{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.956521739130435,
  "eval_steps": 500,
  "global_step": 2180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004576659038901602,
      "grad_norm": 0.795125424861908,
      "learning_rate": 9.174311926605506e-07,
      "loss": 2.0857,
      "step": 1
    },
    {
      "epoch": 0.02288329519450801,
      "grad_norm": 0.7591376900672913,
      "learning_rate": 4.587155963302753e-06,
      "loss": 2.0888,
      "step": 5
    },
    {
      "epoch": 0.04576659038901602,
      "grad_norm": 0.7638061046600342,
      "learning_rate": 9.174311926605506e-06,
      "loss": 2.0744,
      "step": 10
    },
    {
      "epoch": 0.06864988558352403,
      "grad_norm": 0.6205304265022278,
      "learning_rate": 1.3761467889908258e-05,
      "loss": 2.0578,
      "step": 15
    },
    {
      "epoch": 0.09153318077803203,
      "grad_norm": 0.40848588943481445,
      "learning_rate": 1.834862385321101e-05,
      "loss": 2.0541,
      "step": 20
    },
    {
      "epoch": 0.11441647597254005,
      "grad_norm": 0.35839515924453735,
      "learning_rate": 2.2935779816513765e-05,
      "loss": 2.0396,
      "step": 25
    },
    {
      "epoch": 0.13729977116704806,
      "grad_norm": 0.3498051166534424,
      "learning_rate": 2.7522935779816515e-05,
      "loss": 2.0186,
      "step": 30
    },
    {
      "epoch": 0.16018306636155608,
      "grad_norm": 0.34915778040885925,
      "learning_rate": 3.211009174311927e-05,
      "loss": 1.9692,
      "step": 35
    },
    {
      "epoch": 0.18306636155606407,
      "grad_norm": 0.2829169034957886,
      "learning_rate": 3.669724770642202e-05,
      "loss": 1.9478,
      "step": 40
    },
    {
      "epoch": 0.20594965675057209,
      "grad_norm": 0.29471325874328613,
      "learning_rate": 4.1284403669724776e-05,
      "loss": 1.9081,
      "step": 45
    },
    {
      "epoch": 0.2288329519450801,
      "grad_norm": 0.26286041736602783,
      "learning_rate": 4.587155963302753e-05,
      "loss": 1.8852,
      "step": 50
    },
    {
      "epoch": 0.2517162471395881,
      "grad_norm": 0.253055214881897,
      "learning_rate": 5.0458715596330276e-05,
      "loss": 1.8466,
      "step": 55
    },
    {
      "epoch": 0.2745995423340961,
      "grad_norm": 0.25780022144317627,
      "learning_rate": 5.504587155963303e-05,
      "loss": 1.7978,
      "step": 60
    },
    {
      "epoch": 0.2974828375286041,
      "grad_norm": 0.22376658022403717,
      "learning_rate": 5.9633027522935784e-05,
      "loss": 1.7513,
      "step": 65
    },
    {
      "epoch": 0.32036613272311215,
      "grad_norm": 0.20868130028247833,
      "learning_rate": 6.422018348623854e-05,
      "loss": 1.7581,
      "step": 70
    },
    {
      "epoch": 0.34324942791762014,
      "grad_norm": 0.20019210875034332,
      "learning_rate": 6.880733944954129e-05,
      "loss": 1.7225,
      "step": 75
    },
    {
      "epoch": 0.36613272311212813,
      "grad_norm": 0.20799387991428375,
      "learning_rate": 7.339449541284404e-05,
      "loss": 1.7073,
      "step": 80
    },
    {
      "epoch": 0.3890160183066362,
      "grad_norm": 0.19921623170375824,
      "learning_rate": 7.79816513761468e-05,
      "loss": 1.698,
      "step": 85
    },
    {
      "epoch": 0.41189931350114417,
      "grad_norm": 0.19697250425815582,
      "learning_rate": 8.256880733944955e-05,
      "loss": 1.6806,
      "step": 90
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.21829445660114288,
      "learning_rate": 8.715596330275229e-05,
      "loss": 1.661,
      "step": 95
    },
    {
      "epoch": 0.4576659038901602,
      "grad_norm": 0.20831452310085297,
      "learning_rate": 9.174311926605506e-05,
      "loss": 1.6499,
      "step": 100
    },
    {
      "epoch": 0.4805491990846682,
      "grad_norm": 0.2224440723657608,
      "learning_rate": 9.63302752293578e-05,
      "loss": 1.6504,
      "step": 105
    },
    {
      "epoch": 0.5034324942791762,
      "grad_norm": 0.22193986177444458,
      "learning_rate": 0.00010091743119266055,
      "loss": 1.6399,
      "step": 110
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.2367602288722992,
      "learning_rate": 0.00010550458715596329,
      "loss": 1.6236,
      "step": 115
    },
    {
      "epoch": 0.5491990846681922,
      "grad_norm": 0.2362162321805954,
      "learning_rate": 0.00011009174311926606,
      "loss": 1.6227,
      "step": 120
    },
    {
      "epoch": 0.5720823798627003,
      "grad_norm": 0.24902412295341492,
      "learning_rate": 0.00011467889908256881,
      "loss": 1.5984,
      "step": 125
    },
    {
      "epoch": 0.5949656750572082,
      "grad_norm": 0.2517242431640625,
      "learning_rate": 0.00011926605504587157,
      "loss": 1.6003,
      "step": 130
    },
    {
      "epoch": 0.6178489702517163,
      "grad_norm": 0.2526886463165283,
      "learning_rate": 0.00012385321100917432,
      "loss": 1.5961,
      "step": 135
    },
    {
      "epoch": 0.6407322654462243,
      "grad_norm": 0.2544921338558197,
      "learning_rate": 0.00012844036697247707,
      "loss": 1.5782,
      "step": 140
    },
    {
      "epoch": 0.6636155606407322,
      "grad_norm": 0.2370745986700058,
      "learning_rate": 0.00013302752293577983,
      "loss": 1.5817,
      "step": 145
    },
    {
      "epoch": 0.6864988558352403,
      "grad_norm": 0.2654547095298767,
      "learning_rate": 0.00013761467889908258,
      "loss": 1.5683,
      "step": 150
    },
    {
      "epoch": 0.7093821510297483,
      "grad_norm": 0.251737505197525,
      "learning_rate": 0.0001422018348623853,
      "loss": 1.5683,
      "step": 155
    },
    {
      "epoch": 0.7322654462242563,
      "grad_norm": 0.27123987674713135,
      "learning_rate": 0.0001467889908256881,
      "loss": 1.5492,
      "step": 160
    },
    {
      "epoch": 0.7551487414187643,
      "grad_norm": 0.2787420451641083,
      "learning_rate": 0.00015137614678899084,
      "loss": 1.5573,
      "step": 165
    },
    {
      "epoch": 0.7780320366132724,
      "grad_norm": 0.3056250810623169,
      "learning_rate": 0.0001559633027522936,
      "loss": 1.5535,
      "step": 170
    },
    {
      "epoch": 0.8009153318077803,
      "grad_norm": 0.4223385751247406,
      "learning_rate": 0.00016055045871559632,
      "loss": 1.5624,
      "step": 175
    },
    {
      "epoch": 0.8237986270022883,
      "grad_norm": 0.3226919174194336,
      "learning_rate": 0.0001651376146788991,
      "loss": 1.5548,
      "step": 180
    },
    {
      "epoch": 0.8466819221967964,
      "grad_norm": 0.42250385880470276,
      "learning_rate": 0.00016972477064220186,
      "loss": 1.5403,
      "step": 185
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.29832813143730164,
      "learning_rate": 0.00017431192660550458,
      "loss": 1.5406,
      "step": 190
    },
    {
      "epoch": 0.8924485125858124,
      "grad_norm": 0.318697988986969,
      "learning_rate": 0.00017889908256880734,
      "loss": 1.5423,
      "step": 195
    },
    {
      "epoch": 0.9153318077803204,
      "grad_norm": 0.30288153886795044,
      "learning_rate": 0.00018348623853211012,
      "loss": 1.5432,
      "step": 200
    },
    {
      "epoch": 0.9382151029748284,
      "grad_norm": 0.26959261298179626,
      "learning_rate": 0.00018807339449541284,
      "loss": 1.5264,
      "step": 205
    },
    {
      "epoch": 0.9610983981693364,
      "grad_norm": 0.27854207158088684,
      "learning_rate": 0.0001926605504587156,
      "loss": 1.5328,
      "step": 210
    },
    {
      "epoch": 0.9839816933638444,
      "grad_norm": 0.26367124915122986,
      "learning_rate": 0.00019724770642201835,
      "loss": 1.5142,
      "step": 215
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.4457786083221436,
      "eval_runtime": 0.158,
      "eval_samples_per_second": 63.273,
      "eval_steps_per_second": 6.327,
      "step": 219
    },
    {
      "epoch": 1.0045766590389016,
      "grad_norm": 0.24696815013885498,
      "learning_rate": 0.00019999948721966259,
      "loss": 1.5267,
      "step": 220
    },
    {
      "epoch": 1.0274599542334095,
      "grad_norm": 0.2515549063682556,
      "learning_rate": 0.0001999937185012612,
      "loss": 1.5197,
      "step": 225
    },
    {
      "epoch": 1.0503432494279177,
      "grad_norm": 0.26250341534614563,
      "learning_rate": 0.00019998154046002822,
      "loss": 1.4932,
      "step": 230
    },
    {
      "epoch": 1.0732265446224256,
      "grad_norm": 0.2599635422229767,
      "learning_rate": 0.00019996295387654262,
      "loss": 1.5057,
      "step": 235
    },
    {
      "epoch": 1.0961098398169336,
      "grad_norm": 0.2875217795372009,
      "learning_rate": 0.0001999379599421534,
      "loss": 1.5123,
      "step": 240
    },
    {
      "epoch": 1.1189931350114417,
      "grad_norm": 0.2571655511856079,
      "learning_rate": 0.00019990656025890315,
      "loss": 1.5125,
      "step": 245
    },
    {
      "epoch": 1.1418764302059496,
      "grad_norm": 0.2501162588596344,
      "learning_rate": 0.00019986875683942535,
      "loss": 1.5125,
      "step": 250
    },
    {
      "epoch": 1.1647597254004576,
      "grad_norm": 0.2591252028942108,
      "learning_rate": 0.00019982455210681537,
      "loss": 1.491,
      "step": 255
    },
    {
      "epoch": 1.1876430205949657,
      "grad_norm": 0.2666991055011749,
      "learning_rate": 0.00019977394889447524,
      "loss": 1.4931,
      "step": 260
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.3265208601951599,
      "learning_rate": 0.00019971695044593196,
      "loss": 1.5088,
      "step": 265
    },
    {
      "epoch": 1.2334096109839816,
      "grad_norm": 0.3080615699291229,
      "learning_rate": 0.00019965356041462955,
      "loss": 1.4943,
      "step": 270
    },
    {
      "epoch": 1.2562929061784898,
      "grad_norm": 0.2847588062286377,
      "learning_rate": 0.00019958378286369502,
      "loss": 1.4829,
      "step": 275
    },
    {
      "epoch": 1.2791762013729977,
      "grad_norm": 0.2870901823043823,
      "learning_rate": 0.00019950762226567781,
      "loss": 1.4841,
      "step": 280
    },
    {
      "epoch": 1.3020594965675056,
      "grad_norm": 0.27676528692245483,
      "learning_rate": 0.00019942508350226314,
      "loss": 1.4901,
      "step": 285
    },
    {
      "epoch": 1.3249427917620138,
      "grad_norm": 0.26728203892707825,
      "learning_rate": 0.00019933617186395917,
      "loss": 1.4743,
      "step": 290
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 0.2857504189014435,
      "learning_rate": 0.0001992408930497578,
      "loss": 1.4919,
      "step": 295
    },
    {
      "epoch": 1.3707093821510297,
      "grad_norm": 0.36729517579078674,
      "learning_rate": 0.00019913925316676945,
      "loss": 1.4872,
      "step": 300
    },
    {
      "epoch": 1.3935926773455378,
      "grad_norm": 0.2691424489021301,
      "learning_rate": 0.0001990312587298316,
      "loss": 1.4785,
      "step": 305
    },
    {
      "epoch": 1.4164759725400458,
      "grad_norm": 0.3448796570301056,
      "learning_rate": 0.00019891691666109113,
      "loss": 1.4857,
      "step": 310
    },
    {
      "epoch": 1.4393592677345537,
      "grad_norm": 0.29720258712768555,
      "learning_rate": 0.0001987962342895607,
      "loss": 1.4844,
      "step": 315
    },
    {
      "epoch": 1.4622425629290619,
      "grad_norm": 0.25346481800079346,
      "learning_rate": 0.00019866921935064906,
      "loss": 1.4822,
      "step": 320
    },
    {
      "epoch": 1.4851258581235698,
      "grad_norm": 0.2484666258096695,
      "learning_rate": 0.0001985358799856651,
      "loss": 1.4804,
      "step": 325
    },
    {
      "epoch": 1.5080091533180777,
      "grad_norm": 0.2337537705898285,
      "learning_rate": 0.00019839622474129596,
      "loss": 1.4762,
      "step": 330
    },
    {
      "epoch": 1.5308924485125859,
      "grad_norm": 0.26846134662628174,
      "learning_rate": 0.00019825026256905948,
      "loss": 1.4656,
      "step": 335
    },
    {
      "epoch": 1.5537757437070938,
      "grad_norm": 0.239125594496727,
      "learning_rate": 0.00019809800282473013,
      "loss": 1.4802,
      "step": 340
    },
    {
      "epoch": 1.5766590389016018,
      "grad_norm": 0.2293742597103119,
      "learning_rate": 0.00019793945526773947,
      "loss": 1.4603,
      "step": 345
    },
    {
      "epoch": 1.59954233409611,
      "grad_norm": 0.2655136287212372,
      "learning_rate": 0.0001977746300605507,
      "loss": 1.4783,
      "step": 350
    },
    {
      "epoch": 1.6224256292906178,
      "grad_norm": 0.25485122203826904,
      "learning_rate": 0.00019760353776800704,
      "loss": 1.4755,
      "step": 355
    },
    {
      "epoch": 1.6453089244851258,
      "grad_norm": 0.2485138326883316,
      "learning_rate": 0.00019742618935665476,
      "loss": 1.464,
      "step": 360
    },
    {
      "epoch": 1.668192219679634,
      "grad_norm": 0.23952895402908325,
      "learning_rate": 0.0001972425961940401,
      "loss": 1.4652,
      "step": 365
    },
    {
      "epoch": 1.6910755148741419,
      "grad_norm": 0.22710928320884705,
      "learning_rate": 0.00019705277004798073,
      "loss": 1.4585,
      "step": 370
    },
    {
      "epoch": 1.7139588100686498,
      "grad_norm": 0.2363155037164688,
      "learning_rate": 0.00019685672308581152,
      "loss": 1.4725,
      "step": 375
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.23678353428840637,
      "learning_rate": 0.0001966544678736044,
      "loss": 1.4654,
      "step": 380
    },
    {
      "epoch": 1.759725400457666,
      "grad_norm": 0.24644772708415985,
      "learning_rate": 0.00019644601737536338,
      "loss": 1.4473,
      "step": 385
    },
    {
      "epoch": 1.7826086956521738,
      "grad_norm": 0.24988573789596558,
      "learning_rate": 0.00019623138495219292,
      "loss": 1.4658,
      "step": 390
    },
    {
      "epoch": 1.805491990846682,
      "grad_norm": 0.2501707971096039,
      "learning_rate": 0.00019601058436144225,
      "loss": 1.4631,
      "step": 395
    },
    {
      "epoch": 1.82837528604119,
      "grad_norm": 0.22899797558784485,
      "learning_rate": 0.00019578362975582292,
      "loss": 1.4663,
      "step": 400
    },
    {
      "epoch": 1.8512585812356979,
      "grad_norm": 0.2957789897918701,
      "learning_rate": 0.0001955505356825021,
      "loss": 1.4649,
      "step": 405
    },
    {
      "epoch": 1.874141876430206,
      "grad_norm": 0.26199233531951904,
      "learning_rate": 0.00019531131708217005,
      "loss": 1.4568,
      "step": 410
    },
    {
      "epoch": 1.897025171624714,
      "grad_norm": 0.2635655403137207,
      "learning_rate": 0.00019506598928808216,
      "loss": 1.4748,
      "step": 415
    },
    {
      "epoch": 1.919908466819222,
      "grad_norm": 0.23314692080020905,
      "learning_rate": 0.0001948145680250766,
      "loss": 1.4615,
      "step": 420
    },
    {
      "epoch": 1.94279176201373,
      "grad_norm": 0.24421729147434235,
      "learning_rate": 0.000194557069408566,
      "loss": 1.463,
      "step": 425
    },
    {
      "epoch": 1.965675057208238,
      "grad_norm": 0.2970636188983917,
      "learning_rate": 0.00019429350994350483,
      "loss": 1.4702,
      "step": 430
    },
    {
      "epoch": 1.988558352402746,
      "grad_norm": 0.2254825085401535,
      "learning_rate": 0.0001940239065233311,
      "loss": 1.4574,
      "step": 435
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.456493854522705,
      "eval_runtime": 0.1591,
      "eval_samples_per_second": 62.851,
      "eval_steps_per_second": 6.285,
      "step": 438
    },
    {
      "epoch": 2.009153318077803,
      "grad_norm": 0.2410830557346344,
      "learning_rate": 0.00019374827642888398,
      "loss": 1.4467,
      "step": 440
    },
    {
      "epoch": 2.0320366132723113,
      "grad_norm": 0.3021443784236908,
      "learning_rate": 0.00019346663732729572,
      "loss": 1.4415,
      "step": 445
    },
    {
      "epoch": 2.054919908466819,
      "grad_norm": 0.2840650975704193,
      "learning_rate": 0.0001931790072708596,
      "loss": 1.4438,
      "step": 450
    },
    {
      "epoch": 2.077803203661327,
      "grad_norm": 0.25511375069618225,
      "learning_rate": 0.0001928854046958725,
      "loss": 1.4412,
      "step": 455
    },
    {
      "epoch": 2.1006864988558354,
      "grad_norm": 0.25664883852005005,
      "learning_rate": 0.00019258584842145343,
      "loss": 1.4476,
      "step": 460
    },
    {
      "epoch": 2.123569794050343,
      "grad_norm": 0.29023540019989014,
      "learning_rate": 0.00019228035764833718,
      "loss": 1.4393,
      "step": 465
    },
    {
      "epoch": 2.1464530892448512,
      "grad_norm": 0.23540236055850983,
      "learning_rate": 0.00019196895195764362,
      "loss": 1.45,
      "step": 470
    },
    {
      "epoch": 2.1693363844393594,
      "grad_norm": 0.23571573197841644,
      "learning_rate": 0.0001916516513096226,
      "loss": 1.4292,
      "step": 475
    },
    {
      "epoch": 2.192219679633867,
      "grad_norm": 0.22221675515174866,
      "learning_rate": 0.0001913284760423745,
      "loss": 1.4385,
      "step": 480
    },
    {
      "epoch": 2.2151029748283753,
      "grad_norm": 0.24515533447265625,
      "learning_rate": 0.00019099944687054672,
      "loss": 1.438,
      "step": 485
    },
    {
      "epoch": 2.2379862700228834,
      "grad_norm": 0.22849489748477936,
      "learning_rate": 0.00019066458488400584,
      "loss": 1.4365,
      "step": 490
    },
    {
      "epoch": 2.260869565217391,
      "grad_norm": 0.23650361597537994,
      "learning_rate": 0.0001903239115464859,
      "loss": 1.4421,
      "step": 495
    },
    {
      "epoch": 2.2837528604118993,
      "grad_norm": 0.2418275624513626,
      "learning_rate": 0.00018997744869421246,
      "loss": 1.4275,
      "step": 500
    },
    {
      "epoch": 2.3066361556064074,
      "grad_norm": 0.2576926350593567,
      "learning_rate": 0.00018962521853450323,
      "loss": 1.4315,
      "step": 505
    },
    {
      "epoch": 2.329519450800915,
      "grad_norm": 0.26270532608032227,
      "learning_rate": 0.00018926724364434446,
      "loss": 1.4386,
      "step": 510
    },
    {
      "epoch": 2.3524027459954233,
      "grad_norm": 0.2584366798400879,
      "learning_rate": 0.00018890354696894375,
      "loss": 1.4288,
      "step": 515
    },
    {
      "epoch": 2.3752860411899315,
      "grad_norm": 0.2276887744665146,
      "learning_rate": 0.0001885341518202595,
      "loss": 1.4331,
      "step": 520
    },
    {
      "epoch": 2.398169336384439,
      "grad_norm": 0.25574877858161926,
      "learning_rate": 0.00018815908187550667,
      "loss": 1.4343,
      "step": 525
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.2539375424385071,
      "learning_rate": 0.00018777836117563892,
      "loss": 1.4283,
      "step": 530
    },
    {
      "epoch": 2.4439359267734555,
      "grad_norm": 0.23258014023303986,
      "learning_rate": 0.0001873920141238079,
      "loss": 1.4333,
      "step": 535
    },
    {
      "epoch": 2.466819221967963,
      "grad_norm": 0.24933487176895142,
      "learning_rate": 0.00018700006548379898,
      "loss": 1.4245,
      "step": 540
    },
    {
      "epoch": 2.4897025171624714,
      "grad_norm": 0.2398696094751358,
      "learning_rate": 0.00018660254037844388,
      "loss": 1.4373,
      "step": 545
    },
    {
      "epoch": 2.5125858123569795,
      "grad_norm": 0.2470606118440628,
      "learning_rate": 0.0001861994642880105,
      "loss": 1.4316,
      "step": 550
    },
    {
      "epoch": 2.5354691075514877,
      "grad_norm": 0.2763492166996002,
      "learning_rate": 0.0001857908630485696,
      "loss": 1.4229,
      "step": 555
    },
    {
      "epoch": 2.5583524027459954,
      "grad_norm": 0.25183382630348206,
      "learning_rate": 0.00018537676285033887,
      "loss": 1.4235,
      "step": 560
    },
    {
      "epoch": 2.5812356979405036,
      "grad_norm": 0.25710180401802063,
      "learning_rate": 0.00018495719023600414,
      "loss": 1.4203,
      "step": 565
    },
    {
      "epoch": 2.6041189931350113,
      "grad_norm": 0.22346384823322296,
      "learning_rate": 0.0001845321720990181,
      "loss": 1.4274,
      "step": 570
    },
    {
      "epoch": 2.6270022883295194,
      "grad_norm": 0.2290782779455185,
      "learning_rate": 0.00018410173568187647,
      "loss": 1.4304,
      "step": 575
    },
    {
      "epoch": 2.6498855835240276,
      "grad_norm": 0.25173699855804443,
      "learning_rate": 0.00018366590857437184,
      "loss": 1.4331,
      "step": 580
    },
    {
      "epoch": 2.6727688787185357,
      "grad_norm": 0.2373346984386444,
      "learning_rate": 0.00018322471871182528,
      "loss": 1.4257,
      "step": 585
    },
    {
      "epoch": 2.6956521739130435,
      "grad_norm": 0.2577199339866638,
      "learning_rate": 0.00018277819437329576,
      "loss": 1.4382,
      "step": 590
    },
    {
      "epoch": 2.7185354691075516,
      "grad_norm": 0.23805540800094604,
      "learning_rate": 0.00018232636417976744,
      "loss": 1.4365,
      "step": 595
    },
    {
      "epoch": 2.7414187643020593,
      "grad_norm": 0.25108346343040466,
      "learning_rate": 0.00018186925709231532,
      "loss": 1.424,
      "step": 600
    },
    {
      "epoch": 2.7643020594965675,
      "grad_norm": 0.25178954005241394,
      "learning_rate": 0.00018140690241024872,
      "loss": 1.429,
      "step": 605
    },
    {
      "epoch": 2.7871853546910756,
      "grad_norm": 0.2157055288553238,
      "learning_rate": 0.0001809393297692334,
      "loss": 1.4257,
      "step": 610
    },
    {
      "epoch": 2.8100686498855834,
      "grad_norm": 0.22406814992427826,
      "learning_rate": 0.00018046656913939195,
      "loss": 1.4248,
      "step": 615
    },
    {
      "epoch": 2.8329519450800915,
      "grad_norm": 0.22442927956581116,
      "learning_rate": 0.0001799886508233829,
      "loss": 1.4292,
      "step": 620
    },
    {
      "epoch": 2.8558352402745997,
      "grad_norm": 0.22204960882663727,
      "learning_rate": 0.00017950560545445813,
      "loss": 1.4399,
      "step": 625
    },
    {
      "epoch": 2.8787185354691074,
      "grad_norm": 0.2269710898399353,
      "learning_rate": 0.0001790174639944997,
      "loss": 1.4248,
      "step": 630
    },
    {
      "epoch": 2.9016018306636155,
      "grad_norm": 0.23222693800926208,
      "learning_rate": 0.000178524257732035,
      "loss": 1.4261,
      "step": 635
    },
    {
      "epoch": 2.9244851258581237,
      "grad_norm": 0.21854890882968903,
      "learning_rate": 0.00017802601828023138,
      "loss": 1.4172,
      "step": 640
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.2316950112581253,
      "learning_rate": 0.0001775227775748699,
      "loss": 1.4156,
      "step": 645
    },
    {
      "epoch": 2.9702517162471396,
      "grad_norm": 0.21298670768737793,
      "learning_rate": 0.00017701456787229804,
      "loss": 1.4288,
      "step": 650
    },
    {
      "epoch": 2.9931350114416477,
      "grad_norm": 0.21937227249145508,
      "learning_rate": 0.00017650142174736262,
      "loss": 1.4298,
      "step": 655
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.4537367820739746,
      "eval_runtime": 0.1693,
      "eval_samples_per_second": 59.08,
      "eval_steps_per_second": 5.908,
      "step": 657
    },
    {
      "epoch": 3.013729977116705,
      "grad_norm": 0.22709442675113678,
      "learning_rate": 0.0001759833720913214,
      "loss": 1.4058,
      "step": 660
    },
    {
      "epoch": 3.0366132723112127,
      "grad_norm": 0.21663448214530945,
      "learning_rate": 0.00017546045210973507,
      "loss": 1.4108,
      "step": 665
    },
    {
      "epoch": 3.059496567505721,
      "grad_norm": 0.21900029480457306,
      "learning_rate": 0.00017493269532033883,
      "loss": 1.41,
      "step": 670
    },
    {
      "epoch": 3.082379862700229,
      "grad_norm": 0.23556417226791382,
      "learning_rate": 0.00017440013555089393,
      "loss": 1.4084,
      "step": 675
    },
    {
      "epoch": 3.1052631578947367,
      "grad_norm": 0.26198992133140564,
      "learning_rate": 0.0001738628069370195,
      "loss": 1.4108,
      "step": 680
    },
    {
      "epoch": 3.128146453089245,
      "grad_norm": 0.2543962299823761,
      "learning_rate": 0.00017332074392000454,
      "loss": 1.4124,
      "step": 685
    },
    {
      "epoch": 3.151029748283753,
      "grad_norm": 0.23739871382713318,
      "learning_rate": 0.00017277398124460023,
      "loss": 1.4048,
      "step": 690
    },
    {
      "epoch": 3.1739130434782608,
      "grad_norm": 0.23524770140647888,
      "learning_rate": 0.00017222255395679296,
      "loss": 1.4184,
      "step": 695
    },
    {
      "epoch": 3.196796338672769,
      "grad_norm": 0.2273031771183014,
      "learning_rate": 0.000171666497401558,
      "loss": 1.4122,
      "step": 700
    },
    {
      "epoch": 3.219679633867277,
      "grad_norm": 0.23595724999904633,
      "learning_rate": 0.00017110584722059393,
      "loss": 1.407,
      "step": 705
    },
    {
      "epoch": 3.242562929061785,
      "grad_norm": 0.21827439963817596,
      "learning_rate": 0.0001705406393500381,
      "loss": 1.4027,
      "step": 710
    },
    {
      "epoch": 3.265446224256293,
      "grad_norm": 0.22116319835186005,
      "learning_rate": 0.00016997091001816336,
      "loss": 1.4049,
      "step": 715
    },
    {
      "epoch": 3.288329519450801,
      "grad_norm": 0.23044751584529877,
      "learning_rate": 0.00016939669574305566,
      "loss": 1.4095,
      "step": 720
    },
    {
      "epoch": 3.311212814645309,
      "grad_norm": 0.22559754550457,
      "learning_rate": 0.00016881803333027362,
      "loss": 1.3991,
      "step": 725
    },
    {
      "epoch": 3.334096109839817,
      "grad_norm": 0.2260756939649582,
      "learning_rate": 0.0001682349598704892,
      "loss": 1.4123,
      "step": 730
    },
    {
      "epoch": 3.356979405034325,
      "grad_norm": 0.2145521193742752,
      "learning_rate": 0.00016764751273711044,
      "loss": 1.4119,
      "step": 735
    },
    {
      "epoch": 3.379862700228833,
      "grad_norm": 0.23654837906360626,
      "learning_rate": 0.00016705572958388576,
      "loss": 1.4024,
      "step": 740
    },
    {
      "epoch": 3.402745995423341,
      "grad_norm": 0.21862386167049408,
      "learning_rate": 0.0001664596483424906,
      "loss": 1.4104,
      "step": 745
    },
    {
      "epoch": 3.425629290617849,
      "grad_norm": 0.2240535169839859,
      "learning_rate": 0.00016585930722009601,
      "loss": 1.4012,
      "step": 750
    },
    {
      "epoch": 3.448512585812357,
      "grad_norm": 0.23204627633094788,
      "learning_rate": 0.00016525474469691984,
      "loss": 1.4044,
      "step": 755
    },
    {
      "epoch": 3.471395881006865,
      "grad_norm": 0.23645585775375366,
      "learning_rate": 0.00016464599952375998,
      "loss": 1.4135,
      "step": 760
    },
    {
      "epoch": 3.494279176201373,
      "grad_norm": 0.2199520468711853,
      "learning_rate": 0.00016403311071951082,
      "loss": 1.413,
      "step": 765
    },
    {
      "epoch": 3.517162471395881,
      "grad_norm": 0.23638251423835754,
      "learning_rate": 0.000163416117568662,
      "loss": 1.4101,
      "step": 770
    },
    {
      "epoch": 3.540045766590389,
      "grad_norm": 0.2419108748435974,
      "learning_rate": 0.00016279505961878064,
      "loss": 1.399,
      "step": 775
    },
    {
      "epoch": 3.5629290617848968,
      "grad_norm": 0.23601311445236206,
      "learning_rate": 0.0001621699766779763,
      "loss": 1.4084,
      "step": 780
    },
    {
      "epoch": 3.585812356979405,
      "grad_norm": 0.24460086226463318,
      "learning_rate": 0.0001615409088123493,
      "loss": 1.4144,
      "step": 785
    },
    {
      "epoch": 3.608695652173913,
      "grad_norm": 0.21848763525485992,
      "learning_rate": 0.00016090789634342278,
      "loss": 1.3929,
      "step": 790
    },
    {
      "epoch": 3.6315789473684212,
      "grad_norm": 0.22614765167236328,
      "learning_rate": 0.00016027097984555816,
      "loss": 1.399,
      "step": 795
    },
    {
      "epoch": 3.654462242562929,
      "grad_norm": 0.22731591761112213,
      "learning_rate": 0.00015963020014335438,
      "loss": 1.3998,
      "step": 800
    },
    {
      "epoch": 3.677345537757437,
      "grad_norm": 0.21770906448364258,
      "learning_rate": 0.00015898559830903106,
      "loss": 1.4064,
      "step": 805
    },
    {
      "epoch": 3.700228832951945,
      "grad_norm": 0.25158295035362244,
      "learning_rate": 0.0001583372156597961,
      "loss": 1.4046,
      "step": 810
    },
    {
      "epoch": 3.723112128146453,
      "grad_norm": 0.23650844395160675,
      "learning_rate": 0.00015768509375519726,
      "loss": 1.3942,
      "step": 815
    },
    {
      "epoch": 3.745995423340961,
      "grad_norm": 0.22743116319179535,
      "learning_rate": 0.00015702927439445826,
      "loss": 1.4022,
      "step": 820
    },
    {
      "epoch": 3.7688787185354693,
      "grad_norm": 0.23518233001232147,
      "learning_rate": 0.0001563697996137997,
      "loss": 1.3986,
      "step": 825
    },
    {
      "epoch": 3.791762013729977,
      "grad_norm": 0.22372908890247345,
      "learning_rate": 0.00015570671168374438,
      "loss": 1.3937,
      "step": 830
    },
    {
      "epoch": 3.814645308924485,
      "grad_norm": 0.2474096119403839,
      "learning_rate": 0.00015504005310640822,
      "loss": 1.3984,
      "step": 835
    },
    {
      "epoch": 3.837528604118993,
      "grad_norm": 0.23158776760101318,
      "learning_rate": 0.00015436986661277577,
      "loss": 1.3972,
      "step": 840
    },
    {
      "epoch": 3.860411899313501,
      "grad_norm": 0.239366814494133,
      "learning_rate": 0.0001536961951599613,
      "loss": 1.4023,
      "step": 845
    },
    {
      "epoch": 3.883295194508009,
      "grad_norm": 0.26158976554870605,
      "learning_rate": 0.0001530190819284555,
      "loss": 1.3918,
      "step": 850
    },
    {
      "epoch": 3.9061784897025174,
      "grad_norm": 0.2554950416088104,
      "learning_rate": 0.00015233857031935749,
      "loss": 1.3927,
      "step": 855
    },
    {
      "epoch": 3.929061784897025,
      "grad_norm": 0.23338022828102112,
      "learning_rate": 0.00015165470395159313,
      "loss": 1.3929,
      "step": 860
    },
    {
      "epoch": 3.9519450800915332,
      "grad_norm": 0.2260856032371521,
      "learning_rate": 0.00015096752665911913,
      "loss": 1.4055,
      "step": 865
    },
    {
      "epoch": 3.974828375286041,
      "grad_norm": 0.2310597449541092,
      "learning_rate": 0.0001502770824881133,
      "loss": 1.3984,
      "step": 870
    },
    {
      "epoch": 3.997711670480549,
      "grad_norm": 0.22370661795139313,
      "learning_rate": 0.00014958341569415147,
      "loss": 1.3972,
      "step": 875
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.462189197540283,
      "eval_runtime": 0.1584,
      "eval_samples_per_second": 63.121,
      "eval_steps_per_second": 6.312,
      "step": 876
    },
    {
      "epoch": 4.018306636155606,
      "grad_norm": 0.24240370094776154,
      "learning_rate": 0.00014888657073937076,
      "loss": 1.3961,
      "step": 880
    },
    {
      "epoch": 4.0411899313501145,
      "grad_norm": 0.22996193170547485,
      "learning_rate": 0.0001481865922896196,
      "loss": 1.3808,
      "step": 885
    },
    {
      "epoch": 4.064073226544623,
      "grad_norm": 0.21983887255191803,
      "learning_rate": 0.00014748352521159493,
      "loss": 1.3764,
      "step": 890
    },
    {
      "epoch": 4.086956521739131,
      "grad_norm": 0.23078066110610962,
      "learning_rate": 0.00014677741456996617,
      "loss": 1.3826,
      "step": 895
    },
    {
      "epoch": 4.109839816933638,
      "grad_norm": 0.2580978274345398,
      "learning_rate": 0.0001460683056244869,
      "loss": 1.3816,
      "step": 900
    },
    {
      "epoch": 4.132723112128146,
      "grad_norm": 0.22216792404651642,
      "learning_rate": 0.00014535624382709382,
      "loss": 1.3767,
      "step": 905
    },
    {
      "epoch": 4.155606407322654,
      "grad_norm": 0.21901686489582062,
      "learning_rate": 0.00014464127481899312,
      "loss": 1.3951,
      "step": 910
    },
    {
      "epoch": 4.178489702517163,
      "grad_norm": 0.22333645820617676,
      "learning_rate": 0.0001439234444277354,
      "loss": 1.3946,
      "step": 915
    },
    {
      "epoch": 4.201372997711671,
      "grad_norm": 0.2606089413166046,
      "learning_rate": 0.00014320279866427796,
      "loss": 1.3836,
      "step": 920
    },
    {
      "epoch": 4.224256292906179,
      "grad_norm": 0.2596084475517273,
      "learning_rate": 0.00014247938372003582,
      "loss": 1.3849,
      "step": 925
    },
    {
      "epoch": 4.247139588100686,
      "grad_norm": 0.2835715115070343,
      "learning_rate": 0.00014175324596392075,
      "loss": 1.386,
      "step": 930
    },
    {
      "epoch": 4.270022883295194,
      "grad_norm": 0.2396639585494995,
      "learning_rate": 0.0001410244319393694,
      "loss": 1.3848,
      "step": 935
    },
    {
      "epoch": 4.2929061784897025,
      "grad_norm": 0.22530311346054077,
      "learning_rate": 0.00014029298836135988,
      "loss": 1.3839,
      "step": 940
    },
    {
      "epoch": 4.315789473684211,
      "grad_norm": 0.23546090722084045,
      "learning_rate": 0.0001395589621134174,
      "loss": 1.3948,
      "step": 945
    },
    {
      "epoch": 4.338672768878719,
      "grad_norm": 0.24042978882789612,
      "learning_rate": 0.00013882240024460927,
      "loss": 1.3891,
      "step": 950
    },
    {
      "epoch": 4.361556064073227,
      "grad_norm": 0.26460978388786316,
      "learning_rate": 0.00013808334996652904,
      "loss": 1.3821,
      "step": 955
    },
    {
      "epoch": 4.384439359267734,
      "grad_norm": 0.2298058718442917,
      "learning_rate": 0.0001373418586502706,
      "loss": 1.3796,
      "step": 960
    },
    {
      "epoch": 4.407322654462242,
      "grad_norm": 0.21995796263217926,
      "learning_rate": 0.0001365979738233916,
      "loss": 1.3892,
      "step": 965
    },
    {
      "epoch": 4.4302059496567505,
      "grad_norm": 0.22304576635360718,
      "learning_rate": 0.0001358517431668672,
      "loss": 1.3731,
      "step": 970
    },
    {
      "epoch": 4.453089244851259,
      "grad_norm": 0.2750610411167145,
      "learning_rate": 0.0001351032145120337,
      "loss": 1.3732,
      "step": 975
    },
    {
      "epoch": 4.475972540045767,
      "grad_norm": 0.24216437339782715,
      "learning_rate": 0.00013435243583752294,
      "loss": 1.3955,
      "step": 980
    },
    {
      "epoch": 4.498855835240275,
      "grad_norm": 0.23672589659690857,
      "learning_rate": 0.00013359945526618668,
      "loss": 1.3859,
      "step": 985
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 0.21664075553417206,
      "learning_rate": 0.00013284432106201233,
      "loss": 1.3885,
      "step": 990
    },
    {
      "epoch": 4.54462242562929,
      "grad_norm": 0.22267000377178192,
      "learning_rate": 0.00013208708162702922,
      "loss": 1.3811,
      "step": 995
    },
    {
      "epoch": 4.567505720823799,
      "grad_norm": 0.23190075159072876,
      "learning_rate": 0.00013132778549820618,
      "loss": 1.3813,
      "step": 1000
    },
    {
      "epoch": 4.590389016018307,
      "grad_norm": 0.21617767214775085,
      "learning_rate": 0.0001305664813443405,
      "loss": 1.3821,
      "step": 1005
    },
    {
      "epoch": 4.613272311212815,
      "grad_norm": 0.22316230833530426,
      "learning_rate": 0.00012980321796293836,
      "loss": 1.3854,
      "step": 1010
    },
    {
      "epoch": 4.636155606407323,
      "grad_norm": 0.22770841419696808,
      "learning_rate": 0.00012903804427708704,
      "loss": 1.392,
      "step": 1015
    },
    {
      "epoch": 4.65903890160183,
      "grad_norm": 0.25319135189056396,
      "learning_rate": 0.00012827100933231905,
      "loss": 1.384,
      "step": 1020
    },
    {
      "epoch": 4.6819221967963385,
      "grad_norm": 0.24790321290493011,
      "learning_rate": 0.0001275021622934685,
      "loss": 1.3817,
      "step": 1025
    },
    {
      "epoch": 4.704805491990847,
      "grad_norm": 0.24224399030208588,
      "learning_rate": 0.00012673155244151985,
      "loss": 1.3869,
      "step": 1030
    },
    {
      "epoch": 4.727688787185355,
      "grad_norm": 0.2306058555841446,
      "learning_rate": 0.0001259592291704489,
      "loss": 1.3816,
      "step": 1035
    },
    {
      "epoch": 4.750572082379863,
      "grad_norm": 0.2296520620584488,
      "learning_rate": 0.000125185241984057,
      "loss": 1.3855,
      "step": 1040
    },
    {
      "epoch": 4.77345537757437,
      "grad_norm": 0.22120504081249237,
      "learning_rate": 0.00012440964049279787,
      "loss": 1.3839,
      "step": 1045
    },
    {
      "epoch": 4.796338672768878,
      "grad_norm": 0.23121312260627747,
      "learning_rate": 0.00012363247441059776,
      "loss": 1.3816,
      "step": 1050
    },
    {
      "epoch": 4.8192219679633865,
      "grad_norm": 0.22182002663612366,
      "learning_rate": 0.00012285379355166893,
      "loss": 1.3808,
      "step": 1055
    },
    {
      "epoch": 4.842105263157895,
      "grad_norm": 0.22699835896492004,
      "learning_rate": 0.00012207364782731655,
      "loss": 1.3829,
      "step": 1060
    },
    {
      "epoch": 4.864988558352403,
      "grad_norm": 0.24495604634284973,
      "learning_rate": 0.00012129208724273984,
      "loss": 1.3827,
      "step": 1065
    },
    {
      "epoch": 4.887871853546911,
      "grad_norm": 0.235788956284523,
      "learning_rate": 0.00012050916189382646,
      "loss": 1.3898,
      "step": 1070
    },
    {
      "epoch": 4.910755148741419,
      "grad_norm": 0.24433469772338867,
      "learning_rate": 0.00011972492196394187,
      "loss": 1.3913,
      "step": 1075
    },
    {
      "epoch": 4.933638443935926,
      "grad_norm": 0.2331000417470932,
      "learning_rate": 0.00011893941772071249,
      "loss": 1.3906,
      "step": 1080
    },
    {
      "epoch": 4.956521739130435,
      "grad_norm": 0.22595056891441345,
      "learning_rate": 0.0001181526995128038,
      "loss": 1.3981,
      "step": 1085
    },
    {
      "epoch": 4.979405034324943,
      "grad_norm": 0.22172114253044128,
      "learning_rate": 0.00011736481776669306,
      "loss": 1.3895,
      "step": 1090
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3044658601284027,
      "learning_rate": 0.0001165758229834371,
      "loss": 1.3868,
      "step": 1095
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.481019973754883,
      "eval_runtime": 0.1574,
      "eval_samples_per_second": 63.534,
      "eval_steps_per_second": 6.353,
      "step": 1095
    },
    {
      "epoch": 5.022883295194508,
      "grad_norm": 0.24489176273345947,
      "learning_rate": 0.0001157857657354354,
      "loss": 1.376,
      "step": 1100
    },
    {
      "epoch": 5.045766590389016,
      "grad_norm": 0.2253871113061905,
      "learning_rate": 0.00011499469666318858,
      "loss": 1.3759,
      "step": 1105
    },
    {
      "epoch": 5.068649885583524,
      "grad_norm": 0.235123872756958,
      "learning_rate": 0.00011420266647205231,
      "loss": 1.3774,
      "step": 1110
    },
    {
      "epoch": 5.091533180778032,
      "grad_norm": 0.2346729189157486,
      "learning_rate": 0.00011340972592898744,
      "loss": 1.3556,
      "step": 1115
    },
    {
      "epoch": 5.11441647597254,
      "grad_norm": 0.2369905263185501,
      "learning_rate": 0.00011261592585930576,
      "loss": 1.3714,
      "step": 1120
    },
    {
      "epoch": 5.137299771167048,
      "grad_norm": 0.2265693098306656,
      "learning_rate": 0.00011182131714341247,
      "loss": 1.37,
      "step": 1125
    },
    {
      "epoch": 5.160183066361556,
      "grad_norm": 0.21262961626052856,
      "learning_rate": 0.00011102595071354472,
      "loss": 1.3639,
      "step": 1130
    },
    {
      "epoch": 5.183066361556064,
      "grad_norm": 0.2247275561094284,
      "learning_rate": 0.00011022987755050704,
      "loss": 1.3649,
      "step": 1135
    },
    {
      "epoch": 5.2059496567505725,
      "grad_norm": 0.22549380362033844,
      "learning_rate": 0.00010943314868040364,
      "loss": 1.3542,
      "step": 1140
    },
    {
      "epoch": 5.22883295194508,
      "grad_norm": 0.2336389124393463,
      "learning_rate": 0.00010863581517136776,
      "loss": 1.3659,
      "step": 1145
    },
    {
      "epoch": 5.251716247139588,
      "grad_norm": 0.24654635787010193,
      "learning_rate": 0.00010783792813028827,
      "loss": 1.3742,
      "step": 1150
    },
    {
      "epoch": 5.274599542334096,
      "grad_norm": 0.22848239541053772,
      "learning_rate": 0.00010703953869953403,
      "loss": 1.3783,
      "step": 1155
    },
    {
      "epoch": 5.297482837528604,
      "grad_norm": 0.241807758808136,
      "learning_rate": 0.00010624069805367559,
      "loss": 1.3731,
      "step": 1160
    },
    {
      "epoch": 5.320366132723112,
      "grad_norm": 0.23657450079917908,
      "learning_rate": 0.00010544145739620519,
      "loss": 1.3674,
      "step": 1165
    },
    {
      "epoch": 5.34324942791762,
      "grad_norm": 0.2408100813627243,
      "learning_rate": 0.00010464186795625482,
      "loss": 1.3732,
      "step": 1170
    },
    {
      "epoch": 5.366132723112128,
      "grad_norm": 0.23080427944660187,
      "learning_rate": 0.00010384198098531225,
      "loss": 1.3912,
      "step": 1175
    },
    {
      "epoch": 5.389016018306636,
      "grad_norm": 0.23397450149059296,
      "learning_rate": 0.00010304184775393642,
      "loss": 1.381,
      "step": 1180
    },
    {
      "epoch": 5.411899313501144,
      "grad_norm": 0.2350955307483673,
      "learning_rate": 0.00010224151954847064,
      "loss": 1.3768,
      "step": 1185
    },
    {
      "epoch": 5.434782608695652,
      "grad_norm": 0.23939336836338043,
      "learning_rate": 0.00010144104766775572,
      "loss": 1.3731,
      "step": 1190
    },
    {
      "epoch": 5.4576659038901605,
      "grad_norm": 0.23009169101715088,
      "learning_rate": 0.0001006404834198416,
      "loss": 1.3715,
      "step": 1195
    },
    {
      "epoch": 5.480549199084669,
      "grad_norm": 0.2236054390668869,
      "learning_rate": 9.983987811869862e-05,
      "loss": 1.3737,
      "step": 1200
    },
    {
      "epoch": 5.503432494279176,
      "grad_norm": 0.23314499855041504,
      "learning_rate": 9.903928308092865e-05,
      "loss": 1.369,
      "step": 1205
    },
    {
      "epoch": 5.526315789473684,
      "grad_norm": 0.22338473796844482,
      "learning_rate": 9.823874962247564e-05,
      "loss": 1.3717,
      "step": 1210
    },
    {
      "epoch": 5.549199084668192,
      "grad_norm": 0.22037141025066376,
      "learning_rate": 9.743832905533644e-05,
      "loss": 1.3789,
      "step": 1215
    },
    {
      "epoch": 5.5720823798627,
      "grad_norm": 0.2261509895324707,
      "learning_rate": 9.663807268427198e-05,
      "loss": 1.3741,
      "step": 1220
    },
    {
      "epoch": 5.5949656750572085,
      "grad_norm": 0.2503112256526947,
      "learning_rate": 9.583803180351852e-05,
      "loss": 1.3805,
      "step": 1225
    },
    {
      "epoch": 5.617848970251716,
      "grad_norm": 0.22896412014961243,
      "learning_rate": 9.503825769350017e-05,
      "loss": 1.3583,
      "step": 1230
    },
    {
      "epoch": 5.640732265446224,
      "grad_norm": 0.2552075982093811,
      "learning_rate": 9.423880161754158e-05,
      "loss": 1.3767,
      "step": 1235
    },
    {
      "epoch": 5.663615560640732,
      "grad_norm": 0.23098088800907135,
      "learning_rate": 9.343971481858246e-05,
      "loss": 1.3719,
      "step": 1240
    },
    {
      "epoch": 5.68649885583524,
      "grad_norm": 0.23425135016441345,
      "learning_rate": 9.264104851589273e-05,
      "loss": 1.3678,
      "step": 1245
    },
    {
      "epoch": 5.709382151029748,
      "grad_norm": 0.229757159948349,
      "learning_rate": 9.184285390178978e-05,
      "loss": 1.3652,
      "step": 1250
    },
    {
      "epoch": 5.732265446224257,
      "grad_norm": 0.22069285809993744,
      "learning_rate": 9.104518213835692e-05,
      "loss": 1.3628,
      "step": 1255
    },
    {
      "epoch": 5.755148741418765,
      "grad_norm": 0.23451228439807892,
      "learning_rate": 9.024808435416434e-05,
      "loss": 1.3658,
      "step": 1260
    },
    {
      "epoch": 5.778032036613272,
      "grad_norm": 0.2394520491361618,
      "learning_rate": 8.945161164099157e-05,
      "loss": 1.3775,
      "step": 1265
    },
    {
      "epoch": 5.80091533180778,
      "grad_norm": 0.23245683312416077,
      "learning_rate": 8.865581505055291e-05,
      "loss": 1.372,
      "step": 1270
    },
    {
      "epoch": 5.823798627002288,
      "grad_norm": 0.23066289722919464,
      "learning_rate": 8.7860745591225e-05,
      "loss": 1.3739,
      "step": 1275
    },
    {
      "epoch": 5.8466819221967965,
      "grad_norm": 0.23424118757247925,
      "learning_rate": 8.706645422477739e-05,
      "loss": 1.3597,
      "step": 1280
    },
    {
      "epoch": 5.869565217391305,
      "grad_norm": 0.2321302741765976,
      "learning_rate": 8.627299186310603e-05,
      "loss": 1.3732,
      "step": 1285
    },
    {
      "epoch": 5.892448512585812,
      "grad_norm": 0.2273801863193512,
      "learning_rate": 8.548040936496989e-05,
      "loss": 1.3685,
      "step": 1290
    },
    {
      "epoch": 5.91533180778032,
      "grad_norm": 0.2422792911529541,
      "learning_rate": 8.468875753273115e-05,
      "loss": 1.372,
      "step": 1295
    },
    {
      "epoch": 5.938215102974828,
      "grad_norm": 0.22604131698608398,
      "learning_rate": 8.389808710909881e-05,
      "loss": 1.3676,
      "step": 1300
    },
    {
      "epoch": 5.961098398169336,
      "grad_norm": 0.24549026787281036,
      "learning_rate": 8.310844877387637e-05,
      "loss": 1.358,
      "step": 1305
    },
    {
      "epoch": 5.983981693363845,
      "grad_norm": 0.23445729911327362,
      "learning_rate": 8.231989314071317e-05,
      "loss": 1.3692,
      "step": 1310
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.5001282691955566,
      "eval_runtime": 0.1605,
      "eval_samples_per_second": 62.3,
      "eval_steps_per_second": 6.23,
      "step": 1314
    },
    {
      "epoch": 6.004576659038902,
      "grad_norm": 0.2388521283864975,
      "learning_rate": 8.153247075386043e-05,
      "loss": 1.3541,
      "step": 1315
    },
    {
      "epoch": 6.02745995423341,
      "grad_norm": 0.23299774527549744,
      "learning_rate": 8.07462320849313e-05,
      "loss": 1.3572,
      "step": 1320
    },
    {
      "epoch": 6.050343249427917,
      "grad_norm": 0.23086461424827576,
      "learning_rate": 7.996122752966595e-05,
      "loss": 1.3568,
      "step": 1325
    },
    {
      "epoch": 6.073226544622425,
      "grad_norm": 0.23091675341129303,
      "learning_rate": 7.917750740470117e-05,
      "loss": 1.3618,
      "step": 1330
    },
    {
      "epoch": 6.0961098398169336,
      "grad_norm": 0.23161986470222473,
      "learning_rate": 7.839512194434531e-05,
      "loss": 1.363,
      "step": 1335
    },
    {
      "epoch": 6.118993135011442,
      "grad_norm": 0.23545809090137482,
      "learning_rate": 7.761412129735852e-05,
      "loss": 1.3609,
      "step": 1340
    },
    {
      "epoch": 6.14187643020595,
      "grad_norm": 0.23319639265537262,
      "learning_rate": 7.683455552373799e-05,
      "loss": 1.365,
      "step": 1345
    },
    {
      "epoch": 6.164759725400458,
      "grad_norm": 0.23536434769630432,
      "learning_rate": 7.605647459150961e-05,
      "loss": 1.3673,
      "step": 1350
    },
    {
      "epoch": 6.187643020594965,
      "grad_norm": 0.2269584983587265,
      "learning_rate": 7.527992837352501e-05,
      "loss": 1.3635,
      "step": 1355
    },
    {
      "epoch": 6.2105263157894735,
      "grad_norm": 0.22587326169013977,
      "learning_rate": 7.450496664426477e-05,
      "loss": 1.3595,
      "step": 1360
    },
    {
      "epoch": 6.233409610983982,
      "grad_norm": 0.22287671267986298,
      "learning_rate": 7.37316390766482e-05,
      "loss": 1.3752,
      "step": 1365
    },
    {
      "epoch": 6.25629290617849,
      "grad_norm": 0.23099535703659058,
      "learning_rate": 7.295999523884921e-05,
      "loss": 1.3613,
      "step": 1370
    },
    {
      "epoch": 6.279176201372998,
      "grad_norm": 0.22423085570335388,
      "learning_rate": 7.219008459111937e-05,
      "loss": 1.3658,
      "step": 1375
    },
    {
      "epoch": 6.302059496567506,
      "grad_norm": 0.23212574422359467,
      "learning_rate": 7.142195648261747e-05,
      "loss": 1.361,
      "step": 1380
    },
    {
      "epoch": 6.324942791762013,
      "grad_norm": 0.22750739753246307,
      "learning_rate": 7.065566014824643e-05,
      "loss": 1.3602,
      "step": 1385
    },
    {
      "epoch": 6.3478260869565215,
      "grad_norm": 0.2309405356645584,
      "learning_rate": 6.989124470549745e-05,
      "loss": 1.366,
      "step": 1390
    },
    {
      "epoch": 6.37070938215103,
      "grad_norm": 0.24688778817653656,
      "learning_rate": 6.912875915130183e-05,
      "loss": 1.3645,
      "step": 1395
    },
    {
      "epoch": 6.393592677345538,
      "grad_norm": 0.23304879665374756,
      "learning_rate": 6.83682523588902e-05,
      "loss": 1.3519,
      "step": 1400
    },
    {
      "epoch": 6.416475972540046,
      "grad_norm": 0.2324698269367218,
      "learning_rate": 6.760977307466008e-05,
      "loss": 1.358,
      "step": 1405
    },
    {
      "epoch": 6.439359267734554,
      "grad_norm": 0.23698380589485168,
      "learning_rate": 6.685336991505122e-05,
      "loss": 1.3569,
      "step": 1410
    },
    {
      "epoch": 6.462242562929061,
      "grad_norm": 0.2366679459810257,
      "learning_rate": 6.609909136342955e-05,
      "loss": 1.3538,
      "step": 1415
    },
    {
      "epoch": 6.48512585812357,
      "grad_norm": 0.2314102202653885,
      "learning_rate": 6.534698576697939e-05,
      "loss": 1.3607,
      "step": 1420
    },
    {
      "epoch": 6.508009153318078,
      "grad_norm": 0.2393520474433899,
      "learning_rate": 6.459710133360464e-05,
      "loss": 1.3649,
      "step": 1425
    },
    {
      "epoch": 6.530892448512586,
      "grad_norm": 0.23024214804172516,
      "learning_rate": 6.384948612883873e-05,
      "loss": 1.367,
      "step": 1430
    },
    {
      "epoch": 6.553775743707094,
      "grad_norm": 0.23126175999641418,
      "learning_rate": 6.310418807276375e-05,
      "loss": 1.3623,
      "step": 1435
    },
    {
      "epoch": 6.576659038901602,
      "grad_norm": 0.2327370047569275,
      "learning_rate": 6.2361254936939e-05,
      "loss": 1.3532,
      "step": 1440
    },
    {
      "epoch": 6.5995423340961095,
      "grad_norm": 0.22811928391456604,
      "learning_rate": 6.162073434133876e-05,
      "loss": 1.3662,
      "step": 1445
    },
    {
      "epoch": 6.622425629290618,
      "grad_norm": 0.25274351239204407,
      "learning_rate": 6.088267375130023e-05,
      "loss": 1.3537,
      "step": 1450
    },
    {
      "epoch": 6.645308924485126,
      "grad_norm": 0.24889418482780457,
      "learning_rate": 6.01471204744809e-05,
      "loss": 1.3598,
      "step": 1455
    },
    {
      "epoch": 6.668192219679634,
      "grad_norm": 0.23805195093154907,
      "learning_rate": 5.941412165782645e-05,
      "loss": 1.3525,
      "step": 1460
    },
    {
      "epoch": 6.691075514874142,
      "grad_norm": 0.2376497983932495,
      "learning_rate": 5.868372428454861e-05,
      "loss": 1.3662,
      "step": 1465
    },
    {
      "epoch": 6.71395881006865,
      "grad_norm": 0.24028287827968597,
      "learning_rate": 5.79559751711138e-05,
      "loss": 1.3524,
      "step": 1470
    },
    {
      "epoch": 6.7368421052631575,
      "grad_norm": 0.23552508652210236,
      "learning_rate": 5.72309209642422e-05,
      "loss": 1.3435,
      "step": 1475
    },
    {
      "epoch": 6.759725400457666,
      "grad_norm": 0.23250797390937805,
      "learning_rate": 5.650860813791785e-05,
      "loss": 1.3576,
      "step": 1480
    },
    {
      "epoch": 6.782608695652174,
      "grad_norm": 0.23260493576526642,
      "learning_rate": 5.5789082990409945e-05,
      "loss": 1.3648,
      "step": 1485
    },
    {
      "epoch": 6.805491990846682,
      "grad_norm": 0.23065823316574097,
      "learning_rate": 5.507239164130501e-05,
      "loss": 1.3553,
      "step": 1490
    },
    {
      "epoch": 6.82837528604119,
      "grad_norm": 0.22898992896080017,
      "learning_rate": 5.4358580028550896e-05,
      "loss": 1.357,
      "step": 1495
    },
    {
      "epoch": 6.851258581235698,
      "grad_norm": 0.24019262194633484,
      "learning_rate": 5.364769390551225e-05,
      "loss": 1.3561,
      "step": 1500
    },
    {
      "epoch": 6.874141876430206,
      "grad_norm": 0.24061718583106995,
      "learning_rate": 5.293977883803797e-05,
      "loss": 1.3654,
      "step": 1505
    },
    {
      "epoch": 6.897025171624714,
      "grad_norm": 0.2238311618566513,
      "learning_rate": 5.2234880201540284e-05,
      "loss": 1.3542,
      "step": 1510
    },
    {
      "epoch": 6.919908466819222,
      "grad_norm": 0.22695277631282806,
      "learning_rate": 5.1533043178086536e-05,
      "loss": 1.354,
      "step": 1515
    },
    {
      "epoch": 6.94279176201373,
      "grad_norm": 0.22949448227882385,
      "learning_rate": 5.0834312753503124e-05,
      "loss": 1.3613,
      "step": 1520
    },
    {
      "epoch": 6.965675057208238,
      "grad_norm": 0.23366492986679077,
      "learning_rate": 5.0138733714492e-05,
      "loss": 1.3497,
      "step": 1525
    },
    {
      "epoch": 6.988558352402746,
      "grad_norm": 0.2318800687789917,
      "learning_rate": 4.9446350645759885e-05,
      "loss": 1.3509,
      "step": 1530
    },
    {
      "epoch": 7.0,
      "eval_loss": 2.50384783744812,
      "eval_runtime": 0.1625,
      "eval_samples_per_second": 61.536,
      "eval_steps_per_second": 6.154,
      "step": 1533
    },
    {
      "epoch": 7.009153318077804,
      "grad_norm": 0.24519094824790955,
      "learning_rate": 4.8757207927160584e-05,
      "loss": 1.3595,
      "step": 1535
    },
    {
      "epoch": 7.032036613272311,
      "grad_norm": 0.23380842804908752,
      "learning_rate": 4.807134973085036e-05,
      "loss": 1.3402,
      "step": 1540
    },
    {
      "epoch": 7.054919908466819,
      "grad_norm": 0.23818600177764893,
      "learning_rate": 4.738882001845668e-05,
      "loss": 1.3542,
      "step": 1545
    },
    {
      "epoch": 7.077803203661327,
      "grad_norm": 0.2337319701910019,
      "learning_rate": 4.6709662538260267e-05,
      "loss": 1.3603,
      "step": 1550
    },
    {
      "epoch": 7.100686498855835,
      "grad_norm": 0.23182658851146698,
      "learning_rate": 4.603392082239102e-05,
      "loss": 1.3513,
      "step": 1555
    },
    {
      "epoch": 7.1235697940503435,
      "grad_norm": 0.22734318673610687,
      "learning_rate": 4.53616381840377e-05,
      "loss": 1.3515,
      "step": 1560
    },
    {
      "epoch": 7.146453089244852,
      "grad_norm": 0.2282598614692688,
      "learning_rate": 4.469285771467181e-05,
      "loss": 1.3529,
      "step": 1565
    },
    {
      "epoch": 7.169336384439359,
      "grad_norm": 0.23265138268470764,
      "learning_rate": 4.402762228128531e-05,
      "loss": 1.3513,
      "step": 1570
    },
    {
      "epoch": 7.192219679633867,
      "grad_norm": 0.23814916610717773,
      "learning_rate": 4.336597452364309e-05,
      "loss": 1.3459,
      "step": 1575
    },
    {
      "epoch": 7.215102974828375,
      "grad_norm": 0.23314997553825378,
      "learning_rate": 4.2707956851550016e-05,
      "loss": 1.351,
      "step": 1580
    },
    {
      "epoch": 7.237986270022883,
      "grad_norm": 0.2370002567768097,
      "learning_rate": 4.205361144213227e-05,
      "loss": 1.3552,
      "step": 1585
    },
    {
      "epoch": 7.260869565217392,
      "grad_norm": 0.2338656485080719,
      "learning_rate": 4.140298023713416e-05,
      "loss": 1.3598,
      "step": 1590
    },
    {
      "epoch": 7.2837528604119,
      "grad_norm": 0.2293270081281662,
      "learning_rate": 4.075610494022964e-05,
      "loss": 1.3565,
      "step": 1595
    },
    {
      "epoch": 7.306636155606407,
      "grad_norm": 0.23523476719856262,
      "learning_rate": 4.011302701434937e-05,
      "loss": 1.3545,
      "step": 1600
    },
    {
      "epoch": 7.329519450800915,
      "grad_norm": 0.22826853394508362,
      "learning_rate": 3.947378767902284e-05,
      "loss": 1.3411,
      "step": 1605
    },
    {
      "epoch": 7.352402745995423,
      "grad_norm": 0.24477452039718628,
      "learning_rate": 3.8838427907736476e-05,
      "loss": 1.3482,
      "step": 1610
    },
    {
      "epoch": 7.3752860411899315,
      "grad_norm": 0.23173397779464722,
      "learning_rate": 3.8206988425307246e-05,
      "loss": 1.3602,
      "step": 1615
    },
    {
      "epoch": 7.39816933638444,
      "grad_norm": 0.2321847528219223,
      "learning_rate": 3.757950970527249e-05,
      "loss": 1.3572,
      "step": 1620
    },
    {
      "epoch": 7.421052631578947,
      "grad_norm": 0.2273811548948288,
      "learning_rate": 3.695603196729543e-05,
      "loss": 1.3575,
      "step": 1625
    },
    {
      "epoch": 7.443935926773455,
      "grad_norm": 0.22913090884685516,
      "learning_rate": 3.633659517458736e-05,
      "loss": 1.3566,
      "step": 1630
    },
    {
      "epoch": 7.466819221967963,
      "grad_norm": 0.23222309350967407,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 1.3461,
      "step": 1635
    },
    {
      "epoch": 7.489702517162471,
      "grad_norm": 0.2386443018913269,
      "learning_rate": 3.5110002980210975e-05,
      "loss": 1.3335,
      "step": 1640
    },
    {
      "epoch": 7.5125858123569795,
      "grad_norm": 0.23243063688278198,
      "learning_rate": 3.450292619973483e-05,
      "loss": 1.3497,
      "step": 1645
    },
    {
      "epoch": 7.535469107551488,
      "grad_norm": 0.22777137160301208,
      "learning_rate": 3.3900047601872596e-05,
      "loss": 1.3526,
      "step": 1650
    },
    {
      "epoch": 7.558352402745996,
      "grad_norm": 0.23459511995315552,
      "learning_rate": 3.3301405829487195e-05,
      "loss": 1.3484,
      "step": 1655
    },
    {
      "epoch": 7.581235697940503,
      "grad_norm": 0.22907888889312744,
      "learning_rate": 3.270703925387279e-05,
      "loss": 1.3529,
      "step": 1660
    },
    {
      "epoch": 7.604118993135011,
      "grad_norm": 0.22607311606407166,
      "learning_rate": 3.2116985972295076e-05,
      "loss": 1.3557,
      "step": 1665
    },
    {
      "epoch": 7.627002288329519,
      "grad_norm": 0.2280961275100708,
      "learning_rate": 3.153128380554941e-05,
      "loss": 1.3485,
      "step": 1670
    },
    {
      "epoch": 7.649885583524028,
      "grad_norm": 0.22796593606472015,
      "learning_rate": 3.094997029553673e-05,
      "loss": 1.3524,
      "step": 1675
    },
    {
      "epoch": 7.672768878718536,
      "grad_norm": 0.22574880719184875,
      "learning_rate": 3.037308270285709e-05,
      "loss": 1.3415,
      "step": 1680
    },
    {
      "epoch": 7.695652173913043,
      "grad_norm": 0.22538483142852783,
      "learning_rate": 2.9800658004421366e-05,
      "loss": 1.3584,
      "step": 1685
    },
    {
      "epoch": 7.718535469107551,
      "grad_norm": 0.22863738238811493,
      "learning_rate": 2.923273289108115e-05,
      "loss": 1.3555,
      "step": 1690
    },
    {
      "epoch": 7.741418764302059,
      "grad_norm": 0.23403213918209076,
      "learning_rate": 2.8669343765277078e-05,
      "loss": 1.3455,
      "step": 1695
    },
    {
      "epoch": 7.7643020594965675,
      "grad_norm": 0.23343510925769806,
      "learning_rate": 2.8110526738705344e-05,
      "loss": 1.3472,
      "step": 1700
    },
    {
      "epoch": 7.787185354691076,
      "grad_norm": 0.2283450961112976,
      "learning_rate": 2.755631763000318e-05,
      "loss": 1.3594,
      "step": 1705
    },
    {
      "epoch": 7.810068649885584,
      "grad_norm": 0.2360517978668213,
      "learning_rate": 2.7006751962452882e-05,
      "loss": 1.347,
      "step": 1710
    },
    {
      "epoch": 7.832951945080092,
      "grad_norm": 0.2376154214143753,
      "learning_rate": 2.6461864961704975e-05,
      "loss": 1.3532,
      "step": 1715
    },
    {
      "epoch": 7.855835240274599,
      "grad_norm": 0.23671028017997742,
      "learning_rate": 2.592169155352031e-05,
      "loss": 1.3589,
      "step": 1720
    },
    {
      "epoch": 7.878718535469107,
      "grad_norm": 0.23020578920841217,
      "learning_rate": 2.538626636153131e-05,
      "loss": 1.3527,
      "step": 1725
    },
    {
      "epoch": 7.9016018306636155,
      "grad_norm": 0.2411489188671112,
      "learning_rate": 2.485562370502279e-05,
      "loss": 1.3508,
      "step": 1730
    },
    {
      "epoch": 7.924485125858124,
      "grad_norm": 0.23260930180549622,
      "learning_rate": 2.4329797596732252e-05,
      "loss": 1.3446,
      "step": 1735
    },
    {
      "epoch": 7.947368421052632,
      "grad_norm": 0.23165710270404816,
      "learning_rate": 2.3808821740669606e-05,
      "loss": 1.357,
      "step": 1740
    },
    {
      "epoch": 7.970251716247139,
      "grad_norm": 0.240895077586174,
      "learning_rate": 2.3292729529956935e-05,
      "loss": 1.3631,
      "step": 1745
    },
    {
      "epoch": 7.993135011441647,
      "grad_norm": 0.2322818636894226,
      "learning_rate": 2.2781554044688015e-05,
      "loss": 1.3413,
      "step": 1750
    },
    {
      "epoch": 8.0,
      "eval_loss": 2.5123343467712402,
      "eval_runtime": 0.1598,
      "eval_samples_per_second": 62.575,
      "eval_steps_per_second": 6.258,
      "step": 1752
    },
    {
      "epoch": 8.013729977116705,
      "grad_norm": 0.22951894998550415,
      "learning_rate": 2.227532804980813e-05,
      "loss": 1.3476,
      "step": 1755
    },
    {
      "epoch": 8.036613272311213,
      "grad_norm": 0.23239345848560333,
      "learning_rate": 2.1774083993013718e-05,
      "loss": 1.344,
      "step": 1760
    },
    {
      "epoch": 8.05949656750572,
      "grad_norm": 0.23161360621452332,
      "learning_rate": 2.1277854002672683e-05,
      "loss": 1.3465,
      "step": 1765
    },
    {
      "epoch": 8.082379862700229,
      "grad_norm": 0.24183310568332672,
      "learning_rate": 2.078666988576504e-05,
      "loss": 1.3395,
      "step": 1770
    },
    {
      "epoch": 8.105263157894736,
      "grad_norm": 0.23045741021633148,
      "learning_rate": 2.030056312584424e-05,
      "loss": 1.3489,
      "step": 1775
    },
    {
      "epoch": 8.128146453089245,
      "grad_norm": 0.23021478950977325,
      "learning_rate": 1.9819564881018983e-05,
      "loss": 1.3461,
      "step": 1780
    },
    {
      "epoch": 8.151029748283753,
      "grad_norm": 0.22722554206848145,
      "learning_rate": 1.934370598195622e-05,
      "loss": 1.344,
      "step": 1785
    },
    {
      "epoch": 8.173913043478262,
      "grad_norm": 0.23223134875297546,
      "learning_rate": 1.887301692990494e-05,
      "loss": 1.3585,
      "step": 1790
    },
    {
      "epoch": 8.196796338672769,
      "grad_norm": 0.2324499934911728,
      "learning_rate": 1.8407527894741184e-05,
      "loss": 1.3466,
      "step": 1795
    },
    {
      "epoch": 8.219679633867276,
      "grad_norm": 0.23666158318519592,
      "learning_rate": 1.7947268713034127e-05,
      "loss": 1.3353,
      "step": 1800
    },
    {
      "epoch": 8.242562929061785,
      "grad_norm": 0.2395636886358261,
      "learning_rate": 1.7492268886133676e-05,
      "loss": 1.3312,
      "step": 1805
    },
    {
      "epoch": 8.265446224256292,
      "grad_norm": 0.23455668985843658,
      "learning_rate": 1.7042557578279626e-05,
      "loss": 1.3536,
      "step": 1810
    },
    {
      "epoch": 8.288329519450802,
      "grad_norm": 0.24188482761383057,
      "learning_rate": 1.6598163614732154e-05,
      "loss": 1.3592,
      "step": 1815
    },
    {
      "epoch": 8.311212814645309,
      "grad_norm": 0.23694472014904022,
      "learning_rate": 1.6159115479924257e-05,
      "loss": 1.3485,
      "step": 1820
    },
    {
      "epoch": 8.334096109839816,
      "grad_norm": 0.2400536835193634,
      "learning_rate": 1.5725441315636002e-05,
      "loss": 1.3387,
      "step": 1825
    },
    {
      "epoch": 8.356979405034325,
      "grad_norm": 0.2341165691614151,
      "learning_rate": 1.529716891919074e-05,
      "loss": 1.3429,
      "step": 1830
    },
    {
      "epoch": 8.379862700228832,
      "grad_norm": 0.23209024965763092,
      "learning_rate": 1.4874325741673278e-05,
      "loss": 1.3537,
      "step": 1835
    },
    {
      "epoch": 8.402745995423341,
      "grad_norm": 0.2267337441444397,
      "learning_rate": 1.4456938886170412e-05,
      "loss": 1.3432,
      "step": 1840
    },
    {
      "epoch": 8.425629290617849,
      "grad_norm": 0.22982369363307953,
      "learning_rate": 1.4045035106033655e-05,
      "loss": 1.3401,
      "step": 1845
    },
    {
      "epoch": 8.448512585812358,
      "grad_norm": 0.23104478418827057,
      "learning_rate": 1.3638640803164516e-05,
      "loss": 1.3481,
      "step": 1850
    },
    {
      "epoch": 8.471395881006865,
      "grad_norm": 0.22922466695308685,
      "learning_rate": 1.3237782026322055e-05,
      "loss": 1.3548,
      "step": 1855
    },
    {
      "epoch": 8.494279176201372,
      "grad_norm": 0.2303161472082138,
      "learning_rate": 1.2842484469453365e-05,
      "loss": 1.3461,
      "step": 1860
    },
    {
      "epoch": 8.517162471395881,
      "grad_norm": 0.23580402135849,
      "learning_rate": 1.2452773470046541e-05,
      "loss": 1.3517,
      "step": 1865
    },
    {
      "epoch": 8.540045766590389,
      "grad_norm": 0.2284560203552246,
      "learning_rate": 1.2068674007506786e-05,
      "loss": 1.34,
      "step": 1870
    },
    {
      "epoch": 8.562929061784898,
      "grad_norm": 0.23408174514770508,
      "learning_rate": 1.1690210701555104e-05,
      "loss": 1.3382,
      "step": 1875
    },
    {
      "epoch": 8.585812356979405,
      "grad_norm": 0.22775082290172577,
      "learning_rate": 1.1317407810650372e-05,
      "loss": 1.3533,
      "step": 1880
    },
    {
      "epoch": 8.608695652173914,
      "grad_norm": 0.23187576234340668,
      "learning_rate": 1.0950289230434374e-05,
      "loss": 1.342,
      "step": 1885
    },
    {
      "epoch": 8.631578947368421,
      "grad_norm": 0.23138730227947235,
      "learning_rate": 1.058887849220026e-05,
      "loss": 1.3571,
      "step": 1890
    },
    {
      "epoch": 8.654462242562929,
      "grad_norm": 0.2307848185300827,
      "learning_rate": 1.02331987613841e-05,
      "loss": 1.3525,
      "step": 1895
    },
    {
      "epoch": 8.677345537757438,
      "grad_norm": 0.2283322662115097,
      "learning_rate": 9.883272836080116e-06,
      "loss": 1.3472,
      "step": 1900
    },
    {
      "epoch": 8.700228832951945,
      "grad_norm": 0.22654347121715546,
      "learning_rate": 9.539123145579476e-06,
      "loss": 1.3477,
      "step": 1905
    },
    {
      "epoch": 8.723112128146454,
      "grad_norm": 0.23034285008907318,
      "learning_rate": 9.200771748932513e-06,
      "loss": 1.345,
      "step": 1910
    },
    {
      "epoch": 8.745995423340961,
      "grad_norm": 0.23074457049369812,
      "learning_rate": 8.868240333534815e-06,
      "loss": 1.3389,
      "step": 1915
    },
    {
      "epoch": 8.768878718535468,
      "grad_norm": 0.22868825495243073,
      "learning_rate": 8.541550213737171e-06,
      "loss": 1.3417,
      "step": 1920
    },
    {
      "epoch": 8.791762013729977,
      "grad_norm": 0.22654058039188385,
      "learning_rate": 8.220722329479346e-06,
      "loss": 1.3458,
      "step": 1925
    },
    {
      "epoch": 8.814645308924485,
      "grad_norm": 0.22552727162837982,
      "learning_rate": 7.905777244947954e-06,
      "loss": 1.3444,
      "step": 1930
    },
    {
      "epoch": 8.837528604118994,
      "grad_norm": 0.22772538661956787,
      "learning_rate": 7.5967351472582275e-06,
      "loss": 1.355,
      "step": 1935
    },
    {
      "epoch": 8.860411899313501,
      "grad_norm": 0.2261328399181366,
      "learning_rate": 7.293615845160196e-06,
      "loss": 1.3314,
      "step": 1940
    },
    {
      "epoch": 8.883295194508008,
      "grad_norm": 0.23299558460712433,
      "learning_rate": 6.99643876776891e-06,
      "loss": 1.3461,
      "step": 1945
    },
    {
      "epoch": 8.906178489702517,
      "grad_norm": 0.22399510443210602,
      "learning_rate": 6.705222963319191e-06,
      "loss": 1.3412,
      "step": 1950
    },
    {
      "epoch": 8.929061784897025,
      "grad_norm": 0.22913098335266113,
      "learning_rate": 6.419987097944579e-06,
      "loss": 1.3519,
      "step": 1955
    },
    {
      "epoch": 8.951945080091534,
      "grad_norm": 0.22438156604766846,
      "learning_rate": 6.140749454480932e-06,
      "loss": 1.3513,
      "step": 1960
    },
    {
      "epoch": 8.974828375286041,
      "grad_norm": 0.23290272057056427,
      "learning_rate": 5.867527931294614e-06,
      "loss": 1.3495,
      "step": 1965
    },
    {
      "epoch": 8.99771167048055,
      "grad_norm": 0.2285740226507187,
      "learning_rate": 5.6003400411351325e-06,
      "loss": 1.3562,
      "step": 1970
    },
    {
      "epoch": 9.0,
      "eval_loss": 2.517554998397827,
      "eval_runtime": 0.1572,
      "eval_samples_per_second": 63.605,
      "eval_steps_per_second": 6.36,
      "step": 1971
    },
    {
      "epoch": 9.018306636155607,
      "grad_norm": 0.22652767598628998,
      "learning_rate": 5.339202910012708e-06,
      "loss": 1.3499,
      "step": 1975
    },
    {
      "epoch": 9.041189931350115,
      "grad_norm": 0.2265496850013733,
      "learning_rate": 5.0841332761005e-06,
      "loss": 1.3514,
      "step": 1980
    },
    {
      "epoch": 9.064073226544622,
      "grad_norm": 0.22489045560359955,
      "learning_rate": 4.835147488661795e-06,
      "loss": 1.3402,
      "step": 1985
    },
    {
      "epoch": 9.08695652173913,
      "grad_norm": 0.23145288228988647,
      "learning_rate": 4.592261507001993e-06,
      "loss": 1.343,
      "step": 1990
    },
    {
      "epoch": 9.109839816933638,
      "grad_norm": 0.228094220161438,
      "learning_rate": 4.355490899445691e-06,
      "loss": 1.3422,
      "step": 1995
    },
    {
      "epoch": 9.132723112128147,
      "grad_norm": 0.23417501151561737,
      "learning_rate": 4.124850842338779e-06,
      "loss": 1.3512,
      "step": 2000
    },
    {
      "epoch": 9.155606407322654,
      "grad_norm": 0.2259424477815628,
      "learning_rate": 3.900356119075743e-06,
      "loss": 1.3428,
      "step": 2005
    },
    {
      "epoch": 9.178489702517162,
      "grad_norm": 0.22936026751995087,
      "learning_rate": 3.6820211191520125e-06,
      "loss": 1.343,
      "step": 2010
    },
    {
      "epoch": 9.20137299771167,
      "grad_norm": 0.2265051156282425,
      "learning_rate": 3.469859837241651e-06,
      "loss": 1.3517,
      "step": 2015
    },
    {
      "epoch": 9.224256292906178,
      "grad_norm": 0.22797267138957977,
      "learning_rate": 3.263885872300343e-06,
      "loss": 1.3497,
      "step": 2020
    },
    {
      "epoch": 9.247139588100687,
      "grad_norm": 0.2263391762971878,
      "learning_rate": 3.064112426693799e-06,
      "loss": 1.3381,
      "step": 2025
    },
    {
      "epoch": 9.270022883295194,
      "grad_norm": 0.23144488036632538,
      "learning_rate": 2.8705523053513816e-06,
      "loss": 1.3453,
      "step": 2030
    },
    {
      "epoch": 9.292906178489703,
      "grad_norm": 0.2310468852519989,
      "learning_rate": 2.6832179149454793e-06,
      "loss": 1.338,
      "step": 2035
    },
    {
      "epoch": 9.31578947368421,
      "grad_norm": 0.2297661006450653,
      "learning_rate": 2.502121263096224e-06,
      "loss": 1.3424,
      "step": 2040
    },
    {
      "epoch": 9.338672768878718,
      "grad_norm": 0.23045673966407776,
      "learning_rate": 2.3272739576017945e-06,
      "loss": 1.3525,
      "step": 2045
    },
    {
      "epoch": 9.361556064073227,
      "grad_norm": 0.22856095433235168,
      "learning_rate": 2.1586872056944428e-06,
      "loss": 1.3441,
      "step": 2050
    },
    {
      "epoch": 9.384439359267734,
      "grad_norm": 0.23274967074394226,
      "learning_rate": 1.996371813322129e-06,
      "loss": 1.355,
      "step": 2055
    },
    {
      "epoch": 9.407322654462243,
      "grad_norm": 0.22686432301998138,
      "learning_rate": 1.840338184455881e-06,
      "loss": 1.341,
      "step": 2060
    },
    {
      "epoch": 9.43020594965675,
      "grad_norm": 0.22302532196044922,
      "learning_rate": 1.6905963204229436e-06,
      "loss": 1.3573,
      "step": 2065
    },
    {
      "epoch": 9.453089244851258,
      "grad_norm": 0.2239583283662796,
      "learning_rate": 1.5471558192656777e-06,
      "loss": 1.3357,
      "step": 2070
    },
    {
      "epoch": 9.475972540045767,
      "grad_norm": 0.22839844226837158,
      "learning_rate": 1.4100258751264195e-06,
      "loss": 1.3387,
      "step": 2075
    },
    {
      "epoch": 9.498855835240274,
      "grad_norm": 0.22954653203487396,
      "learning_rate": 1.2792152776580968e-06,
      "loss": 1.3428,
      "step": 2080
    },
    {
      "epoch": 9.521739130434783,
      "grad_norm": 0.22747966647148132,
      "learning_rate": 1.1547324114608904e-06,
      "loss": 1.3396,
      "step": 2085
    },
    {
      "epoch": 9.54462242562929,
      "grad_norm": 0.2284754067659378,
      "learning_rate": 1.036585255544764e-06,
      "loss": 1.3472,
      "step": 2090
    },
    {
      "epoch": 9.5675057208238,
      "grad_norm": 0.22196947038173676,
      "learning_rate": 9.247813828180407e-07,
      "loss": 1.3454,
      "step": 2095
    },
    {
      "epoch": 9.590389016018307,
      "grad_norm": 0.22999419271945953,
      "learning_rate": 8.193279596020121e-07,
      "loss": 1.3431,
      "step": 2100
    },
    {
      "epoch": 9.613272311212814,
      "grad_norm": 0.23028169572353363,
      "learning_rate": 7.202317451716067e-07,
      "loss": 1.3277,
      "step": 2105
    },
    {
      "epoch": 9.636155606407323,
      "grad_norm": 0.22792372107505798,
      "learning_rate": 6.274990913221035e-07,
      "loss": 1.3374,
      "step": 2110
    },
    {
      "epoch": 9.65903890160183,
      "grad_norm": 0.22512663900852203,
      "learning_rate": 5.411359419620232e-07,
      "loss": 1.3503,
      "step": 2115
    },
    {
      "epoch": 9.68192219679634,
      "grad_norm": 0.22875413298606873,
      "learning_rate": 4.6114783273213393e-07,
      "loss": 1.3451,
      "step": 2120
    },
    {
      "epoch": 9.704805491990847,
      "grad_norm": 0.22629615664482117,
      "learning_rate": 3.8753989065064557e-07,
      "loss": 1.3315,
      "step": 2125
    },
    {
      "epoch": 9.727688787185354,
      "grad_norm": 0.22211916744709015,
      "learning_rate": 3.203168337845508e-07,
      "loss": 1.35,
      "step": 2130
    },
    {
      "epoch": 9.750572082379863,
      "grad_norm": 0.22795149683952332,
      "learning_rate": 2.594829709472446e-07,
      "loss": 1.3314,
      "step": 2135
    },
    {
      "epoch": 9.77345537757437,
      "grad_norm": 0.231648787856102,
      "learning_rate": 2.05042201422323e-07,
      "loss": 1.3568,
      "step": 2140
    },
    {
      "epoch": 9.79633867276888,
      "grad_norm": 0.22703665494918823,
      "learning_rate": 1.5699801471364962e-07,
      "loss": 1.3411,
      "step": 2145
    },
    {
      "epoch": 9.819221967963387,
      "grad_norm": 0.2311718612909317,
      "learning_rate": 1.1535349032167908e-07,
      "loss": 1.3456,
      "step": 2150
    },
    {
      "epoch": 9.842105263157894,
      "grad_norm": 0.2319943606853485,
      "learning_rate": 8.011129754611491e-08,
      "loss": 1.3507,
      "step": 2155
    },
    {
      "epoch": 9.864988558352403,
      "grad_norm": 0.2376013547182083,
      "learning_rate": 5.127369531473525e-08,
      "loss": 1.3372,
      "step": 2160
    },
    {
      "epoch": 9.88787185354691,
      "grad_norm": 0.2285187989473343,
      "learning_rate": 2.884253203869758e-08,
      "loss": 1.3469,
      "step": 2165
    },
    {
      "epoch": 9.91075514874142,
      "grad_norm": 0.22958822548389435,
      "learning_rate": 1.2819245493955744e-08,
      "loss": 1.3467,
      "step": 2170
    },
    {
      "epoch": 9.933638443935926,
      "grad_norm": 0.22622478008270264,
      "learning_rate": 3.2048627292113887e-09,
      "loss": 1.3452,
      "step": 2175
    },
    {
      "epoch": 9.956521739130435,
      "grad_norm": 0.2237871289253235,
      "learning_rate": 0.0,
      "loss": 1.3538,
      "step": 2180
    },
    {
      "epoch": 9.956521739130435,
      "eval_loss": 2.5189335346221924,
      "eval_runtime": 0.1831,
      "eval_samples_per_second": 54.615,
      "eval_steps_per_second": 5.461,
      "step": 2180
    },
    {
      "epoch": 9.956521739130435,
      "step": 2180,
      "total_flos": 2.4191218891641324e+18,
      "train_loss": 1.418152675716155,
      "train_runtime": 3547.1741,
      "train_samples_per_second": 39.395,
      "train_steps_per_second": 0.615
    }
  ],
  "logging_steps": 5,
  "max_steps": 2180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4191218891641324e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}