{
  "best_metric": 0.8592745065689087,
  "epoch": 1.0057919744736934,
  "global_step": 5800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 1.9169,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.00011999999999999999,
      "loss": 1.6001,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.00017999999999999998,
      "loss": 1.2273,
      "step": 60
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.00023999999999999998,
      "loss": 1.1652,
      "step": 80
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0003,
      "loss": 1.1155,
      "step": 100
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002996511222235143,
      "loss": 1.0974,
      "step": 120
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002993022444470287,
      "loss": 1.0811,
      "step": 140
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00029895336667054307,
      "loss": 1.0819,
      "step": 160
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002986044888940574,
      "loss": 1.0658,
      "step": 180
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002982556111175718,
      "loss": 1.0597,
      "step": 200
    },
    {
      "epoch": 0.03,
      "eval_loss": 1.0481364727020264,
      "eval_runtime": 50.5238,
      "eval_samples_per_second": 39.585,
      "eval_steps_per_second": 1.663,
      "step": 200
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002979067333410861,
      "loss": 1.0552,
      "step": 220
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002975578555646005,
      "loss": 1.0405,
      "step": 240
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002972089777881149,
      "loss": 1.0433,
      "step": 260
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002968601000116292,
      "loss": 1.0386,
      "step": 280
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002965112222351436,
      "loss": 1.0273,
      "step": 300
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.000296162344458658,
      "loss": 1.04,
      "step": 320
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002958134666821723,
      "loss": 1.0148,
      "step": 340
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002954645889056867,
      "loss": 1.0144,
      "step": 360
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.000295115711129201,
      "loss": 1.0088,
      "step": 380
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002947668333527154,
      "loss": 1.0048,
      "step": 400
    },
    {
      "epoch": 0.07,
      "eval_loss": 1.0047168731689453,
      "eval_runtime": 49.9795,
      "eval_samples_per_second": 40.016,
      "eval_steps_per_second": 1.681,
      "step": 400
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002944179555762298,
      "loss": 1.0181,
      "step": 420
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002940690777997441,
      "loss": 1.0054,
      "step": 440
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002937202000232585,
      "loss": 1.0125,
      "step": 460
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00029337132224677284,
      "loss": 1.0066,
      "step": 480
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002930224444702872,
      "loss": 1.011,
      "step": 500
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002926735666938016,
      "loss": 0.9941,
      "step": 520
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00029232468891731594,
      "loss": 0.9812,
      "step": 540
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002919758111408303,
      "loss": 1.0008,
      "step": 560
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00029162693336434465,
      "loss": 0.9934,
      "step": 580
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00029127805558785904,
      "loss": 0.9811,
      "step": 600
    },
    {
      "epoch": 0.1,
      "eval_loss": 0.9793989062309265,
      "eval_runtime": 50.7216,
      "eval_samples_per_second": 39.431,
      "eval_steps_per_second": 1.656,
      "step": 600
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002909291778113734,
      "loss": 0.9917,
      "step": 620
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00029058030003488775,
      "loss": 0.977,
      "step": 640
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002902314222584021,
      "loss": 0.9777,
      "step": 660
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00028988254448191646,
      "loss": 0.9782,
      "step": 680
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00028953366670543085,
      "loss": 0.9656,
      "step": 700
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002891847889289452,
      "loss": 0.9595,
      "step": 720
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00028883591115245956,
      "loss": 0.9712,
      "step": 740
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00028848703337597395,
      "loss": 0.9784,
      "step": 760
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002881381555994883,
      "loss": 0.9714,
      "step": 780
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00028778927782300266,
      "loss": 0.9728,
      "step": 800
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.9619740843772888,
      "eval_runtime": 49.3289,
      "eval_samples_per_second": 40.544,
      "eval_steps_per_second": 1.703,
      "step": 800
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.000287440400046517,
      "loss": 0.9684,
      "step": 820
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002870915222700314,
      "loss": 0.955,
      "step": 840
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00028674264449354576,
      "loss": 0.9642,
      "step": 860
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002863937667170601,
      "loss": 0.9758,
      "step": 880
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00028604488894057447,
      "loss": 0.9601,
      "step": 900
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002856960111640888,
      "loss": 0.9597,
      "step": 920
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002853471333876032,
      "loss": 0.9615,
      "step": 940
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00028499825561111757,
      "loss": 0.9562,
      "step": 960
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002846493778346319,
      "loss": 0.9543,
      "step": 980
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002843005000581463,
      "loss": 0.9593,
      "step": 1000
    },
    {
      "epoch": 0.17,
      "eval_loss": 0.9471739530563354,
      "eval_runtime": 49.1237,
      "eval_samples_per_second": 40.714,
      "eval_steps_per_second": 1.71,
      "step": 1000
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002839516222816606,
      "loss": 0.9498,
      "step": 1020
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.000283602744505175,
      "loss": 0.9416,
      "step": 1040
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002832538667286894,
      "loss": 0.9476,
      "step": 1060
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002829049889522037,
      "loss": 0.9486,
      "step": 1080
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002825561111757181,
      "loss": 0.9581,
      "step": 1100
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002822072333992325,
      "loss": 0.9429,
      "step": 1120
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002818583556227468,
      "loss": 0.9513,
      "step": 1140
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00028150947784626114,
      "loss": 0.9489,
      "step": 1160
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002811606000697755,
      "loss": 0.9384,
      "step": 1180
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002808117222932899,
      "loss": 0.9496,
      "step": 1200
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.9367659687995911,
      "eval_runtime": 48.936,
      "eval_samples_per_second": 40.87,
      "eval_steps_per_second": 1.717,
      "step": 1200
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00028046284451680424,
      "loss": 0.941,
      "step": 1220
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002801139667403186,
      "loss": 0.94,
      "step": 1240
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00027976508896383295,
      "loss": 0.9349,
      "step": 1260
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00027941621118734734,
      "loss": 0.9378,
      "step": 1280
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002790673334108617,
      "loss": 0.9361,
      "step": 1300
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00027871845563437605,
      "loss": 0.9426,
      "step": 1320
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00027836957785789043,
      "loss": 0.9427,
      "step": 1340
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00027802070008140476,
      "loss": 0.9339,
      "step": 1360
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00027767182230491915,
      "loss": 0.9313,
      "step": 1380
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00027732294452843353,
      "loss": 0.9308,
      "step": 1400
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.9267547726631165,
      "eval_runtime": 49.2408,
      "eval_samples_per_second": 40.617,
      "eval_steps_per_second": 1.706,
      "step": 1400
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00027697406675194786,
      "loss": 0.9299,
      "step": 1420
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00027662518897546225,
      "loss": 0.9334,
      "step": 1440
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002762763111989766,
      "loss": 0.9264,
      "step": 1460
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00027592743342249096,
      "loss": 0.9175,
      "step": 1480
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00027557855564600535,
      "loss": 0.9376,
      "step": 1500
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002752296778695197,
      "loss": 0.9273,
      "step": 1520
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00027488080009303406,
      "loss": 0.9319,
      "step": 1540
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00027453192231654844,
      "loss": 0.9261,
      "step": 1560
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002741830445400628,
      "loss": 0.9246,
      "step": 1580
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00027383416676357716,
      "loss": 0.9249,
      "step": 1600
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.9193373322486877,
      "eval_runtime": 49.234,
      "eval_samples_per_second": 40.622,
      "eval_steps_per_second": 1.706,
      "step": 1600
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002734852889870915,
      "loss": 0.9114,
      "step": 1620
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00027313641121060587,
      "loss": 0.9314,
      "step": 1640
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00027278753343412026,
      "loss": 0.9225,
      "step": 1660
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002724386556576346,
      "loss": 0.9265,
      "step": 1680
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002720897778811489,
      "loss": 0.9223,
      "step": 1700
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002717409001046633,
      "loss": 0.9105,
      "step": 1720
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002713920223281777,
      "loss": 0.9241,
      "step": 1740
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.000271043144551692,
      "loss": 0.9176,
      "step": 1760
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002706942667752064,
      "loss": 0.9254,
      "step": 1780
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00027034538899872073,
      "loss": 0.9225,
      "step": 1800
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.9116300344467163,
      "eval_runtime": 49.2973,
      "eval_samples_per_second": 40.57,
      "eval_steps_per_second": 1.704,
      "step": 1800
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002699965112222351,
      "loss": 0.9097,
      "step": 1820
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002696476334457495,
      "loss": 0.9224,
      "step": 1840
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002692987556692638,
      "loss": 0.921,
      "step": 1860
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002689498778927782,
      "loss": 0.9143,
      "step": 1880
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00026860100011629254,
      "loss": 0.9144,
      "step": 1900
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002682521223398069,
      "loss": 0.9129,
      "step": 1920
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002679032445633213,
      "loss": 0.9124,
      "step": 1940
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00026755436678683564,
      "loss": 0.9107,
      "step": 1960
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00026720548901035,
      "loss": 0.9123,
      "step": 1980
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002668566112338644,
      "loss": 0.9081,
      "step": 2000
    },
    {
      "epoch": 0.35,
      "eval_loss": 0.9067729115486145,
      "eval_runtime": 49.6863,
      "eval_samples_per_second": 40.253,
      "eval_steps_per_second": 1.691,
      "step": 2000
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00026650773345737874,
      "loss": 0.9066,
      "step": 2020
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002661588556808931,
      "loss": 0.9063,
      "step": 2040
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00026580997790440745,
      "loss": 0.913,
      "step": 2060
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00026546110012792183,
      "loss": 0.9131,
      "step": 2080
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002651122223514362,
      "loss": 0.9071,
      "step": 2100
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00026476334457495055,
      "loss": 0.9076,
      "step": 2120
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00026441446679846493,
      "loss": 0.9184,
      "step": 2140
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00026406558902197926,
      "loss": 0.9082,
      "step": 2160
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00026371671124549365,
      "loss": 0.9151,
      "step": 2180
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00026336783346900803,
      "loss": 0.9214,
      "step": 2200
    },
    {
      "epoch": 0.38,
      "eval_loss": 0.901930034160614,
      "eval_runtime": 49.1887,
      "eval_samples_per_second": 40.66,
      "eval_steps_per_second": 1.708,
      "step": 2200
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00026301895569252236,
      "loss": 0.9065,
      "step": 2220
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002626700779160367,
      "loss": 0.9105,
      "step": 2240
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002623212001395511,
      "loss": 0.8953,
      "step": 2260
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00026197232236306546,
      "loss": 0.9041,
      "step": 2280
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002616234445865798,
      "loss": 0.8988,
      "step": 2300
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002612745668100942,
      "loss": 0.9124,
      "step": 2320
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002609256890336085,
      "loss": 0.9042,
      "step": 2340
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002605768112571229,
      "loss": 0.9154,
      "step": 2360
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.00026022793348063727,
      "loss": 0.9015,
      "step": 2380
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002598790557041516,
      "loss": 0.9015,
      "step": 2400
    },
    {
      "epoch": 0.42,
      "eval_loss": 0.8977901935577393,
      "eval_runtime": 50.3787,
      "eval_samples_per_second": 39.699,
      "eval_steps_per_second": 1.667,
      "step": 2400
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.000259530177927666,
      "loss": 0.8981,
      "step": 2420
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00025918130015118037,
      "loss": 0.8962,
      "step": 2440
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002588324223746947,
      "loss": 0.8933,
      "step": 2460
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002584835445982091,
      "loss": 0.8962,
      "step": 2480
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002581346668217234,
      "loss": 0.8951,
      "step": 2500
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002577857890452378,
      "loss": 0.898,
      "step": 2520
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002574369112687522,
      "loss": 0.9002,
      "step": 2540
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002570880334922665,
      "loss": 0.896,
      "step": 2560
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002567391557157809,
      "loss": 0.9062,
      "step": 2580
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002563902779392952,
      "loss": 0.8962,
      "step": 2600
    },
    {
      "epoch": 0.45,
      "eval_loss": 0.8936859369277954,
      "eval_runtime": 49.4977,
      "eval_samples_per_second": 40.406,
      "eval_steps_per_second": 1.697,
      "step": 2600
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002560414001628096,
      "loss": 0.8948,
      "step": 2620
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.000255692522386324,
      "loss": 0.9002,
      "step": 2640
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002553436446098383,
      "loss": 0.8956,
      "step": 2660
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002549947668333527,
      "loss": 0.8972,
      "step": 2680
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002546458890568671,
      "loss": 0.8883,
      "step": 2700
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002542970112803814,
      "loss": 0.9009,
      "step": 2720
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00025394813350389575,
      "loss": 0.8969,
      "step": 2740
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00025359925572741014,
      "loss": 0.8963,
      "step": 2760
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00025325037795092447,
      "loss": 0.8868,
      "step": 2780
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00025290150017443885,
      "loss": 0.8945,
      "step": 2800
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.8895061612129211,
      "eval_runtime": 49.4815,
      "eval_samples_per_second": 40.419,
      "eval_steps_per_second": 1.698,
      "step": 2800
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00025255262239795323,
      "loss": 0.8915,
      "step": 2820
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00025220374462146756,
      "loss": 0.8965,
      "step": 2840
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00025185486684498195,
      "loss": 0.8934,
      "step": 2860
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00025150598906849633,
      "loss": 0.8904,
      "step": 2880
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00025115711129201066,
      "loss": 0.8974,
      "step": 2900
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00025080823351552505,
      "loss": 0.8851,
      "step": 2920
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002504593557390394,
      "loss": 0.8883,
      "step": 2940
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00025011047796255376,
      "loss": 0.8898,
      "step": 2960
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00024976160018606815,
      "loss": 0.8916,
      "step": 2980
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002494127224095825,
      "loss": 0.8891,
      "step": 3000
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.885992169380188,
      "eval_runtime": 50.6986,
      "eval_samples_per_second": 39.449,
      "eval_steps_per_second": 1.657,
      "step": 3000
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00024906384463309686,
      "loss": 0.8866,
      "step": 3020
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002487149668566112,
      "loss": 0.887,
      "step": 3040
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002483660890801256,
      "loss": 0.8756,
      "step": 3060
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00024801721130363996,
      "loss": 0.8814,
      "step": 3080
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002476683335271543,
      "loss": 0.8766,
      "step": 3100
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00024731945575066867,
      "loss": 0.887,
      "step": 3120
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00024697057797418306,
      "loss": 0.8901,
      "step": 3140
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002466217001976974,
      "loss": 0.8807,
      "step": 3160
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00024627282242121177,
      "loss": 0.8927,
      "step": 3180
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002459239446447261,
      "loss": 0.89,
      "step": 3200
    },
    {
      "epoch": 0.55,
      "eval_loss": 0.8836617469787598,
      "eval_runtime": 49.351,
      "eval_samples_per_second": 40.526,
      "eval_steps_per_second": 1.702,
      "step": 3200
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00024557506686824043,
      "loss": 0.892,
      "step": 3220
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00024522618909175487,
      "loss": 0.8862,
      "step": 3240
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002448773113152692,
      "loss": 0.8843,
      "step": 3260
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00024452843353878353,
      "loss": 0.8812,
      "step": 3280
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002441795557622979,
      "loss": 0.8905,
      "step": 3300
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002438306779858123,
      "loss": 0.8858,
      "step": 3320
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00024348180020932665,
      "loss": 0.8768,
      "step": 3340
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.000243132922432841,
      "loss": 0.8765,
      "step": 3360
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00024278404465635537,
      "loss": 0.8889,
      "step": 3380
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00024243516687986975,
      "loss": 0.883,
      "step": 3400
    },
    {
      "epoch": 0.59,
      "eval_loss": 0.8804351091384888,
      "eval_runtime": 50.8292,
      "eval_samples_per_second": 39.347,
      "eval_steps_per_second": 1.653,
      "step": 3400
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002420862891033841,
      "loss": 0.8861,
      "step": 3420
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00024173741132689847,
      "loss": 0.8889,
      "step": 3440
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00024138853355041282,
      "loss": 0.897,
      "step": 3460
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00024103965577392715,
      "loss": 0.8901,
      "step": 3480
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00024069077799744156,
      "loss": 0.8895,
      "step": 3500
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00024034190022095592,
      "loss": 0.8863,
      "step": 3520
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00023999302244447025,
      "loss": 0.8852,
      "step": 3540
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002396441446679846,
      "loss": 0.8842,
      "step": 3560
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.000239295266891499,
      "loss": 0.8843,
      "step": 3580
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00023894638911501335,
      "loss": 0.8803,
      "step": 3600
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.8769158720970154,
      "eval_runtime": 49.0911,
      "eval_samples_per_second": 40.741,
      "eval_steps_per_second": 1.711,
      "step": 3600
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002385975113385277,
      "loss": 0.8869,
      "step": 3620
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00023824863356204206,
      "loss": 0.8786,
      "step": 3640
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00023789975578555642,
      "loss": 0.8761,
      "step": 3660
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002375508780090708,
      "loss": 0.8811,
      "step": 3680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00023720200023258516,
      "loss": 0.8815,
      "step": 3700
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00023685312245609952,
      "loss": 0.8782,
      "step": 3720
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00023650424467961387,
      "loss": 0.8868,
      "step": 3740
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00023615536690312826,
      "loss": 0.8804,
      "step": 3760
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00023580648912664262,
      "loss": 0.8754,
      "step": 3780
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00023545761135015697,
      "loss": 0.8826,
      "step": 3800
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.8761597871780396,
      "eval_runtime": 49.3308,
      "eval_samples_per_second": 40.543,
      "eval_steps_per_second": 1.703,
      "step": 3800
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00023510873357367133,
      "loss": 0.8835,
      "step": 3820
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00023475985579718571,
      "loss": 0.8655,
      "step": 3840
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00023441097802070007,
      "loss": 0.8839,
      "step": 3860
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00023406210024421443,
      "loss": 0.8814,
      "step": 3880
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00023371322246772879,
      "loss": 0.8839,
      "step": 3900
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00023336434469124314,
      "loss": 0.8794,
      "step": 3920
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00023301546691475753,
      "loss": 0.8755,
      "step": 3940
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00023266658913827188,
      "loss": 0.8771,
      "step": 3960
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00023231771136178624,
      "loss": 0.8731,
      "step": 3980
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002319688335853006,
      "loss": 0.8764,
      "step": 4000
    },
    {
      "epoch": 0.69,
      "eval_loss": 0.8736156225204468,
      "eval_runtime": 50.6647,
      "eval_samples_per_second": 39.475,
      "eval_steps_per_second": 1.658,
      "step": 4000
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023161995580881498,
      "loss": 0.8821,
      "step": 4020
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023127107803232934,
      "loss": 0.8791,
      "step": 4040
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023092220025584367,
      "loss": 0.8841,
      "step": 4060
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00023057332247935803,
      "loss": 0.8718,
      "step": 4080
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00023022444470287238,
      "loss": 0.8731,
      "step": 4100
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00022987556692638677,
      "loss": 0.8805,
      "step": 4120
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00022952668914990112,
      "loss": 0.8821,
      "step": 4140
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00022917781137341548,
      "loss": 0.8669,
      "step": 4160
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00022882893359692984,
      "loss": 0.8701,
      "step": 4180
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00022848005582044422,
      "loss": 0.862,
      "step": 4200
    },
    {
      "epoch": 0.73,
      "eval_loss": 0.8710012435913086,
      "eval_runtime": 49.64,
      "eval_samples_per_second": 40.29,
      "eval_steps_per_second": 1.692,
      "step": 4200
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00022813117804395858,
      "loss": 0.8716,
      "step": 4220
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00022778230026747294,
      "loss": 0.8708,
      "step": 4240
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002274334224909873,
      "loss": 0.8668,
      "step": 4260
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00022708454471450168,
      "loss": 0.8691,
      "step": 4280
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00022673566693801603,
      "loss": 0.874,
      "step": 4300
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002263867891615304,
      "loss": 0.8688,
      "step": 4320
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00022603791138504475,
      "loss": 0.8714,
      "step": 4340
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002256890336085591,
      "loss": 0.8694,
      "step": 4360
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002253401558320735,
      "loss": 0.8733,
      "step": 4380
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00022499127805558785,
      "loss": 0.8798,
      "step": 4400
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.869438886642456,
      "eval_runtime": 50.0009,
      "eval_samples_per_second": 39.999,
      "eval_steps_per_second": 1.68,
      "step": 4400
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002246424002791022,
      "loss": 0.8599,
      "step": 4420
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00022429352250261656,
      "loss": 0.8733,
      "step": 4440
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00022394464472613094,
      "loss": 0.8748,
      "step": 4460
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002235957669496453,
      "loss": 0.8764,
      "step": 4480
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00022324688917315966,
      "loss": 0.865,
      "step": 4500
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00022289801139667402,
      "loss": 0.8663,
      "step": 4520
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00022254913362018835,
      "loss": 0.8738,
      "step": 4540
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00022220025584370276,
      "loss": 0.8611,
      "step": 4560
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002218513780672171,
      "loss": 0.8683,
      "step": 4580
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00022150250029073144,
      "loss": 0.8725,
      "step": 4600
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.8671180605888367,
      "eval_runtime": 50.2687,
      "eval_samples_per_second": 39.786,
      "eval_steps_per_second": 1.671,
      "step": 4600
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002211536225142458,
      "loss": 0.8686,
      "step": 4620
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00022080474473776019,
      "loss": 0.8674,
      "step": 4640
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00022045586696127454,
      "loss": 0.8657,
      "step": 4660
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002201069891847889,
      "loss": 0.8607,
      "step": 4680
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00021975811140830326,
      "loss": 0.8672,
      "step": 4700
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00021940923363181764,
      "loss": 0.8603,
      "step": 4720
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.000219060355855332,
      "loss": 0.8672,
      "step": 4740
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00021871147807884635,
      "loss": 0.8659,
      "step": 4760
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002183626003023607,
      "loss": 0.8638,
      "step": 4780
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00021801372252587507,
      "loss": 0.8776,
      "step": 4800
    },
    {
      "epoch": 0.83,
      "eval_loss": 0.8659985065460205,
      "eval_runtime": 49.0999,
      "eval_samples_per_second": 40.733,
      "eval_steps_per_second": 1.711,
      "step": 4800
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00021766484474938945,
      "loss": 0.8714,
      "step": 4820
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002173159669729038,
      "loss": 0.8786,
      "step": 4840
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00021696708919641817,
      "loss": 0.8674,
      "step": 4860
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00021661821141993252,
      "loss": 0.8705,
      "step": 4880
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002162693336434469,
      "loss": 0.8716,
      "step": 4900
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00021592045586696126,
      "loss": 0.8573,
      "step": 4920
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00021557157809047562,
      "loss": 0.8596,
      "step": 4940
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00021522270031398998,
      "loss": 0.8605,
      "step": 4960
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00021487382253750436,
      "loss": 0.8612,
      "step": 4980
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00021452494476101872,
      "loss": 0.8686,
      "step": 5000
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.8649623394012451,
      "eval_runtime": 49.5648,
      "eval_samples_per_second": 40.351,
      "eval_steps_per_second": 1.695,
      "step": 5000
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00021417606698453308,
      "loss": 0.869,
      "step": 5020
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00021382718920804743,
      "loss": 0.8712,
      "step": 5040
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00021347831143156176,
      "loss": 0.8633,
      "step": 5060
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00021312943365507618,
      "loss": 0.8611,
      "step": 5080
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00021278055587859053,
      "loss": 0.8714,
      "step": 5100
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00021243167810210486,
      "loss": 0.8684,
      "step": 5120
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00021208280032561922,
      "loss": 0.8619,
      "step": 5140
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002117339225491336,
      "loss": 0.8589,
      "step": 5160
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00021138504477264796,
      "loss": 0.8633,
      "step": 5180
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00021103616699616232,
      "loss": 0.8642,
      "step": 5200
    },
    {
      "epoch": 0.9,
      "eval_loss": 0.8629763126373291,
      "eval_runtime": 49.7031,
      "eval_samples_per_second": 40.239,
      "eval_steps_per_second": 1.69,
      "step": 5200
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00021068728921967667,
      "loss": 0.8627,
      "step": 5220
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00021033841144319103,
      "loss": 0.8631,
      "step": 5240
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00020998953366670542,
      "loss": 0.8685,
      "step": 5260
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00020964065589021977,
      "loss": 0.8663,
      "step": 5280
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00020929177811373413,
      "loss": 0.8676,
      "step": 5300
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002089429003372485,
      "loss": 0.8655,
      "step": 5320
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00020859402256076287,
      "loss": 0.8586,
      "step": 5340
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00020824514478427723,
      "loss": 0.8663,
      "step": 5360
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00020789626700779159,
      "loss": 0.855,
      "step": 5380
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00020754738923130594,
      "loss": 0.8652,
      "step": 5400
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.8607897758483887,
      "eval_runtime": 51.0242,
      "eval_samples_per_second": 39.197,
      "eval_steps_per_second": 1.646,
      "step": 5400
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00020719851145482033,
      "loss": 0.8681,
      "step": 5420
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00020684963367833468,
      "loss": 0.8593,
      "step": 5440
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00020650075590184904,
      "loss": 0.8738,
      "step": 5460
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002061518781253634,
      "loss": 0.8658,
      "step": 5480
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00020580300034887775,
      "loss": 0.8593,
      "step": 5500
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00020545412257239214,
      "loss": 0.8563,
      "step": 5520
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002051052447959065,
      "loss": 0.8613,
      "step": 5540
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00020475636701942085,
      "loss": 0.8579,
      "step": 5560
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002044074892429352,
      "loss": 0.8568,
      "step": 5580
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002040586114664496,
      "loss": 0.8549,
      "step": 5600
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.8596189022064209,
      "eval_runtime": 50.2751,
      "eval_samples_per_second": 39.781,
      "eval_steps_per_second": 1.671,
      "step": 5600
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00020370973368996395,
      "loss": 0.8571,
      "step": 5620
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00020336085591347828,
      "loss": 0.8639,
      "step": 5640
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00020301197813699264,
      "loss": 0.8569,
      "step": 5660
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.000202663100360507,
      "loss": 0.8611,
      "step": 5680
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00020231422258402138,
      "loss": 0.8603,
      "step": 5700
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00020196534480753574,
      "loss": 0.8662,
      "step": 5720
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002016164670310501,
      "loss": 0.8684,
      "step": 5740
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00020126758925456445,
      "loss": 0.8732,
      "step": 5760
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00020091871147807883,
      "loss": 0.8532,
      "step": 5780
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002005698337015932,
      "loss": 0.8673,
      "step": 5800
    },
    {
      "epoch": 1.01,
      "eval_loss": 0.8592745065689087,
      "eval_runtime": 49.5912,
      "eval_samples_per_second": 40.33,
      "eval_steps_per_second": 1.694,
      "step": 5800
    }
  ],
  "max_steps": 17298,
  "num_train_epochs": 3,
  "total_flos": 7.068073796791435e+18,
  "trial_name": null,
  "trial_params": null
}