{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10688,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004678362573099415,
      "grad_norm": 0.2359442412853241,
      "learning_rate": 0.0001991762613498081,
      "loss": 0.8207,
      "step": 50
    },
    {
      "epoch": 0.00935672514619883,
      "grad_norm": 0.1787925511598587,
      "learning_rate": 0.00019824019470186279,
      "loss": 0.7803,
      "step": 100
    },
    {
      "epoch": 0.014035087719298246,
      "grad_norm": 0.19425372779369354,
      "learning_rate": 0.00019730412805391744,
      "loss": 0.752,
      "step": 150
    },
    {
      "epoch": 0.01871345029239766,
      "grad_norm": 0.1763751357793808,
      "learning_rate": 0.0001963680614059721,
      "loss": 0.7502,
      "step": 200
    },
    {
      "epoch": 0.023391812865497075,
      "grad_norm": 0.208680659532547,
      "learning_rate": 0.00019543199475802677,
      "loss": 0.7493,
      "step": 250
    },
    {
      "epoch": 0.028070175438596492,
      "grad_norm": 0.1655270755290985,
      "learning_rate": 0.00019449592811008145,
      "loss": 0.7551,
      "step": 300
    },
    {
      "epoch": 0.03274853801169591,
      "grad_norm": 0.1512351781129837,
      "learning_rate": 0.0001935598614621361,
      "loss": 0.7418,
      "step": 350
    },
    {
      "epoch": 0.03742690058479532,
      "grad_norm": 0.19491629302501678,
      "learning_rate": 0.00019262379481419078,
      "loss": 0.7278,
      "step": 400
    },
    {
      "epoch": 0.042105263157894736,
      "grad_norm": 0.1969548463821411,
      "learning_rate": 0.00019168772816624546,
      "loss": 0.7365,
      "step": 450
    },
    {
      "epoch": 0.04678362573099415,
      "grad_norm": 0.21336689591407776,
      "learning_rate": 0.00019075166151830012,
      "loss": 0.7834,
      "step": 500
    },
    {
      "epoch": 0.05146198830409357,
      "grad_norm": 0.23860010504722595,
      "learning_rate": 0.0001898155948703548,
      "loss": 0.7523,
      "step": 550
    },
    {
      "epoch": 0.056140350877192984,
      "grad_norm": 0.16545253992080688,
      "learning_rate": 0.00018887952822240945,
      "loss": 0.7365,
      "step": 600
    },
    {
      "epoch": 0.0608187134502924,
      "grad_norm": 0.17125707864761353,
      "learning_rate": 0.00018794346157446413,
      "loss": 0.7455,
      "step": 650
    },
    {
      "epoch": 0.06549707602339182,
      "grad_norm": 0.20272980630397797,
      "learning_rate": 0.00018700739492651878,
      "loss": 0.7322,
      "step": 700
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 0.22747305035591125,
      "learning_rate": 0.00018607132827857346,
      "loss": 0.7369,
      "step": 750
    },
    {
      "epoch": 0.07485380116959064,
      "grad_norm": 0.1757550686597824,
      "learning_rate": 0.00018513526163062811,
      "loss": 0.7377,
      "step": 800
    },
    {
      "epoch": 0.07953216374269007,
      "grad_norm": 0.2243908941745758,
      "learning_rate": 0.00018419919498268277,
      "loss": 0.7519,
      "step": 850
    },
    {
      "epoch": 0.08421052631578947,
      "grad_norm": 0.2035253494977951,
      "learning_rate": 0.00018326312833473745,
      "loss": 0.7228,
      "step": 900
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.18074704706668854,
      "learning_rate": 0.0001823270616867921,
      "loss": 0.7152,
      "step": 950
    },
    {
      "epoch": 0.0935672514619883,
      "grad_norm": 0.20341496169567108,
      "learning_rate": 0.00018139099503884678,
      "loss": 0.7386,
      "step": 1000
    },
    {
      "epoch": 0.09824561403508772,
      "grad_norm": 0.20967456698417664,
      "learning_rate": 0.00018045492839090143,
      "loss": 0.7212,
      "step": 1050
    },
    {
      "epoch": 0.10292397660818714,
      "grad_norm": 0.20394659042358398,
      "learning_rate": 0.0001795188617429561,
      "loss": 0.7487,
      "step": 1100
    },
    {
      "epoch": 0.10760233918128655,
      "grad_norm": 0.19915203750133514,
      "learning_rate": 0.00017858279509501076,
      "loss": 0.7678,
      "step": 1150
    },
    {
      "epoch": 0.11228070175438597,
      "grad_norm": 0.2663570046424866,
      "learning_rate": 0.00017764672844706544,
      "loss": 0.7409,
      "step": 1200
    },
    {
      "epoch": 0.11695906432748537,
      "grad_norm": 0.24071092903614044,
      "learning_rate": 0.0001767106617991201,
      "loss": 0.7216,
      "step": 1250
    },
    {
      "epoch": 0.1216374269005848,
      "grad_norm": 0.1935265064239502,
      "learning_rate": 0.00017577459515117478,
      "loss": 0.7306,
      "step": 1300
    },
    {
      "epoch": 0.12631578947368421,
      "grad_norm": 0.2921196222305298,
      "learning_rate": 0.00017483852850322943,
      "loss": 0.7306,
      "step": 1350
    },
    {
      "epoch": 0.13099415204678364,
      "grad_norm": 0.24034500122070312,
      "learning_rate": 0.0001739024618552841,
      "loss": 0.725,
      "step": 1400
    },
    {
      "epoch": 0.13567251461988303,
      "grad_norm": 0.2127804309129715,
      "learning_rate": 0.00017296639520733876,
      "loss": 0.7266,
      "step": 1450
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 0.2236451804637909,
      "learning_rate": 0.00017203032855939341,
      "loss": 0.7101,
      "step": 1500
    },
    {
      "epoch": 0.14502923976608187,
      "grad_norm": 0.22269318997859955,
      "learning_rate": 0.0001710942619114481,
      "loss": 0.7292,
      "step": 1550
    },
    {
      "epoch": 0.1497076023391813,
      "grad_norm": 0.2598223090171814,
      "learning_rate": 0.00017015819526350277,
      "loss": 0.7073,
      "step": 1600
    },
    {
      "epoch": 0.1543859649122807,
      "grad_norm": 0.21632170677185059,
      "learning_rate": 0.00016922212861555743,
      "loss": 0.7451,
      "step": 1650
    },
    {
      "epoch": 0.15906432748538013,
      "grad_norm": 0.2735025882720947,
      "learning_rate": 0.0001682860619676121,
      "loss": 0.7234,
      "step": 1700
    },
    {
      "epoch": 0.16374269005847952,
      "grad_norm": 0.2583097219467163,
      "learning_rate": 0.0001673499953196668,
      "loss": 0.7145,
      "step": 1750
    },
    {
      "epoch": 0.16842105263157894,
      "grad_norm": 0.300374299287796,
      "learning_rate": 0.00016641392867172144,
      "loss": 0.7382,
      "step": 1800
    },
    {
      "epoch": 0.17309941520467836,
      "grad_norm": 0.25665584206581116,
      "learning_rate": 0.00016547786202377612,
      "loss": 0.7111,
      "step": 1850
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.17964990437030792,
      "learning_rate": 0.00016454179537583077,
      "loss": 0.7428,
      "step": 1900
    },
    {
      "epoch": 0.1824561403508772,
      "grad_norm": 0.2846137583255768,
      "learning_rate": 0.00016360572872788545,
      "loss": 0.7071,
      "step": 1950
    },
    {
      "epoch": 0.1871345029239766,
      "grad_norm": 0.23494410514831543,
      "learning_rate": 0.0001626696620799401,
      "loss": 0.7313,
      "step": 2000
    },
    {
      "epoch": 0.19181286549707602,
      "grad_norm": 0.2580121159553528,
      "learning_rate": 0.00016173359543199476,
      "loss": 0.7091,
      "step": 2050
    },
    {
      "epoch": 0.19649122807017544,
      "grad_norm": 0.2654828727245331,
      "learning_rate": 0.00016079752878404944,
      "loss": 0.7171,
      "step": 2100
    },
    {
      "epoch": 0.20116959064327486,
      "grad_norm": 0.22615693509578705,
      "learning_rate": 0.0001598614621361041,
      "loss": 0.7189,
      "step": 2150
    },
    {
      "epoch": 0.20584795321637428,
      "grad_norm": 0.19022518396377563,
      "learning_rate": 0.00015892539548815877,
      "loss": 0.716,
      "step": 2200
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.23111920058727264,
      "learning_rate": 0.00015798932884021342,
      "loss": 0.7446,
      "step": 2250
    },
    {
      "epoch": 0.2152046783625731,
      "grad_norm": 0.22850322723388672,
      "learning_rate": 0.0001570532621922681,
      "loss": 0.7243,
      "step": 2300
    },
    {
      "epoch": 0.2198830409356725,
      "grad_norm": 0.24424132704734802,
      "learning_rate": 0.00015611719554432276,
      "loss": 0.7072,
      "step": 2350
    },
    {
      "epoch": 0.22456140350877193,
      "grad_norm": 0.31915757060050964,
      "learning_rate": 0.00015518112889637743,
      "loss": 0.7098,
      "step": 2400
    },
    {
      "epoch": 0.22923976608187135,
      "grad_norm": 0.21317467093467712,
      "learning_rate": 0.0001542450622484321,
      "loss": 0.7472,
      "step": 2450
    },
    {
      "epoch": 0.23391812865497075,
      "grad_norm": 0.2181481122970581,
      "learning_rate": 0.00015330899560048677,
      "loss": 0.7295,
      "step": 2500
    },
    {
      "epoch": 0.23859649122807017,
      "grad_norm": 0.17121131718158722,
      "learning_rate": 0.00015237292895254142,
      "loss": 0.6705,
      "step": 2550
    },
    {
      "epoch": 0.2432748538011696,
      "grad_norm": 0.21892113983631134,
      "learning_rate": 0.0001514368623045961,
      "loss": 0.7284,
      "step": 2600
    },
    {
      "epoch": 0.247953216374269,
      "grad_norm": 0.2405460923910141,
      "learning_rate": 0.00015050079565665075,
      "loss": 0.7128,
      "step": 2650
    },
    {
      "epoch": 0.25263157894736843,
      "grad_norm": 0.30338019132614136,
      "learning_rate": 0.0001495647290087054,
      "loss": 0.7355,
      "step": 2700
    },
    {
      "epoch": 0.2573099415204678,
      "grad_norm": 0.288082480430603,
      "learning_rate": 0.00014862866236076009,
      "loss": 0.7038,
      "step": 2750
    },
    {
      "epoch": 0.26198830409356727,
      "grad_norm": 0.25093698501586914,
      "learning_rate": 0.00014769259571281477,
      "loss": 0.7207,
      "step": 2800
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.21849419176578522,
      "learning_rate": 0.00014675652906486942,
      "loss": 0.7001,
      "step": 2850
    },
    {
      "epoch": 0.27134502923976606,
      "grad_norm": 0.27592793107032776,
      "learning_rate": 0.0001458204624169241,
      "loss": 0.6953,
      "step": 2900
    },
    {
      "epoch": 0.2760233918128655,
      "grad_norm": 0.22119542956352234,
      "learning_rate": 0.00014488439576897878,
      "loss": 0.6827,
      "step": 2950
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 0.2122359573841095,
      "learning_rate": 0.00014394832912103343,
      "loss": 0.7163,
      "step": 3000
    },
    {
      "epoch": 0.28538011695906434,
      "grad_norm": 0.24913975596427917,
      "learning_rate": 0.0001430122624730881,
      "loss": 0.7182,
      "step": 3050
    },
    {
      "epoch": 0.29005847953216374,
      "grad_norm": 0.2344418615102768,
      "learning_rate": 0.00014207619582514276,
      "loss": 0.6989,
      "step": 3100
    },
    {
      "epoch": 0.29473684210526313,
      "grad_norm": 0.2660394608974457,
      "learning_rate": 0.00014114012917719744,
      "loss": 0.7247,
      "step": 3150
    },
    {
      "epoch": 0.2994152046783626,
      "grad_norm": 0.21425625681877136,
      "learning_rate": 0.0001402040625292521,
      "loss": 0.6875,
      "step": 3200
    },
    {
      "epoch": 0.30409356725146197,
      "grad_norm": 0.2691395580768585,
      "learning_rate": 0.00013926799588130675,
      "loss": 0.7112,
      "step": 3250
    },
    {
      "epoch": 0.3087719298245614,
      "grad_norm": 0.28755274415016174,
      "learning_rate": 0.00013833192923336143,
      "loss": 0.7212,
      "step": 3300
    },
    {
      "epoch": 0.3134502923976608,
      "grad_norm": 0.22710080444812775,
      "learning_rate": 0.00013739586258541608,
      "loss": 0.6994,
      "step": 3350
    },
    {
      "epoch": 0.31812865497076026,
      "grad_norm": 0.2685355246067047,
      "learning_rate": 0.00013645979593747076,
      "loss": 0.6774,
      "step": 3400
    },
    {
      "epoch": 0.32280701754385965,
      "grad_norm": 0.28679826855659485,
      "learning_rate": 0.0001355237292895254,
      "loss": 0.7081,
      "step": 3450
    },
    {
      "epoch": 0.32748538011695905,
      "grad_norm": 0.25061824917793274,
      "learning_rate": 0.0001345876626415801,
      "loss": 0.7078,
      "step": 3500
    },
    {
      "epoch": 0.3321637426900585,
      "grad_norm": 0.1860637366771698,
      "learning_rate": 0.00013365159599363475,
      "loss": 0.7008,
      "step": 3550
    },
    {
      "epoch": 0.3368421052631579,
      "grad_norm": 0.21458761394023895,
      "learning_rate": 0.00013271552934568943,
      "loss": 0.765,
      "step": 3600
    },
    {
      "epoch": 0.34152046783625734,
      "grad_norm": 0.3054530918598175,
      "learning_rate": 0.00013177946269774408,
      "loss": 0.6864,
      "step": 3650
    },
    {
      "epoch": 0.34619883040935673,
      "grad_norm": 0.27829650044441223,
      "learning_rate": 0.00013084339604979876,
      "loss": 0.6915,
      "step": 3700
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 0.2334837019443512,
      "learning_rate": 0.0001299073294018534,
      "loss": 0.712,
      "step": 3750
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.257228285074234,
      "learning_rate": 0.0001289712627539081,
      "loss": 0.7008,
      "step": 3800
    },
    {
      "epoch": 0.36023391812865496,
      "grad_norm": 0.24542610347270966,
      "learning_rate": 0.00012803519610596274,
      "loss": 0.6876,
      "step": 3850
    },
    {
      "epoch": 0.3649122807017544,
      "grad_norm": 0.2166256308555603,
      "learning_rate": 0.0001270991294580174,
      "loss": 0.7191,
      "step": 3900
    },
    {
      "epoch": 0.3695906432748538,
      "grad_norm": 0.2670128345489502,
      "learning_rate": 0.00012616306281007208,
      "loss": 0.6869,
      "step": 3950
    },
    {
      "epoch": 0.3742690058479532,
      "grad_norm": 0.23986391723155975,
      "learning_rate": 0.00012522699616212673,
      "loss": 0.6916,
      "step": 4000
    },
    {
      "epoch": 0.37894736842105264,
      "grad_norm": 0.2251734882593155,
      "learning_rate": 0.0001242909295141814,
      "loss": 0.7147,
      "step": 4050
    },
    {
      "epoch": 0.38362573099415204,
      "grad_norm": 0.21978424489498138,
      "learning_rate": 0.0001233548628662361,
      "loss": 0.6995,
      "step": 4100
    },
    {
      "epoch": 0.3883040935672515,
      "grad_norm": 0.24759012460708618,
      "learning_rate": 0.00012241879621829074,
      "loss": 0.7101,
      "step": 4150
    },
    {
      "epoch": 0.3929824561403509,
      "grad_norm": 0.24026772379875183,
      "learning_rate": 0.00012148272957034541,
      "loss": 0.7108,
      "step": 4200
    },
    {
      "epoch": 0.39766081871345027,
      "grad_norm": 0.2681400179862976,
      "learning_rate": 0.00012054666292240009,
      "loss": 0.7008,
      "step": 4250
    },
    {
      "epoch": 0.4023391812865497,
      "grad_norm": 0.28573960065841675,
      "learning_rate": 0.00011961059627445474,
      "loss": 0.668,
      "step": 4300
    },
    {
      "epoch": 0.4070175438596491,
      "grad_norm": 0.20998354256153107,
      "learning_rate": 0.00011867452962650942,
      "loss": 0.7022,
      "step": 4350
    },
    {
      "epoch": 0.41169590643274856,
      "grad_norm": 0.24716372787952423,
      "learning_rate": 0.00011773846297856407,
      "loss": 0.6983,
      "step": 4400
    },
    {
      "epoch": 0.41637426900584795,
      "grad_norm": 0.25760316848754883,
      "learning_rate": 0.00011680239633061874,
      "loss": 0.6914,
      "step": 4450
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.24404512345790863,
      "learning_rate": 0.00011586632968267342,
      "loss": 0.6941,
      "step": 4500
    },
    {
      "epoch": 0.4257309941520468,
      "grad_norm": 0.26983878016471863,
      "learning_rate": 0.00011493026303472807,
      "loss": 0.6881,
      "step": 4550
    },
    {
      "epoch": 0.4304093567251462,
      "grad_norm": 0.28383776545524597,
      "learning_rate": 0.00011399419638678275,
      "loss": 0.7449,
      "step": 4600
    },
    {
      "epoch": 0.43508771929824563,
      "grad_norm": 0.3157394230365753,
      "learning_rate": 0.0001130581297388374,
      "loss": 0.7065,
      "step": 4650
    },
    {
      "epoch": 0.439766081871345,
      "grad_norm": 0.2585708796977997,
      "learning_rate": 0.00011212206309089208,
      "loss": 0.6915,
      "step": 4700
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.26849091053009033,
      "learning_rate": 0.00011118599644294674,
      "loss": 0.7182,
      "step": 4750
    },
    {
      "epoch": 0.44912280701754387,
      "grad_norm": 0.32337141036987305,
      "learning_rate": 0.00011024992979500142,
      "loss": 0.7003,
      "step": 4800
    },
    {
      "epoch": 0.45380116959064326,
      "grad_norm": 0.2619641423225403,
      "learning_rate": 0.00010931386314705607,
      "loss": 0.693,
      "step": 4850
    },
    {
      "epoch": 0.4584795321637427,
      "grad_norm": 0.22884933650493622,
      "learning_rate": 0.00010837779649911075,
      "loss": 0.6987,
      "step": 4900
    },
    {
      "epoch": 0.4631578947368421,
      "grad_norm": 0.2512333393096924,
      "learning_rate": 0.0001074417298511654,
      "loss": 0.6628,
      "step": 4950
    },
    {
      "epoch": 0.4678362573099415,
      "grad_norm": 0.2615059018135071,
      "learning_rate": 0.00010650566320322008,
      "loss": 0.7093,
      "step": 5000
    },
    {
      "epoch": 0.47251461988304094,
      "grad_norm": 0.2515311539173126,
      "learning_rate": 0.00010556959655527475,
      "loss": 0.7151,
      "step": 5050
    },
    {
      "epoch": 0.47719298245614034,
      "grad_norm": 0.3073877990245819,
      "learning_rate": 0.0001046335299073294,
      "loss": 0.7217,
      "step": 5100
    },
    {
      "epoch": 0.4818713450292398,
      "grad_norm": 0.2709997296333313,
      "learning_rate": 0.00010369746325938408,
      "loss": 0.6995,
      "step": 5150
    },
    {
      "epoch": 0.4865497076023392,
      "grad_norm": 0.25649893283843994,
      "learning_rate": 0.00010276139661143873,
      "loss": 0.7028,
      "step": 5200
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 0.2822120487689972,
      "learning_rate": 0.00010182532996349341,
      "loss": 0.6855,
      "step": 5250
    },
    {
      "epoch": 0.495906432748538,
      "grad_norm": 0.2896905541419983,
      "learning_rate": 0.00010088926331554807,
      "loss": 0.6737,
      "step": 5300
    },
    {
      "epoch": 0.5005847953216375,
      "grad_norm": 0.23256748914718628,
      "learning_rate": 9.995319666760273e-05,
      "loss": 0.7149,
      "step": 5350
    },
    {
      "epoch": 0.5052631578947369,
      "grad_norm": 0.3130648136138916,
      "learning_rate": 9.90171300196574e-05,
      "loss": 0.7355,
      "step": 5400
    },
    {
      "epoch": 0.5099415204678363,
      "grad_norm": 0.24791193008422852,
      "learning_rate": 9.808106337171206e-05,
      "loss": 0.6764,
      "step": 5450
    },
    {
      "epoch": 0.5146198830409356,
      "grad_norm": 0.28329938650131226,
      "learning_rate": 9.714499672376673e-05,
      "loss": 0.7058,
      "step": 5500
    },
    {
      "epoch": 0.519298245614035,
      "grad_norm": 0.3048527240753174,
      "learning_rate": 9.62089300758214e-05,
      "loss": 0.6993,
      "step": 5550
    },
    {
      "epoch": 0.5239766081871345,
      "grad_norm": 0.2914963960647583,
      "learning_rate": 9.527286342787606e-05,
      "loss": 0.6802,
      "step": 5600
    },
    {
      "epoch": 0.5286549707602339,
      "grad_norm": 0.3110167682170868,
      "learning_rate": 9.433679677993073e-05,
      "loss": 0.6779,
      "step": 5650
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.30692344903945923,
      "learning_rate": 9.340073013198541e-05,
      "loss": 0.7139,
      "step": 5700
    },
    {
      "epoch": 0.5380116959064327,
      "grad_norm": 0.2898903787136078,
      "learning_rate": 9.246466348404008e-05,
      "loss": 0.6889,
      "step": 5750
    },
    {
      "epoch": 0.5426900584795321,
      "grad_norm": 0.2612452507019043,
      "learning_rate": 9.152859683609474e-05,
      "loss": 0.7074,
      "step": 5800
    },
    {
      "epoch": 0.5473684210526316,
      "grad_norm": 0.32114362716674805,
      "learning_rate": 9.059253018814941e-05,
      "loss": 0.7213,
      "step": 5850
    },
    {
      "epoch": 0.552046783625731,
      "grad_norm": 0.2521548271179199,
      "learning_rate": 8.965646354020407e-05,
      "loss": 0.6989,
      "step": 5900
    },
    {
      "epoch": 0.5567251461988304,
      "grad_norm": 0.25725868344306946,
      "learning_rate": 8.872039689225873e-05,
      "loss": 0.6849,
      "step": 5950
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.3528817892074585,
      "learning_rate": 8.77843302443134e-05,
      "loss": 0.6836,
      "step": 6000
    },
    {
      "epoch": 0.5660818713450292,
      "grad_norm": 0.2749318778514862,
      "learning_rate": 8.684826359636806e-05,
      "loss": 0.6688,
      "step": 6050
    },
    {
      "epoch": 0.5707602339181287,
      "grad_norm": 0.23874488472938538,
      "learning_rate": 8.591219694842273e-05,
      "loss": 0.6975,
      "step": 6100
    },
    {
      "epoch": 0.5754385964912281,
      "grad_norm": 0.268975168466568,
      "learning_rate": 8.497613030047739e-05,
      "loss": 0.704,
      "step": 6150
    },
    {
      "epoch": 0.5801169590643275,
      "grad_norm": 0.2570686638355255,
      "learning_rate": 8.404006365253206e-05,
      "loss": 0.6826,
      "step": 6200
    },
    {
      "epoch": 0.5847953216374269,
      "grad_norm": 0.26320692896842957,
      "learning_rate": 8.310399700458673e-05,
      "loss": 0.6997,
      "step": 6250
    },
    {
      "epoch": 0.5894736842105263,
      "grad_norm": 0.2496190369129181,
      "learning_rate": 8.21679303566414e-05,
      "loss": 0.7193,
      "step": 6300
    },
    {
      "epoch": 0.5941520467836258,
      "grad_norm": 0.29877322912216187,
      "learning_rate": 8.123186370869607e-05,
      "loss": 0.7071,
      "step": 6350
    },
    {
      "epoch": 0.5988304093567252,
      "grad_norm": 0.3319626748561859,
      "learning_rate": 8.029579706075074e-05,
      "loss": 0.6961,
      "step": 6400
    },
    {
      "epoch": 0.6035087719298246,
      "grad_norm": 0.30963343381881714,
      "learning_rate": 7.93597304128054e-05,
      "loss": 0.6892,
      "step": 6450
    },
    {
      "epoch": 0.6081871345029239,
      "grad_norm": 0.2534225583076477,
      "learning_rate": 7.842366376486007e-05,
      "loss": 0.6887,
      "step": 6500
    },
    {
      "epoch": 0.6128654970760234,
      "grad_norm": 0.23774613440036774,
      "learning_rate": 7.748759711691472e-05,
      "loss": 0.6858,
      "step": 6550
    },
    {
      "epoch": 0.6175438596491228,
      "grad_norm": 0.2526530623435974,
      "learning_rate": 7.655153046896939e-05,
      "loss": 0.6729,
      "step": 6600
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.21034641563892365,
      "learning_rate": 7.561546382102406e-05,
      "loss": 0.6555,
      "step": 6650
    },
    {
      "epoch": 0.6269005847953216,
      "grad_norm": 0.2742060422897339,
      "learning_rate": 7.467939717307872e-05,
      "loss": 0.6543,
      "step": 6700
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.2967930734157562,
      "learning_rate": 7.374333052513339e-05,
      "loss": 0.7058,
      "step": 6750
    },
    {
      "epoch": 0.6362573099415205,
      "grad_norm": 0.26209503412246704,
      "learning_rate": 7.280726387718805e-05,
      "loss": 0.7027,
      "step": 6800
    },
    {
      "epoch": 0.6409356725146199,
      "grad_norm": 0.2824839949607849,
      "learning_rate": 7.187119722924272e-05,
      "loss": 0.7019,
      "step": 6850
    },
    {
      "epoch": 0.6456140350877193,
      "grad_norm": 0.23571640253067017,
      "learning_rate": 7.093513058129739e-05,
      "loss": 0.6787,
      "step": 6900
    },
    {
      "epoch": 0.6502923976608187,
      "grad_norm": 0.27061739563941956,
      "learning_rate": 6.999906393335207e-05,
      "loss": 0.7513,
      "step": 6950
    },
    {
      "epoch": 0.6549707602339181,
      "grad_norm": 0.2449525147676468,
      "learning_rate": 6.906299728540673e-05,
      "loss": 0.7028,
      "step": 7000
    },
    {
      "epoch": 0.6596491228070176,
      "grad_norm": 0.294810026884079,
      "learning_rate": 6.81269306374614e-05,
      "loss": 0.6994,
      "step": 7050
    },
    {
      "epoch": 0.664327485380117,
      "grad_norm": 0.3177715837955475,
      "learning_rate": 6.719086398951607e-05,
      "loss": 0.6799,
      "step": 7100
    },
    {
      "epoch": 0.6690058479532164,
      "grad_norm": 0.30433663725852966,
      "learning_rate": 6.625479734157072e-05,
      "loss": 0.6938,
      "step": 7150
    },
    {
      "epoch": 0.6736842105263158,
      "grad_norm": 0.2912716567516327,
      "learning_rate": 6.531873069362538e-05,
      "loss": 0.7484,
      "step": 7200
    },
    {
      "epoch": 0.6783625730994152,
      "grad_norm": 0.22346681356430054,
      "learning_rate": 6.438266404568005e-05,
      "loss": 0.673,
      "step": 7250
    },
    {
      "epoch": 0.6830409356725147,
      "grad_norm": 0.31418198347091675,
      "learning_rate": 6.344659739773472e-05,
      "loss": 0.7023,
      "step": 7300
    },
    {
      "epoch": 0.6877192982456141,
      "grad_norm": 0.27822884917259216,
      "learning_rate": 6.251053074978938e-05,
      "loss": 0.6923,
      "step": 7350
    },
    {
      "epoch": 0.6923976608187135,
      "grad_norm": 0.3008076250553131,
      "learning_rate": 6.157446410184405e-05,
      "loss": 0.6966,
      "step": 7400
    },
    {
      "epoch": 0.6970760233918128,
      "grad_norm": 0.27879807353019714,
      "learning_rate": 6.063839745389872e-05,
      "loss": 0.6674,
      "step": 7450
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.29556870460510254,
      "learning_rate": 5.970233080595339e-05,
      "loss": 0.6949,
      "step": 7500
    },
    {
      "epoch": 0.7064327485380117,
      "grad_norm": 0.31239598989486694,
      "learning_rate": 5.8766264158008055e-05,
      "loss": 0.7039,
      "step": 7550
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.2879217863082886,
      "learning_rate": 5.783019751006272e-05,
      "loss": 0.7058,
      "step": 7600
    },
    {
      "epoch": 0.7157894736842105,
      "grad_norm": 0.2703321576118469,
      "learning_rate": 5.689413086211739e-05,
      "loss": 0.6766,
      "step": 7650
    },
    {
      "epoch": 0.7204678362573099,
      "grad_norm": 0.2751176655292511,
      "learning_rate": 5.5958064214172054e-05,
      "loss": 0.6735,
      "step": 7700
    },
    {
      "epoch": 0.7251461988304093,
      "grad_norm": 0.31331416964530945,
      "learning_rate": 5.5021997566226714e-05,
      "loss": 0.6852,
      "step": 7750
    },
    {
      "epoch": 0.7298245614035088,
      "grad_norm": 0.2925184369087219,
      "learning_rate": 5.408593091828138e-05,
      "loss": 0.7044,
      "step": 7800
    },
    {
      "epoch": 0.7345029239766082,
      "grad_norm": 0.2615364193916321,
      "learning_rate": 5.3149864270336046e-05,
      "loss": 0.7049,
      "step": 7850
    },
    {
      "epoch": 0.7391812865497076,
      "grad_norm": 0.33412986993789673,
      "learning_rate": 5.221379762239071e-05,
      "loss": 0.6653,
      "step": 7900
    },
    {
      "epoch": 0.743859649122807,
      "grad_norm": 0.30998972058296204,
      "learning_rate": 5.127773097444538e-05,
      "loss": 0.7081,
      "step": 7950
    },
    {
      "epoch": 0.7485380116959064,
      "grad_norm": 0.2581268548965454,
      "learning_rate": 5.034166432650005e-05,
      "loss": 0.7073,
      "step": 8000
    },
    {
      "epoch": 0.7532163742690059,
      "grad_norm": 0.23186282813549042,
      "learning_rate": 4.940559767855472e-05,
      "loss": 0.7033,
      "step": 8050
    },
    {
      "epoch": 0.7578947368421053,
      "grad_norm": 0.28655052185058594,
      "learning_rate": 4.8469531030609384e-05,
      "loss": 0.6702,
      "step": 8100
    },
    {
      "epoch": 0.7625730994152047,
      "grad_norm": 0.2882712185382843,
      "learning_rate": 4.753346438266405e-05,
      "loss": 0.6704,
      "step": 8150
    },
    {
      "epoch": 0.7672514619883041,
      "grad_norm": 0.28367742896080017,
      "learning_rate": 4.659739773471871e-05,
      "loss": 0.7011,
      "step": 8200
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 0.2787933051586151,
      "learning_rate": 4.5661331086773376e-05,
      "loss": 0.6957,
      "step": 8250
    },
    {
      "epoch": 0.776608187134503,
      "grad_norm": 0.3134559690952301,
      "learning_rate": 4.472526443882805e-05,
      "loss": 0.7143,
      "step": 8300
    },
    {
      "epoch": 0.7812865497076024,
      "grad_norm": 0.2954425811767578,
      "learning_rate": 4.3789197790882716e-05,
      "loss": 0.6948,
      "step": 8350
    },
    {
      "epoch": 0.7859649122807018,
      "grad_norm": 0.24117419123649597,
      "learning_rate": 4.285313114293738e-05,
      "loss": 0.6859,
      "step": 8400
    },
    {
      "epoch": 0.7906432748538011,
      "grad_norm": 0.32657894492149353,
      "learning_rate": 4.191706449499205e-05,
      "loss": 0.691,
      "step": 8450
    },
    {
      "epoch": 0.7953216374269005,
      "grad_norm": 0.3736024796962738,
      "learning_rate": 4.098099784704671e-05,
      "loss": 0.6687,
      "step": 8500
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2669476866722107,
      "learning_rate": 4.0044931199101374e-05,
      "loss": 0.71,
      "step": 8550
    },
    {
      "epoch": 0.8046783625730994,
      "grad_norm": 0.25700825452804565,
      "learning_rate": 3.910886455115604e-05,
      "loss": 0.6938,
      "step": 8600
    },
    {
      "epoch": 0.8093567251461988,
      "grad_norm": 0.22253499925136566,
      "learning_rate": 3.817279790321071e-05,
      "loss": 0.7247,
      "step": 8650
    },
    {
      "epoch": 0.8140350877192982,
      "grad_norm": 0.34762486815452576,
      "learning_rate": 3.723673125526538e-05,
      "loss": 0.6928,
      "step": 8700
    },
    {
      "epoch": 0.8187134502923976,
      "grad_norm": 0.22078180313110352,
      "learning_rate": 3.6300664607320046e-05,
      "loss": 0.7218,
      "step": 8750
    },
    {
      "epoch": 0.8233918128654971,
      "grad_norm": 0.2702735364437103,
      "learning_rate": 3.5364597959374705e-05,
      "loss": 0.707,
      "step": 8800
    },
    {
      "epoch": 0.8280701754385965,
      "grad_norm": 0.302895724773407,
      "learning_rate": 3.442853131142937e-05,
      "loss": 0.6811,
      "step": 8850
    },
    {
      "epoch": 0.8327485380116959,
      "grad_norm": 0.28420576453208923,
      "learning_rate": 3.349246466348404e-05,
      "loss": 0.6807,
      "step": 8900
    },
    {
      "epoch": 0.8374269005847953,
      "grad_norm": 0.259122371673584,
      "learning_rate": 3.255639801553871e-05,
      "loss": 0.6807,
      "step": 8950
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.2519673705101013,
      "learning_rate": 3.162033136759338e-05,
      "loss": 0.7102,
      "step": 9000
    },
    {
      "epoch": 0.8467836257309942,
      "grad_norm": 0.21814559400081635,
      "learning_rate": 3.0684264719648044e-05,
      "loss": 0.6916,
      "step": 9050
    },
    {
      "epoch": 0.8514619883040936,
      "grad_norm": 0.22553086280822754,
      "learning_rate": 2.9748198071702703e-05,
      "loss": 0.6623,
      "step": 9100
    },
    {
      "epoch": 0.856140350877193,
      "grad_norm": 0.25638681650161743,
      "learning_rate": 2.8812131423757373e-05,
      "loss": 0.6994,
      "step": 9150
    },
    {
      "epoch": 0.8608187134502924,
      "grad_norm": 0.2546500265598297,
      "learning_rate": 2.787606477581204e-05,
      "loss": 0.704,
      "step": 9200
    },
    {
      "epoch": 0.8654970760233918,
      "grad_norm": 0.2655654549598694,
      "learning_rate": 2.6939998127866705e-05,
      "loss": 0.6795,
      "step": 9250
    },
    {
      "epoch": 0.8701754385964913,
      "grad_norm": 0.27429452538490295,
      "learning_rate": 2.600393147992137e-05,
      "loss": 0.6848,
      "step": 9300
    },
    {
      "epoch": 0.8748538011695907,
      "grad_norm": 0.2434183955192566,
      "learning_rate": 2.506786483197604e-05,
      "loss": 0.6924,
      "step": 9350
    },
    {
      "epoch": 0.87953216374269,
      "grad_norm": 0.2134724110364914,
      "learning_rate": 2.4131798184030704e-05,
      "loss": 0.7131,
      "step": 9400
    },
    {
      "epoch": 0.8842105263157894,
      "grad_norm": 0.3318798243999481,
      "learning_rate": 2.319573153608537e-05,
      "loss": 0.6652,
      "step": 9450
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.2777431011199951,
      "learning_rate": 2.2259664888140037e-05,
      "loss": 0.6911,
      "step": 9500
    },
    {
      "epoch": 0.8935672514619883,
      "grad_norm": 0.29116806387901306,
      "learning_rate": 2.1323598240194703e-05,
      "loss": 0.6855,
      "step": 9550
    },
    {
      "epoch": 0.8982456140350877,
      "grad_norm": 0.28743380308151245,
      "learning_rate": 2.038753159224937e-05,
      "loss": 0.7057,
      "step": 9600
    },
    {
      "epoch": 0.9029239766081871,
      "grad_norm": 0.30142566561698914,
      "learning_rate": 1.9451464944304036e-05,
      "loss": 0.7153,
      "step": 9650
    },
    {
      "epoch": 0.9076023391812865,
      "grad_norm": 0.26111066341400146,
      "learning_rate": 1.8515398296358702e-05,
      "loss": 0.6843,
      "step": 9700
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.25944146513938904,
      "learning_rate": 1.7579331648413368e-05,
      "loss": 0.6955,
      "step": 9750
    },
    {
      "epoch": 0.9169590643274854,
      "grad_norm": 0.3067375719547272,
      "learning_rate": 1.6643265000468034e-05,
      "loss": 0.7066,
      "step": 9800
    },
    {
      "epoch": 0.9216374269005848,
      "grad_norm": 0.2763468623161316,
      "learning_rate": 1.57071983525227e-05,
      "loss": 0.6799,
      "step": 9850
    },
    {
      "epoch": 0.9263157894736842,
      "grad_norm": 0.35574787855148315,
      "learning_rate": 1.4771131704577367e-05,
      "loss": 0.6601,
      "step": 9900
    },
    {
      "epoch": 0.9309941520467836,
      "grad_norm": 0.2564525902271271,
      "learning_rate": 1.3835065056632032e-05,
      "loss": 0.666,
      "step": 9950
    },
    {
      "epoch": 0.935672514619883,
      "grad_norm": 0.2216213494539261,
      "learning_rate": 1.28989984086867e-05,
      "loss": 0.6848,
      "step": 10000
    },
    {
      "epoch": 0.9403508771929825,
      "grad_norm": 0.24584440886974335,
      "learning_rate": 1.1962931760741366e-05,
      "loss": 0.6414,
      "step": 10050
    },
    {
      "epoch": 0.9450292397660819,
      "grad_norm": 0.28561264276504517,
      "learning_rate": 1.102686511279603e-05,
      "loss": 0.6891,
      "step": 10100
    },
    {
      "epoch": 0.9497076023391813,
      "grad_norm": 0.3367581367492676,
      "learning_rate": 1.0090798464850698e-05,
      "loss": 0.6703,
      "step": 10150
    },
    {
      "epoch": 0.9543859649122807,
      "grad_norm": 0.28867948055267334,
      "learning_rate": 9.154731816905365e-06,
      "loss": 0.7203,
      "step": 10200
    },
    {
      "epoch": 0.9590643274853801,
      "grad_norm": 0.24281640350818634,
      "learning_rate": 8.21866516896003e-06,
      "loss": 0.6786,
      "step": 10250
    },
    {
      "epoch": 0.9637426900584796,
      "grad_norm": 0.26338231563568115,
      "learning_rate": 7.282598521014697e-06,
      "loss": 0.6712,
      "step": 10300
    },
    {
      "epoch": 0.968421052631579,
      "grad_norm": 0.2599546015262604,
      "learning_rate": 6.346531873069363e-06,
      "loss": 0.6857,
      "step": 10350
    },
    {
      "epoch": 0.9730994152046784,
      "grad_norm": 0.2876617908477783,
      "learning_rate": 5.410465225124029e-06,
      "loss": 0.6699,
      "step": 10400
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.366769015789032,
      "learning_rate": 4.474398577178695e-06,
      "loss": 0.6677,
      "step": 10450
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.2691202163696289,
      "learning_rate": 3.5383319292333616e-06,
      "loss": 0.6572,
      "step": 10500
    },
    {
      "epoch": 0.9871345029239766,
      "grad_norm": 0.2817225158214569,
      "learning_rate": 2.602265281288028e-06,
      "loss": 0.6772,
      "step": 10550
    },
    {
      "epoch": 0.991812865497076,
      "grad_norm": 0.29911836981773376,
      "learning_rate": 1.666198633342694e-06,
      "loss": 0.7113,
      "step": 10600
    },
    {
      "epoch": 0.9964912280701754,
      "grad_norm": 0.25651147961616516,
      "learning_rate": 7.301319853973604e-07,
      "loss": 0.6651,
      "step": 10650
    }
  ],
  "logging_steps": 50,
  "max_steps": 10688,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.962227515107451e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}