      "step": 3960
    },
    {
      "epoch": 0.30203895313451007,
      "grad_norm": 0.7118554306232184,
      "learning_rate": 2e-05,
      "loss": 0.6040401458740234,
      "step": 3970
    },
    {
      "epoch": 0.3027997565429093,
      "grad_norm": 0.9071491300673238,
      "learning_rate": 2e-05,
      "loss": 0.6258529186248779,
      "step": 3980
    },
    {
      "epoch": 0.3035605599513086,
      "grad_norm": 0.9038005131444103,
      "learning_rate": 2e-05,
      "loss": 0.6107657432556153,
      "step": 3990
    },
    {
      "epoch": 0.30432136335970783,
      "grad_norm": 0.9133874285340832,
      "learning_rate": 2e-05,
      "loss": 0.6253015518188476,
      "step": 4000
    },
    {
      "epoch": 0.3050821667681071,
      "grad_norm": 1.1022244895474538,
      "learning_rate": 2e-05,
      "loss": 0.5970585823059082,
      "step": 4010
    },
    {
      "epoch": 0.3058429701765064,
      "grad_norm": 0.982933929694812,
      "learning_rate": 2e-05,
      "loss": 0.610454511642456,
      "step": 4020
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 6.471445748327421,
      "learning_rate": 2e-05,
      "loss": 0.5650215148925781,
      "step": 4030
    },
    {
      "epoch": 0.30736457699330494,
      "grad_norm": 1.1515240977943098,
      "learning_rate": 2e-05,
      "loss": 0.6278285026550293,
      "step": 4040
    },
    {
      "epoch": 0.3081253804017042,
      "grad_norm": 2.457199029245135,
      "learning_rate": 2e-05,
      "loss": 0.6022915840148926,
      "step": 4050
    },
    {
      "epoch": 0.30888618381010347,
      "grad_norm": 0.7407605190468444,
      "learning_rate": 2e-05,
      "loss": 0.6401365280151368,
      "step": 4060
    },
    {
      "epoch": 0.30964698721850276,
      "grad_norm": 0.8695411167792898,
      "learning_rate": 2e-05,
      "loss": 0.5914569854736328,
      "step": 4070
    },
    {
      "epoch": 0.310407790626902,
      "grad_norm": 1.1123525656325708,
      "learning_rate": 2e-05,
      "loss": 0.6278794288635254,
      "step": 4080
    },
    {
      "epoch": 0.3111685940353013,
      "grad_norm": 0.8089054567791197,
      "learning_rate": 2e-05,
      "loss": 0.5776612281799316,
      "step": 4090
    },
    {
      "epoch": 0.3119293974437006,
      "grad_norm": 1.44682150408943,
      "learning_rate": 2e-05,
      "loss": 0.6120601177215577,
      "step": 4100
    },
    {
      "epoch": 0.3126902008520998,
      "grad_norm": 1.2477506247557861,
      "learning_rate": 2e-05,
      "loss": 0.6243569374084472,
      "step": 4110
    },
    {
      "epoch": 0.3134510042604991,
      "grad_norm": 1.452737770482452,
      "learning_rate": 2e-05,
      "loss": 0.6241071224212646,
      "step": 4120
    },
    {
      "epoch": 0.31421180766889834,
      "grad_norm": 0.7363949927071634,
      "learning_rate": 2e-05,
      "loss": 0.615998649597168,
      "step": 4130
    },
    {
      "epoch": 0.3149726110772976,
      "grad_norm": 0.9904461219747637,
      "learning_rate": 2e-05,
      "loss": 0.5982001304626465,
      "step": 4140
    },
    {
      "epoch": 0.3157334144856969,
      "grad_norm": 2.6307432080351836,
      "learning_rate": 2e-05,
      "loss": 0.6170999526977539,
      "step": 4150
    },
    {
      "epoch": 0.31649421789409615,
      "grad_norm": 1.3103833934588613,
      "learning_rate": 2e-05,
      "loss": 0.616608715057373,
      "step": 4160
    },
    {
      "epoch": 0.31725502130249544,
      "grad_norm": 1.642585084301309,
      "learning_rate": 2e-05,
      "loss": 0.6181431770324707,
      "step": 4170
    },
    {
      "epoch": 0.3180158247108947,
      "grad_norm": 1.3358344049655098,
      "learning_rate": 2e-05,
      "loss": 0.5898998260498047,
      "step": 4180
    },
    {
      "epoch": 0.31877662811929397,
      "grad_norm": 2.033036923101361,
      "learning_rate": 2e-05,
      "loss": 0.6010503768920898,
      "step": 4190
    },
    {
      "epoch": 0.31953743152769326,
      "grad_norm": 1.0048016779941806,
      "learning_rate": 2e-05,
      "loss": 0.6192368984222412,
      "step": 4200
    },
    {
      "epoch": 0.3202982349360925,
      "grad_norm": 1.5318421777607685,
      "learning_rate": 2e-05,
      "loss": 0.6008861541748047,
      "step": 4210
    },
    {
      "epoch": 0.3210590383444918,
      "grad_norm": 1.0073799174677105,
      "learning_rate": 2e-05,
      "loss": 0.6188548088073731,
      "step": 4220
    },
    {
      "epoch": 0.3218198417528911,
      "grad_norm": 0.8283343412424335,
      "learning_rate": 2e-05,
      "loss": 0.6003226280212403,
      "step": 4230
    },
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 0.7571380810077775,
      "learning_rate": 2e-05,
      "loss": 0.6002717971801758,
      "step": 4240
    },
    {
      "epoch": 0.3233414485696896,
      "grad_norm": 0.8338609352847893,
      "learning_rate": 2e-05,
      "loss": 0.6231700897216796,
      "step": 4250
    },
    {
      "epoch": 0.32410225197808884,
      "grad_norm": 0.740619566155922,
      "learning_rate": 2e-05,
      "loss": 0.6033841133117676,
      "step": 4260
    },
    {
      "epoch": 0.32486305538648813,
      "grad_norm": 0.8956374906539315,
      "learning_rate": 2e-05,
      "loss": 0.6189849853515625,
      "step": 4270
    },
    {
      "epoch": 0.3256238587948874,
      "grad_norm": 1.409026605977312,
      "learning_rate": 2e-05,
      "loss": 0.6028616905212403,
      "step": 4280
    },
    {
      "epoch": 0.32638466220328666,
      "grad_norm": 2.453626744827294,
      "learning_rate": 2e-05,
      "loss": 0.6057555675506592,
      "step": 4290
    },
    {
      "epoch": 0.32714546561168595,
      "grad_norm": 1.3959026927627025,
      "learning_rate": 2e-05,
      "loss": 0.6028185844421386,
      "step": 4300
    },
    {
      "epoch": 0.3279062690200852,
      "grad_norm": 0.7577927271333064,
      "learning_rate": 2e-05,
      "loss": 0.6127864360809326,
      "step": 4310
    },
    {
      "epoch": 0.3286670724284845,
      "grad_norm": 0.9447217279629392,
      "learning_rate": 2e-05,
      "loss": 0.6024003982543945,
      "step": 4320
    },
    {
      "epoch": 0.32942787583688377,
      "grad_norm": 0.9272576340694912,
      "learning_rate": 2e-05,
      "loss": 0.6185897827148438,
      "step": 4330
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.7119916555178322,
      "learning_rate": 2e-05,
      "loss": 0.6182665824890137,
      "step": 4340
    },
    {
      "epoch": 0.3309494826536823,
      "grad_norm": 5.95345406970356,
      "learning_rate": 2e-05,
      "loss": 0.6071166038513184,
      "step": 4350
    },
    {
      "epoch": 0.3317102860620816,
      "grad_norm": 1.4215648692146343,
      "learning_rate": 2e-05,
      "loss": 0.6039448738098144,
      "step": 4360
    },
    {
      "epoch": 0.3324710894704808,
      "grad_norm": 0.7570623558291363,
      "learning_rate": 2e-05,
      "loss": 0.5986960887908935,
      "step": 4370
    },
    {
      "epoch": 0.3332318928788801,
      "grad_norm": 1.0228573337145783,
      "learning_rate": 2e-05,
      "loss": 0.6171799659729004,
      "step": 4380
    },
    {
      "epoch": 0.33399269628727934,
      "grad_norm": 0.9028857839163633,
      "learning_rate": 2e-05,
      "loss": 0.6072957515716553,
      "step": 4390
    },
    {
      "epoch": 0.33475349969567864,
      "grad_norm": 0.7415941430991778,
      "learning_rate": 2e-05,
      "loss": 0.6106678962707519,
      "step": 4400
    },
    {
      "epoch": 0.3355143031040779,
      "grad_norm": 0.7411236910221118,
      "learning_rate": 2e-05,
      "loss": 0.5931798934936523,
      "step": 4410
    },
    {
      "epoch": 0.33627510651247716,
      "grad_norm": 1.1774708021652436,
      "learning_rate": 2e-05,
      "loss": 0.6152263164520264,
      "step": 4420
    },
    {
      "epoch": 0.33703590992087645,
      "grad_norm": 0.8505680675812167,
      "learning_rate": 2e-05,
      "loss": 0.6033729553222656,
      "step": 4430
    },
    {
      "epoch": 0.3377967133292757,
      "grad_norm": 0.7891820178944161,
      "learning_rate": 2e-05,
      "loss": 0.6123302459716797,
      "step": 4440
    },
    {
      "epoch": 0.338557516737675,
      "grad_norm": 0.866079289805027,
      "learning_rate": 2e-05,
      "loss": 0.6025050163269043,
      "step": 4450
    },
    {
      "epoch": 0.33931832014607427,
      "grad_norm": 0.7241976779424538,
      "learning_rate": 2e-05,
      "loss": 0.612579345703125,
      "step": 4460
    },
    {
      "epoch": 0.3400791235544735,
      "grad_norm": 0.7966629670827999,
      "learning_rate": 2e-05,
      "loss": 0.5974034786224365,
      "step": 4470
    },
    {
      "epoch": 0.3408399269628728,
      "grad_norm": 0.7767550667528098,
      "learning_rate": 2e-05,
      "loss": 0.6011282920837402,
      "step": 4480
    },
    {
      "epoch": 0.3416007303712721,
      "grad_norm": 1.8420521999393737,
      "learning_rate": 2e-05,
      "loss": 0.6026616573333741,
      "step": 4490
    },
    {
      "epoch": 0.3423615337796713,
      "grad_norm": 1.3139173936334982,
      "learning_rate": 2e-05,
      "loss": 0.6053057670593261,
      "step": 4500
    },
    {
      "epoch": 0.3431223371880706,
      "grad_norm": 1.022314820431628,
      "learning_rate": 2e-05,
      "loss": 0.583978271484375,
      "step": 4510
    },
    {
      "epoch": 0.34388314059646985,
      "grad_norm": 0.7165802493976553,
      "learning_rate": 2e-05,
      "loss": 0.6061666488647461,
      "step": 4520
    },
    {
      "epoch": 0.34464394400486914,
      "grad_norm": 0.747468299503708,
      "learning_rate": 2e-05,
      "loss": 0.5799343109130859,
      "step": 4530
    },
    {
      "epoch": 0.34540474741326843,
      "grad_norm": 0.8132856915849325,
      "learning_rate": 2e-05,
      "loss": 0.6095840454101562,
      "step": 4540
    },
    {
      "epoch": 0.34616555082166767,
      "grad_norm": 0.7027943277436396,
      "learning_rate": 2e-05,
      "loss": 0.6092688560485839,
      "step": 4550
    },
    {
      "epoch": 0.34692635423006696,
      "grad_norm": 1.1086210965087457,
      "learning_rate": 2e-05,
      "loss": 0.5932141304016113,
      "step": 4560
    },
    {
      "epoch": 0.3476871576384662,
      "grad_norm": 0.8852791797840133,
      "learning_rate": 2e-05,
      "loss": 0.5923095703125,
      "step": 4570
    },
    {
      "epoch": 0.3484479610468655,
      "grad_norm": 1.6353079953312049,
      "learning_rate": 2e-05,
      "loss": 0.6154132843017578,
      "step": 4580
    },
    {
      "epoch": 0.3492087644552648,
      "grad_norm": 1.1789394247020066,
      "learning_rate": 2e-05,
      "loss": 0.6273578643798828,
      "step": 4590
    },
    {
      "epoch": 0.349969567863664,
      "grad_norm": 1.4102805493635073,
      "learning_rate": 2e-05,
      "loss": 0.6174351692199707,
      "step": 4600
    },
    {
      "epoch": 0.3507303712720633,
      "grad_norm": 3.4182331641961428,
      "learning_rate": 2e-05,
      "loss": 0.5923943996429444,
      "step": 4610
    },
    {
      "epoch": 0.3514911746804626,
      "grad_norm": 0.9392991206242777,
      "learning_rate": 2e-05,
      "loss": 0.5950442790985108,
      "step": 4620
    },
    {
      "epoch": 0.35225197808886183,
      "grad_norm": 0.84844491388597,
      "learning_rate": 2e-05,
      "loss": 0.5990563869476319,
      "step": 4630
    },
    {
      "epoch": 0.3530127814972611,
      "grad_norm": 0.7558327897640099,
      "learning_rate": 2e-05,
      "loss": 0.5872565746307373,
      "step": 4640
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 0.786357619659807,
      "learning_rate": 2e-05,
      "loss": 0.5881314277648926,
      "step": 4650
    },
    {
      "epoch": 0.35453438831405965,
      "grad_norm": 0.7946945273066692,
      "learning_rate": 2e-05,
      "loss": 0.5885752677917481,
      "step": 4660
    },
    {
      "epoch": 0.35529519172245894,
      "grad_norm": 0.8509812786655356,
      "learning_rate": 2e-05,
      "loss": 0.5973938941955567,
      "step": 4670
    },
    {
      "epoch": 0.35605599513085817,
      "grad_norm": 1.192996958876806,
      "learning_rate": 2e-05,
      "loss": 0.5827082633972168,
      "step": 4680
    },
    {
      "epoch": 0.35681679853925746,
      "grad_norm": 1.0699080043046383,
      "learning_rate": 2e-05,
      "loss": 0.6129659652709961,
      "step": 4690
    },
    {
      "epoch": 0.3575776019476567,
      "grad_norm": 0.8727409547698386,
      "learning_rate": 2e-05,
      "loss": 0.58746337890625,
      "step": 4700
    },
    {
      "epoch": 0.358338405356056,
      "grad_norm": 5.43225571997352,
      "learning_rate": 2e-05,
      "loss": 0.6099301338195801,
      "step": 4710
    },
    {
      "epoch": 0.3590992087644553,
      "grad_norm": 0.9887457028131443,
      "learning_rate": 2e-05,
      "loss": 0.607042646408081,
      "step": 4720
    },
    {
      "epoch": 0.3598600121728545,
      "grad_norm": 1.2046804745263027,
      "learning_rate": 2e-05,
      "loss": 0.5955885887145996,
      "step": 4730
    },
    {
      "epoch": 0.3606208155812538,
      "grad_norm": 0.7639829739280138,
      "learning_rate": 2e-05,
      "loss": 0.6083533763885498,
      "step": 4740
    },
    {
      "epoch": 0.3613816189896531,
      "grad_norm": 0.7267500781357502,
      "learning_rate": 2e-05,
      "loss": 0.60111403465271,
      "step": 4750
    },
    {
      "epoch": 0.36214242239805233,
      "grad_norm": 0.8710289774828294,
      "learning_rate": 2e-05,
      "loss": 0.6345517635345459,
      "step": 4760
    },
    {
      "epoch": 0.3629032258064516,
      "grad_norm": 1.0637000129245222,
      "learning_rate": 2e-05,
      "loss": 0.6069129943847656,
      "step": 4770
    },
    {
      "epoch": 0.36366402921485086,
      "grad_norm": 1.8024452267660673,
      "learning_rate": 2e-05,
      "loss": 0.6435677528381347,
      "step": 4780
    },
    {
      "epoch": 0.36442483262325015,
      "grad_norm": 0.8346426924240165,
      "learning_rate": 2e-05,
      "loss": 0.5989354133605957,
      "step": 4790
    },
    {
      "epoch": 0.36518563603164944,
      "grad_norm": 0.6792142781631723,
      "learning_rate": 2e-05,
      "loss": 0.6012043952941895,
      "step": 4800
    },
    {
      "epoch": 0.3659464394400487,
      "grad_norm": 0.8409837452079166,
      "learning_rate": 2e-05,
      "loss": 0.6044680595397949,
      "step": 4810
    },
    {
      "epoch": 0.36670724284844797,
      "grad_norm": 0.8644602675103753,
      "learning_rate": 2e-05,
      "loss": 0.6201369762420654,
      "step": 4820
    },
    {
      "epoch": 0.36746804625684726,
      "grad_norm": 1.580222949710994,
      "learning_rate": 2e-05,
      "loss": 0.5904644012451172,
      "step": 4830
    },
    {
      "epoch": 0.3682288496652465,
      "grad_norm": 1.1928080164601287,
      "learning_rate": 2e-05,
      "loss": 0.6086296081542969,
      "step": 4840
    },
    {
      "epoch": 0.3689896530736458,
      "grad_norm": 0.689264603926416,
      "learning_rate": 2e-05,
      "loss": 0.5967980861663819,
      "step": 4850
    },
    {
      "epoch": 0.369750456482045,
      "grad_norm": 0.8126776376611288,
      "learning_rate": 2e-05,
      "loss": 0.5855744361877442,
      "step": 4860
    },
    {
      "epoch": 0.3705112598904443,
      "grad_norm": 1.092356779204631,
      "learning_rate": 2e-05,
      "loss": 0.6178097724914551,
      "step": 4870
    },
    {
      "epoch": 0.3712720632988436,
      "grad_norm": 0.9622832608857461,
      "learning_rate": 2e-05,
      "loss": 0.617055082321167,
      "step": 4880
    },
    {
      "epoch": 0.37203286670724284,
      "grad_norm": 0.7618052699113795,
      "learning_rate": 2e-05,
      "loss": 0.602293586730957,
      "step": 4890
    },
    {
      "epoch": 0.37279367011564213,
      "grad_norm": 1.3489861276357102,
      "learning_rate": 2e-05,
      "loss": 0.5918971538543701,
      "step": 4900
    },
    {
      "epoch": 0.37355447352404136,
      "grad_norm": 1.7707644440212729,
      "learning_rate": 2e-05,
      "loss": 0.625973892211914,
      "step": 4910
    },
    {
      "epoch": 0.37431527693244065,
      "grad_norm": 15.014705035598531,
      "learning_rate": 2e-05,
      "loss": 0.6238719463348389,
      "step": 4920
    },
    {
      "epoch": 0.37507608034083995,
      "grad_norm": 1.1670299633789776,
      "learning_rate": 2e-05,
      "loss": 0.6075002193450928,
      "step": 4930
    },
    {
      "epoch": 0.3758368837492392,
      "grad_norm": 0.8549851050612812,
      "learning_rate": 2e-05,
      "loss": 0.6109541893005371,
      "step": 4940
    },
    {
      "epoch": 0.37659768715763847,
      "grad_norm": 0.911106365991643,
      "learning_rate": 2e-05,
      "loss": 0.6021104335784913,
      "step": 4950
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.8650943358256169,
      "learning_rate": 2e-05,
      "loss": 0.5897356033325195,
      "step": 4960
    },
    {
      "epoch": 0.378119293974437,
      "grad_norm": 1.2102016223127976,
      "learning_rate": 2e-05,
      "loss": 0.6001923561096192,
      "step": 4970
    },
    {
      "epoch": 0.3788800973828363,
      "grad_norm": 1.159865181760775,
      "learning_rate": 2e-05,
      "loss": 0.6132318496704101,
      "step": 4980
    },
    {
      "epoch": 0.3796409007912355,
      "grad_norm": 1.9290598997736847,
      "learning_rate": 2e-05,
      "loss": 0.6039675712585449,
      "step": 4990
    },
    {
      "epoch": 0.3804017041996348,
      "grad_norm": 0.8225860734261352,
      "learning_rate": 2e-05,
      "loss": 0.5970118522644043,
      "step": 5000
    },
    {
      "epoch": 0.3811625076080341,
      "grad_norm": 0.8026146605102198,
      "learning_rate": 2e-05,
      "loss": 0.6008700847625732,
      "step": 5010
    },
    {
      "epoch": 0.38192331101643334,
      "grad_norm": 0.9141026787845802,
      "learning_rate": 2e-05,
      "loss": 0.6239216804504395,
      "step": 5020
    },
    {
      "epoch": 0.38268411442483263,
      "grad_norm": 3.558820995930603,
      "learning_rate": 2e-05,
      "loss": 0.6232359886169434,
      "step": 5030
    },
    {
      "epoch": 0.38344491783323187,
      "grad_norm": 1.4087261604377643,
      "learning_rate": 2e-05,
      "loss": 0.5870172023773194,
      "step": 5040
    },
    {
      "epoch": 0.38420572124163116,
      "grad_norm": 0.8660084567723209,
      "learning_rate": 2e-05,
      "loss": 0.6123439311981201,
      "step": 5050
    },
    {
      "epoch": 0.38496652465003045,
      "grad_norm": 0.8358289715878096,
      "learning_rate": 2e-05,
      "loss": 0.5848734855651856,
      "step": 5060
    },
    {
      "epoch": 0.3857273280584297,
      "grad_norm": 0.9371082715485197,
      "learning_rate": 2e-05,
      "loss": 0.5995091915130615,
      "step": 5070
    },
    {
      "epoch": 0.386488131466829,
      "grad_norm": 0.8722665925481548,
      "learning_rate": 2e-05,
      "loss": 0.5924876689910888,
      "step": 5080
    },
    {
      "epoch": 0.38724893487522827,
      "grad_norm": 0.9824630701618475,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.5860276222229004,
      "step": 5090
    },
    {
      "epoch": 0.3880097382836275,
      "grad_norm": 0.7415429692689308,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.6085256099700928,
      "step": 5100
    },
    {
      "epoch": 0.3887705416920268,
      "grad_norm": 5.573223281455862,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5799697875976563,
      "step": 5110
    },
    {
      "epoch": 0.38953134510042603,
      "grad_norm": 0.9999050919189829,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.593186330795288,
      "step": 5120
    },
    {
      "epoch": 0.3902921485088253,
      "grad_norm": 1.6253646841983163,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.6109166622161866,
      "step": 5130
    },
    {
      "epoch": 0.3910529519172246,
      "grad_norm": 1.3235143393745812,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5991012573242187,
      "step": 5140
    },
    {
      "epoch": 0.39181375532562385,
      "grad_norm": 0.974272822537509,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.6065539360046387,
      "step": 5150
    },
    {
      "epoch": 0.39257455873402314,
      "grad_norm": 0.8621033345279884,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.6009471893310547,
      "step": 5160
    },
    {
      "epoch": 0.3933353621424224,
      "grad_norm": 0.956778479354448,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.6029147148132324,
      "step": 5170
    },
    {
      "epoch": 0.39409616555082166,
      "grad_norm": 0.8208349799111286,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5948306083679199,
      "step": 5180
    },
    {
      "epoch": 0.39485696895922096,
      "grad_norm": 7.737809262820633,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.594374418258667,
      "step": 5190
    },
    {
      "epoch": 0.3956177723676202,
      "grad_norm": 1.1138660030334389,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.5811373710632324,
      "step": 5200
    },
    {
      "epoch": 0.3963785757760195,
      "grad_norm": 1.1455390739337663,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5855268001556396,
      "step": 5210
    },
    {
      "epoch": 0.3971393791844188,
      "grad_norm": 0.8028835177617913,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.6003912925720215,
      "step": 5220
    },
    {
      "epoch": 0.397900182592818,
      "grad_norm": 1.6843154389024695,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5828447341918945,
      "step": 5230
    },
    {
      "epoch": 0.3986609860012173,
      "grad_norm": 1.0284961790196756,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.5892212867736817,
      "step": 5240
    },
    {
      "epoch": 0.39942178940961653,
      "grad_norm": 0.8376699967768668,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.5816598415374756,
      "step": 5250
    },
    {
      "epoch": 0.4001825928180158,
      "grad_norm": 0.8616950302522994,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.5883991241455078,
      "step": 5260
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 1.3365096226533872,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.5804986476898193,
      "step": 5270
    },
    {
      "epoch": 0.40170419963481435,
      "grad_norm": 0.8566802135418893,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.5921107292175293,
      "step": 5280
    },
    {
      "epoch": 0.40246500304321364,
      "grad_norm": 3.731050302760531,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.5780891418457031,
      "step": 5290
    },
    {
      "epoch": 0.4032258064516129,
      "grad_norm": 1.5776428046510773,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.5826613426208496,
      "step": 5300
    },
    {
      "epoch": 0.40398660986001217,
      "grad_norm": 0.7179721998545378,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.5878949642181397,
      "step": 5310
    },
    {
      "epoch": 0.40474741326841146,
      "grad_norm": 0.8763442265772037,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.607383918762207,
      "step": 5320
    },
    {
      "epoch": 0.4055082166768107,
      "grad_norm": 1.0394330266762095,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.5794002056121826,
      "step": 5330
    },
    {
      "epoch": 0.40626902008521,
      "grad_norm": 1.5257881699975921,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5627182960510254,
      "step": 5340
    },
    {
      "epoch": 0.4070298234936093,
      "grad_norm": 0.7285114702926738,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.5978140830993652,
      "step": 5350
    },
    {
      "epoch": 0.4077906269020085,
      "grad_norm": 0.8838079171472663,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.5772541046142579,
      "step": 5360
    },
    {
      "epoch": 0.4085514303104078,
      "grad_norm": 0.7780642750572746,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5725757598876953,
      "step": 5370
    },
    {
      "epoch": 0.40931223371880704,
      "grad_norm": 1.263118197256777,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.6037075996398926,
      "step": 5380
    },
    {
      "epoch": 0.41007303712720633,
      "grad_norm": 0.8147518115748165,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.5896283149719238,
      "step": 5390
    },
    {
      "epoch": 0.4108338405356056,
      "grad_norm": 0.6799993301623181,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.5590286254882812,
      "step": 5400
    },
    {
      "epoch": 0.41159464394400486,
      "grad_norm": 0.9198245616153391,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.5855107784271241,
      "step": 5410
    },
    {
      "epoch": 0.41235544735240415,
      "grad_norm": 1.0052061596928377,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.5983226776123047,
      "step": 5420
    },
    {
      "epoch": 0.4131162507608034,
      "grad_norm": 0.7961295706578796,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.5928987026214599,
      "step": 5430
    },
    {
      "epoch": 0.4138770541692027,
      "grad_norm": 0.873286932998829,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.5997305870056152,
      "step": 5440
    },
    {
      "epoch": 0.41463785757760196,
      "grad_norm": 0.9567172283651997,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.5914229869842529,
      "step": 5450
    },
    {
      "epoch": 0.4153986609860012,
      "grad_norm": 0.7924248044585457,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.5997265815734864,
      "step": 5460
    },
    {
      "epoch": 0.4161594643944005,
      "grad_norm": 0.9803062676735439,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.5905625820159912,
      "step": 5470
    },
    {
      "epoch": 0.4169202678027998,
      "grad_norm": 0.8415249586166841,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.5703700065612793,
      "step": 5480
    },
    {
      "epoch": 0.417681071211199,
      "grad_norm": 0.7357449826755322,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.6008023738861084,
      "step": 5490
    },
    {
      "epoch": 0.4184418746195983,
      "grad_norm": 0.749059941547076,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.6048541069030762,
      "step": 5500
    },
    {
      "epoch": 0.41920267802799754,
      "grad_norm": 1.353316706089092,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.5650211334228515,
      "step": 5510
    },
    {
      "epoch": 0.41996348143639683,
      "grad_norm": 0.8679079360187725,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.5596850872039795,
      "step": 5520
    },
    {
      "epoch": 0.4207242848447961,
      "grad_norm": 1.0528618019029898,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.5501175880432129,
      "step": 5530
    },
    {
      "epoch": 0.42148508825319536,
      "grad_norm": 0.7609752729059599,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.584521484375,
      "step": 5540
    },
    {
      "epoch": 0.42224589166159465,
      "grad_norm": 0.7999483927195585,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.5674933433532715,
      "step": 5550
    },
    {
      "epoch": 0.4230066950699939,
      "grad_norm": 1.344323281674346,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.5871614456176758,
      "step": 5560
    },
    {
      "epoch": 0.4237674984783932,
      "grad_norm": 1.742821426081335,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.5725449562072754,
      "step": 5570
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 1.9687557143362788,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.5592889308929443,
      "step": 5580
    },
    {
      "epoch": 0.4252891052951917,
      "grad_norm": 0.7212969572790084,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.5469254493713379,
      "step": 5590
    },
    {
      "epoch": 0.4258216676810712,
      "step": 5597,
      "total_flos": 2650292248117248.0,
      "train_loss": 0.05331031803746553,
      "train_runtime": 7364.5894,
      "train_samples_per_second": 9.12,
      "train_steps_per_second": 0.76
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2650292248117248.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}