{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1692,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01773049645390071,
      "grad_norm": 1.0585654973983765,
      "learning_rate": 4.970449172576833e-05,
      "loss": 0.4344,
      "step": 10
    },
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 0.5365344285964966,
      "learning_rate": 4.940898345153664e-05,
      "loss": 0.357,
      "step": 20
    },
    {
      "epoch": 0.05319148936170213,
      "grad_norm": 1.0602149963378906,
      "learning_rate": 4.911347517730497e-05,
      "loss": 0.3397,
      "step": 30
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 2.197633743286133,
      "learning_rate": 4.8817966903073283e-05,
      "loss": 0.3472,
      "step": 40
    },
    {
      "epoch": 0.08865248226950355,
      "grad_norm": 1.7486677169799805,
      "learning_rate": 4.852245862884161e-05,
      "loss": 0.3175,
      "step": 50
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 2.9779016971588135,
      "learning_rate": 4.822695035460993e-05,
      "loss": 0.3174,
      "step": 60
    },
    {
      "epoch": 0.12411347517730496,
      "grad_norm": 1.5208463668823242,
      "learning_rate": 4.793144208037825e-05,
      "loss": 0.3061,
      "step": 70
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 2.6333279609680176,
      "learning_rate": 4.763593380614658e-05,
      "loss": 0.3196,
      "step": 80
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 2.9724740982055664,
      "learning_rate": 4.734042553191489e-05,
      "loss": 0.2976,
      "step": 90
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 1.83737313747406,
      "learning_rate": 4.704491725768322e-05,
      "loss": 0.2861,
      "step": 100
    },
    {
      "epoch": 0.1950354609929078,
      "grad_norm": 2.455568313598633,
      "learning_rate": 4.674940898345154e-05,
      "loss": 0.3028,
      "step": 110
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 1.068047046661377,
      "learning_rate": 4.645390070921986e-05,
      "loss": 0.29,
      "step": 120
    },
    {
      "epoch": 0.23049645390070922,
      "grad_norm": 1.9365094900131226,
      "learning_rate": 4.615839243498818e-05,
      "loss": 0.2338,
      "step": 130
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 4.382534027099609,
      "learning_rate": 4.58628841607565e-05,
      "loss": 0.2383,
      "step": 140
    },
    {
      "epoch": 0.26595744680851063,
      "grad_norm": 1.8602826595306396,
      "learning_rate": 4.556737588652483e-05,
      "loss": 0.2698,
      "step": 150
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 4.105195999145508,
      "learning_rate": 4.527186761229315e-05,
      "loss": 0.2895,
      "step": 160
    },
    {
      "epoch": 0.30141843971631205,
      "grad_norm": 2.267707347869873,
      "learning_rate": 4.497635933806147e-05,
      "loss": 0.3412,
      "step": 170
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.7341336011886597,
      "learning_rate": 4.468085106382979e-05,
      "loss": 0.2304,
      "step": 180
    },
    {
      "epoch": 0.33687943262411346,
      "grad_norm": 4.169299602508545,
      "learning_rate": 4.438534278959811e-05,
      "loss": 0.2063,
      "step": 190
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 4.8010663986206055,
      "learning_rate": 4.4089834515366435e-05,
      "loss": 0.2919,
      "step": 200
    },
    {
      "epoch": 0.3723404255319149,
      "grad_norm": 1.2835743427276611,
      "learning_rate": 4.3794326241134755e-05,
      "loss": 0.2172,
      "step": 210
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 2.9821650981903076,
      "learning_rate": 4.3498817966903076e-05,
      "loss": 0.2404,
      "step": 220
    },
    {
      "epoch": 0.4078014184397163,
      "grad_norm": 4.525284767150879,
      "learning_rate": 4.3203309692671396e-05,
      "loss": 0.2233,
      "step": 230
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 2.9262726306915283,
      "learning_rate": 4.2907801418439716e-05,
      "loss": 0.2247,
      "step": 240
    },
    {
      "epoch": 0.4432624113475177,
      "grad_norm": 1.120538592338562,
      "learning_rate": 4.2612293144208036e-05,
      "loss": 0.2,
      "step": 250
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 3.67739200592041,
      "learning_rate": 4.231678486997636e-05,
      "loss": 0.2774,
      "step": 260
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 2.233776330947876,
      "learning_rate": 4.2021276595744684e-05,
      "loss": 0.2339,
      "step": 270
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 3.464088201522827,
      "learning_rate": 4.1725768321513004e-05,
      "loss": 0.2443,
      "step": 280
    },
    {
      "epoch": 0.5141843971631206,
      "grad_norm": 3.5438666343688965,
      "learning_rate": 4.1430260047281324e-05,
      "loss": 0.2949,
      "step": 290
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 2.237935781478882,
      "learning_rate": 4.1134751773049644e-05,
      "loss": 0.212,
      "step": 300
    },
    {
      "epoch": 0.549645390070922,
      "grad_norm": 2.065207004547119,
      "learning_rate": 4.083924349881797e-05,
      "loss": 0.1957,
      "step": 310
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 1.892553448677063,
      "learning_rate": 4.0543735224586285e-05,
      "loss": 0.2182,
      "step": 320
    },
    {
      "epoch": 0.5851063829787234,
      "grad_norm": 1.1844576597213745,
      "learning_rate": 4.024822695035461e-05,
      "loss": 0.258,
      "step": 330
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 1.786897897720337,
      "learning_rate": 3.995271867612293e-05,
      "loss": 0.2151,
      "step": 340
    },
    {
      "epoch": 0.6205673758865248,
      "grad_norm": 4.924149036407471,
      "learning_rate": 3.965721040189125e-05,
      "loss": 0.2095,
      "step": 350
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 3.0387415885925293,
      "learning_rate": 3.936170212765958e-05,
      "loss": 0.2339,
      "step": 360
    },
    {
      "epoch": 0.6560283687943262,
      "grad_norm": 0.9237979054450989,
      "learning_rate": 3.906619385342789e-05,
      "loss": 0.2731,
      "step": 370
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 1.2017602920532227,
      "learning_rate": 3.877068557919622e-05,
      "loss": 0.2323,
      "step": 380
    },
    {
      "epoch": 0.6914893617021277,
      "grad_norm": 4.876067161560059,
      "learning_rate": 3.847517730496454e-05,
      "loss": 0.2312,
      "step": 390
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 2.4507224559783936,
      "learning_rate": 3.817966903073286e-05,
      "loss": 0.2548,
      "step": 400
    },
    {
      "epoch": 0.7269503546099291,
      "grad_norm": 3.6143534183502197,
      "learning_rate": 3.788416075650119e-05,
      "loss": 0.2564,
      "step": 410
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 3.5805842876434326,
      "learning_rate": 3.75886524822695e-05,
      "loss": 0.2275,
      "step": 420
    },
    {
      "epoch": 0.7624113475177305,
      "grad_norm": 3.5016636848449707,
      "learning_rate": 3.729314420803783e-05,
      "loss": 0.2724,
      "step": 430
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 4.354279518127441,
      "learning_rate": 3.699763593380615e-05,
      "loss": 0.1945,
      "step": 440
    },
    {
      "epoch": 0.7978723404255319,
      "grad_norm": 1.473893642425537,
      "learning_rate": 3.670212765957447e-05,
      "loss": 0.2098,
      "step": 450
    },
    {
      "epoch": 0.8156028368794326,
      "grad_norm": 3.5749127864837646,
      "learning_rate": 3.6406619385342796e-05,
      "loss": 0.2062,
      "step": 460
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 1.5818848609924316,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.226,
      "step": 470
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.9678372144699097,
      "learning_rate": 3.5815602836879437e-05,
      "loss": 0.2087,
      "step": 480
    },
    {
      "epoch": 0.8687943262411347,
      "grad_norm": 3.465823173522949,
      "learning_rate": 3.552009456264776e-05,
      "loss": 0.2473,
      "step": 490
    },
    {
      "epoch": 0.8865248226950354,
      "grad_norm": 2.4084010124206543,
      "learning_rate": 3.522458628841608e-05,
      "loss": 0.2732,
      "step": 500
    },
    {
      "epoch": 0.9042553191489362,
      "grad_norm": 2.5898969173431396,
      "learning_rate": 3.49290780141844e-05,
      "loss": 0.1784,
      "step": 510
    },
    {
      "epoch": 0.9219858156028369,
      "grad_norm": 2.0226316452026367,
      "learning_rate": 3.463356973995272e-05,
      "loss": 0.1835,
      "step": 520
    },
    {
      "epoch": 0.9397163120567376,
      "grad_norm": 2.84747576713562,
      "learning_rate": 3.4338061465721045e-05,
      "loss": 0.2101,
      "step": 530
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 1.7862353324890137,
      "learning_rate": 3.4042553191489365e-05,
      "loss": 0.1742,
      "step": 540
    },
    {
      "epoch": 0.975177304964539,
      "grad_norm": 5.6382155418396,
      "learning_rate": 3.3747044917257685e-05,
      "loss": 0.1596,
      "step": 550
    },
    {
      "epoch": 0.9929078014184397,
      "grad_norm": 2.4179556369781494,
      "learning_rate": 3.3451536643026005e-05,
      "loss": 0.2326,
      "step": 560
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9245283018867925,
      "eval_f1": 0.4970414201183432,
      "eval_loss": 0.1933222860097885,
      "eval_precision": 0.7924528301886793,
      "eval_recall": 0.3620689655172414,
      "eval_runtime": 18.2335,
      "eval_samples_per_second": 247.073,
      "eval_steps_per_second": 30.932,
      "step": 564
    },
    {
      "epoch": 1.0106382978723405,
      "grad_norm": 1.8270294666290283,
      "learning_rate": 3.3156028368794326e-05,
      "loss": 0.1533,
      "step": 570
    },
    {
      "epoch": 1.0283687943262412,
      "grad_norm": 2.191605567932129,
      "learning_rate": 3.2860520094562646e-05,
      "loss": 0.1668,
      "step": 580
    },
    {
      "epoch": 1.0460992907801419,
      "grad_norm": 5.722381591796875,
      "learning_rate": 3.256501182033097e-05,
      "loss": 0.1999,
      "step": 590
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 5.328127384185791,
      "learning_rate": 3.226950354609929e-05,
      "loss": 0.1799,
      "step": 600
    },
    {
      "epoch": 1.0815602836879432,
      "grad_norm": 3.9988811016082764,
      "learning_rate": 3.1973995271867614e-05,
      "loss": 0.1787,
      "step": 610
    },
    {
      "epoch": 1.099290780141844,
      "grad_norm": 2.0374648571014404,
      "learning_rate": 3.1678486997635934e-05,
      "loss": 0.1728,
      "step": 620
    },
    {
      "epoch": 1.1170212765957448,
      "grad_norm": 1.21236252784729,
      "learning_rate": 3.1382978723404254e-05,
      "loss": 0.1884,
      "step": 630
    },
    {
      "epoch": 1.1347517730496455,
      "grad_norm": 1.166555404663086,
      "learning_rate": 3.108747044917258e-05,
      "loss": 0.1946,
      "step": 640
    },
    {
      "epoch": 1.1524822695035462,
      "grad_norm": 1.6630871295928955,
      "learning_rate": 3.0791962174940895e-05,
      "loss": 0.1791,
      "step": 650
    },
    {
      "epoch": 1.1702127659574468,
      "grad_norm": 1.3428844213485718,
      "learning_rate": 3.0496453900709222e-05,
      "loss": 0.1607,
      "step": 660
    },
    {
      "epoch": 1.1879432624113475,
      "grad_norm": 2.1707041263580322,
      "learning_rate": 3.0200945626477545e-05,
      "loss": 0.1888,
      "step": 670
    },
    {
      "epoch": 1.2056737588652482,
      "grad_norm": 2.2633979320526123,
      "learning_rate": 2.9905437352245862e-05,
      "loss": 0.159,
      "step": 680
    },
    {
      "epoch": 1.2234042553191489,
      "grad_norm": 2.3307528495788574,
      "learning_rate": 2.9609929078014186e-05,
      "loss": 0.145,
      "step": 690
    },
    {
      "epoch": 1.2411347517730495,
      "grad_norm": 5.760026931762695,
      "learning_rate": 2.9314420803782506e-05,
      "loss": 0.1352,
      "step": 700
    },
    {
      "epoch": 1.2588652482269502,
      "grad_norm": 2.038200855255127,
      "learning_rate": 2.901891252955083e-05,
      "loss": 0.1637,
      "step": 710
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 3.5735347270965576,
      "learning_rate": 2.8723404255319154e-05,
      "loss": 0.125,
      "step": 720
    },
    {
      "epoch": 1.2943262411347518,
      "grad_norm": 1.2548505067825317,
      "learning_rate": 2.842789598108747e-05,
      "loss": 0.1571,
      "step": 730
    },
    {
      "epoch": 1.3120567375886525,
      "grad_norm": 3.3048200607299805,
      "learning_rate": 2.8132387706855794e-05,
      "loss": 0.1298,
      "step": 740
    },
    {
      "epoch": 1.3297872340425532,
      "grad_norm": 1.5077462196350098,
      "learning_rate": 2.7836879432624114e-05,
      "loss": 0.1593,
      "step": 750
    },
    {
      "epoch": 1.3475177304964538,
      "grad_norm": 2.6117773056030273,
      "learning_rate": 2.7541371158392438e-05,
      "loss": 0.2071,
      "step": 760
    },
    {
      "epoch": 1.3652482269503547,
      "grad_norm": 3.2539641857147217,
      "learning_rate": 2.7245862884160755e-05,
      "loss": 0.1661,
      "step": 770
    },
    {
      "epoch": 1.3829787234042552,
      "grad_norm": 2.994198799133301,
      "learning_rate": 2.695035460992908e-05,
      "loss": 0.1904,
      "step": 780
    },
    {
      "epoch": 1.400709219858156,
      "grad_norm": 2.030109405517578,
      "learning_rate": 2.6654846335697402e-05,
      "loss": 0.1583,
      "step": 790
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 1.1124039888381958,
      "learning_rate": 2.6359338061465723e-05,
      "loss": 0.2241,
      "step": 800
    },
    {
      "epoch": 1.4361702127659575,
      "grad_norm": 1.3373945951461792,
      "learning_rate": 2.6063829787234046e-05,
      "loss": 0.1659,
      "step": 810
    },
    {
      "epoch": 1.4539007092198581,
      "grad_norm": 4.298996448516846,
      "learning_rate": 2.5768321513002363e-05,
      "loss": 0.1597,
      "step": 820
    },
    {
      "epoch": 1.4716312056737588,
      "grad_norm": 2.2879459857940674,
      "learning_rate": 2.5472813238770687e-05,
      "loss": 0.1806,
      "step": 830
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 5.475185394287109,
      "learning_rate": 2.5177304964539007e-05,
      "loss": 0.2264,
      "step": 840
    },
    {
      "epoch": 1.5070921985815602,
      "grad_norm": 3.100987672805786,
      "learning_rate": 2.488179669030733e-05,
      "loss": 0.1517,
      "step": 850
    },
    {
      "epoch": 1.524822695035461,
      "grad_norm": 1.2688792943954468,
      "learning_rate": 2.458628841607565e-05,
      "loss": 0.1257,
      "step": 860
    },
    {
      "epoch": 1.5425531914893615,
      "grad_norm": 2.6867027282714844,
      "learning_rate": 2.429078014184397e-05,
      "loss": 0.1489,
      "step": 870
    },
    {
      "epoch": 1.5602836879432624,
      "grad_norm": 2.5701425075531006,
      "learning_rate": 2.3995271867612295e-05,
      "loss": 0.2134,
      "step": 880
    },
    {
      "epoch": 1.5780141843971631,
      "grad_norm": 5.73723840713501,
      "learning_rate": 2.3699763593380615e-05,
      "loss": 0.191,
      "step": 890
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 3.5702483654022217,
      "learning_rate": 2.340425531914894e-05,
      "loss": 0.17,
      "step": 900
    },
    {
      "epoch": 1.6134751773049647,
      "grad_norm": 4.70973014831543,
      "learning_rate": 2.310874704491726e-05,
      "loss": 0.1357,
      "step": 910
    },
    {
      "epoch": 1.6312056737588652,
      "grad_norm": 1.4206900596618652,
      "learning_rate": 2.281323877068558e-05,
      "loss": 0.1264,
      "step": 920
    },
    {
      "epoch": 1.648936170212766,
      "grad_norm": 3.2121095657348633,
      "learning_rate": 2.25177304964539e-05,
      "loss": 0.1409,
      "step": 930
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 5.003275394439697,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1993,
      "step": 940
    },
    {
      "epoch": 1.6843971631205674,
      "grad_norm": 8.793657302856445,
      "learning_rate": 2.1926713947990547e-05,
      "loss": 0.097,
      "step": 950
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 2.719095230102539,
      "learning_rate": 2.1631205673758867e-05,
      "loss": 0.1739,
      "step": 960
    },
    {
      "epoch": 1.7198581560283688,
      "grad_norm": 4.192145347595215,
      "learning_rate": 2.1335697399527187e-05,
      "loss": 0.1819,
      "step": 970
    },
    {
      "epoch": 1.7375886524822695,
      "grad_norm": 3.239504098892212,
      "learning_rate": 2.1040189125295508e-05,
      "loss": 0.1775,
      "step": 980
    },
    {
      "epoch": 1.7553191489361701,
      "grad_norm": 1.6205350160598755,
      "learning_rate": 2.074468085106383e-05,
      "loss": 0.1455,
      "step": 990
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 2.7471115589141846,
      "learning_rate": 2.0449172576832152e-05,
      "loss": 0.2198,
      "step": 1000
    },
    {
      "epoch": 1.7907801418439715,
      "grad_norm": 2.1342594623565674,
      "learning_rate": 2.0153664302600475e-05,
      "loss": 0.1577,
      "step": 1010
    },
    {
      "epoch": 1.8085106382978724,
      "grad_norm": 2.6111040115356445,
      "learning_rate": 1.9858156028368796e-05,
      "loss": 0.1907,
      "step": 1020
    },
    {
      "epoch": 1.826241134751773,
      "grad_norm": 1.9335544109344482,
      "learning_rate": 1.9562647754137116e-05,
      "loss": 0.1812,
      "step": 1030
    },
    {
      "epoch": 1.8439716312056738,
      "grad_norm": 3.2039554119110107,
      "learning_rate": 1.926713947990544e-05,
      "loss": 0.1604,
      "step": 1040
    },
    {
      "epoch": 1.8617021276595744,
      "grad_norm": 4.588708400726318,
      "learning_rate": 1.897163120567376e-05,
      "loss": 0.1576,
      "step": 1050
    },
    {
      "epoch": 1.8794326241134751,
      "grad_norm": 2.468317985534668,
      "learning_rate": 1.867612293144208e-05,
      "loss": 0.2102,
      "step": 1060
    },
    {
      "epoch": 1.897163120567376,
      "grad_norm": 2.883124828338623,
      "learning_rate": 1.83806146572104e-05,
      "loss": 0.1883,
      "step": 1070
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 1.7469326257705688,
      "learning_rate": 1.8085106382978724e-05,
      "loss": 0.1598,
      "step": 1080
    },
    {
      "epoch": 1.9326241134751774,
      "grad_norm": 3.9554150104522705,
      "learning_rate": 1.7789598108747048e-05,
      "loss": 0.171,
      "step": 1090
    },
    {
      "epoch": 1.950354609929078,
      "grad_norm": 7.007796287536621,
      "learning_rate": 1.7494089834515368e-05,
      "loss": 0.1908,
      "step": 1100
    },
    {
      "epoch": 1.9680851063829787,
      "grad_norm": 2.2939341068267822,
      "learning_rate": 1.7198581560283688e-05,
      "loss": 0.2282,
      "step": 1110
    },
    {
      "epoch": 1.9858156028368794,
      "grad_norm": 1.1014021635055542,
      "learning_rate": 1.690307328605201e-05,
      "loss": 0.1599,
      "step": 1120
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.93007769145394,
      "eval_f1": 0.5827814569536424,
      "eval_loss": 0.18422859907150269,
      "eval_precision": 0.7560137457044673,
      "eval_recall": 0.47413793103448276,
      "eval_runtime": 18.3061,
      "eval_samples_per_second": 246.093,
      "eval_steps_per_second": 30.809,
      "step": 1128
    },
    {
      "epoch": 2.00354609929078,
      "grad_norm": 1.147599458694458,
      "learning_rate": 1.6607565011820332e-05,
      "loss": 0.1652,
      "step": 1130
    },
    {
      "epoch": 2.021276595744681,
      "grad_norm": 1.893280029296875,
      "learning_rate": 1.6312056737588656e-05,
      "loss": 0.1345,
      "step": 1140
    },
    {
      "epoch": 2.0390070921985815,
      "grad_norm": 3.6475846767425537,
      "learning_rate": 1.6016548463356976e-05,
      "loss": 0.1214,
      "step": 1150
    },
    {
      "epoch": 2.0567375886524824,
      "grad_norm": 6.175610542297363,
      "learning_rate": 1.5721040189125296e-05,
      "loss": 0.131,
      "step": 1160
    },
    {
      "epoch": 2.074468085106383,
      "grad_norm": 2.870246171951294,
      "learning_rate": 1.5425531914893617e-05,
      "loss": 0.1379,
      "step": 1170
    },
    {
      "epoch": 2.0921985815602837,
      "grad_norm": 1.1721495389938354,
      "learning_rate": 1.5130023640661939e-05,
      "loss": 0.1239,
      "step": 1180
    },
    {
      "epoch": 2.1099290780141846,
      "grad_norm": 5.15436315536499,
      "learning_rate": 1.483451536643026e-05,
      "loss": 0.097,
      "step": 1190
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.1995478719472885,
      "learning_rate": 1.4539007092198581e-05,
      "loss": 0.1416,
      "step": 1200
    },
    {
      "epoch": 2.145390070921986,
      "grad_norm": 4.829070091247559,
      "learning_rate": 1.4243498817966905e-05,
      "loss": 0.1075,
      "step": 1210
    },
    {
      "epoch": 2.1631205673758864,
      "grad_norm": 4.2572479248046875,
      "learning_rate": 1.3947990543735227e-05,
      "loss": 0.1233,
      "step": 1220
    },
    {
      "epoch": 2.1808510638297873,
      "grad_norm": 3.297318458557129,
      "learning_rate": 1.3652482269503547e-05,
      "loss": 0.1171,
      "step": 1230
    },
    {
      "epoch": 2.198581560283688,
      "grad_norm": 2.2237844467163086,
      "learning_rate": 1.3356973995271869e-05,
      "loss": 0.0868,
      "step": 1240
    },
    {
      "epoch": 2.2163120567375887,
      "grad_norm": 4.605975151062012,
      "learning_rate": 1.3061465721040189e-05,
      "loss": 0.1458,
      "step": 1250
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 6.8933610916137695,
      "learning_rate": 1.2765957446808511e-05,
      "loss": 0.1242,
      "step": 1260
    },
    {
      "epoch": 2.25177304964539,
      "grad_norm": 2.5259997844696045,
      "learning_rate": 1.2470449172576833e-05,
      "loss": 0.1131,
      "step": 1270
    },
    {
      "epoch": 2.269503546099291,
      "grad_norm": 5.3347296714782715,
      "learning_rate": 1.2174940898345153e-05,
      "loss": 0.1334,
      "step": 1280
    },
    {
      "epoch": 2.2872340425531914,
      "grad_norm": 3.3900346755981445,
      "learning_rate": 1.1879432624113477e-05,
      "loss": 0.1162,
      "step": 1290
    },
    {
      "epoch": 2.3049645390070923,
      "grad_norm": 3.4547274112701416,
      "learning_rate": 1.1583924349881797e-05,
      "loss": 0.1586,
      "step": 1300
    },
    {
      "epoch": 2.3226950354609928,
      "grad_norm": 3.0310921669006348,
      "learning_rate": 1.1288416075650119e-05,
      "loss": 0.1088,
      "step": 1310
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 4.557296276092529,
      "learning_rate": 1.0992907801418441e-05,
      "loss": 0.0988,
      "step": 1320
    },
    {
      "epoch": 2.3581560283687946,
      "grad_norm": 18.163665771484375,
      "learning_rate": 1.0697399527186761e-05,
      "loss": 0.1055,
      "step": 1330
    },
    {
      "epoch": 2.375886524822695,
      "grad_norm": 3.0445291996002197,
      "learning_rate": 1.0401891252955083e-05,
      "loss": 0.1547,
      "step": 1340
    },
    {
      "epoch": 2.393617021276596,
      "grad_norm": 2.7467446327209473,
      "learning_rate": 1.0106382978723404e-05,
      "loss": 0.1248,
      "step": 1350
    },
    {
      "epoch": 2.4113475177304964,
      "grad_norm": 0.8503823280334473,
      "learning_rate": 9.810874704491727e-06,
      "loss": 0.0915,
      "step": 1360
    },
    {
      "epoch": 2.4290780141843973,
      "grad_norm": 1.9473341703414917,
      "learning_rate": 9.515366430260048e-06,
      "loss": 0.1267,
      "step": 1370
    },
    {
      "epoch": 2.4468085106382977,
      "grad_norm": 2.457197427749634,
      "learning_rate": 9.219858156028368e-06,
      "loss": 0.1228,
      "step": 1380
    },
    {
      "epoch": 2.4645390070921986,
      "grad_norm": 2.5377461910247803,
      "learning_rate": 8.924349881796691e-06,
      "loss": 0.085,
      "step": 1390
    },
    {
      "epoch": 2.482269503546099,
      "grad_norm": 4.247258186340332,
      "learning_rate": 8.628841607565012e-06,
      "loss": 0.1345,
      "step": 1400
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.234439373016357,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1396,
      "step": 1410
    },
    {
      "epoch": 2.5177304964539005,
      "grad_norm": 2.6999399662017822,
      "learning_rate": 8.037825059101656e-06,
      "loss": 0.0934,
      "step": 1420
    },
    {
      "epoch": 2.5354609929078014,
      "grad_norm": 3.0950117111206055,
      "learning_rate": 7.742316784869976e-06,
      "loss": 0.088,
      "step": 1430
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 6.523775577545166,
      "learning_rate": 7.446808510638298e-06,
      "loss": 0.1234,
      "step": 1440
    },
    {
      "epoch": 2.5709219858156027,
      "grad_norm": 0.6427033543586731,
      "learning_rate": 7.151300236406621e-06,
      "loss": 0.0849,
      "step": 1450
    },
    {
      "epoch": 2.5886524822695036,
      "grad_norm": 4.271224498748779,
      "learning_rate": 6.855791962174941e-06,
      "loss": 0.158,
      "step": 1460
    },
    {
      "epoch": 2.6063829787234045,
      "grad_norm": 1.9938910007476807,
      "learning_rate": 6.560283687943262e-06,
      "loss": 0.1214,
      "step": 1470
    },
    {
      "epoch": 2.624113475177305,
      "grad_norm": 1.0301127433776855,
      "learning_rate": 6.264775413711583e-06,
      "loss": 0.1217,
      "step": 1480
    },
    {
      "epoch": 2.6418439716312054,
      "grad_norm": 1.3055609464645386,
      "learning_rate": 5.969267139479906e-06,
      "loss": 0.0761,
      "step": 1490
    },
    {
      "epoch": 2.6595744680851063,
      "grad_norm": 4.245100021362305,
      "learning_rate": 5.673758865248227e-06,
      "loss": 0.0884,
      "step": 1500
    },
    {
      "epoch": 2.6773049645390072,
      "grad_norm": 0.3163594901561737,
      "learning_rate": 5.378250591016549e-06,
      "loss": 0.081,
      "step": 1510
    },
    {
      "epoch": 2.6950354609929077,
      "grad_norm": 2.392202615737915,
      "learning_rate": 5.08274231678487e-06,
      "loss": 0.1288,
      "step": 1520
    },
    {
      "epoch": 2.7127659574468086,
      "grad_norm": 4.6757493019104,
      "learning_rate": 4.787234042553191e-06,
      "loss": 0.1134,
      "step": 1530
    },
    {
      "epoch": 2.7304964539007095,
      "grad_norm": 14.765192985534668,
      "learning_rate": 4.491725768321513e-06,
      "loss": 0.1706,
      "step": 1540
    },
    {
      "epoch": 2.74822695035461,
      "grad_norm": 5.195065021514893,
      "learning_rate": 4.1962174940898345e-06,
      "loss": 0.1669,
      "step": 1550
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 2.5050220489501953,
      "learning_rate": 3.9007092198581565e-06,
      "loss": 0.0757,
      "step": 1560
    },
    {
      "epoch": 2.7836879432624113,
      "grad_norm": 2.086787223815918,
      "learning_rate": 3.605200945626478e-06,
      "loss": 0.1797,
      "step": 1570
    },
    {
      "epoch": 2.801418439716312,
      "grad_norm": 5.924837589263916,
      "learning_rate": 3.309692671394799e-06,
      "loss": 0.1602,
      "step": 1580
    },
    {
      "epoch": 2.8191489361702127,
      "grad_norm": 4.369917392730713,
      "learning_rate": 3.0141843971631207e-06,
      "loss": 0.08,
      "step": 1590
    },
    {
      "epoch": 2.8368794326241136,
      "grad_norm": 1.1417466402053833,
      "learning_rate": 2.7186761229314422e-06,
      "loss": 0.1123,
      "step": 1600
    },
    {
      "epoch": 2.854609929078014,
      "grad_norm": 0.7693955302238464,
      "learning_rate": 2.4231678486997638e-06,
      "loss": 0.1303,
      "step": 1610
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 3.25712251663208,
      "learning_rate": 2.1276595744680853e-06,
      "loss": 0.0748,
      "step": 1620
    },
    {
      "epoch": 2.8900709219858154,
      "grad_norm": 3.0831634998321533,
      "learning_rate": 1.8321513002364066e-06,
      "loss": 0.0703,
      "step": 1630
    },
    {
      "epoch": 2.9078014184397163,
      "grad_norm": 2.351789712905884,
      "learning_rate": 1.5366430260047282e-06,
      "loss": 0.0501,
      "step": 1640
    },
    {
      "epoch": 2.925531914893617,
      "grad_norm": 2.154916286468506,
      "learning_rate": 1.2411347517730497e-06,
      "loss": 0.0907,
      "step": 1650
    },
    {
      "epoch": 2.9432624113475176,
      "grad_norm": 6.570488452911377,
      "learning_rate": 9.456264775413712e-07,
      "loss": 0.1337,
      "step": 1660
    },
    {
      "epoch": 2.9609929078014185,
      "grad_norm": 4.879380702972412,
      "learning_rate": 6.501182033096927e-07,
      "loss": 0.1201,
      "step": 1670
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 4.488287448883057,
      "learning_rate": 3.546099290780142e-07,
      "loss": 0.1064,
      "step": 1680
    },
    {
      "epoch": 2.99645390070922,
      "grad_norm": 3.126821279525757,
      "learning_rate": 5.91016548463357e-08,
      "loss": 0.1375,
      "step": 1690
    }
  ],
  "logging_steps": 10,
  "max_steps": 1692,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7160790169147392.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}