{
  "best_metric": 0.7255927324295044,
  "best_model_checkpoint": "/mnt/bn/qingyi-bn-lq/llama/saved-belle-7b/checkpoint-12200",
  "epoch": 2.93204859161174,
  "global_step": 12400,
  "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": 1.8908,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.00011999999999999999,
      "loss": 1.5545,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.00017999999999999998,
      "loss": 1.1252,
      "step": 60
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00023999999999999998,
      "loss": 1.054,
      "step": 80
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0003,
      "loss": 1.0137,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002995233177087471,
      "loss": 1.0046,
      "step": 120
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002990466354174942,
      "loss": 0.9867,
      "step": 140
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00029856995312624134,
      "loss": 0.9612,
      "step": 160
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002980932708349884,
      "loss": 0.9588,
      "step": 180
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002976165885437356,
      "loss": 0.9551,
      "step": 200
    },
    {
      "epoch": 0.05,
      "eval_loss": 0.9459459185600281,
      "eval_runtime": 19.4211,
      "eval_samples_per_second": 102.981,
      "eval_steps_per_second": 3.244,
      "step": 200
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002971399062524827,
      "loss": 0.9516,
      "step": 220
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00029666322396122984,
      "loss": 0.937,
      "step": 240
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002961865416699769,
      "loss": 0.936,
      "step": 260
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00029570985937872406,
      "loss": 0.9305,
      "step": 280
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00029523317708747115,
      "loss": 0.9146,
      "step": 300
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002947564947962183,
      "loss": 0.9226,
      "step": 320
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00029427981250496543,
      "loss": 0.9108,
      "step": 340
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002938031302137125,
      "loss": 0.9129,
      "step": 360
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00029332644792245965,
      "loss": 0.9063,
      "step": 380
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002928497656312068,
      "loss": 0.8996,
      "step": 400
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.9003962874412537,
      "eval_runtime": 19.5716,
      "eval_samples_per_second": 102.189,
      "eval_steps_per_second": 3.219,
      "step": 400
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00029237308333995393,
      "loss": 0.898,
      "step": 420
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.000291896401048701,
      "loss": 0.8936,
      "step": 440
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00029141971875744815,
      "loss": 0.8932,
      "step": 460
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00029094303646619524,
      "loss": 0.8779,
      "step": 480
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002904663541749424,
      "loss": 0.8871,
      "step": 500
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00028998967188368946,
      "loss": 0.8929,
      "step": 520
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002895129895924366,
      "loss": 0.8878,
      "step": 540
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00028903630730118374,
      "loss": 0.8818,
      "step": 560
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00028855962500993083,
      "loss": 0.8826,
      "step": 580
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00028808294271867797,
      "loss": 0.879,
      "step": 600
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.8738257884979248,
      "eval_runtime": 19.3526,
      "eval_samples_per_second": 103.345,
      "eval_steps_per_second": 3.255,
      "step": 600
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002876062604274251,
      "loss": 0.87,
      "step": 620
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00028712957813617224,
      "loss": 0.8715,
      "step": 640
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00028665289584491933,
      "loss": 0.8724,
      "step": 660
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00028617621355366647,
      "loss": 0.8741,
      "step": 680
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00028569953126241355,
      "loss": 0.8705,
      "step": 700
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002852228489711607,
      "loss": 0.8702,
      "step": 720
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00028474616667990783,
      "loss": 0.8618,
      "step": 740
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002842694843886549,
      "loss": 0.8617,
      "step": 760
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00028379280209740206,
      "loss": 0.8677,
      "step": 780
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002833161198061492,
      "loss": 0.853,
      "step": 800
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.8541846871376038,
      "eval_runtime": 19.49,
      "eval_samples_per_second": 102.617,
      "eval_steps_per_second": 3.232,
      "step": 800
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002828394375148963,
      "loss": 0.8549,
      "step": 820
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002823627552236434,
      "loss": 0.847,
      "step": 840
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00028188607293239056,
      "loss": 0.8585,
      "step": 860
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00028140939064113764,
      "loss": 0.8568,
      "step": 880
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002809327083498848,
      "loss": 0.8482,
      "step": 900
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00028045602605863187,
      "loss": 0.845,
      "step": 920
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.000279979343767379,
      "loss": 0.8548,
      "step": 940
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00027950266147612615,
      "loss": 0.8372,
      "step": 960
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002790259791848733,
      "loss": 0.8423,
      "step": 980
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00027854929689362037,
      "loss": 0.8433,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.8396860361099243,
      "eval_runtime": 19.3459,
      "eval_samples_per_second": 103.381,
      "eval_steps_per_second": 3.257,
      "step": 1000
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002780726146023675,
      "loss": 0.848,
      "step": 1020
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002775959323111146,
      "loss": 0.8383,
      "step": 1040
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00027711925001986173,
      "loss": 0.8385,
      "step": 1060
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002766425677286089,
      "loss": 0.8308,
      "step": 1080
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00027616588543735596,
      "loss": 0.8244,
      "step": 1100
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002756892031461031,
      "loss": 0.835,
      "step": 1120
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002752125208548502,
      "loss": 0.8337,
      "step": 1140
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002747358385635974,
      "loss": 0.8348,
      "step": 1160
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00027425915627234446,
      "loss": 0.8353,
      "step": 1180
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002737824739810916,
      "loss": 0.8294,
      "step": 1200
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.8274422287940979,
      "eval_runtime": 19.4187,
      "eval_samples_per_second": 102.993,
      "eval_steps_per_second": 3.244,
      "step": 1200
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002733057916898387,
      "loss": 0.8337,
      "step": 1220
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002728291093985858,
      "loss": 0.8435,
      "step": 1240
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00027235242710733296,
      "loss": 0.8347,
      "step": 1260
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00027187574481608005,
      "loss": 0.8258,
      "step": 1280
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002713990625248272,
      "loss": 0.8304,
      "step": 1300
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002709223802335743,
      "loss": 0.8264,
      "step": 1320
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002704456979423214,
      "loss": 0.8313,
      "step": 1340
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00026996901565106855,
      "loss": 0.814,
      "step": 1360
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002694923333598157,
      "loss": 0.8223,
      "step": 1380
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002690156510685628,
      "loss": 0.8159,
      "step": 1400
    },
    {
      "epoch": 0.33,
      "eval_loss": 0.8179089426994324,
      "eval_runtime": 19.4736,
      "eval_samples_per_second": 102.703,
      "eval_steps_per_second": 3.235,
      "step": 1400
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002685389687773099,
      "loss": 0.8218,
      "step": 1420
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.000268062286486057,
      "loss": 0.808,
      "step": 1440
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00026758560419480414,
      "loss": 0.8253,
      "step": 1460
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002671089219035513,
      "loss": 0.8174,
      "step": 1480
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00026663223961229836,
      "loss": 0.8157,
      "step": 1500
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002661555573210455,
      "loss": 0.8142,
      "step": 1520
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002656788750297926,
      "loss": 0.8112,
      "step": 1540
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00026520219273853973,
      "loss": 0.8232,
      "step": 1560
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00026472551044728687,
      "loss": 0.8254,
      "step": 1580
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.000264248828156034,
      "loss": 0.8059,
      "step": 1600
    },
    {
      "epoch": 0.38,
      "eval_loss": 0.8101135492324829,
      "eval_runtime": 19.5846,
      "eval_samples_per_second": 102.121,
      "eval_steps_per_second": 3.217,
      "step": 1600
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002637721458647811,
      "loss": 0.8062,
      "step": 1620
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00026329546357352823,
      "loss": 0.805,
      "step": 1640
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002628187812822753,
      "loss": 0.8109,
      "step": 1660
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00026234209899102245,
      "loss": 0.801,
      "step": 1680
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002618654166997696,
      "loss": 0.8043,
      "step": 1700
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002613887344085167,
      "loss": 0.8002,
      "step": 1720
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002609120521172638,
      "loss": 0.8152,
      "step": 1740
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00026043536982601096,
      "loss": 0.8052,
      "step": 1760
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002599586875347581,
      "loss": 0.8136,
      "step": 1780
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002594820052435052,
      "loss": 0.8044,
      "step": 1800
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.8030326962471008,
      "eval_runtime": 19.4835,
      "eval_samples_per_second": 102.651,
      "eval_steps_per_second": 3.234,
      "step": 1800
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002590053229522523,
      "loss": 0.7995,
      "step": 1820
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002585286406609994,
      "loss": 0.7958,
      "step": 1840
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00025805195836974654,
      "loss": 0.8034,
      "step": 1860
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00025757527607849363,
      "loss": 0.8016,
      "step": 1880
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00025709859378724077,
      "loss": 0.8048,
      "step": 1900
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002566219114959879,
      "loss": 0.8004,
      "step": 1920
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00025614522920473505,
      "loss": 0.8041,
      "step": 1940
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00025566854691348213,
      "loss": 0.7908,
      "step": 1960
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.00025519186462222927,
      "loss": 0.7958,
      "step": 1980
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002547151823309764,
      "loss": 0.8013,
      "step": 2000
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.7965430021286011,
      "eval_runtime": 19.4852,
      "eval_samples_per_second": 102.642,
      "eval_steps_per_second": 3.233,
      "step": 2000
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002542385000397235,
      "loss": 0.803,
      "step": 2020
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00025376181774847064,
      "loss": 0.7966,
      "step": 2040
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002532851354572177,
      "loss": 0.7946,
      "step": 2060
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00025280845316596486,
      "loss": 0.8023,
      "step": 2080
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00025233177087471194,
      "loss": 0.7953,
      "step": 2100
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002518550885834591,
      "loss": 0.8053,
      "step": 2120
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002513784062922062,
      "loss": 0.7883,
      "step": 2140
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00025090172400095336,
      "loss": 0.7984,
      "step": 2160
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.00025042504170970045,
      "loss": 0.7962,
      "step": 2180
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002499483594184476,
      "loss": 0.7847,
      "step": 2200
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.7915623784065247,
      "eval_runtime": 19.5509,
      "eval_samples_per_second": 102.297,
      "eval_steps_per_second": 3.222,
      "step": 2200
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002494716771271947,
      "loss": 0.7917,
      "step": 2220
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002489949948359418,
      "loss": 0.7942,
      "step": 2240
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00024851831254468895,
      "loss": 0.7921,
      "step": 2260
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00024804163025343603,
      "loss": 0.7971,
      "step": 2280
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002475649479621832,
      "loss": 0.7919,
      "step": 2300
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002470882656709303,
      "loss": 0.7917,
      "step": 2320
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00024661158337967745,
      "loss": 0.8024,
      "step": 2340
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00024613490108842454,
      "loss": 0.7761,
      "step": 2360
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002456582187971717,
      "loss": 0.7958,
      "step": 2380
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00024518153650591876,
      "loss": 0.7855,
      "step": 2400
    },
    {
      "epoch": 0.57,
      "eval_loss": 0.7870249152183533,
      "eval_runtime": 19.5953,
      "eval_samples_per_second": 102.065,
      "eval_steps_per_second": 3.215,
      "step": 2400
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002447048542146659,
      "loss": 0.784,
      "step": 2420
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00024422817192341304,
      "loss": 0.7926,
      "step": 2440
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00024375148963216013,
      "loss": 0.7845,
      "step": 2460
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00024327480734090726,
      "loss": 0.782,
      "step": 2480
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.00024279812504965438,
      "loss": 0.7808,
      "step": 2500
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00024232144275840152,
      "loss": 0.7926,
      "step": 2520
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00024184476046714863,
      "loss": 0.7795,
      "step": 2540
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00024136807817589574,
      "loss": 0.7888,
      "step": 2560
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00024089139588464288,
      "loss": 0.7888,
      "step": 2580
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00024041471359339,
      "loss": 0.7863,
      "step": 2600
    },
    {
      "epoch": 0.61,
      "eval_loss": 0.7825512290000916,
      "eval_runtime": 19.4274,
      "eval_samples_per_second": 102.948,
      "eval_steps_per_second": 3.243,
      "step": 2600
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002399380313021371,
      "loss": 0.7881,
      "step": 2620
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00023946134901088422,
      "loss": 0.7841,
      "step": 2640
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00023898466671963133,
      "loss": 0.7849,
      "step": 2660
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.00023850798442837844,
      "loss": 0.7809,
      "step": 2680
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002380313021371256,
      "loss": 0.7757,
      "step": 2700
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00023755461984587272,
      "loss": 0.7787,
      "step": 2720
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00023707793755461983,
      "loss": 0.7766,
      "step": 2740
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00023660125526336694,
      "loss": 0.7867,
      "step": 2760
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00023612457297211405,
      "loss": 0.7767,
      "step": 2780
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002356478906808612,
      "loss": 0.7806,
      "step": 2800
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.7781409621238708,
      "eval_runtime": 20.131,
      "eval_samples_per_second": 99.349,
      "eval_steps_per_second": 3.13,
      "step": 2800
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002351712083896083,
      "loss": 0.7774,
      "step": 2820
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00023469452609835542,
      "loss": 0.7782,
      "step": 2840
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00023421784380710253,
      "loss": 0.7773,
      "step": 2860
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00023374116151584964,
      "loss": 0.7845,
      "step": 2880
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002332644792245968,
      "loss": 0.7879,
      "step": 2900
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00023278779693334392,
      "loss": 0.7801,
      "step": 2920
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023231111464209103,
      "loss": 0.7713,
      "step": 2940
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023183443235083814,
      "loss": 0.7742,
      "step": 2960
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00023135775005958526,
      "loss": 0.7783,
      "step": 2980
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002308810677683324,
      "loss": 0.7698,
      "step": 3000
    },
    {
      "epoch": 0.71,
      "eval_loss": 0.7747411131858826,
      "eval_runtime": 20.0968,
      "eval_samples_per_second": 99.519,
      "eval_steps_per_second": 3.135,
      "step": 3000
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002304043854770795,
      "loss": 0.7696,
      "step": 3020
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00022992770318582662,
      "loss": 0.7744,
      "step": 3040
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00022945102089457373,
      "loss": 0.7687,
      "step": 3060
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00022897433860332084,
      "loss": 0.7765,
      "step": 3080
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.000228497656312068,
      "loss": 0.7709,
      "step": 3100
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00022802097402081512,
      "loss": 0.773,
      "step": 3120
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00022754429172956224,
      "loss": 0.7862,
      "step": 3140
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00022706760943830935,
      "loss": 0.7668,
      "step": 3160
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00022659092714705646,
      "loss": 0.7816,
      "step": 3180
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.00022611424485580357,
      "loss": 0.7831,
      "step": 3200
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.7719215154647827,
      "eval_runtime": 19.6387,
      "eval_samples_per_second": 101.84,
      "eval_steps_per_second": 3.208,
      "step": 3200
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002256375625645507,
      "loss": 0.7723,
      "step": 3220
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00022516088027329782,
      "loss": 0.7727,
      "step": 3240
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00022468419798204493,
      "loss": 0.7719,
      "step": 3260
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00022420751569079207,
      "loss": 0.7796,
      "step": 3280
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002237308333995392,
      "loss": 0.7685,
      "step": 3300
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00022325415110828633,
      "loss": 0.7725,
      "step": 3320
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00022277746881703344,
      "loss": 0.7638,
      "step": 3340
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00022230078652578055,
      "loss": 0.7771,
      "step": 3360
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00022182410423452766,
      "loss": 0.7689,
      "step": 3380
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00022134742194327477,
      "loss": 0.7797,
      "step": 3400
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.768983006477356,
      "eval_runtime": 19.4428,
      "eval_samples_per_second": 102.866,
      "eval_steps_per_second": 3.24,
      "step": 3400
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002208707396520219,
      "loss": 0.7734,
      "step": 3420
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00022039405736076903,
      "loss": 0.7719,
      "step": 3440
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00021991737506951614,
      "loss": 0.767,
      "step": 3460
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00021944069277826328,
      "loss": 0.7758,
      "step": 3480
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002189640104870104,
      "loss": 0.7768,
      "step": 3500
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00021848732819575753,
      "loss": 0.7641,
      "step": 3520
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00021801064590450464,
      "loss": 0.7694,
      "step": 3540
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00021753396361325175,
      "loss": 0.7835,
      "step": 3560
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00021705728132199886,
      "loss": 0.7642,
      "step": 3580
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00021658059903074598,
      "loss": 0.7719,
      "step": 3600
    },
    {
      "epoch": 0.85,
      "eval_loss": 0.7660636305809021,
      "eval_runtime": 19.5996,
      "eval_samples_per_second": 102.043,
      "eval_steps_per_second": 3.214,
      "step": 3600
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002161039167394931,
      "loss": 0.7723,
      "step": 3620
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00021562723444824023,
      "loss": 0.76,
      "step": 3640
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00021515055215698734,
      "loss": 0.7643,
      "step": 3660
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00021467386986573448,
      "loss": 0.7599,
      "step": 3680
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002141971875744816,
      "loss": 0.7623,
      "step": 3700
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002137205052832287,
      "loss": 0.7621,
      "step": 3720
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00021324382299197584,
      "loss": 0.7691,
      "step": 3740
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00021276714070072295,
      "loss": 0.7665,
      "step": 3760
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00021229045840947007,
      "loss": 0.7742,
      "step": 3780
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00021181377611821718,
      "loss": 0.7624,
      "step": 3800
    },
    {
      "epoch": 0.9,
      "eval_loss": 0.7643172740936279,
      "eval_runtime": 19.487,
      "eval_samples_per_second": 102.633,
      "eval_steps_per_second": 3.233,
      "step": 3800
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002113370938269643,
      "loss": 0.7726,
      "step": 3820
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002108604115357114,
      "loss": 0.7559,
      "step": 3840
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00021038372924445857,
      "loss": 0.7634,
      "step": 3860
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00020990704695320568,
      "loss": 0.765,
      "step": 3880
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002094303646619528,
      "loss": 0.7649,
      "step": 3900
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002089536823706999,
      "loss": 0.763,
      "step": 3920
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00020847700007944705,
      "loss": 0.7679,
      "step": 3940
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00020800031778819416,
      "loss": 0.7644,
      "step": 3960
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00020752363549694127,
      "loss": 0.7655,
      "step": 3980
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00020704695320568838,
      "loss": 0.7681,
      "step": 4000
    },
    {
      "epoch": 0.95,
      "eval_loss": 0.7610963582992554,
      "eval_runtime": 19.5269,
      "eval_samples_per_second": 102.423,
      "eval_steps_per_second": 3.226,
      "step": 4000
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002065702709144355,
      "loss": 0.7623,
      "step": 4020
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002060935886231826,
      "loss": 0.7625,
      "step": 4040
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00020561690633192977,
      "loss": 0.7524,
      "step": 4060
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00020514022404067688,
      "loss": 0.764,
      "step": 4080
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.000204663541749424,
      "loss": 0.7513,
      "step": 4100
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002041868594581711,
      "loss": 0.753,
      "step": 4120
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00020371017716691822,
      "loss": 0.7602,
      "step": 4140
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00020323349487566536,
      "loss": 0.7701,
      "step": 4160
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00020275681258441247,
      "loss": 0.7602,
      "step": 4180
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00020228013029315958,
      "loss": 0.7598,
      "step": 4200
    },
    {
      "epoch": 0.99,
      "eval_loss": 0.760128915309906,
      "eval_runtime": 19.4387,
      "eval_samples_per_second": 102.888,
      "eval_steps_per_second": 3.241,
      "step": 4200
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002018034480019067,
      "loss": 0.7579,
      "step": 4220
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00020132676571065384,
      "loss": 0.7628,
      "step": 4240
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00020085008341940097,
      "loss": 0.7551,
      "step": 4260
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002003734011281481,
      "loss": 0.7582,
      "step": 4280
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0001998967188368952,
      "loss": 0.7623,
      "step": 4300
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0001994200365456423,
      "loss": 0.7504,
      "step": 4320
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00019894335425438942,
      "loss": 0.7587,
      "step": 4340
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00019846667196313654,
      "loss": 0.7528,
      "step": 4360
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00019798998967188367,
      "loss": 0.754,
      "step": 4380
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00019751330738063079,
      "loss": 0.759,
      "step": 4400
    },
    {
      "epoch": 1.04,
      "eval_loss": 0.7575392127037048,
      "eval_runtime": 19.5275,
      "eval_samples_per_second": 102.42,
      "eval_steps_per_second": 3.226,
      "step": 4400
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0001970366250893779,
      "loss": 0.7592,
      "step": 4420
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00019655994279812504,
      "loss": 0.7548,
      "step": 4440
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00019608326050687218,
      "loss": 0.7632,
      "step": 4460
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001956065782156193,
      "loss": 0.7472,
      "step": 4480
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0001951298959243664,
      "loss": 0.7496,
      "step": 4500
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001946532136331135,
      "loss": 0.7549,
      "step": 4520
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00019417653134186063,
      "loss": 0.77,
      "step": 4540
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00019369984905060774,
      "loss": 0.759,
      "step": 4560
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00019322316675935488,
      "loss": 0.7554,
      "step": 4580
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.000192746484468102,
      "loss": 0.7577,
      "step": 4600
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.7568497061729431,
      "eval_runtime": 19.53,
      "eval_samples_per_second": 102.406,
      "eval_steps_per_second": 3.226,
      "step": 4600
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0001922698021768491,
      "loss": 0.7617,
      "step": 4620
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00019179311988559624,
      "loss": 0.7551,
      "step": 4640
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00019131643759434335,
      "loss": 0.7482,
      "step": 4660
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0001908397553030905,
      "loss": 0.7516,
      "step": 4680
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0001903630730118376,
      "loss": 0.7555,
      "step": 4700
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00018988639072058472,
      "loss": 0.7605,
      "step": 4720
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00018940970842933183,
      "loss": 0.7506,
      "step": 4740
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00018893302613807894,
      "loss": 0.7622,
      "step": 4760
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00018845634384682605,
      "loss": 0.75,
      "step": 4780
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001879796615555732,
      "loss": 0.7572,
      "step": 4800
    },
    {
      "epoch": 1.13,
      "eval_loss": 0.7548028826713562,
      "eval_runtime": 19.5411,
      "eval_samples_per_second": 102.349,
      "eval_steps_per_second": 3.224,
      "step": 4800
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00018750297926432033,
      "loss": 0.7427,
      "step": 4820
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00018702629697306744,
      "loss": 0.7489,
      "step": 4840
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00018654961468181455,
      "loss": 0.755,
      "step": 4860
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00018607293239056167,
      "loss": 0.7517,
      "step": 4880
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0001855962500993088,
      "loss": 0.7529,
      "step": 4900
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00018511956780805592,
      "loss": 0.7498,
      "step": 4920
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00018464288551680303,
      "loss": 0.756,
      "step": 4940
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00018416620322555014,
      "loss": 0.7492,
      "step": 4960
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00018368952093429725,
      "loss": 0.7491,
      "step": 4980
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00018321283864304437,
      "loss": 0.7585,
      "step": 5000
    },
    {
      "epoch": 1.18,
      "eval_loss": 0.7538104057312012,
      "eval_runtime": 19.6106,
      "eval_samples_per_second": 101.986,
      "eval_steps_per_second": 3.213,
      "step": 5000
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00018273615635179153,
      "loss": 0.7531,
      "step": 5020
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00018225947406053865,
      "loss": 0.7511,
      "step": 5040
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00018178279176928576,
      "loss": 0.7541,
      "step": 5060
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00018130610947803287,
      "loss": 0.7465,
      "step": 5080
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00018082942718678,
      "loss": 0.7403,
      "step": 5100
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00018035274489552712,
      "loss": 0.749,
      "step": 5120
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00017987606260427423,
      "loss": 0.7548,
      "step": 5140
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00017939938031302134,
      "loss": 0.7443,
      "step": 5160
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00017892269802176846,
      "loss": 0.7461,
      "step": 5180
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00017844601573051557,
      "loss": 0.7511,
      "step": 5200
    },
    {
      "epoch": 1.23,
      "eval_loss": 0.7509217262268066,
      "eval_runtime": 19.5437,
      "eval_samples_per_second": 102.335,
      "eval_steps_per_second": 3.224,
      "step": 5200
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00017796933343926274,
      "loss": 0.7562,
      "step": 5220
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00017749265114800985,
      "loss": 0.7489,
      "step": 5240
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.00017701596885675696,
      "loss": 0.7499,
      "step": 5260
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00017653928656550407,
      "loss": 0.7519,
      "step": 5280
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00017606260427425118,
      "loss": 0.7536,
      "step": 5300
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00017558592198299832,
      "loss": 0.7536,
      "step": 5320
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00017510923969174544,
      "loss": 0.7492,
      "step": 5340
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00017463255740049255,
      "loss": 0.7454,
      "step": 5360
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00017415587510923966,
      "loss": 0.7528,
      "step": 5380
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0001736791928179868,
      "loss": 0.7409,
      "step": 5400
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.7497395873069763,
      "eval_runtime": 19.5671,
      "eval_samples_per_second": 102.212,
      "eval_steps_per_second": 3.22,
      "step": 5400
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00017320251052673394,
      "loss": 0.7434,
      "step": 5420
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00017272582823548105,
      "loss": 0.7543,
      "step": 5440
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00017224914594422816,
      "loss": 0.7457,
      "step": 5460
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00017177246365297527,
      "loss": 0.7439,
      "step": 5480
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0001712957813617224,
      "loss": 0.7412,
      "step": 5500
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0001708190990704695,
      "loss": 0.7409,
      "step": 5520
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00017034241677921664,
      "loss": 0.7473,
      "step": 5540
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00016986573448796375,
      "loss": 0.7486,
      "step": 5560
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00016938905219671086,
      "loss": 0.7439,
      "step": 5580
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.000168912369905458,
      "loss": 0.7524,
      "step": 5600
    },
    {
      "epoch": 1.32,
      "eval_loss": 0.7480019330978394,
      "eval_runtime": 19.5018,
      "eval_samples_per_second": 102.555,
      "eval_steps_per_second": 3.23,
      "step": 5600
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00016843568761420514,
      "loss": 0.7464,
      "step": 5620
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00016795900532295225,
      "loss": 0.7511,
      "step": 5640
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00016748232303169936,
      "loss": 0.7423,
      "step": 5660
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00016700564074044648,
      "loss": 0.7422,
      "step": 5680
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0001665289584491936,
      "loss": 0.742,
      "step": 5700
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0001660522761579407,
      "loss": 0.7421,
      "step": 5720
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00016557559386668784,
      "loss": 0.749,
      "step": 5740
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00016509891157543495,
      "loss": 0.7432,
      "step": 5760
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001646222292841821,
      "loss": 0.7426,
      "step": 5780
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001641455469929292,
      "loss": 0.7543,
      "step": 5800
    },
    {
      "epoch": 1.37,
      "eval_loss": 0.7470090389251709,
      "eval_runtime": 19.5563,
      "eval_samples_per_second": 102.269,
      "eval_steps_per_second": 3.221,
      "step": 5800
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00016366886470167632,
      "loss": 0.7451,
      "step": 5820
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00016319218241042346,
      "loss": 0.7481,
      "step": 5840
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00016271550011917057,
      "loss": 0.7381,
      "step": 5860
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00016223881782791768,
      "loss": 0.7461,
      "step": 5880
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0001617621355366648,
      "loss": 0.7467,
      "step": 5900
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0001612854532454119,
      "loss": 0.745,
      "step": 5920
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00016080877095415902,
      "loss": 0.745,
      "step": 5940
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00016033208866290615,
      "loss": 0.7386,
      "step": 5960
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0001598554063716533,
      "loss": 0.7363,
      "step": 5980
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0001593787240804004,
      "loss": 0.7412,
      "step": 6000
    },
    {
      "epoch": 1.42,
      "eval_loss": 0.7454522848129272,
      "eval_runtime": 19.555,
      "eval_samples_per_second": 102.276,
      "eval_steps_per_second": 3.222,
      "step": 6000
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00015890204178914752,
      "loss": 0.7501,
      "step": 6020
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00015842535949789463,
      "loss": 0.7528,
      "step": 6040
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00015794867720664177,
      "loss": 0.7373,
      "step": 6060
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00015747199491538888,
      "loss": 0.7451,
      "step": 6080
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.000156995312624136,
      "loss": 0.7384,
      "step": 6100
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0001565186303328831,
      "loss": 0.7471,
      "step": 6120
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00015604194804163022,
      "loss": 0.7454,
      "step": 6140
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00015556526575037733,
      "loss": 0.7415,
      "step": 6160
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0001550885834591245,
      "loss": 0.7514,
      "step": 6180
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0001546119011678716,
      "loss": 0.7343,
      "step": 6200
    },
    {
      "epoch": 1.47,
      "eval_loss": 0.7457332611083984,
      "eval_runtime": 19.5673,
      "eval_samples_per_second": 102.212,
      "eval_steps_per_second": 3.22,
      "step": 6200
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00015413521887661872,
      "loss": 0.7452,
      "step": 6220
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00015365853658536583,
      "loss": 0.7456,
      "step": 6240
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00015318185429411297,
      "loss": 0.7326,
      "step": 6260
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00015270517200286008,
      "loss": 0.7431,
      "step": 6280
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001522284897116072,
      "loss": 0.7419,
      "step": 6300
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0001517518074203543,
      "loss": 0.7375,
      "step": 6320
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.00015127512512910142,
      "loss": 0.7419,
      "step": 6340
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001507984428378486,
      "loss": 0.7431,
      "step": 6360
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0001503217605465957,
      "loss": 0.7412,
      "step": 6380
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00014984507825534278,
      "loss": 0.7447,
      "step": 6400
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.7441338896751404,
      "eval_runtime": 19.4509,
      "eval_samples_per_second": 102.823,
      "eval_steps_per_second": 3.239,
      "step": 6400
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00014936839596408992,
      "loss": 0.7436,
      "step": 6420
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00014889171367283704,
      "loss": 0.7402,
      "step": 6440
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00014841503138158415,
      "loss": 0.7454,
      "step": 6460
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0001479383490903313,
      "loss": 0.738,
      "step": 6480
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001474616667990784,
      "loss": 0.7396,
      "step": 6500
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.00014698498450782554,
      "loss": 0.7333,
      "step": 6520
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.00014650830221657265,
      "loss": 0.7482,
      "step": 6540
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.00014603161992531976,
      "loss": 0.7376,
      "step": 6560
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00014555493763406687,
      "loss": 0.7369,
      "step": 6580
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00014507825534281401,
      "loss": 0.7347,
      "step": 6600
    },
    {
      "epoch": 1.56,
      "eval_loss": 0.7425362467765808,
      "eval_runtime": 19.5248,
      "eval_samples_per_second": 102.434,
      "eval_steps_per_second": 3.227,
      "step": 6600
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00014460157305156113,
      "loss": 0.7446,
      "step": 6620
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00014412489076030824,
      "loss": 0.7343,
      "step": 6640
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00014364820846905535,
      "loss": 0.7468,
      "step": 6660
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0001431715261778025,
      "loss": 0.749,
      "step": 6680
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0001426948438865496,
      "loss": 0.7401,
      "step": 6700
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0001422181615952967,
      "loss": 0.7364,
      "step": 6720
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00014174147930404385,
      "loss": 0.7442,
      "step": 6740
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00014126479701279096,
      "loss": 0.7385,
      "step": 6760
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00014078811472153808,
      "loss": 0.7412,
      "step": 6780
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00014031143243028522,
      "loss": 0.7377,
      "step": 6800
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.7418386936187744,
      "eval_runtime": 19.5679,
      "eval_samples_per_second": 102.208,
      "eval_steps_per_second": 3.22,
      "step": 6800
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00013983475013903233,
      "loss": 0.7432,
      "step": 6820
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00013935806784777944,
      "loss": 0.7379,
      "step": 6840
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00013888138555652655,
      "loss": 0.7346,
      "step": 6860
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00013840470326527366,
      "loss": 0.7373,
      "step": 6880
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0001379280209740208,
      "loss": 0.7403,
      "step": 6900
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00013745133868276792,
      "loss": 0.7477,
      "step": 6920
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00013697465639151506,
      "loss": 0.7343,
      "step": 6940
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00013649797410026217,
      "loss": 0.7419,
      "step": 6960
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00013602129180900928,
      "loss": 0.7327,
      "step": 6980
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00013554460951775642,
      "loss": 0.7398,
      "step": 7000
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.7402775883674622,
      "eval_runtime": 19.5554,
      "eval_samples_per_second": 102.274,
      "eval_steps_per_second": 3.222,
      "step": 7000
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00013506792722650353,
      "loss": 0.7311,
      "step": 7020
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00013459124493525064,
      "loss": 0.7319,
      "step": 7040
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.00013411456264399775,
      "loss": 0.7315,
      "step": 7060
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0001336378803527449,
      "loss": 0.7329,
      "step": 7080
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.000133161198061492,
      "loss": 0.7471,
      "step": 7100
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00013268451577023912,
      "loss": 0.7446,
      "step": 7120
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00013220783347898623,
      "loss": 0.7359,
      "step": 7140
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00013173115118773337,
      "loss": 0.7348,
      "step": 7160
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00013125446889648048,
      "loss": 0.7331,
      "step": 7180
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00013077778660522762,
      "loss": 0.7385,
      "step": 7200
    },
    {
      "epoch": 1.7,
      "eval_loss": 0.7401012182235718,
      "eval_runtime": 19.7831,
      "eval_samples_per_second": 101.096,
      "eval_steps_per_second": 3.185,
      "step": 7200
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00013030110431397473,
      "loss": 0.744,
      "step": 7220
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00012982442202272185,
      "loss": 0.7327,
      "step": 7240
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00012934773973146896,
      "loss": 0.7384,
      "step": 7260
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0001288710574402161,
      "loss": 0.7399,
      "step": 7280
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0001283943751489632,
      "loss": 0.7376,
      "step": 7300
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00012791769285771032,
      "loss": 0.7416,
      "step": 7320
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00012744101056645743,
      "loss": 0.7299,
      "step": 7340
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00012696432827520455,
      "loss": 0.7389,
      "step": 7360
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00012648764598395168,
      "loss": 0.7295,
      "step": 7380
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0001260109636926988,
      "loss": 0.7389,
      "step": 7400
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.7385362386703491,
      "eval_runtime": 19.6728,
      "eval_samples_per_second": 101.663,
      "eval_steps_per_second": 3.202,
      "step": 7400
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00012553428140144594,
      "loss": 0.7346,
      "step": 7420
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00012505759911019305,
      "loss": 0.7357,
      "step": 7440
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00012458091681894016,
      "loss": 0.7295,
      "step": 7460
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0001241042345276873,
      "loss": 0.7418,
      "step": 7480
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0001236275522364344,
      "loss": 0.7248,
      "step": 7500
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00012315086994518152,
      "loss": 0.7326,
      "step": 7520
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.00012267418765392864,
      "loss": 0.7422,
      "step": 7540
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00012219750536267577,
      "loss": 0.7376,
      "step": 7560
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00012172082307142289,
      "loss": 0.7358,
      "step": 7580
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00012124414078017001,
      "loss": 0.7337,
      "step": 7600
    },
    {
      "epoch": 1.8,
      "eval_loss": 0.737734854221344,
      "eval_runtime": 19.8317,
      "eval_samples_per_second": 100.849,
      "eval_steps_per_second": 3.177,
      "step": 7600
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00012076745848891712,
      "loss": 0.7318,
      "step": 7620
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00012029077619766424,
      "loss": 0.7356,
      "step": 7640
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00011981409390641138,
      "loss": 0.7355,
      "step": 7660
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00011933741161515849,
      "loss": 0.74,
      "step": 7680
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0001188607293239056,
      "loss": 0.7342,
      "step": 7700
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00011838404703265273,
      "loss": 0.7368,
      "step": 7720
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00011790736474139984,
      "loss": 0.7337,
      "step": 7740
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.00011743068245014698,
      "loss": 0.7317,
      "step": 7760
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00011695400015889409,
      "loss": 0.738,
      "step": 7780
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0001164773178676412,
      "loss": 0.7375,
      "step": 7800
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.7366506457328796,
      "eval_runtime": 19.9586,
      "eval_samples_per_second": 100.208,
      "eval_steps_per_second": 3.157,
      "step": 7800
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00011600063557638833,
      "loss": 0.7349,
      "step": 7820
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00011552395328513544,
      "loss": 0.733,
      "step": 7840
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00011504727099388258,
      "loss": 0.7277,
      "step": 7860
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00011457058870262969,
      "loss": 0.7235,
      "step": 7880
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001140939064113768,
      "loss": 0.7405,
      "step": 7900
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00011361722412012393,
      "loss": 0.7378,
      "step": 7920
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00011314054182887104,
      "loss": 0.7292,
      "step": 7940
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00011266385953761818,
      "loss": 0.7427,
      "step": 7960
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.00011218717724636529,
      "loss": 0.7313,
      "step": 7980
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0001117104949551124,
      "loss": 0.7252,
      "step": 8000
    },
    {
      "epoch": 1.89,
      "eval_loss": 0.736083984375,
      "eval_runtime": 19.7958,
      "eval_samples_per_second": 101.031,
      "eval_steps_per_second": 3.182,
      "step": 8000
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00011123381266385953,
      "loss": 0.7268,
      "step": 8020
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00011075713037260666,
      "loss": 0.729,
      "step": 8040
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00011028044808135377,
      "loss": 0.7358,
      "step": 8060
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010980376579010089,
      "loss": 0.7408,
      "step": 8080
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.000109327083498848,
      "loss": 0.73,
      "step": 8100
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010887423532215777,
      "loss": 0.7298,
      "step": 8120
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0001083975530309049,
      "loss": 0.7324,
      "step": 8140
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010792087073965201,
      "loss": 0.7296,
      "step": 8160
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010744418844839912,
      "loss": 0.7346,
      "step": 8180
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010696750615714626,
      "loss": 0.7281,
      "step": 8200
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.7352190613746643,
      "eval_runtime": 19.6635,
      "eval_samples_per_second": 101.711,
      "eval_steps_per_second": 3.204,
      "step": 8200
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010649082386589337,
      "loss": 0.7377,
      "step": 8220
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0001060141415746405,
      "loss": 0.7281,
      "step": 8240
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010553745928338761,
      "loss": 0.7251,
      "step": 8260
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010506077699213472,
      "loss": 0.7331,
      "step": 8280
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010458409470088186,
      "loss": 0.7432,
      "step": 8300
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010410741240962897,
      "loss": 0.7366,
      "step": 8320
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0001036307301183761,
      "loss": 0.7334,
      "step": 8340
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010315404782712321,
      "loss": 0.7351,
      "step": 8360
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.00010267736553587032,
      "loss": 0.7355,
      "step": 8380
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010220068324461746,
      "loss": 0.7228,
      "step": 8400
    },
    {
      "epoch": 1.99,
      "eval_loss": 0.7341500520706177,
      "eval_runtime": 19.6196,
      "eval_samples_per_second": 101.939,
      "eval_steps_per_second": 3.211,
      "step": 8400
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010172400095336457,
      "loss": 0.7451,
      "step": 8420
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010124731866211169,
      "loss": 0.7356,
      "step": 8440
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00010077063637085881,
      "loss": 0.7255,
      "step": 8460
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.00010029395407960592,
      "loss": 0.7267,
      "step": 8480
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.981727178835306e-05,
      "loss": 0.7291,
      "step": 8500
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.934058949710018e-05,
      "loss": 0.7294,
      "step": 8520
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.886390720584729e-05,
      "loss": 0.7377,
      "step": 8540
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.838722491459441e-05,
      "loss": 0.7324,
      "step": 8560
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.791054262334154e-05,
      "loss": 0.7286,
      "step": 8580
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.743386033208867e-05,
      "loss": 0.7286,
      "step": 8600
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.734474241733551,
      "eval_runtime": 19.5642,
      "eval_samples_per_second": 102.228,
      "eval_steps_per_second": 3.22,
      "step": 8600
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.695717804083578e-05,
      "loss": 0.7304,
      "step": 8620
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.648049574958289e-05,
      "loss": 0.7348,
      "step": 8640
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.600381345833002e-05,
      "loss": 0.7261,
      "step": 8660
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.552713116707714e-05,
      "loss": 0.7313,
      "step": 8680
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.505044887582425e-05,
      "loss": 0.7379,
      "step": 8700
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.457376658457138e-05,
      "loss": 0.7203,
      "step": 8720
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.409708429331849e-05,
      "loss": 0.7306,
      "step": 8740
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.36204020020656e-05,
      "loss": 0.7332,
      "step": 8760
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.314371971081274e-05,
      "loss": 0.7228,
      "step": 8780
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.266703741955985e-05,
      "loss": 0.731,
      "step": 8800
    },
    {
      "epoch": 2.08,
      "eval_loss": 0.7332338690757751,
      "eval_runtime": 19.7114,
      "eval_samples_per_second": 101.464,
      "eval_steps_per_second": 3.196,
      "step": 8800
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.219035512830698e-05,
      "loss": 0.7267,
      "step": 8820
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.171367283705409e-05,
      "loss": 0.7285,
      "step": 8840
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.12369905458012e-05,
      "loss": 0.7214,
      "step": 8860
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.076030825454834e-05,
      "loss": 0.7204,
      "step": 8880
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.028362596329546e-05,
      "loss": 0.7253,
      "step": 8900
    },
    {
      "epoch": 2.11,
      "learning_rate": 8.980694367204258e-05,
      "loss": 0.7253,
      "step": 8920
    },
    {
      "epoch": 2.11,
      "learning_rate": 8.933026138078969e-05,
      "loss": 0.7238,
      "step": 8940
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.88535790895368e-05,
      "loss": 0.7286,
      "step": 8960
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.837689679828394e-05,
      "loss": 0.7385,
      "step": 8980
    },
    {
      "epoch": 2.13,
      "learning_rate": 8.790021450703106e-05,
      "loss": 0.7237,
      "step": 9000
    },
    {
      "epoch": 2.13,
      "eval_loss": 0.7329864501953125,
      "eval_runtime": 19.7024,
      "eval_samples_per_second": 101.51,
      "eval_steps_per_second": 3.198,
      "step": 9000
    },
    {
      "epoch": 2.13,
      "learning_rate": 8.742353221577817e-05,
      "loss": 0.7311,
      "step": 9020
    },
    {
      "epoch": 2.14,
      "learning_rate": 8.69468499245253e-05,
      "loss": 0.7374,
      "step": 9040
    },
    {
      "epoch": 2.14,
      "learning_rate": 8.64701676332724e-05,
      "loss": 0.7194,
      "step": 9060
    },
    {
      "epoch": 2.15,
      "learning_rate": 8.599348534201955e-05,
      "loss": 0.7237,
      "step": 9080
    },
    {
      "epoch": 2.15,
      "learning_rate": 8.551680305076666e-05,
      "loss": 0.7287,
      "step": 9100
    },
    {
      "epoch": 2.16,
      "learning_rate": 8.504012075951377e-05,
      "loss": 0.7385,
      "step": 9120
    },
    {
      "epoch": 2.16,
      "learning_rate": 8.45634384682609e-05,
      "loss": 0.7319,
      "step": 9140
    },
    {
      "epoch": 2.17,
      "learning_rate": 8.408675617700802e-05,
      "loss": 0.7278,
      "step": 9160
    },
    {
      "epoch": 2.17,
      "learning_rate": 8.361007388575515e-05,
      "loss": 0.7293,
      "step": 9180
    },
    {
      "epoch": 2.18,
      "learning_rate": 8.313339159450226e-05,
      "loss": 0.7232,
      "step": 9200
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.7326176762580872,
      "eval_runtime": 20.1581,
      "eval_samples_per_second": 99.215,
      "eval_steps_per_second": 3.125,
      "step": 9200
    },
    {
      "epoch": 2.18,
      "learning_rate": 8.265670930324937e-05,
      "loss": 0.7281,
      "step": 9220
    },
    {
      "epoch": 2.18,
      "learning_rate": 8.21800270119965e-05,
      "loss": 0.728,
      "step": 9240
    },
    {
      "epoch": 2.19,
      "learning_rate": 8.170334472074362e-05,
      "loss": 0.728,
      "step": 9260
    },
    {
      "epoch": 2.19,
      "learning_rate": 8.122666242949073e-05,
      "loss": 0.7221,
      "step": 9280
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.074998013823786e-05,
      "loss": 0.7242,
      "step": 9300
    },
    {
      "epoch": 2.2,
      "learning_rate": 8.027329784698497e-05,
      "loss": 0.7306,
      "step": 9320
    },
    {
      "epoch": 2.21,
      "learning_rate": 7.979661555573208e-05,
      "loss": 0.7218,
      "step": 9340
    },
    {
      "epoch": 2.21,
      "learning_rate": 7.931993326447922e-05,
      "loss": 0.7289,
      "step": 9360
    },
    {
      "epoch": 2.22,
      "learning_rate": 7.884325097322634e-05,
      "loss": 0.7177,
      "step": 9380
    },
    {
      "epoch": 2.22,
      "learning_rate": 7.836656868197346e-05,
      "loss": 0.7265,
      "step": 9400
    },
    {
      "epoch": 2.22,
      "eval_loss": 0.7311453819274902,
      "eval_runtime": 19.9076,
      "eval_samples_per_second": 100.464,
      "eval_steps_per_second": 3.165,
      "step": 9400
    },
    {
      "epoch": 2.23,
      "learning_rate": 7.788988639072057e-05,
      "loss": 0.7269,
      "step": 9420
    },
    {
      "epoch": 2.23,
      "learning_rate": 7.741320409946769e-05,
      "loss": 0.7275,
      "step": 9440
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.693652180821483e-05,
      "loss": 0.7317,
      "step": 9460
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.645983951696194e-05,
      "loss": 0.7344,
      "step": 9480
    },
    {
      "epoch": 2.25,
      "learning_rate": 7.598315722570906e-05,
      "loss": 0.7263,
      "step": 9500
    },
    {
      "epoch": 2.25,
      "learning_rate": 7.550647493445617e-05,
      "loss": 0.7299,
      "step": 9520
    },
    {
      "epoch": 2.26,
      "learning_rate": 7.502979264320329e-05,
      "loss": 0.724,
      "step": 9540
    },
    {
      "epoch": 2.26,
      "learning_rate": 7.455311035195041e-05,
      "loss": 0.7266,
      "step": 9560
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.407642806069754e-05,
      "loss": 0.7299,
      "step": 9580
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.359974576944465e-05,
      "loss": 0.7236,
      "step": 9600
    },
    {
      "epoch": 2.27,
      "eval_loss": 0.7311366200447083,
      "eval_runtime": 20.0053,
      "eval_samples_per_second": 99.973,
      "eval_steps_per_second": 3.149,
      "step": 9600
    },
    {
      "epoch": 2.27,
      "learning_rate": 7.314689759275442e-05,
      "loss": 0.7252,
      "step": 9620
    },
    {
      "epoch": 2.28,
      "learning_rate": 7.267021530150154e-05,
      "loss": 0.7252,
      "step": 9640
    },
    {
      "epoch": 2.28,
      "learning_rate": 7.219353301024865e-05,
      "loss": 0.7188,
      "step": 9660
    },
    {
      "epoch": 2.29,
      "learning_rate": 7.171685071899578e-05,
      "loss": 0.7243,
      "step": 9680
    },
    {
      "epoch": 2.29,
      "learning_rate": 7.12401684277429e-05,
      "loss": 0.7298,
      "step": 9700
    },
    {
      "epoch": 2.3,
      "learning_rate": 7.076348613649002e-05,
      "loss": 0.7325,
      "step": 9720
    },
    {
      "epoch": 2.3,
      "learning_rate": 7.028680384523714e-05,
      "loss": 0.7286,
      "step": 9740
    },
    {
      "epoch": 2.31,
      "learning_rate": 6.981012155398426e-05,
      "loss": 0.7201,
      "step": 9760
    },
    {
      "epoch": 2.31,
      "learning_rate": 6.933343926273138e-05,
      "loss": 0.7184,
      "step": 9780
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.885675697147851e-05,
      "loss": 0.7291,
      "step": 9800
    },
    {
      "epoch": 2.32,
      "eval_loss": 0.7308618426322937,
      "eval_runtime": 19.7965,
      "eval_samples_per_second": 101.028,
      "eval_steps_per_second": 3.182,
      "step": 9800
    },
    {
      "epoch": 2.32,
      "learning_rate": 6.838007468022563e-05,
      "loss": 0.7318,
      "step": 9820
    },
    {
      "epoch": 2.33,
      "learning_rate": 6.790339238897274e-05,
      "loss": 0.7227,
      "step": 9840
    },
    {
      "epoch": 2.33,
      "learning_rate": 6.742671009771986e-05,
      "loss": 0.7377,
      "step": 9860
    },
    {
      "epoch": 2.34,
      "learning_rate": 6.695002780646698e-05,
      "loss": 0.7367,
      "step": 9880
    },
    {
      "epoch": 2.34,
      "learning_rate": 6.647334551521411e-05,
      "loss": 0.7218,
      "step": 9900
    },
    {
      "epoch": 2.35,
      "learning_rate": 6.599666322396122e-05,
      "loss": 0.7282,
      "step": 9920
    },
    {
      "epoch": 2.35,
      "learning_rate": 6.551998093270835e-05,
      "loss": 0.7231,
      "step": 9940
    },
    {
      "epoch": 2.36,
      "learning_rate": 6.504329864145546e-05,
      "loss": 0.7257,
      "step": 9960
    },
    {
      "epoch": 2.36,
      "learning_rate": 6.456661635020258e-05,
      "loss": 0.7275,
      "step": 9980
    },
    {
      "epoch": 2.36,
      "learning_rate": 6.40899340589497e-05,
      "loss": 0.725,
      "step": 10000
    },
    {
      "epoch": 2.36,
      "eval_loss": 0.7301817536354065,
      "eval_runtime": 19.7914,
      "eval_samples_per_second": 101.054,
      "eval_steps_per_second": 3.183,
      "step": 10000
    },
    {
      "epoch": 2.37,
      "learning_rate": 6.361325176769682e-05,
      "loss": 0.72,
      "step": 10020
    },
    {
      "epoch": 2.37,
      "learning_rate": 6.313656947644395e-05,
      "loss": 0.7267,
      "step": 10040
    },
    {
      "epoch": 2.38,
      "learning_rate": 6.265988718519107e-05,
      "loss": 0.7276,
      "step": 10060
    },
    {
      "epoch": 2.38,
      "learning_rate": 6.218320489393818e-05,
      "loss": 0.7262,
      "step": 10080
    },
    {
      "epoch": 2.39,
      "learning_rate": 6.17065226026853e-05,
      "loss": 0.7149,
      "step": 10100
    },
    {
      "epoch": 2.39,
      "learning_rate": 6.122984031143242e-05,
      "loss": 0.7305,
      "step": 10120
    },
    {
      "epoch": 2.4,
      "learning_rate": 6.075315802017954e-05,
      "loss": 0.7314,
      "step": 10140
    },
    {
      "epoch": 2.4,
      "learning_rate": 6.027647572892667e-05,
      "loss": 0.7154,
      "step": 10160
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.9799793437673786e-05,
      "loss": 0.7263,
      "step": 10180
    },
    {
      "epoch": 2.41,
      "learning_rate": 5.93231111464209e-05,
      "loss": 0.7203,
      "step": 10200
    },
    {
      "epoch": 2.41,
      "eval_loss": 0.7294782996177673,
      "eval_runtime": 19.7824,
      "eval_samples_per_second": 101.1,
      "eval_steps_per_second": 3.185,
      "step": 10200
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.8846428855168024e-05,
      "loss": 0.7208,
      "step": 10220
    },
    {
      "epoch": 2.42,
      "learning_rate": 5.836974656391514e-05,
      "loss": 0.7266,
      "step": 10240
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.789306427266227e-05,
      "loss": 0.7285,
      "step": 10260
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.741638198140939e-05,
      "loss": 0.7215,
      "step": 10280
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.6939699690156506e-05,
      "loss": 0.7203,
      "step": 10300
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.6463017398903625e-05,
      "loss": 0.7314,
      "step": 10320
    },
    {
      "epoch": 2.44,
      "learning_rate": 5.5986335107650744e-05,
      "loss": 0.7394,
      "step": 10340
    },
    {
      "epoch": 2.45,
      "learning_rate": 5.550965281639787e-05,
      "loss": 0.7138,
      "step": 10360
    },
    {
      "epoch": 2.45,
      "learning_rate": 5.503297052514498e-05,
      "loss": 0.721,
      "step": 10380
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.455628823389211e-05,
      "loss": 0.7199,
      "step": 10400
    },
    {
      "epoch": 2.46,
      "eval_loss": 0.728507936000824,
      "eval_runtime": 19.7761,
      "eval_samples_per_second": 101.132,
      "eval_steps_per_second": 3.186,
      "step": 10400
    },
    {
      "epoch": 2.46,
      "learning_rate": 5.4079605942639226e-05,
      "loss": 0.7228,
      "step": 10420
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.3602923651386345e-05,
      "loss": 0.7193,
      "step": 10440
    },
    {
      "epoch": 2.47,
      "learning_rate": 5.3126241360133464e-05,
      "loss": 0.7269,
      "step": 10460
    },
    {
      "epoch": 2.48,
      "learning_rate": 5.264955906888058e-05,
      "loss": 0.729,
      "step": 10480
    },
    {
      "epoch": 2.48,
      "learning_rate": 5.217287677762771e-05,
      "loss": 0.7193,
      "step": 10500
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.169619448637483e-05,
      "loss": 0.7158,
      "step": 10520
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.121951219512195e-05,
      "loss": 0.7158,
      "step": 10540
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.0742829903869065e-05,
      "loss": 0.7177,
      "step": 10560
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.0266147612616184e-05,
      "loss": 0.7187,
      "step": 10580
    },
    {
      "epoch": 2.51,
      "learning_rate": 4.978946532136331e-05,
      "loss": 0.7185,
      "step": 10600
    },
    {
      "epoch": 2.51,
      "eval_loss": 0.7283052802085876,
      "eval_runtime": 20.2682,
      "eval_samples_per_second": 98.677,
      "eval_steps_per_second": 3.108,
      "step": 10600
    },
    {
      "epoch": 2.51,
      "learning_rate": 4.931278303011042e-05,
      "loss": 0.7264,
      "step": 10620
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.883610073885755e-05,
      "loss": 0.7208,
      "step": 10640
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.835941844760467e-05,
      "loss": 0.7275,
      "step": 10660
    },
    {
      "epoch": 2.53,
      "learning_rate": 4.7882736156351786e-05,
      "loss": 0.7205,
      "step": 10680
    },
    {
      "epoch": 2.53,
      "learning_rate": 4.740605386509891e-05,
      "loss": 0.7213,
      "step": 10700
    },
    {
      "epoch": 2.53,
      "learning_rate": 4.692937157384602e-05,
      "loss": 0.7324,
      "step": 10720
    },
    {
      "epoch": 2.54,
      "learning_rate": 4.645268928259315e-05,
      "loss": 0.7197,
      "step": 10740
    },
    {
      "epoch": 2.54,
      "learning_rate": 4.597600699134027e-05,
      "loss": 0.7162,
      "step": 10760
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.5499324700087394e-05,
      "loss": 0.7223,
      "step": 10780
    },
    {
      "epoch": 2.55,
      "learning_rate": 4.5022642408834506e-05,
      "loss": 0.7249,
      "step": 10800
    },
    {
      "epoch": 2.55,
      "eval_loss": 0.7278863191604614,
      "eval_runtime": 19.7684,
      "eval_samples_per_second": 101.171,
      "eval_steps_per_second": 3.187,
      "step": 10800
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.4545960117581625e-05,
      "loss": 0.7245,
      "step": 10820
    },
    {
      "epoch": 2.56,
      "learning_rate": 4.406927782632875e-05,
      "loss": 0.7298,
      "step": 10840
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.359259553507587e-05,
      "loss": 0.7172,
      "step": 10860
    },
    {
      "epoch": 2.57,
      "learning_rate": 4.3115913243822995e-05,
      "loss": 0.7183,
      "step": 10880
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.263923095257011e-05,
      "loss": 0.7172,
      "step": 10900
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.2162548661317226e-05,
      "loss": 0.7166,
      "step": 10920
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.168586637006435e-05,
      "loss": 0.7303,
      "step": 10940
    },
    {
      "epoch": 2.59,
      "learning_rate": 4.1209184078811464e-05,
      "loss": 0.716,
      "step": 10960
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.073250178755859e-05,
      "loss": 0.7199,
      "step": 10980
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.025581949630571e-05,
      "loss": 0.7227,
      "step": 11000
    },
    {
      "epoch": 2.6,
      "eval_loss": 0.7274474501609802,
      "eval_runtime": 19.9546,
      "eval_samples_per_second": 100.228,
      "eval_steps_per_second": 3.157,
      "step": 11000
    },
    {
      "epoch": 2.61,
      "learning_rate": 3.9779137205052834e-05,
      "loss": 0.7134,
      "step": 11020
    },
    {
      "epoch": 2.61,
      "learning_rate": 3.930245491379995e-05,
      "loss": 0.7354,
      "step": 11040
    },
    {
      "epoch": 2.62,
      "learning_rate": 3.8825772622547065e-05,
      "loss": 0.7269,
      "step": 11060
    },
    {
      "epoch": 2.62,
      "learning_rate": 3.834909033129419e-05,
      "loss": 0.7261,
      "step": 11080
    },
    {
      "epoch": 2.62,
      "learning_rate": 3.787240804004131e-05,
      "loss": 0.735,
      "step": 11100
    },
    {
      "epoch": 2.63,
      "learning_rate": 3.739572574878843e-05,
      "loss": 0.716,
      "step": 11120
    },
    {
      "epoch": 2.63,
      "learning_rate": 3.691904345753555e-05,
      "loss": 0.721,
      "step": 11140
    },
    {
      "epoch": 2.64,
      "learning_rate": 3.644236116628267e-05,
      "loss": 0.7201,
      "step": 11160
    },
    {
      "epoch": 2.64,
      "learning_rate": 3.596567887502979e-05,
      "loss": 0.7231,
      "step": 11180
    },
    {
      "epoch": 2.65,
      "learning_rate": 3.548899658377691e-05,
      "loss": 0.7172,
      "step": 11200
    },
    {
      "epoch": 2.65,
      "eval_loss": 0.7270590662956238,
      "eval_runtime": 19.753,
      "eval_samples_per_second": 101.251,
      "eval_steps_per_second": 3.189,
      "step": 11200
    },
    {
      "epoch": 2.65,
      "learning_rate": 3.501231429252403e-05,
      "loss": 0.7296,
      "step": 11220
    },
    {
      "epoch": 2.66,
      "learning_rate": 3.453563200127115e-05,
      "loss": 0.7239,
      "step": 11240
    },
    {
      "epoch": 2.66,
      "learning_rate": 3.405894971001827e-05,
      "loss": 0.7215,
      "step": 11260
    },
    {
      "epoch": 2.67,
      "learning_rate": 3.358226741876539e-05,
      "loss": 0.7176,
      "step": 11280
    },
    {
      "epoch": 2.67,
      "learning_rate": 3.310558512751251e-05,
      "loss": 0.7277,
      "step": 11300
    },
    {
      "epoch": 2.68,
      "learning_rate": 3.262890283625963e-05,
      "loss": 0.7237,
      "step": 11320
    },
    {
      "epoch": 2.68,
      "learning_rate": 3.215222054500675e-05,
      "loss": 0.7167,
      "step": 11340
    },
    {
      "epoch": 2.69,
      "learning_rate": 3.167553825375387e-05,
      "loss": 0.7184,
      "step": 11360
    },
    {
      "epoch": 2.69,
      "learning_rate": 3.119885596250099e-05,
      "loss": 0.7238,
      "step": 11380
    },
    {
      "epoch": 2.7,
      "learning_rate": 3.072217367124811e-05,
      "loss": 0.7188,
      "step": 11400
    },
    {
      "epoch": 2.7,
      "eval_loss": 0.7263159155845642,
      "eval_runtime": 19.6317,
      "eval_samples_per_second": 101.876,
      "eval_steps_per_second": 3.209,
      "step": 11400
    },
    {
      "epoch": 2.7,
      "learning_rate": 3.0245491379995232e-05,
      "loss": 0.7146,
      "step": 11420
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.9768809088742348e-05,
      "loss": 0.7307,
      "step": 11440
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.929212679748947e-05,
      "loss": 0.721,
      "step": 11460
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.881544450623659e-05,
      "loss": 0.7293,
      "step": 11480
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.833876221498371e-05,
      "loss": 0.7245,
      "step": 11500
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.7862079923730833e-05,
      "loss": 0.7264,
      "step": 11520
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.7385397632477952e-05,
      "loss": 0.722,
      "step": 11540
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.6908715341225068e-05,
      "loss": 0.7195,
      "step": 11560
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.643203304997219e-05,
      "loss": 0.7181,
      "step": 11580
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.5955350758719312e-05,
      "loss": 0.7225,
      "step": 11600
    },
    {
      "epoch": 2.74,
      "eval_loss": 0.7265506386756897,
      "eval_runtime": 19.5252,
      "eval_samples_per_second": 102.432,
      "eval_steps_per_second": 3.227,
      "step": 11600
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.547866846746643e-05,
      "loss": 0.7151,
      "step": 11620
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.5001986176213553e-05,
      "loss": 0.7211,
      "step": 11640
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.4525303884960672e-05,
      "loss": 0.7231,
      "step": 11660
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.404862159370779e-05,
      "loss": 0.7236,
      "step": 11680
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.357193930245491e-05,
      "loss": 0.7161,
      "step": 11700
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.3095257011202032e-05,
      "loss": 0.7248,
      "step": 11720
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.261857471994915e-05,
      "loss": 0.7195,
      "step": 11740
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.2141892428696274e-05,
      "loss": 0.718,
      "step": 11760
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.1665210137443392e-05,
      "loss": 0.7161,
      "step": 11780
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.118852784619051e-05,
      "loss": 0.7204,
      "step": 11800
    },
    {
      "epoch": 2.79,
      "eval_loss": 0.7261104583740234,
      "eval_runtime": 20.0617,
      "eval_samples_per_second": 99.692,
      "eval_steps_per_second": 3.14,
      "step": 11800
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.071184555493763e-05,
      "loss": 0.716,
      "step": 11820
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.0235163263684753e-05,
      "loss": 0.7211,
      "step": 11840
    },
    {
      "epoch": 2.8,
      "learning_rate": 1.975848097243187e-05,
      "loss": 0.7242,
      "step": 11860
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.9281798681178994e-05,
      "loss": 0.7129,
      "step": 11880
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.8828950504488756e-05,
      "loss": 0.7233,
      "step": 11900
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.8352268213235875e-05,
      "loss": 0.7286,
      "step": 11920
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.7875585921982997e-05,
      "loss": 0.7147,
      "step": 11940
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.7398903630730116e-05,
      "loss": 0.7303,
      "step": 11960
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.692222133947724e-05,
      "loss": 0.7126,
      "step": 11980
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.6445539048224358e-05,
      "loss": 0.7174,
      "step": 12000
    },
    {
      "epoch": 2.84,
      "eval_loss": 0.7259587645530701,
      "eval_runtime": 20.6636,
      "eval_samples_per_second": 96.788,
      "eval_steps_per_second": 3.049,
      "step": 12000
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.5968856756971476e-05,
      "loss": 0.7147,
      "step": 12020
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.54921744657186e-05,
      "loss": 0.7184,
      "step": 12040
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.5015492174465718e-05,
      "loss": 0.7218,
      "step": 12060
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.4538809883212837e-05,
      "loss": 0.7172,
      "step": 12080
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.4062127591959957e-05,
      "loss": 0.7326,
      "step": 12100
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.3585445300707078e-05,
      "loss": 0.726,
      "step": 12120
    },
    {
      "epoch": 2.87,
      "learning_rate": 1.3108763009454197e-05,
      "loss": 0.711,
      "step": 12140
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.2632080718201317e-05,
      "loss": 0.7199,
      "step": 12160
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.215539842694844e-05,
      "loss": 0.7256,
      "step": 12180
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.1678716135695557e-05,
      "loss": 0.7183,
      "step": 12200
    },
    {
      "epoch": 2.88,
      "eval_loss": 0.7255927324295044,
      "eval_runtime": 20.0566,
      "eval_samples_per_second": 99.718,
      "eval_steps_per_second": 3.141,
      "step": 12200
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.1202033844442679e-05,
      "loss": 0.7244,
      "step": 12220
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.07253515531898e-05,
      "loss": 0.717,
      "step": 12240
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.0248669261936918e-05,
      "loss": 0.7224,
      "step": 12260
    },
    {
      "epoch": 2.9,
      "learning_rate": 9.771986970684039e-06,
      "loss": 0.7124,
      "step": 12280
    },
    {
      "epoch": 2.91,
      "learning_rate": 9.295304679431158e-06,
      "loss": 0.7285,
      "step": 12300
    },
    {
      "epoch": 2.91,
      "learning_rate": 8.818622388178278e-06,
      "loss": 0.7337,
      "step": 12320
    },
    {
      "epoch": 2.92,
      "learning_rate": 8.341940096925399e-06,
      "loss": 0.716,
      "step": 12340
    },
    {
      "epoch": 2.92,
      "learning_rate": 7.865257805672518e-06,
      "loss": 0.7212,
      "step": 12360
    },
    {
      "epoch": 2.93,
      "learning_rate": 7.3885755144196385e-06,
      "loss": 0.7262,
      "step": 12380
    },
    {
      "epoch": 2.93,
      "learning_rate": 6.911893223166759e-06,
      "loss": 0.7151,
      "step": 12400
    },
    {
      "epoch": 2.93,
      "eval_loss": 0.7256051301956177,
      "eval_runtime": 19.7012,
      "eval_samples_per_second": 101.516,
      "eval_steps_per_second": 3.198,
      "step": 12400
    }
  ],
  "max_steps": 12687,
  "num_train_epochs": 3,
  "total_flos": 1.6118284402370281e+19,
  "trial_name": null,
  "trial_params": null
}