{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6166281755196303,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004618937644341801,
      "grad_norm": 11.009225845336914,
      "learning_rate": 5.773672055427253e-07,
      "loss": 2.4977,
      "step": 10
    },
    {
      "epoch": 0.009237875288683603,
      "grad_norm": 9.019889831542969,
      "learning_rate": 1.1547344110854505e-06,
      "loss": 2.7655,
      "step": 20
    },
    {
      "epoch": 0.013856812933025405,
      "grad_norm": 11.212933540344238,
      "learning_rate": 1.7321016166281756e-06,
      "loss": 2.7241,
      "step": 30
    },
    {
      "epoch": 0.018475750577367205,
      "grad_norm": 9.982237815856934,
      "learning_rate": 2.309468822170901e-06,
      "loss": 2.8357,
      "step": 40
    },
    {
      "epoch": 0.023094688221709007,
      "grad_norm": 16.251842498779297,
      "learning_rate": 2.886836027713626e-06,
      "loss": 2.1159,
      "step": 50
    },
    {
      "epoch": 0.02771362586605081,
      "grad_norm": 7.950500011444092,
      "learning_rate": 3.464203233256351e-06,
      "loss": 1.498,
      "step": 60
    },
    {
      "epoch": 0.03233256351039261,
      "grad_norm": 4.816328048706055,
      "learning_rate": 4.041570438799077e-06,
      "loss": 0.9835,
      "step": 70
    },
    {
      "epoch": 0.03695150115473441,
      "grad_norm": 9.529770851135254,
      "learning_rate": 4.618937644341802e-06,
      "loss": 1.0009,
      "step": 80
    },
    {
      "epoch": 0.04157043879907621,
      "grad_norm": 12.515681266784668,
      "learning_rate": 5.196304849884527e-06,
      "loss": 1.0264,
      "step": 90
    },
    {
      "epoch": 0.046189376443418015,
      "grad_norm": 5.393396854400635,
      "learning_rate": 5.773672055427252e-06,
      "loss": 0.8218,
      "step": 100
    },
    {
      "epoch": 0.050808314087759814,
      "grad_norm": 2.777273416519165,
      "learning_rate": 6.351039260969978e-06,
      "loss": 0.8719,
      "step": 110
    },
    {
      "epoch": 0.05542725173210162,
      "grad_norm": 3.8990070819854736,
      "learning_rate": 6.928406466512702e-06,
      "loss": 0.8592,
      "step": 120
    },
    {
      "epoch": 0.06004618937644342,
      "grad_norm": 4.342905521392822,
      "learning_rate": 7.505773672055427e-06,
      "loss": 0.7982,
      "step": 130
    },
    {
      "epoch": 0.06466512702078522,
      "grad_norm": 2.4967517852783203,
      "learning_rate": 8.083140877598153e-06,
      "loss": 0.7203,
      "step": 140
    },
    {
      "epoch": 0.06928406466512702,
      "grad_norm": 5.491476058959961,
      "learning_rate": 8.660508083140878e-06,
      "loss": 0.7647,
      "step": 150
    },
    {
      "epoch": 0.07390300230946882,
      "grad_norm": 4.13068962097168,
      "learning_rate": 9.237875288683604e-06,
      "loss": 0.7378,
      "step": 160
    },
    {
      "epoch": 0.07852193995381063,
      "grad_norm": 3.9301445484161377,
      "learning_rate": 9.815242494226329e-06,
      "loss": 0.7355,
      "step": 170
    },
    {
      "epoch": 0.08314087759815242,
      "grad_norm": 2.891854763031006,
      "learning_rate": 1.0392609699769053e-05,
      "loss": 0.6911,
      "step": 180
    },
    {
      "epoch": 0.08775981524249422,
      "grad_norm": 5.377829074859619,
      "learning_rate": 1.0969976905311778e-05,
      "loss": 0.7972,
      "step": 190
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 3.1744544506073,
      "learning_rate": 1.1547344110854504e-05,
      "loss": 0.6792,
      "step": 200
    },
    {
      "epoch": 0.09699769053117784,
      "grad_norm": 4.408601760864258,
      "learning_rate": 1.2124711316397229e-05,
      "loss": 0.7378,
      "step": 210
    },
    {
      "epoch": 0.10161662817551963,
      "grad_norm": 3.3633627891540527,
      "learning_rate": 1.2702078521939955e-05,
      "loss": 0.7131,
      "step": 220
    },
    {
      "epoch": 0.10623556581986143,
      "grad_norm": 5.786050319671631,
      "learning_rate": 1.3279445727482678e-05,
      "loss": 0.6806,
      "step": 230
    },
    {
      "epoch": 0.11085450346420324,
      "grad_norm": 6.865295886993408,
      "learning_rate": 1.3856812933025404e-05,
      "loss": 0.7117,
      "step": 240
    },
    {
      "epoch": 0.11547344110854503,
      "grad_norm": 3.714210033416748,
      "learning_rate": 1.4434180138568129e-05,
      "loss": 0.6354,
      "step": 250
    },
    {
      "epoch": 0.12009237875288684,
      "grad_norm": 5.501813888549805,
      "learning_rate": 1.5011547344110854e-05,
      "loss": 0.6266,
      "step": 260
    },
    {
      "epoch": 0.12471131639722864,
      "grad_norm": 6.622759819030762,
      "learning_rate": 1.558891454965358e-05,
      "loss": 0.6894,
      "step": 270
    },
    {
      "epoch": 0.12933025404157045,
      "grad_norm": 3.647956609725952,
      "learning_rate": 1.6166281755196306e-05,
      "loss": 0.66,
      "step": 280
    },
    {
      "epoch": 0.13394919168591224,
      "grad_norm": 4.103866100311279,
      "learning_rate": 1.674364896073903e-05,
      "loss": 0.665,
      "step": 290
    },
    {
      "epoch": 0.13856812933025403,
      "grad_norm": 3.84169340133667,
      "learning_rate": 1.7321016166281756e-05,
      "loss": 0.6676,
      "step": 300
    },
    {
      "epoch": 0.14318706697459585,
      "grad_norm": 8.042781829833984,
      "learning_rate": 1.789838337182448e-05,
      "loss": 0.6236,
      "step": 310
    },
    {
      "epoch": 0.14780600461893764,
      "grad_norm": 4.35403299331665,
      "learning_rate": 1.8475750577367208e-05,
      "loss": 0.6283,
      "step": 320
    },
    {
      "epoch": 0.15242494226327943,
      "grad_norm": 3.7114272117614746,
      "learning_rate": 1.9053117782909933e-05,
      "loss": 0.6472,
      "step": 330
    },
    {
      "epoch": 0.15704387990762125,
      "grad_norm": 2.5825653076171875,
      "learning_rate": 1.9630484988452657e-05,
      "loss": 0.6139,
      "step": 340
    },
    {
      "epoch": 0.16166281755196305,
      "grad_norm": 4.71831750869751,
      "learning_rate": 2.0207852193995382e-05,
      "loss": 0.6019,
      "step": 350
    },
    {
      "epoch": 0.16628175519630484,
      "grad_norm": 7.264115333557129,
      "learning_rate": 2.0785219399538107e-05,
      "loss": 0.6484,
      "step": 360
    },
    {
      "epoch": 0.17090069284064666,
      "grad_norm": 5.484978199005127,
      "learning_rate": 2.1362586605080835e-05,
      "loss": 0.6423,
      "step": 370
    },
    {
      "epoch": 0.17551963048498845,
      "grad_norm": 4.004003524780273,
      "learning_rate": 2.1939953810623556e-05,
      "loss": 0.6167,
      "step": 380
    },
    {
      "epoch": 0.18013856812933027,
      "grad_norm": 3.858074426651001,
      "learning_rate": 2.251732101616628e-05,
      "loss": 0.6029,
      "step": 390
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 4.84448766708374,
      "learning_rate": 2.309468822170901e-05,
      "loss": 0.5511,
      "step": 400
    },
    {
      "epoch": 0.18937644341801385,
      "grad_norm": 3.419329881668091,
      "learning_rate": 2.3672055427251733e-05,
      "loss": 0.5919,
      "step": 410
    },
    {
      "epoch": 0.19399538106235567,
      "grad_norm": 2.367598295211792,
      "learning_rate": 2.4249422632794458e-05,
      "loss": 0.5407,
      "step": 420
    },
    {
      "epoch": 0.19861431870669746,
      "grad_norm": 4.016935348510742,
      "learning_rate": 2.4826789838337182e-05,
      "loss": 0.5829,
      "step": 430
    },
    {
      "epoch": 0.20323325635103925,
      "grad_norm": 4.788145065307617,
      "learning_rate": 2.540415704387991e-05,
      "loss": 0.593,
      "step": 440
    },
    {
      "epoch": 0.20785219399538107,
      "grad_norm": 3.618800401687622,
      "learning_rate": 2.5981524249422635e-05,
      "loss": 0.56,
      "step": 450
    },
    {
      "epoch": 0.21247113163972287,
      "grad_norm": 3.6879987716674805,
      "learning_rate": 2.6558891454965356e-05,
      "loss": 0.5448,
      "step": 460
    },
    {
      "epoch": 0.21709006928406466,
      "grad_norm": 3.6736364364624023,
      "learning_rate": 2.7136258660508084e-05,
      "loss": 0.5707,
      "step": 470
    },
    {
      "epoch": 0.22170900692840648,
      "grad_norm": 3.954604387283325,
      "learning_rate": 2.771362586605081e-05,
      "loss": 0.5891,
      "step": 480
    },
    {
      "epoch": 0.22632794457274827,
      "grad_norm": 6.050353050231934,
      "learning_rate": 2.8290993071593537e-05,
      "loss": 0.6231,
      "step": 490
    },
    {
      "epoch": 0.23094688221709006,
      "grad_norm": 3.728318214416504,
      "learning_rate": 2.8868360277136258e-05,
      "loss": 0.6294,
      "step": 500
    },
    {
      "epoch": 0.23556581986143188,
      "grad_norm": 5.047093391418457,
      "learning_rate": 2.9445727482678986e-05,
      "loss": 0.5795,
      "step": 510
    },
    {
      "epoch": 0.24018475750577367,
      "grad_norm": 4.392754077911377,
      "learning_rate": 3.0023094688221707e-05,
      "loss": 0.5634,
      "step": 520
    },
    {
      "epoch": 0.24480369515011546,
      "grad_norm": 3.634922504425049,
      "learning_rate": 3.060046189376444e-05,
      "loss": 0.5372,
      "step": 530
    },
    {
      "epoch": 0.24942263279445728,
      "grad_norm": 3.2436106204986572,
      "learning_rate": 3.117782909930716e-05,
      "loss": 0.5914,
      "step": 540
    },
    {
      "epoch": 0.2540415704387991,
      "grad_norm": 4.555751323699951,
      "learning_rate": 3.175519630484989e-05,
      "loss": 0.5767,
      "step": 550
    },
    {
      "epoch": 0.2586605080831409,
      "grad_norm": 3.359252452850342,
      "learning_rate": 3.233256351039261e-05,
      "loss": 0.5569,
      "step": 560
    },
    {
      "epoch": 0.2632794457274827,
      "grad_norm": 3.362272262573242,
      "learning_rate": 3.290993071593534e-05,
      "loss": 0.6248,
      "step": 570
    },
    {
      "epoch": 0.2678983833718245,
      "grad_norm": 3.1555299758911133,
      "learning_rate": 3.348729792147806e-05,
      "loss": 0.5629,
      "step": 580
    },
    {
      "epoch": 0.27251732101616627,
      "grad_norm": 4.786733627319336,
      "learning_rate": 3.4064665127020787e-05,
      "loss": 0.5942,
      "step": 590
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 4.290501117706299,
      "learning_rate": 3.464203233256351e-05,
      "loss": 0.5694,
      "step": 600
    },
    {
      "epoch": 0.2817551963048499,
      "grad_norm": 2.2575786113739014,
      "learning_rate": 3.5219399538106236e-05,
      "loss": 0.5459,
      "step": 610
    },
    {
      "epoch": 0.2863741339491917,
      "grad_norm": 4.7077484130859375,
      "learning_rate": 3.579676674364896e-05,
      "loss": 0.5686,
      "step": 620
    },
    {
      "epoch": 0.2909930715935335,
      "grad_norm": 3.487034559249878,
      "learning_rate": 3.6374133949191685e-05,
      "loss": 0.5185,
      "step": 630
    },
    {
      "epoch": 0.2956120092378753,
      "grad_norm": 4.230603218078613,
      "learning_rate": 3.6951501154734416e-05,
      "loss": 0.5548,
      "step": 640
    },
    {
      "epoch": 0.3002309468822171,
      "grad_norm": 3.5367519855499268,
      "learning_rate": 3.7528868360277134e-05,
      "loss": 0.5658,
      "step": 650
    },
    {
      "epoch": 0.30484988452655887,
      "grad_norm": 3.3856465816497803,
      "learning_rate": 3.8106235565819866e-05,
      "loss": 0.5459,
      "step": 660
    },
    {
      "epoch": 0.3094688221709007,
      "grad_norm": 3.4272944927215576,
      "learning_rate": 3.868360277136259e-05,
      "loss": 0.5823,
      "step": 670
    },
    {
      "epoch": 0.3140877598152425,
      "grad_norm": 2.7848410606384277,
      "learning_rate": 3.9260969976905315e-05,
      "loss": 0.5481,
      "step": 680
    },
    {
      "epoch": 0.3187066974595843,
      "grad_norm": 2.8584043979644775,
      "learning_rate": 3.983833718244804e-05,
      "loss": 0.4907,
      "step": 690
    },
    {
      "epoch": 0.3233256351039261,
      "grad_norm": 3.9872729778289795,
      "learning_rate": 4.0415704387990764e-05,
      "loss": 0.5703,
      "step": 700
    },
    {
      "epoch": 0.3279445727482679,
      "grad_norm": 4.052183151245117,
      "learning_rate": 4.099307159353349e-05,
      "loss": 0.6229,
      "step": 710
    },
    {
      "epoch": 0.3325635103926097,
      "grad_norm": 3.6171059608459473,
      "learning_rate": 4.1570438799076213e-05,
      "loss": 0.5022,
      "step": 720
    },
    {
      "epoch": 0.3371824480369515,
      "grad_norm": 3.4360239505767822,
      "learning_rate": 4.214780600461894e-05,
      "loss": 0.5642,
      "step": 730
    },
    {
      "epoch": 0.3418013856812933,
      "grad_norm": 3.328563928604126,
      "learning_rate": 4.272517321016167e-05,
      "loss": 0.543,
      "step": 740
    },
    {
      "epoch": 0.3464203233256351,
      "grad_norm": 2.494722843170166,
      "learning_rate": 4.330254041570439e-05,
      "loss": 0.5056,
      "step": 750
    },
    {
      "epoch": 0.3510392609699769,
      "grad_norm": 2.9856996536254883,
      "learning_rate": 4.387990762124711e-05,
      "loss": 0.5215,
      "step": 760
    },
    {
      "epoch": 0.3556581986143187,
      "grad_norm": 3.6510608196258545,
      "learning_rate": 4.445727482678984e-05,
      "loss": 0.5418,
      "step": 770
    },
    {
      "epoch": 0.36027713625866054,
      "grad_norm": 2.837355136871338,
      "learning_rate": 4.503464203233256e-05,
      "loss": 0.5197,
      "step": 780
    },
    {
      "epoch": 0.3648960739030023,
      "grad_norm": 2.4272849559783936,
      "learning_rate": 4.561200923787529e-05,
      "loss": 0.5151,
      "step": 790
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 2.2198147773742676,
      "learning_rate": 4.618937644341802e-05,
      "loss": 0.5037,
      "step": 800
    },
    {
      "epoch": 0.3741339491916859,
      "grad_norm": 3.082343816757202,
      "learning_rate": 4.676674364896074e-05,
      "loss": 0.5411,
      "step": 810
    },
    {
      "epoch": 0.3787528868360277,
      "grad_norm": 2.739429235458374,
      "learning_rate": 4.7344110854503466e-05,
      "loss": 0.5034,
      "step": 820
    },
    {
      "epoch": 0.3833718244803695,
      "grad_norm": 2.8211052417755127,
      "learning_rate": 4.792147806004619e-05,
      "loss": 0.5077,
      "step": 830
    },
    {
      "epoch": 0.38799076212471134,
      "grad_norm": 4.462055683135986,
      "learning_rate": 4.8498845265588916e-05,
      "loss": 0.4883,
      "step": 840
    },
    {
      "epoch": 0.39260969976905313,
      "grad_norm": 2.8090898990631104,
      "learning_rate": 4.907621247113165e-05,
      "loss": 0.4918,
      "step": 850
    },
    {
      "epoch": 0.3972286374133949,
      "grad_norm": 2.787473440170288,
      "learning_rate": 4.9653579676674365e-05,
      "loss": 0.4736,
      "step": 860
    },
    {
      "epoch": 0.4018475750577367,
      "grad_norm": 3.198763132095337,
      "learning_rate": 4.999996750557592e-05,
      "loss": 0.5243,
      "step": 870
    },
    {
      "epoch": 0.4064665127020785,
      "grad_norm": 3.8778421878814697,
      "learning_rate": 4.9999601944275146e-05,
      "loss": 0.5206,
      "step": 880
    },
    {
      "epoch": 0.4110854503464203,
      "grad_norm": 4.326559066772461,
      "learning_rate": 4.999883020960263e-05,
      "loss": 0.5998,
      "step": 890
    },
    {
      "epoch": 0.41570438799076215,
      "grad_norm": 2.1009411811828613,
      "learning_rate": 4.99976523140969e-05,
      "loss": 0.5097,
      "step": 900
    },
    {
      "epoch": 0.42032332563510394,
      "grad_norm": 3.0838584899902344,
      "learning_rate": 4.999606827689545e-05,
      "loss": 0.4699,
      "step": 910
    },
    {
      "epoch": 0.42494226327944573,
      "grad_norm": 3.067430019378662,
      "learning_rate": 4.999407812373445e-05,
      "loss": 0.5223,
      "step": 920
    },
    {
      "epoch": 0.4295612009237875,
      "grad_norm": 2.580418109893799,
      "learning_rate": 4.9991681886948297e-05,
      "loss": 0.5396,
      "step": 930
    },
    {
      "epoch": 0.4341801385681293,
      "grad_norm": 2.748032808303833,
      "learning_rate": 4.9988879605469106e-05,
      "loss": 0.4965,
      "step": 940
    },
    {
      "epoch": 0.4387990762124711,
      "grad_norm": 2.841571569442749,
      "learning_rate": 4.99856713248261e-05,
      "loss": 0.472,
      "step": 950
    },
    {
      "epoch": 0.44341801385681295,
      "grad_norm": 1.8557236194610596,
      "learning_rate": 4.9982057097144826e-05,
      "loss": 0.4878,
      "step": 960
    },
    {
      "epoch": 0.44803695150115475,
      "grad_norm": 5.707876205444336,
      "learning_rate": 4.9978036981146344e-05,
      "loss": 0.4883,
      "step": 970
    },
    {
      "epoch": 0.45265588914549654,
      "grad_norm": 3.867081880569458,
      "learning_rate": 4.9973611042146254e-05,
      "loss": 0.4929,
      "step": 980
    },
    {
      "epoch": 0.45727482678983833,
      "grad_norm": 2.291074275970459,
      "learning_rate": 4.996877935205365e-05,
      "loss": 0.5046,
      "step": 990
    },
    {
      "epoch": 0.4618937644341801,
      "grad_norm": 2.332185745239258,
      "learning_rate": 4.996354198936993e-05,
      "loss": 0.4539,
      "step": 1000
    },
    {
      "epoch": 0.4665127020785219,
      "grad_norm": 3.7381327152252197,
      "learning_rate": 4.995789903918755e-05,
      "loss": 0.5296,
      "step": 1010
    },
    {
      "epoch": 0.47113163972286376,
      "grad_norm": 2.239875078201294,
      "learning_rate": 4.99518505931886e-05,
      "loss": 0.5004,
      "step": 1020
    },
    {
      "epoch": 0.47575057736720555,
      "grad_norm": 3.7051751613616943,
      "learning_rate": 4.9945396749643356e-05,
      "loss": 0.5319,
      "step": 1030
    },
    {
      "epoch": 0.48036951501154734,
      "grad_norm": 2.5135374069213867,
      "learning_rate": 4.993853761340867e-05,
      "loss": 0.4952,
      "step": 1040
    },
    {
      "epoch": 0.48498845265588914,
      "grad_norm": 2.6355159282684326,
      "learning_rate": 4.993127329592625e-05,
      "loss": 0.4807,
      "step": 1050
    },
    {
      "epoch": 0.4896073903002309,
      "grad_norm": 5.0833563804626465,
      "learning_rate": 4.9923603915220874e-05,
      "loss": 0.5269,
      "step": 1060
    },
    {
      "epoch": 0.4942263279445728,
      "grad_norm": 2.4188873767852783,
      "learning_rate": 4.9915529595898445e-05,
      "loss": 0.5183,
      "step": 1070
    },
    {
      "epoch": 0.49884526558891457,
      "grad_norm": 2.623342514038086,
      "learning_rate": 4.9907050469144e-05,
      "loss": 0.4928,
      "step": 1080
    },
    {
      "epoch": 0.5034642032332564,
      "grad_norm": 2.629373073577881,
      "learning_rate": 4.9898166672719547e-05,
      "loss": 0.5144,
      "step": 1090
    },
    {
      "epoch": 0.5080831408775982,
      "grad_norm": 1.6594828367233276,
      "learning_rate": 4.9888878350961857e-05,
      "loss": 0.4733,
      "step": 1100
    },
    {
      "epoch": 0.5127020785219399,
      "grad_norm": 2.773136854171753,
      "learning_rate": 4.987918565478008e-05,
      "loss": 0.5054,
      "step": 1110
    },
    {
      "epoch": 0.5173210161662818,
      "grad_norm": 1.9525351524353027,
      "learning_rate": 4.9869088741653335e-05,
      "loss": 0.4454,
      "step": 1120
    },
    {
      "epoch": 0.5219399538106235,
      "grad_norm": 2.84041166305542,
      "learning_rate": 4.985858777562812e-05,
      "loss": 0.4871,
      "step": 1130
    },
    {
      "epoch": 0.5265588914549654,
      "grad_norm": 2.2563703060150146,
      "learning_rate": 4.9847682927315656e-05,
      "loss": 0.4605,
      "step": 1140
    },
    {
      "epoch": 0.5311778290993071,
      "grad_norm": 3.4986164569854736,
      "learning_rate": 4.983637437388913e-05,
      "loss": 0.5281,
      "step": 1150
    },
    {
      "epoch": 0.535796766743649,
      "grad_norm": 2.7081637382507324,
      "learning_rate": 4.98246622990808e-05,
      "loss": 0.4568,
      "step": 1160
    },
    {
      "epoch": 0.5404157043879908,
      "grad_norm": 3.4630649089813232,
      "learning_rate": 4.981254689317902e-05,
      "loss": 0.5351,
      "step": 1170
    },
    {
      "epoch": 0.5450346420323325,
      "grad_norm": 2.5266454219818115,
      "learning_rate": 4.9800028353025125e-05,
      "loss": 0.4378,
      "step": 1180
    },
    {
      "epoch": 0.5496535796766744,
      "grad_norm": 2.003904104232788,
      "learning_rate": 4.978710688201026e-05,
      "loss": 0.4539,
      "step": 1190
    },
    {
      "epoch": 0.5542725173210161,
      "grad_norm": 3.799851655960083,
      "learning_rate": 4.9773782690072055e-05,
      "loss": 0.5349,
      "step": 1200
    },
    {
      "epoch": 0.558891454965358,
      "grad_norm": 2.7120118141174316,
      "learning_rate": 4.976005599369126e-05,
      "loss": 0.4938,
      "step": 1210
    },
    {
      "epoch": 0.5635103926096998,
      "grad_norm": 2.687408208847046,
      "learning_rate": 4.974592701588816e-05,
      "loss": 0.4473,
      "step": 1220
    },
    {
      "epoch": 0.5681293302540416,
      "grad_norm": 2.5710184574127197,
      "learning_rate": 4.9731395986218976e-05,
      "loss": 0.4854,
      "step": 1230
    },
    {
      "epoch": 0.5727482678983834,
      "grad_norm": 3.5843756198883057,
      "learning_rate": 4.9716463140772176e-05,
      "loss": 0.5108,
      "step": 1240
    },
    {
      "epoch": 0.5773672055427251,
      "grad_norm": 3.0808424949645996,
      "learning_rate": 4.970112872216459e-05,
      "loss": 0.4664,
      "step": 1250
    },
    {
      "epoch": 0.581986143187067,
      "grad_norm": 2.7078258991241455,
      "learning_rate": 4.968539297953748e-05,
      "loss": 0.5728,
      "step": 1260
    },
    {
      "epoch": 0.5866050808314087,
      "grad_norm": 2.1983377933502197,
      "learning_rate": 4.96692561685525e-05,
      "loss": 0.4551,
      "step": 1270
    },
    {
      "epoch": 0.5912240184757506,
      "grad_norm": 3.719346046447754,
      "learning_rate": 4.9652718551387545e-05,
      "loss": 0.4358,
      "step": 1280
    },
    {
      "epoch": 0.5958429561200924,
      "grad_norm": 3.773397207260132,
      "learning_rate": 4.963578039673248e-05,
      "loss": 0.3883,
      "step": 1290
    },
    {
      "epoch": 0.6004618937644342,
      "grad_norm": 2.14461350440979,
      "learning_rate": 4.961844197978479e-05,
      "loss": 0.4403,
      "step": 1300
    },
    {
      "epoch": 0.605080831408776,
      "grad_norm": 3.8363335132598877,
      "learning_rate": 4.960070358224508e-05,
      "loss": 0.4753,
      "step": 1310
    },
    {
      "epoch": 0.6096997690531177,
      "grad_norm": 2.10432505607605,
      "learning_rate": 4.9582565492312537e-05,
      "loss": 0.4728,
      "step": 1320
    },
    {
      "epoch": 0.6143187066974596,
      "grad_norm": 2.5199790000915527,
      "learning_rate": 4.9564028004680215e-05,
      "loss": 0.4544,
      "step": 1330
    },
    {
      "epoch": 0.6189376443418014,
      "grad_norm": 3.8586044311523438,
      "learning_rate": 4.9545091420530265e-05,
      "loss": 0.4649,
      "step": 1340
    },
    {
      "epoch": 0.6235565819861432,
      "grad_norm": 1.9125807285308838,
      "learning_rate": 4.9525756047529023e-05,
      "loss": 0.4609,
      "step": 1350
    },
    {
      "epoch": 0.628175519630485,
      "grad_norm": 2.1655430793762207,
      "learning_rate": 4.950602219982207e-05,
      "loss": 0.4832,
      "step": 1360
    },
    {
      "epoch": 0.6327944572748267,
      "grad_norm": 4.029664993286133,
      "learning_rate": 4.948589019802901e-05,
      "loss": 0.4751,
      "step": 1370
    },
    {
      "epoch": 0.6374133949191686,
      "grad_norm": 2.2781596183776855,
      "learning_rate": 4.94653603692384e-05,
      "loss": 0.4691,
      "step": 1380
    },
    {
      "epoch": 0.6420323325635104,
      "grad_norm": 1.623131513595581,
      "learning_rate": 4.9444433047002327e-05,
      "loss": 0.4397,
      "step": 1390
    },
    {
      "epoch": 0.6466512702078522,
      "grad_norm": 2.9153196811676025,
      "learning_rate": 4.9423108571331055e-05,
      "loss": 0.4737,
      "step": 1400
    },
    {
      "epoch": 0.651270207852194,
      "grad_norm": 3.0765182971954346,
      "learning_rate": 4.940138728868746e-05,
      "loss": 0.4329,
      "step": 1410
    },
    {
      "epoch": 0.6558891454965358,
      "grad_norm": 3.2848074436187744,
      "learning_rate": 4.937926955198144e-05,
      "loss": 0.4488,
      "step": 1420
    },
    {
      "epoch": 0.6605080831408776,
      "grad_norm": 2.2458810806274414,
      "learning_rate": 4.935675572056413e-05,
      "loss": 0.4645,
      "step": 1430
    },
    {
      "epoch": 0.6651270207852193,
      "grad_norm": 2.3431642055511475,
      "learning_rate": 4.933384616022211e-05,
      "loss": 0.4492,
      "step": 1440
    },
    {
      "epoch": 0.6697459584295612,
      "grad_norm": 2.6324145793914795,
      "learning_rate": 4.931054124317144e-05,
      "loss": 0.4978,
      "step": 1450
    },
    {
      "epoch": 0.674364896073903,
      "grad_norm": 2.575723171234131,
      "learning_rate": 4.928684134805162e-05,
      "loss": 0.4442,
      "step": 1460
    },
    {
      "epoch": 0.6789838337182448,
      "grad_norm": 2.842388868331909,
      "learning_rate": 4.926274685991944e-05,
      "loss": 0.4412,
      "step": 1470
    },
    {
      "epoch": 0.6836027713625866,
      "grad_norm": 1.8333570957183838,
      "learning_rate": 4.9238258170242703e-05,
      "loss": 0.4655,
      "step": 1480
    },
    {
      "epoch": 0.6882217090069284,
      "grad_norm": 2.276472806930542,
      "learning_rate": 4.921337567689389e-05,
      "loss": 0.3944,
      "step": 1490
    },
    {
      "epoch": 0.6928406466512702,
      "grad_norm": 2.947540760040283,
      "learning_rate": 4.91880997841437e-05,
      "loss": 0.4545,
      "step": 1500
    },
    {
      "epoch": 0.6974595842956121,
      "grad_norm": 2.092616558074951,
      "learning_rate": 4.916243090265444e-05,
      "loss": 0.4202,
      "step": 1510
    },
    {
      "epoch": 0.7020785219399538,
      "grad_norm": 2.4105489253997803,
      "learning_rate": 4.91363694494734e-05,
      "loss": 0.4488,
      "step": 1520
    },
    {
      "epoch": 0.7066974595842956,
      "grad_norm": 2.499821662902832,
      "learning_rate": 4.910991584802606e-05,
      "loss": 0.4599,
      "step": 1530
    },
    {
      "epoch": 0.7113163972286374,
      "grad_norm": 1.7518564462661743,
      "learning_rate": 4.9083070528109195e-05,
      "loss": 0.4469,
      "step": 1540
    },
    {
      "epoch": 0.7159353348729792,
      "grad_norm": 1.6236387491226196,
      "learning_rate": 4.905583392588392e-05,
      "loss": 0.4756,
      "step": 1550
    },
    {
      "epoch": 0.7205542725173211,
      "grad_norm": 3.137925624847412,
      "learning_rate": 4.902820648386858e-05,
      "loss": 0.4793,
      "step": 1560
    },
    {
      "epoch": 0.7251732101616628,
      "grad_norm": 2.584268093109131,
      "learning_rate": 4.900018865093158e-05,
      "loss": 0.4445,
      "step": 1570
    },
    {
      "epoch": 0.7297921478060047,
      "grad_norm": 2.1350574493408203,
      "learning_rate": 4.897178088228406e-05,
      "loss": 0.4422,
      "step": 1580
    },
    {
      "epoch": 0.7344110854503464,
      "grad_norm": 2.1932408809661865,
      "learning_rate": 4.8942983639472554e-05,
      "loss": 0.4386,
      "step": 1590
    },
    {
      "epoch": 0.7390300230946882,
      "grad_norm": 2.0968682765960693,
      "learning_rate": 4.891379739037142e-05,
      "loss": 0.4321,
      "step": 1600
    },
    {
      "epoch": 0.74364896073903,
      "grad_norm": 1.718479037284851,
      "learning_rate": 4.888422260917532e-05,
      "loss": 0.4422,
      "step": 1610
    },
    {
      "epoch": 0.7482678983833718,
      "grad_norm": 2.742122173309326,
      "learning_rate": 4.885425977639143e-05,
      "loss": 0.4118,
      "step": 1620
    },
    {
      "epoch": 0.7528868360277137,
      "grad_norm": 3.560037136077881,
      "learning_rate": 4.8823909378831694e-05,
      "loss": 0.4352,
      "step": 1630
    },
    {
      "epoch": 0.7575057736720554,
      "grad_norm": 1.751957654953003,
      "learning_rate": 4.8793171909604904e-05,
      "loss": 0.4222,
      "step": 1640
    },
    {
      "epoch": 0.7621247113163973,
      "grad_norm": 2.7816951274871826,
      "learning_rate": 4.876204786810867e-05,
      "loss": 0.4458,
      "step": 1650
    },
    {
      "epoch": 0.766743648960739,
      "grad_norm": 1.7703163623809814,
      "learning_rate": 4.873053776002133e-05,
      "loss": 0.4387,
      "step": 1660
    },
    {
      "epoch": 0.7713625866050808,
      "grad_norm": 1.7305679321289062,
      "learning_rate": 4.8698642097293684e-05,
      "loss": 0.4203,
      "step": 1670
    },
    {
      "epoch": 0.7759815242494227,
      "grad_norm": 2.028597116470337,
      "learning_rate": 4.866636139814076e-05,
      "loss": 0.4445,
      "step": 1680
    },
    {
      "epoch": 0.7806004618937644,
      "grad_norm": 2.1649012565612793,
      "learning_rate": 4.863369618703331e-05,
      "loss": 0.4392,
      "step": 1690
    },
    {
      "epoch": 0.7852193995381063,
      "grad_norm": 1.9613428115844727,
      "learning_rate": 4.860064699468935e-05,
      "loss": 0.4128,
      "step": 1700
    },
    {
      "epoch": 0.789838337182448,
      "grad_norm": 1.6601072549819946,
      "learning_rate": 4.85672143580655e-05,
      "loss": 0.4211,
      "step": 1710
    },
    {
      "epoch": 0.7944572748267898,
      "grad_norm": 1.360724925994873,
      "learning_rate": 4.85333988203483e-05,
      "loss": 0.4641,
      "step": 1720
    },
    {
      "epoch": 0.7990762124711316,
      "grad_norm": 1.8491480350494385,
      "learning_rate": 4.84992009309453e-05,
      "loss": 0.4466,
      "step": 1730
    },
    {
      "epoch": 0.8036951501154734,
      "grad_norm": 3.4997713565826416,
      "learning_rate": 4.8464621245476254e-05,
      "loss": 0.3786,
      "step": 1740
    },
    {
      "epoch": 0.8083140877598153,
      "grad_norm": 1.9356039762496948,
      "learning_rate": 4.8429660325764e-05,
      "loss": 0.3948,
      "step": 1750
    },
    {
      "epoch": 0.812933025404157,
      "grad_norm": 1.8650673627853394,
      "learning_rate": 4.8394318739825364e-05,
      "loss": 0.3733,
      "step": 1760
    },
    {
      "epoch": 0.8175519630484989,
      "grad_norm": 2.526487112045288,
      "learning_rate": 4.835859706186193e-05,
      "loss": 0.4702,
      "step": 1770
    },
    {
      "epoch": 0.8221709006928406,
      "grad_norm": 2.5216803550720215,
      "learning_rate": 4.832249587225073e-05,
      "loss": 0.4268,
      "step": 1780
    },
    {
      "epoch": 0.8267898383371824,
      "grad_norm": 3.427783966064453,
      "learning_rate": 4.828601575753476e-05,
      "loss": 0.4198,
      "step": 1790
    },
    {
      "epoch": 0.8314087759815243,
      "grad_norm": 2.734025001525879,
      "learning_rate": 4.824915731041352e-05,
      "loss": 0.4305,
      "step": 1800
    },
    {
      "epoch": 0.836027713625866,
      "grad_norm": 1.9482347965240479,
      "learning_rate": 4.821192112973333e-05,
      "loss": 0.4373,
      "step": 1810
    },
    {
      "epoch": 0.8406466512702079,
      "grad_norm": 2.4370124340057373,
      "learning_rate": 4.817430782047762e-05,
      "loss": 0.4477,
      "step": 1820
    },
    {
      "epoch": 0.8452655889145496,
      "grad_norm": 2.8961851596832275,
      "learning_rate": 4.8136317993757114e-05,
      "loss": 0.4611,
      "step": 1830
    },
    {
      "epoch": 0.8498845265588915,
      "grad_norm": 2.3259379863739014,
      "learning_rate": 4.8097952266799876e-05,
      "loss": 0.4297,
      "step": 1840
    },
    {
      "epoch": 0.8545034642032333,
      "grad_norm": 3.403167963027954,
      "learning_rate": 4.805921126294129e-05,
      "loss": 0.4028,
      "step": 1850
    },
    {
      "epoch": 0.859122401847575,
      "grad_norm": 1.5918769836425781,
      "learning_rate": 4.802009561161396e-05,
      "loss": 0.3982,
      "step": 1860
    },
    {
      "epoch": 0.8637413394919169,
      "grad_norm": 2.220611333847046,
      "learning_rate": 4.798060594833743e-05,
      "loss": 0.4213,
      "step": 1870
    },
    {
      "epoch": 0.8683602771362586,
      "grad_norm": 1.634158968925476,
      "learning_rate": 4.794074291470791e-05,
      "loss": 0.4085,
      "step": 1880
    },
    {
      "epoch": 0.8729792147806005,
      "grad_norm": 2.6168339252471924,
      "learning_rate": 4.790050715838783e-05,
      "loss": 0.437,
      "step": 1890
    },
    {
      "epoch": 0.8775981524249422,
      "grad_norm": 2.0392565727233887,
      "learning_rate": 4.7859899333095296e-05,
      "loss": 0.4407,
      "step": 1900
    },
    {
      "epoch": 0.8822170900692841,
      "grad_norm": 2.446300506591797,
      "learning_rate": 4.7818920098593516e-05,
      "loss": 0.4415,
      "step": 1910
    },
    {
      "epoch": 0.8868360277136259,
      "grad_norm": 2.7988600730895996,
      "learning_rate": 4.777757012068004e-05,
      "loss": 0.433,
      "step": 1920
    },
    {
      "epoch": 0.8914549653579676,
      "grad_norm": 1.9571830034255981,
      "learning_rate": 4.773585007117597e-05,
      "loss": 0.3957,
      "step": 1930
    },
    {
      "epoch": 0.8960739030023095,
      "grad_norm": 3.4647884368896484,
      "learning_rate": 4.7693760627915015e-05,
      "loss": 0.4442,
      "step": 1940
    },
    {
      "epoch": 0.9006928406466512,
      "grad_norm": 2.8123703002929688,
      "learning_rate": 4.765130247473252e-05,
      "loss": 0.502,
      "step": 1950
    },
    {
      "epoch": 0.9053117782909931,
      "grad_norm": 2.4884793758392334,
      "learning_rate": 4.760847630145431e-05,
      "loss": 0.4532,
      "step": 1960
    },
    {
      "epoch": 0.9099307159353349,
      "grad_norm": 2.372985601425171,
      "learning_rate": 4.756528280388551e-05,
      "loss": 0.3951,
      "step": 1970
    },
    {
      "epoch": 0.9145496535796767,
      "grad_norm": 2.322315216064453,
      "learning_rate": 4.752172268379923e-05,
      "loss": 0.4442,
      "step": 1980
    },
    {
      "epoch": 0.9191685912240185,
      "grad_norm": 2.875863552093506,
      "learning_rate": 4.7477796648925186e-05,
      "loss": 0.4512,
      "step": 1990
    },
    {
      "epoch": 0.9237875288683602,
      "grad_norm": 2.0720949172973633,
      "learning_rate": 4.743350541293817e-05,
      "loss": 0.4076,
      "step": 2000
    },
    {
      "epoch": 0.9284064665127021,
      "grad_norm": 2.1072168350219727,
      "learning_rate": 4.738884969544646e-05,
      "loss": 0.3686,
      "step": 2010
    },
    {
      "epoch": 0.9330254041570438,
      "grad_norm": 2.536849021911621,
      "learning_rate": 4.734383022198013e-05,
      "loss": 0.3822,
      "step": 2020
    },
    {
      "epoch": 0.9376443418013857,
      "grad_norm": 2.638981580734253,
      "learning_rate": 4.72984477239793e-05,
      "loss": 0.3934,
      "step": 2030
    },
    {
      "epoch": 0.9422632794457275,
      "grad_norm": 2.287445545196533,
      "learning_rate": 4.725270293878218e-05,
      "loss": 0.4234,
      "step": 2040
    },
    {
      "epoch": 0.9468822170900693,
      "grad_norm": 3.737877130508423,
      "learning_rate": 4.720659660961316e-05,
      "loss": 0.3855,
      "step": 2050
    },
    {
      "epoch": 0.9515011547344111,
      "grad_norm": 1.3333017826080322,
      "learning_rate": 4.7160129485570706e-05,
      "loss": 0.4159,
      "step": 2060
    },
    {
      "epoch": 0.9561200923787528,
      "grad_norm": 3.8679983615875244,
      "learning_rate": 4.711330232161515e-05,
      "loss": 0.3882,
      "step": 2070
    },
    {
      "epoch": 0.9607390300230947,
      "grad_norm": 2.0616493225097656,
      "learning_rate": 4.706611587855651e-05,
      "loss": 0.4304,
      "step": 2080
    },
    {
      "epoch": 0.9653579676674365,
      "grad_norm": 1.910409688949585,
      "learning_rate": 4.7018570923042054e-05,
      "loss": 0.4648,
      "step": 2090
    },
    {
      "epoch": 0.9699769053117783,
      "grad_norm": 2.010974645614624,
      "learning_rate": 4.697066822754388e-05,
      "loss": 0.407,
      "step": 2100
    },
    {
      "epoch": 0.9745958429561201,
      "grad_norm": 2.0885708332061768,
      "learning_rate": 4.692240857034634e-05,
      "loss": 0.4084,
      "step": 2110
    },
    {
      "epoch": 0.9792147806004619,
      "grad_norm": 2.3431713581085205,
      "learning_rate": 4.6873792735533436e-05,
      "loss": 0.436,
      "step": 2120
    },
    {
      "epoch": 0.9838337182448037,
      "grad_norm": 2.349169969558716,
      "learning_rate": 4.682482151297605e-05,
      "loss": 0.4102,
      "step": 2130
    },
    {
      "epoch": 0.9884526558891455,
      "grad_norm": 2.2386903762817383,
      "learning_rate": 4.6775495698319114e-05,
      "loss": 0.4126,
      "step": 2140
    },
    {
      "epoch": 0.9930715935334873,
      "grad_norm": 1.9783356189727783,
      "learning_rate": 4.672581609296868e-05,
      "loss": 0.4056,
      "step": 2150
    },
    {
      "epoch": 0.9976905311778291,
      "grad_norm": 1.545580506324768,
      "learning_rate": 4.667578350407891e-05,
      "loss": 0.4159,
      "step": 2160
    },
    {
      "epoch": 1.002309468822171,
      "grad_norm": 2.3541977405548096,
      "learning_rate": 4.662539874453896e-05,
      "loss": 0.3425,
      "step": 2170
    },
    {
      "epoch": 1.0069284064665127,
      "grad_norm": 1.4054622650146484,
      "learning_rate": 4.657466263295978e-05,
      "loss": 0.2923,
      "step": 2180
    },
    {
      "epoch": 1.0115473441108545,
      "grad_norm": 2.2558600902557373,
      "learning_rate": 4.652357599366078e-05,
      "loss": 0.3432,
      "step": 2190
    },
    {
      "epoch": 1.0161662817551964,
      "grad_norm": 2.1105244159698486,
      "learning_rate": 4.6472139656656504e-05,
      "loss": 0.3093,
      "step": 2200
    },
    {
      "epoch": 1.0207852193995381,
      "grad_norm": 1.7107679843902588,
      "learning_rate": 4.642035445764305e-05,
      "loss": 0.2854,
      "step": 2210
    },
    {
      "epoch": 1.0254041570438799,
      "grad_norm": 2.075529098510742,
      "learning_rate": 4.636822123798458e-05,
      "loss": 0.3232,
      "step": 2220
    },
    {
      "epoch": 1.0300230946882216,
      "grad_norm": 2.467179536819458,
      "learning_rate": 4.6315740844699606e-05,
      "loss": 0.3597,
      "step": 2230
    },
    {
      "epoch": 1.0346420323325636,
      "grad_norm": 1.628585934638977,
      "learning_rate": 4.626291413044723e-05,
      "loss": 0.3165,
      "step": 2240
    },
    {
      "epoch": 1.0392609699769053,
      "grad_norm": 1.6538641452789307,
      "learning_rate": 4.620974195351331e-05,
      "loss": 0.3332,
      "step": 2250
    },
    {
      "epoch": 1.043879907621247,
      "grad_norm": 1.2917189598083496,
      "learning_rate": 4.615622517779648e-05,
      "loss": 0.3207,
      "step": 2260
    },
    {
      "epoch": 1.048498845265589,
      "grad_norm": 1.6541030406951904,
      "learning_rate": 4.6102364672794186e-05,
      "loss": 0.3206,
      "step": 2270
    },
    {
      "epoch": 1.0531177829099307,
      "grad_norm": 2.9960944652557373,
      "learning_rate": 4.6048161313588454e-05,
      "loss": 0.3283,
      "step": 2280
    },
    {
      "epoch": 1.0577367205542725,
      "grad_norm": 1.576786994934082,
      "learning_rate": 4.599361598083175e-05,
      "loss": 0.3108,
      "step": 2290
    },
    {
      "epoch": 1.0623556581986142,
      "grad_norm": 2.9151947498321533,
      "learning_rate": 4.593872956073267e-05,
      "loss": 0.3134,
      "step": 2300
    },
    {
      "epoch": 1.0669745958429562,
      "grad_norm": 1.6089911460876465,
      "learning_rate": 4.588350294504149e-05,
      "loss": 0.3059,
      "step": 2310
    },
    {
      "epoch": 1.071593533487298,
      "grad_norm": 2.043606996536255,
      "learning_rate": 4.582793703103574e-05,
      "loss": 0.3063,
      "step": 2320
    },
    {
      "epoch": 1.0762124711316396,
      "grad_norm": 1.3332914113998413,
      "learning_rate": 4.577203272150557e-05,
      "loss": 0.3358,
      "step": 2330
    },
    {
      "epoch": 1.0808314087759816,
      "grad_norm": 1.6624664068222046,
      "learning_rate": 4.571579092473911e-05,
      "loss": 0.327,
      "step": 2340
    },
    {
      "epoch": 1.0854503464203233,
      "grad_norm": 1.5694808959960938,
      "learning_rate": 4.565921255450773e-05,
      "loss": 0.3128,
      "step": 2350
    },
    {
      "epoch": 1.090069284064665,
      "grad_norm": 2.037529468536377,
      "learning_rate": 4.5602298530051165e-05,
      "loss": 0.3295,
      "step": 2360
    },
    {
      "epoch": 1.0946882217090068,
      "grad_norm": 2.3830647468566895,
      "learning_rate": 4.5545049776062577e-05,
      "loss": 0.3218,
      "step": 2370
    },
    {
      "epoch": 1.0993071593533488,
      "grad_norm": 1.4938024282455444,
      "learning_rate": 4.548746722267356e-05,
      "loss": 0.3154,
      "step": 2380
    },
    {
      "epoch": 1.1039260969976905,
      "grad_norm": 2.7618491649627686,
      "learning_rate": 4.542955180543901e-05,
      "loss": 0.2857,
      "step": 2390
    },
    {
      "epoch": 1.1085450346420322,
      "grad_norm": 1.6438803672790527,
      "learning_rate": 4.53713044653219e-05,
      "loss": 0.2997,
      "step": 2400
    },
    {
      "epoch": 1.1131639722863742,
      "grad_norm": 1.8261016607284546,
      "learning_rate": 4.531272614867806e-05,
      "loss": 0.3157,
      "step": 2410
    },
    {
      "epoch": 1.117782909930716,
      "grad_norm": 1.5871988534927368,
      "learning_rate": 4.525381780724073e-05,
      "loss": 0.3208,
      "step": 2420
    },
    {
      "epoch": 1.1224018475750577,
      "grad_norm": 2.379218339920044,
      "learning_rate": 4.519458039810515e-05,
      "loss": 0.2979,
      "step": 2430
    },
    {
      "epoch": 1.1270207852193996,
      "grad_norm": 1.5023610591888428,
      "learning_rate": 4.513501488371294e-05,
      "loss": 0.2979,
      "step": 2440
    },
    {
      "epoch": 1.1316397228637414,
      "grad_norm": 1.8525944948196411,
      "learning_rate": 4.5075122231836575e-05,
      "loss": 0.3316,
      "step": 2450
    },
    {
      "epoch": 1.136258660508083,
      "grad_norm": 2.108273983001709,
      "learning_rate": 4.501490341556354e-05,
      "loss": 0.3246,
      "step": 2460
    },
    {
      "epoch": 1.140877598152425,
      "grad_norm": 1.5220661163330078,
      "learning_rate": 4.495435941328062e-05,
      "loss": 0.3256,
      "step": 2470
    },
    {
      "epoch": 1.1454965357967668,
      "grad_norm": 1.5954303741455078,
      "learning_rate": 4.489349120865792e-05,
      "loss": 0.3325,
      "step": 2480
    },
    {
      "epoch": 1.1501154734411085,
      "grad_norm": 2.012314558029175,
      "learning_rate": 4.483229979063295e-05,
      "loss": 0.328,
      "step": 2490
    },
    {
      "epoch": 1.1547344110854503,
      "grad_norm": 2.3194642066955566,
      "learning_rate": 4.4770786153394526e-05,
      "loss": 0.2955,
      "step": 2500
    },
    {
      "epoch": 1.1593533487297922,
      "grad_norm": 2.1321589946746826,
      "learning_rate": 4.470895129636662e-05,
      "loss": 0.3145,
      "step": 2510
    },
    {
      "epoch": 1.163972286374134,
      "grad_norm": 1.6388169527053833,
      "learning_rate": 4.464679622419211e-05,
      "loss": 0.324,
      "step": 2520
    },
    {
      "epoch": 1.1685912240184757,
      "grad_norm": 1.7595680952072144,
      "learning_rate": 4.45843219467165e-05,
      "loss": 0.3178,
      "step": 2530
    },
    {
      "epoch": 1.1732101616628174,
      "grad_norm": 1.8043133020401,
      "learning_rate": 4.452152947897146e-05,
      "loss": 0.2641,
      "step": 2540
    },
    {
      "epoch": 1.1778290993071594,
      "grad_norm": 2.0495808124542236,
      "learning_rate": 4.445841984115836e-05,
      "loss": 0.3144,
      "step": 2550
    },
    {
      "epoch": 1.1824480369515011,
      "grad_norm": 2.4559502601623535,
      "learning_rate": 4.4394994058631724e-05,
      "loss": 0.2987,
      "step": 2560
    },
    {
      "epoch": 1.1870669745958429,
      "grad_norm": 1.9398882389068604,
      "learning_rate": 4.433125316188249e-05,
      "loss": 0.314,
      "step": 2570
    },
    {
      "epoch": 1.1916859122401848,
      "grad_norm": 2.5720202922821045,
      "learning_rate": 4.426719818652137e-05,
      "loss": 0.3346,
      "step": 2580
    },
    {
      "epoch": 1.1963048498845266,
      "grad_norm": 2.9271321296691895,
      "learning_rate": 4.4202830173261936e-05,
      "loss": 0.3118,
      "step": 2590
    },
    {
      "epoch": 1.2009237875288683,
      "grad_norm": 4.06657600402832,
      "learning_rate": 4.413815016790376e-05,
      "loss": 0.3118,
      "step": 2600
    },
    {
      "epoch": 1.2055427251732103,
      "grad_norm": 1.627180814743042,
      "learning_rate": 4.4073159221315396e-05,
      "loss": 0.3239,
      "step": 2610
    },
    {
      "epoch": 1.210161662817552,
      "grad_norm": 3.904036283493042,
      "learning_rate": 4.4007858389417354e-05,
      "loss": 0.3295,
      "step": 2620
    },
    {
      "epoch": 1.2147806004618937,
      "grad_norm": 1.2671165466308594,
      "learning_rate": 4.3942248733164884e-05,
      "loss": 0.285,
      "step": 2630
    },
    {
      "epoch": 1.2193995381062355,
      "grad_norm": 2.1351537704467773,
      "learning_rate": 4.3876331318530755e-05,
      "loss": 0.315,
      "step": 2640
    },
    {
      "epoch": 1.2240184757505774,
      "grad_norm": 2.152552366256714,
      "learning_rate": 4.381010721648798e-05,
      "loss": 0.3065,
      "step": 2650
    },
    {
      "epoch": 1.2286374133949192,
      "grad_norm": 2.0803706645965576,
      "learning_rate": 4.374357750299236e-05,
      "loss": 0.3419,
      "step": 2660
    },
    {
      "epoch": 1.233256351039261,
      "grad_norm": 1.8164416551589966,
      "learning_rate": 4.3676743258965015e-05,
      "loss": 0.3044,
      "step": 2670
    },
    {
      "epoch": 1.2378752886836029,
      "grad_norm": 1.7581324577331543,
      "learning_rate": 4.3609605570274856e-05,
      "loss": 0.3248,
      "step": 2680
    },
    {
      "epoch": 1.2424942263279446,
      "grad_norm": 2.031179904937744,
      "learning_rate": 4.3542165527720915e-05,
      "loss": 0.3262,
      "step": 2690
    },
    {
      "epoch": 1.2471131639722863,
      "grad_norm": 1.826240062713623,
      "learning_rate": 4.347442422701459e-05,
      "loss": 0.2911,
      "step": 2700
    },
    {
      "epoch": 1.251732101616628,
      "grad_norm": 1.4430904388427734,
      "learning_rate": 4.3406382768761935e-05,
      "loss": 0.305,
      "step": 2710
    },
    {
      "epoch": 1.25635103926097,
      "grad_norm": 1.7261760234832764,
      "learning_rate": 4.333804225844567e-05,
      "loss": 0.3197,
      "step": 2720
    },
    {
      "epoch": 1.2609699769053118,
      "grad_norm": 1.8625311851501465,
      "learning_rate": 4.326940380640731e-05,
      "loss": 0.3265,
      "step": 2730
    },
    {
      "epoch": 1.2655889145496535,
      "grad_norm": 2.4488320350646973,
      "learning_rate": 4.3200468527829055e-05,
      "loss": 0.294,
      "step": 2740
    },
    {
      "epoch": 1.2702078521939955,
      "grad_norm": 1.8369277715682983,
      "learning_rate": 4.313123754271574e-05,
      "loss": 0.3141,
      "step": 2750
    },
    {
      "epoch": 1.2748267898383372,
      "grad_norm": 1.7010807991027832,
      "learning_rate": 4.3061711975876564e-05,
      "loss": 0.3119,
      "step": 2760
    },
    {
      "epoch": 1.279445727482679,
      "grad_norm": 1.5624935626983643,
      "learning_rate": 4.2991892956906865e-05,
      "loss": 0.3199,
      "step": 2770
    },
    {
      "epoch": 1.2840646651270209,
      "grad_norm": 2.6501541137695312,
      "learning_rate": 4.2921781620169776e-05,
      "loss": 0.2817,
      "step": 2780
    },
    {
      "epoch": 1.2886836027713626,
      "grad_norm": 1.5064769983291626,
      "learning_rate": 4.285137910477774e-05,
      "loss": 0.3012,
      "step": 2790
    },
    {
      "epoch": 1.2933025404157044,
      "grad_norm": 1.9986552000045776,
      "learning_rate": 4.278068655457408e-05,
      "loss": 0.2705,
      "step": 2800
    },
    {
      "epoch": 1.2979214780600463,
      "grad_norm": 1.8197659254074097,
      "learning_rate": 4.27097051181143e-05,
      "loss": 0.2814,
      "step": 2810
    },
    {
      "epoch": 1.302540415704388,
      "grad_norm": 1.5221188068389893,
      "learning_rate": 4.2638435948647564e-05,
      "loss": 0.3154,
      "step": 2820
    },
    {
      "epoch": 1.3071593533487298,
      "grad_norm": 1.5869982242584229,
      "learning_rate": 4.256688020409786e-05,
      "loss": 0.2716,
      "step": 2830
    },
    {
      "epoch": 1.3117782909930715,
      "grad_norm": 2.0754973888397217,
      "learning_rate": 4.249503904704522e-05,
      "loss": 0.2901,
      "step": 2840
    },
    {
      "epoch": 1.3163972286374133,
      "grad_norm": 2.163522243499756,
      "learning_rate": 4.242291364470682e-05,
      "loss": 0.2968,
      "step": 2850
    },
    {
      "epoch": 1.3210161662817552,
      "grad_norm": 3.169379472732544,
      "learning_rate": 4.235050516891804e-05,
      "loss": 0.337,
      "step": 2860
    },
    {
      "epoch": 1.325635103926097,
      "grad_norm": 2.4248273372650146,
      "learning_rate": 4.227781479611341e-05,
      "loss": 0.3052,
      "step": 2870
    },
    {
      "epoch": 1.3302540415704387,
      "grad_norm": 2.0661935806274414,
      "learning_rate": 4.220484370730746e-05,
      "loss": 0.2999,
      "step": 2880
    },
    {
      "epoch": 1.3348729792147807,
      "grad_norm": 2.3294215202331543,
      "learning_rate": 4.213159308807563e-05,
      "loss": 0.3077,
      "step": 2890
    },
    {
      "epoch": 1.3394919168591224,
      "grad_norm": 1.8532743453979492,
      "learning_rate": 4.205806412853488e-05,
      "loss": 0.3034,
      "step": 2900
    },
    {
      "epoch": 1.3441108545034641,
      "grad_norm": 1.987816333770752,
      "learning_rate": 4.198425802332446e-05,
      "loss": 0.3049,
      "step": 2910
    },
    {
      "epoch": 1.348729792147806,
      "grad_norm": 2.121181011199951,
      "learning_rate": 4.191017597158645e-05,
      "loss": 0.2842,
      "step": 2920
    },
    {
      "epoch": 1.3533487297921478,
      "grad_norm": 1.72597336769104,
      "learning_rate": 4.183581917694627e-05,
      "loss": 0.2924,
      "step": 2930
    },
    {
      "epoch": 1.3579676674364896,
      "grad_norm": 2.3341033458709717,
      "learning_rate": 4.1761188847493174e-05,
      "loss": 0.3303,
      "step": 2940
    },
    {
      "epoch": 1.3625866050808315,
      "grad_norm": 1.3130851984024048,
      "learning_rate": 4.168628619576056e-05,
      "loss": 0.3022,
      "step": 2950
    },
    {
      "epoch": 1.3672055427251733,
      "grad_norm": 1.6701748371124268,
      "learning_rate": 4.161111243870631e-05,
      "loss": 0.2885,
      "step": 2960
    },
    {
      "epoch": 1.371824480369515,
      "grad_norm": 1.4572288990020752,
      "learning_rate": 4.1535668797693e-05,
      "loss": 0.3114,
      "step": 2970
    },
    {
      "epoch": 1.376443418013857,
      "grad_norm": 1.6978940963745117,
      "learning_rate": 4.145995649846808e-05,
      "loss": 0.2594,
      "step": 2980
    },
    {
      "epoch": 1.3810623556581987,
      "grad_norm": 1.2742643356323242,
      "learning_rate": 4.138397677114393e-05,
      "loss": 0.3137,
      "step": 2990
    },
    {
      "epoch": 1.3856812933025404,
      "grad_norm": 2.4312751293182373,
      "learning_rate": 4.1307730850177865e-05,
      "loss": 0.2967,
      "step": 3000
    },
    {
      "epoch": 1.3903002309468822,
      "grad_norm": 2.715585708618164,
      "learning_rate": 4.123121997435215e-05,
      "loss": 0.3043,
      "step": 3010
    },
    {
      "epoch": 1.394919168591224,
      "grad_norm": 1.99354887008667,
      "learning_rate": 4.115444538675378e-05,
      "loss": 0.2914,
      "step": 3020
    },
    {
      "epoch": 1.3995381062355658,
      "grad_norm": 1.861504316329956,
      "learning_rate": 4.1077408334754335e-05,
      "loss": 0.2637,
      "step": 3030
    },
    {
      "epoch": 1.4041570438799076,
      "grad_norm": 1.8581842184066772,
      "learning_rate": 4.1000110069989724e-05,
      "loss": 0.2907,
      "step": 3040
    },
    {
      "epoch": 1.4087759815242493,
      "grad_norm": 1.6948480606079102,
      "learning_rate": 4.092255184833981e-05,
      "loss": 0.3158,
      "step": 3050
    },
    {
      "epoch": 1.4133949191685913,
      "grad_norm": 2.080018997192383,
      "learning_rate": 4.0844734929908025e-05,
      "loss": 0.3259,
      "step": 3060
    },
    {
      "epoch": 1.418013856812933,
      "grad_norm": 1.1076098680496216,
      "learning_rate": 4.076666057900091e-05,
      "loss": 0.2839,
      "step": 3070
    },
    {
      "epoch": 1.4226327944572748,
      "grad_norm": 2.909213066101074,
      "learning_rate": 4.068833006410757e-05,
      "loss": 0.2865,
      "step": 3080
    },
    {
      "epoch": 1.4272517321016167,
      "grad_norm": 1.6407644748687744,
      "learning_rate": 4.0609744657879024e-05,
      "loss": 0.2718,
      "step": 3090
    },
    {
      "epoch": 1.4318706697459584,
      "grad_norm": 2.367949962615967,
      "learning_rate": 4.0530905637107584e-05,
      "loss": 0.3163,
      "step": 3100
    },
    {
      "epoch": 1.4364896073903002,
      "grad_norm": 1.6912730932235718,
      "learning_rate": 4.0451814282706085e-05,
      "loss": 0.3377,
      "step": 3110
    },
    {
      "epoch": 1.4411085450346421,
      "grad_norm": 1.321198582649231,
      "learning_rate": 4.037247187968706e-05,
      "loss": 0.2846,
      "step": 3120
    },
    {
      "epoch": 1.4457274826789839,
      "grad_norm": 2.0895841121673584,
      "learning_rate": 4.02928797171419e-05,
      "loss": 0.3111,
      "step": 3130
    },
    {
      "epoch": 1.4503464203233256,
      "grad_norm": 1.6102547645568848,
      "learning_rate": 4.021303908821987e-05,
      "loss": 0.2987,
      "step": 3140
    },
    {
      "epoch": 1.4549653579676676,
      "grad_norm": 1.1567573547363281,
      "learning_rate": 4.013295129010712e-05,
      "loss": 0.3045,
      "step": 3150
    },
    {
      "epoch": 1.4595842956120093,
      "grad_norm": 3.8533406257629395,
      "learning_rate": 4.005261762400562e-05,
      "loss": 0.3208,
      "step": 3160
    },
    {
      "epoch": 1.464203233256351,
      "grad_norm": 2.1979851722717285,
      "learning_rate": 3.997203939511198e-05,
      "loss": 0.2706,
      "step": 3170
    },
    {
      "epoch": 1.4688221709006928,
      "grad_norm": 1.7719027996063232,
      "learning_rate": 3.989121791259629e-05,
      "loss": 0.2948,
      "step": 3180
    },
    {
      "epoch": 1.4734411085450345,
      "grad_norm": 2.193924903869629,
      "learning_rate": 3.98101544895808e-05,
      "loss": 0.2977,
      "step": 3190
    },
    {
      "epoch": 1.4780600461893765,
      "grad_norm": 1.6634687185287476,
      "learning_rate": 3.972885044311867e-05,
      "loss": 0.2933,
      "step": 3200
    },
    {
      "epoch": 1.4826789838337182,
      "grad_norm": 1.2648128271102905,
      "learning_rate": 3.964730709417245e-05,
      "loss": 0.2989,
      "step": 3210
    },
    {
      "epoch": 1.48729792147806,
      "grad_norm": 8.130200386047363,
      "learning_rate": 3.956552576759272e-05,
      "loss": 0.3293,
      "step": 3220
    },
    {
      "epoch": 1.491916859122402,
      "grad_norm": 2.254474639892578,
      "learning_rate": 3.948350779209653e-05,
      "loss": 0.3134,
      "step": 3230
    },
    {
      "epoch": 1.4965357967667436,
      "grad_norm": 1.8100476264953613,
      "learning_rate": 3.940125450024581e-05,
      "loss": 0.3405,
      "step": 3240
    },
    {
      "epoch": 1.5011547344110854,
      "grad_norm": 2.5194456577301025,
      "learning_rate": 3.931876722842569e-05,
      "loss": 0.3073,
      "step": 3250
    },
    {
      "epoch": 1.5057736720554273,
      "grad_norm": 2.842329502105713,
      "learning_rate": 3.9236047316822866e-05,
      "loss": 0.2643,
      "step": 3260
    },
    {
      "epoch": 1.510392609699769,
      "grad_norm": 2.350539445877075,
      "learning_rate": 3.9153096109403736e-05,
      "loss": 0.2983,
      "step": 3270
    },
    {
      "epoch": 1.5150115473441108,
      "grad_norm": 2.8748178482055664,
      "learning_rate": 3.906991495389263e-05,
      "loss": 0.2785,
      "step": 3280
    },
    {
      "epoch": 1.5196304849884528,
      "grad_norm": 2.3582661151885986,
      "learning_rate": 3.898650520174989e-05,
      "loss": 0.2953,
      "step": 3290
    },
    {
      "epoch": 1.5242494226327945,
      "grad_norm": 2.257141590118408,
      "learning_rate": 3.890286820814989e-05,
      "loss": 0.278,
      "step": 3300
    },
    {
      "epoch": 1.5288683602771362,
      "grad_norm": 3.344249963760376,
      "learning_rate": 3.881900533195906e-05,
      "loss": 0.2853,
      "step": 3310
    },
    {
      "epoch": 1.5334872979214782,
      "grad_norm": 2.0517385005950928,
      "learning_rate": 3.873491793571378e-05,
      "loss": 0.2794,
      "step": 3320
    },
    {
      "epoch": 1.5381062355658197,
      "grad_norm": 2.145541191101074,
      "learning_rate": 3.8650607385598256e-05,
      "loss": 0.3133,
      "step": 3330
    },
    {
      "epoch": 1.5427251732101617,
      "grad_norm": 1.6390986442565918,
      "learning_rate": 3.856607505142232e-05,
      "loss": 0.2763,
      "step": 3340
    },
    {
      "epoch": 1.5473441108545036,
      "grad_norm": 2.163872241973877,
      "learning_rate": 3.8481322306599154e-05,
      "loss": 0.2951,
      "step": 3350
    },
    {
      "epoch": 1.5519630484988451,
      "grad_norm": 1.493821382522583,
      "learning_rate": 3.839635052812302e-05,
      "loss": 0.2754,
      "step": 3360
    },
    {
      "epoch": 1.556581986143187,
      "grad_norm": 1.9270718097686768,
      "learning_rate": 3.831116109654684e-05,
      "loss": 0.2972,
      "step": 3370
    },
    {
      "epoch": 1.5612009237875288,
      "grad_norm": 2.00677227973938,
      "learning_rate": 3.822575539595981e-05,
      "loss": 0.2637,
      "step": 3380
    },
    {
      "epoch": 1.5658198614318706,
      "grad_norm": 3.3901729583740234,
      "learning_rate": 3.8140134813964866e-05,
      "loss": 0.2946,
      "step": 3390
    },
    {
      "epoch": 1.5704387990762125,
      "grad_norm": 2.659478187561035,
      "learning_rate": 3.8054300741656186e-05,
      "loss": 0.33,
      "step": 3400
    },
    {
      "epoch": 1.5750577367205543,
      "grad_norm": 2.5138628482818604,
      "learning_rate": 3.796825457359655e-05,
      "loss": 0.2866,
      "step": 3410
    },
    {
      "epoch": 1.579676674364896,
      "grad_norm": 1.9975327253341675,
      "learning_rate": 3.7881997707794716e-05,
      "loss": 0.3171,
      "step": 3420
    },
    {
      "epoch": 1.584295612009238,
      "grad_norm": 1.2960349321365356,
      "learning_rate": 3.779553154568266e-05,
      "loss": 0.2952,
      "step": 3430
    },
    {
      "epoch": 1.5889145496535797,
      "grad_norm": 1.4498865604400635,
      "learning_rate": 3.7708857492092874e-05,
      "loss": 0.2759,
      "step": 3440
    },
    {
      "epoch": 1.5935334872979214,
      "grad_norm": 1.7237732410430908,
      "learning_rate": 3.762197695523547e-05,
      "loss": 0.2781,
      "step": 3450
    },
    {
      "epoch": 1.5981524249422634,
      "grad_norm": 1.2863229513168335,
      "learning_rate": 3.7534891346675367e-05,
      "loss": 0.2587,
      "step": 3460
    },
    {
      "epoch": 1.6027713625866051,
      "grad_norm": 2.125469446182251,
      "learning_rate": 3.744760208130928e-05,
      "loss": 0.3124,
      "step": 3470
    },
    {
      "epoch": 1.6073903002309469,
      "grad_norm": 1.4202874898910522,
      "learning_rate": 3.73601105773428e-05,
      "loss": 0.298,
      "step": 3480
    },
    {
      "epoch": 1.6120092378752888,
      "grad_norm": 2.05185604095459,
      "learning_rate": 3.7272418256267347e-05,
      "loss": 0.2934,
      "step": 3490
    },
    {
      "epoch": 1.6166281755196303,
      "grad_norm": 1.3896102905273438,
      "learning_rate": 3.7184526542837016e-05,
      "loss": 0.2804,
      "step": 3500
    }
  ],
  "logging_steps": 10,
  "max_steps": 8660,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.605425304646451e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}