TULU-3-SFT-LoRA / checkpoint-14676 /trainer_state.json
bnjmnmarie's picture
Upload folder using huggingface_hub
47bac12 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9997796332117235,
"eval_steps": 500,
"global_step": 14676,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003406636340506077,
"grad_norm": 0.8387678265571594,
"learning_rate": 5.4421768707483e-06,
"loss": 0.8629,
"step": 25
},
{
"epoch": 0.006813272681012154,
"grad_norm": 0.5580975413322449,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.8351,
"step": 50
},
{
"epoch": 0.010219909021518231,
"grad_norm": 0.4770251214504242,
"learning_rate": 1.6780045351473924e-05,
"loss": 0.8023,
"step": 75
},
{
"epoch": 0.013626545362024308,
"grad_norm": 0.35437077283859253,
"learning_rate": 2.2448979591836737e-05,
"loss": 0.7808,
"step": 100
},
{
"epoch": 0.017033181702530386,
"grad_norm": 0.37850669026374817,
"learning_rate": 2.811791383219955e-05,
"loss": 0.7554,
"step": 125
},
{
"epoch": 0.020439818043036462,
"grad_norm": 0.40117064118385315,
"learning_rate": 3.378684807256236e-05,
"loss": 0.7419,
"step": 150
},
{
"epoch": 0.02384645438354254,
"grad_norm": 0.4868236780166626,
"learning_rate": 3.945578231292517e-05,
"loss": 0.751,
"step": 175
},
{
"epoch": 0.027253090724048617,
"grad_norm": 0.3966948390007019,
"learning_rate": 4.512471655328798e-05,
"loss": 0.7251,
"step": 200
},
{
"epoch": 0.030659727064554693,
"grad_norm": 0.3908109664916992,
"learning_rate": 5.0793650793650794e-05,
"loss": 0.7088,
"step": 225
},
{
"epoch": 0.03406636340506077,
"grad_norm": 0.3687989115715027,
"learning_rate": 5.646258503401361e-05,
"loss": 0.7115,
"step": 250
},
{
"epoch": 0.03747299974556685,
"grad_norm": 0.3919059634208679,
"learning_rate": 6.213151927437642e-05,
"loss": 0.7026,
"step": 275
},
{
"epoch": 0.040879636086072924,
"grad_norm": 0.42019009590148926,
"learning_rate": 6.780045351473924e-05,
"loss": 0.6967,
"step": 300
},
{
"epoch": 0.044286272426579,
"grad_norm": 0.4229620695114136,
"learning_rate": 7.346938775510205e-05,
"loss": 0.7148,
"step": 325
},
{
"epoch": 0.04769290876708508,
"grad_norm": 0.39575278759002686,
"learning_rate": 7.913832199546486e-05,
"loss": 0.744,
"step": 350
},
{
"epoch": 0.051099545107591154,
"grad_norm": 0.45732468366622925,
"learning_rate": 8.480725623582767e-05,
"loss": 0.7216,
"step": 375
},
{
"epoch": 0.054506181448097234,
"grad_norm": 0.3956912159919739,
"learning_rate": 9.047619047619048e-05,
"loss": 0.6953,
"step": 400
},
{
"epoch": 0.05791281778860331,
"grad_norm": 0.3948104679584503,
"learning_rate": 9.61451247165533e-05,
"loss": 0.7235,
"step": 425
},
{
"epoch": 0.061319454129109385,
"grad_norm": 0.3913336396217346,
"learning_rate": 9.99438004917457e-05,
"loss": 0.6676,
"step": 450
},
{
"epoch": 0.06472609046961547,
"grad_norm": 0.3908584713935852,
"learning_rate": 9.9768177028451e-05,
"loss": 0.7158,
"step": 475
},
{
"epoch": 0.06813272681012154,
"grad_norm": 0.4225063920021057,
"learning_rate": 9.959255356515631e-05,
"loss": 0.7129,
"step": 500
},
{
"epoch": 0.07153936315062762,
"grad_norm": 2.2388832569122314,
"learning_rate": 9.941693010186162e-05,
"loss": 0.7199,
"step": 525
},
{
"epoch": 0.0749459994911337,
"grad_norm": 0.39503997564315796,
"learning_rate": 9.924130663856692e-05,
"loss": 0.7298,
"step": 550
},
{
"epoch": 0.07835263583163977,
"grad_norm": 2.1647109985351562,
"learning_rate": 9.906568317527221e-05,
"loss": 0.7499,
"step": 575
},
{
"epoch": 0.08175927217214585,
"grad_norm": 0.36966434121131897,
"learning_rate": 9.889005971197752e-05,
"loss": 0.7212,
"step": 600
},
{
"epoch": 0.08516590851265193,
"grad_norm": 0.36990946531295776,
"learning_rate": 9.871443624868283e-05,
"loss": 0.7214,
"step": 625
},
{
"epoch": 0.088572544853158,
"grad_norm": 0.4158572852611542,
"learning_rate": 9.853881278538813e-05,
"loss": 0.6942,
"step": 650
},
{
"epoch": 0.09197918119366408,
"grad_norm": 0.3846476972103119,
"learning_rate": 9.836318932209344e-05,
"loss": 0.7218,
"step": 675
},
{
"epoch": 0.09538581753417016,
"grad_norm": 0.33537471294403076,
"learning_rate": 9.818756585879874e-05,
"loss": 0.7115,
"step": 700
},
{
"epoch": 0.09879245387467624,
"grad_norm": 0.3672342896461487,
"learning_rate": 9.801194239550405e-05,
"loss": 0.7225,
"step": 725
},
{
"epoch": 0.10219909021518231,
"grad_norm": 0.3498263955116272,
"learning_rate": 9.783631893220935e-05,
"loss": 0.7124,
"step": 750
},
{
"epoch": 0.1056057265556884,
"grad_norm": 0.3860284388065338,
"learning_rate": 9.766069546891466e-05,
"loss": 0.7018,
"step": 775
},
{
"epoch": 0.10901236289619447,
"grad_norm": 0.33633533120155334,
"learning_rate": 9.748507200561996e-05,
"loss": 0.6962,
"step": 800
},
{
"epoch": 0.11241899923670054,
"grad_norm": 0.3424709439277649,
"learning_rate": 9.730944854232526e-05,
"loss": 0.7068,
"step": 825
},
{
"epoch": 0.11582563557720663,
"grad_norm": 0.3627208173274994,
"learning_rate": 9.713382507903056e-05,
"loss": 0.6738,
"step": 850
},
{
"epoch": 0.1192322719177127,
"grad_norm": 0.3304712176322937,
"learning_rate": 9.695820161573587e-05,
"loss": 0.7004,
"step": 875
},
{
"epoch": 0.12263890825821877,
"grad_norm": 0.37575623393058777,
"learning_rate": 9.678257815244117e-05,
"loss": 0.711,
"step": 900
},
{
"epoch": 0.12604554459872486,
"grad_norm": 0.37238940596580505,
"learning_rate": 9.660695468914648e-05,
"loss": 0.7172,
"step": 925
},
{
"epoch": 0.12945218093923094,
"grad_norm": 0.39354655146598816,
"learning_rate": 9.643133122585177e-05,
"loss": 0.6949,
"step": 950
},
{
"epoch": 0.132858817279737,
"grad_norm": 0.34536346793174744,
"learning_rate": 9.625570776255708e-05,
"loss": 0.6871,
"step": 975
},
{
"epoch": 0.1362654536202431,
"grad_norm": 0.3518439829349518,
"learning_rate": 9.608008429926238e-05,
"loss": 0.7168,
"step": 1000
},
{
"epoch": 0.13967208996074917,
"grad_norm": 0.3994406759738922,
"learning_rate": 9.590446083596769e-05,
"loss": 0.722,
"step": 1025
},
{
"epoch": 0.14307872630125523,
"grad_norm": 0.38314729928970337,
"learning_rate": 9.5728837372673e-05,
"loss": 0.6858,
"step": 1050
},
{
"epoch": 0.14648536264176132,
"grad_norm": 0.3678247630596161,
"learning_rate": 9.55532139093783e-05,
"loss": 0.6895,
"step": 1075
},
{
"epoch": 0.1498919989822674,
"grad_norm": 0.37229785323143005,
"learning_rate": 9.537759044608359e-05,
"loss": 0.6858,
"step": 1100
},
{
"epoch": 0.15329863532277346,
"grad_norm": 0.34786805510520935,
"learning_rate": 9.52019669827889e-05,
"loss": 0.7055,
"step": 1125
},
{
"epoch": 0.15670527166327955,
"grad_norm": 0.34139758348464966,
"learning_rate": 9.50263435194942e-05,
"loss": 0.6959,
"step": 1150
},
{
"epoch": 0.16011190800378564,
"grad_norm": 0.33563101291656494,
"learning_rate": 9.485072005619952e-05,
"loss": 0.6696,
"step": 1175
},
{
"epoch": 0.1635185443442917,
"grad_norm": 0.4068642556667328,
"learning_rate": 9.467509659290483e-05,
"loss": 0.7035,
"step": 1200
},
{
"epoch": 0.16692518068479778,
"grad_norm": 0.3011302053928375,
"learning_rate": 9.449947312961012e-05,
"loss": 0.6759,
"step": 1225
},
{
"epoch": 0.17033181702530387,
"grad_norm": 0.3068943917751312,
"learning_rate": 9.432384966631543e-05,
"loss": 0.6874,
"step": 1250
},
{
"epoch": 0.17373845336580993,
"grad_norm": 0.3331654369831085,
"learning_rate": 9.414822620302073e-05,
"loss": 0.6886,
"step": 1275
},
{
"epoch": 0.177145089706316,
"grad_norm": 0.3533156216144562,
"learning_rate": 9.397260273972604e-05,
"loss": 0.7059,
"step": 1300
},
{
"epoch": 0.1805517260468221,
"grad_norm": 0.3277696669101715,
"learning_rate": 9.379697927643134e-05,
"loss": 0.6793,
"step": 1325
},
{
"epoch": 0.18395836238732816,
"grad_norm": 0.6513714790344238,
"learning_rate": 9.362135581313663e-05,
"loss": 0.6741,
"step": 1350
},
{
"epoch": 0.18736499872783424,
"grad_norm": 0.3339422047138214,
"learning_rate": 9.344573234984194e-05,
"loss": 0.7229,
"step": 1375
},
{
"epoch": 0.19077163506834033,
"grad_norm": 0.27368980646133423,
"learning_rate": 9.327010888654725e-05,
"loss": 0.6889,
"step": 1400
},
{
"epoch": 0.1941782714088464,
"grad_norm": 0.3697713017463684,
"learning_rate": 9.309448542325255e-05,
"loss": 0.7091,
"step": 1425
},
{
"epoch": 0.19758490774935247,
"grad_norm": 0.30846071243286133,
"learning_rate": 9.291886195995786e-05,
"loss": 0.6929,
"step": 1450
},
{
"epoch": 0.20099154408985856,
"grad_norm": 0.32923388481140137,
"learning_rate": 9.274323849666315e-05,
"loss": 0.6828,
"step": 1475
},
{
"epoch": 0.20439818043036462,
"grad_norm": 0.34551820158958435,
"learning_rate": 9.256761503336846e-05,
"loss": 0.7007,
"step": 1500
},
{
"epoch": 0.2078048167708707,
"grad_norm": 0.33527928590774536,
"learning_rate": 9.239199157007376e-05,
"loss": 0.7209,
"step": 1525
},
{
"epoch": 0.2112114531113768,
"grad_norm": 0.30302199721336365,
"learning_rate": 9.221636810677907e-05,
"loss": 0.6802,
"step": 1550
},
{
"epoch": 0.21461808945188285,
"grad_norm": 0.2833121120929718,
"learning_rate": 9.204074464348437e-05,
"loss": 0.6841,
"step": 1575
},
{
"epoch": 0.21802472579238893,
"grad_norm": 0.3293180465698242,
"learning_rate": 9.186512118018968e-05,
"loss": 0.6902,
"step": 1600
},
{
"epoch": 0.22143136213289502,
"grad_norm": 0.3296971023082733,
"learning_rate": 9.168949771689498e-05,
"loss": 0.6796,
"step": 1625
},
{
"epoch": 0.22483799847340108,
"grad_norm": 0.29069727659225464,
"learning_rate": 9.151387425360029e-05,
"loss": 0.7244,
"step": 1650
},
{
"epoch": 0.22824463481390717,
"grad_norm": 0.3884179890155792,
"learning_rate": 9.13382507903056e-05,
"loss": 0.6951,
"step": 1675
},
{
"epoch": 0.23165127115441325,
"grad_norm": 0.30107781291007996,
"learning_rate": 9.11626273270109e-05,
"loss": 0.6873,
"step": 1700
},
{
"epoch": 0.2350579074949193,
"grad_norm": 0.501139760017395,
"learning_rate": 9.098700386371619e-05,
"loss": 0.7114,
"step": 1725
},
{
"epoch": 0.2384645438354254,
"grad_norm": 0.3498416841030121,
"learning_rate": 9.08113804004215e-05,
"loss": 0.6801,
"step": 1750
},
{
"epoch": 0.24187118017593148,
"grad_norm": 0.34361162781715393,
"learning_rate": 9.06357569371268e-05,
"loss": 0.6702,
"step": 1775
},
{
"epoch": 0.24527781651643754,
"grad_norm": 0.39064115285873413,
"learning_rate": 9.046013347383211e-05,
"loss": 0.6805,
"step": 1800
},
{
"epoch": 0.24868445285694363,
"grad_norm": 0.29779884219169617,
"learning_rate": 9.028451001053742e-05,
"loss": 0.6929,
"step": 1825
},
{
"epoch": 0.2520910891974497,
"grad_norm": 0.3189874291419983,
"learning_rate": 9.010888654724272e-05,
"loss": 0.6877,
"step": 1850
},
{
"epoch": 0.2554977255379558,
"grad_norm": 0.33130019903182983,
"learning_rate": 8.993326308394801e-05,
"loss": 0.7176,
"step": 1875
},
{
"epoch": 0.2589043618784619,
"grad_norm": 0.3092060983181,
"learning_rate": 8.975763962065332e-05,
"loss": 0.6872,
"step": 1900
},
{
"epoch": 0.2623109982189679,
"grad_norm": 0.2977277338504791,
"learning_rate": 8.958201615735862e-05,
"loss": 0.6854,
"step": 1925
},
{
"epoch": 0.265717634559474,
"grad_norm": 0.3292854130268097,
"learning_rate": 8.940639269406393e-05,
"loss": 0.6726,
"step": 1950
},
{
"epoch": 0.2691242708999801,
"grad_norm": 0.30629754066467285,
"learning_rate": 8.923076923076924e-05,
"loss": 0.6928,
"step": 1975
},
{
"epoch": 0.2725309072404862,
"grad_norm": 0.38389158248901367,
"learning_rate": 8.905514576747454e-05,
"loss": 0.6804,
"step": 2000
},
{
"epoch": 0.27593754358099226,
"grad_norm": 0.29301708936691284,
"learning_rate": 8.887952230417985e-05,
"loss": 0.6815,
"step": 2025
},
{
"epoch": 0.27934417992149835,
"grad_norm": 0.3136097490787506,
"learning_rate": 8.870389884088515e-05,
"loss": 0.7038,
"step": 2050
},
{
"epoch": 0.2827508162620044,
"grad_norm": 0.4756239950656891,
"learning_rate": 8.852827537759046e-05,
"loss": 0.6772,
"step": 2075
},
{
"epoch": 0.28615745260251046,
"grad_norm": 0.3121085464954376,
"learning_rate": 8.835265191429576e-05,
"loss": 0.6604,
"step": 2100
},
{
"epoch": 0.28956408894301655,
"grad_norm": 0.3641608655452728,
"learning_rate": 8.817702845100106e-05,
"loss": 0.6645,
"step": 2125
},
{
"epoch": 0.29297072528352264,
"grad_norm": 0.34608688950538635,
"learning_rate": 8.800140498770636e-05,
"loss": 0.6702,
"step": 2150
},
{
"epoch": 0.2963773616240287,
"grad_norm": 0.3482177257537842,
"learning_rate": 8.782578152441167e-05,
"loss": 0.7293,
"step": 2175
},
{
"epoch": 0.2997839979645348,
"grad_norm": 0.34053704142570496,
"learning_rate": 8.765015806111697e-05,
"loss": 0.6724,
"step": 2200
},
{
"epoch": 0.30319063430504084,
"grad_norm": 0.32497864961624146,
"learning_rate": 8.747453459782228e-05,
"loss": 0.6511,
"step": 2225
},
{
"epoch": 0.3065972706455469,
"grad_norm": 0.3069411814212799,
"learning_rate": 8.729891113452757e-05,
"loss": 0.6849,
"step": 2250
},
{
"epoch": 0.310003906986053,
"grad_norm": 0.28222939372062683,
"learning_rate": 8.712328767123288e-05,
"loss": 0.66,
"step": 2275
},
{
"epoch": 0.3134105433265591,
"grad_norm": 0.3223375082015991,
"learning_rate": 8.694766420793818e-05,
"loss": 0.6819,
"step": 2300
},
{
"epoch": 0.3168171796670652,
"grad_norm": 0.2859184443950653,
"learning_rate": 8.677204074464349e-05,
"loss": 0.7061,
"step": 2325
},
{
"epoch": 0.32022381600757127,
"grad_norm": 0.30723094940185547,
"learning_rate": 8.659641728134879e-05,
"loss": 0.678,
"step": 2350
},
{
"epoch": 0.3236304523480773,
"grad_norm": 0.3178076446056366,
"learning_rate": 8.642079381805409e-05,
"loss": 0.6797,
"step": 2375
},
{
"epoch": 0.3270370886885834,
"grad_norm": 0.3042934536933899,
"learning_rate": 8.624517035475939e-05,
"loss": 0.6654,
"step": 2400
},
{
"epoch": 0.3304437250290895,
"grad_norm": 0.3517778813838959,
"learning_rate": 8.60695468914647e-05,
"loss": 0.6727,
"step": 2425
},
{
"epoch": 0.33385036136959556,
"grad_norm": 0.2900087833404541,
"learning_rate": 8.589392342817002e-05,
"loss": 0.6886,
"step": 2450
},
{
"epoch": 0.33725699771010165,
"grad_norm": 0.3086068630218506,
"learning_rate": 8.571829996487532e-05,
"loss": 0.6885,
"step": 2475
},
{
"epoch": 0.34066363405060773,
"grad_norm": 0.3216247856616974,
"learning_rate": 8.554267650158061e-05,
"loss": 0.6753,
"step": 2500
},
{
"epoch": 0.34407027039111376,
"grad_norm": 0.3239280581474304,
"learning_rate": 8.536705303828592e-05,
"loss": 0.7092,
"step": 2525
},
{
"epoch": 0.34747690673161985,
"grad_norm": 0.2760443091392517,
"learning_rate": 8.519142957499122e-05,
"loss": 0.7039,
"step": 2550
},
{
"epoch": 0.35088354307212594,
"grad_norm": 0.3144066631793976,
"learning_rate": 8.501580611169653e-05,
"loss": 0.6988,
"step": 2575
},
{
"epoch": 0.354290179412632,
"grad_norm": 0.32952046394348145,
"learning_rate": 8.484018264840184e-05,
"loss": 0.6911,
"step": 2600
},
{
"epoch": 0.3576968157531381,
"grad_norm": 2.4748454093933105,
"learning_rate": 8.466455918510714e-05,
"loss": 0.6748,
"step": 2625
},
{
"epoch": 0.3611034520936442,
"grad_norm": 0.32749420404434204,
"learning_rate": 8.448893572181243e-05,
"loss": 0.6775,
"step": 2650
},
{
"epoch": 0.3645100884341502,
"grad_norm": 0.28332722187042236,
"learning_rate": 8.431331225851774e-05,
"loss": 0.6788,
"step": 2675
},
{
"epoch": 0.3679167247746563,
"grad_norm": 0.29811403155326843,
"learning_rate": 8.413768879522305e-05,
"loss": 0.6954,
"step": 2700
},
{
"epoch": 0.3713233611151624,
"grad_norm": 0.3304058313369751,
"learning_rate": 8.396206533192835e-05,
"loss": 0.6759,
"step": 2725
},
{
"epoch": 0.3747299974556685,
"grad_norm": 0.32398712635040283,
"learning_rate": 8.378644186863366e-05,
"loss": 0.6886,
"step": 2750
},
{
"epoch": 0.37813663379617457,
"grad_norm": 0.2734803259372711,
"learning_rate": 8.361081840533895e-05,
"loss": 0.6839,
"step": 2775
},
{
"epoch": 0.38154327013668066,
"grad_norm": 0.2840772271156311,
"learning_rate": 8.343519494204425e-05,
"loss": 0.6828,
"step": 2800
},
{
"epoch": 0.3849499064771867,
"grad_norm": 0.39391207695007324,
"learning_rate": 8.325957147874956e-05,
"loss": 0.6935,
"step": 2825
},
{
"epoch": 0.3883565428176928,
"grad_norm": 0.27879825234413147,
"learning_rate": 8.308394801545487e-05,
"loss": 0.7002,
"step": 2850
},
{
"epoch": 0.39176317915819886,
"grad_norm": 0.30057233572006226,
"learning_rate": 8.290832455216017e-05,
"loss": 0.6784,
"step": 2875
},
{
"epoch": 0.39516981549870495,
"grad_norm": 0.31198152899742126,
"learning_rate": 8.273270108886548e-05,
"loss": 0.6896,
"step": 2900
},
{
"epoch": 0.39857645183921103,
"grad_norm": 0.2815667390823364,
"learning_rate": 8.255707762557078e-05,
"loss": 0.6649,
"step": 2925
},
{
"epoch": 0.4019830881797171,
"grad_norm": 0.35836973786354065,
"learning_rate": 8.238145416227609e-05,
"loss": 0.6601,
"step": 2950
},
{
"epoch": 0.4053897245202232,
"grad_norm": 0.289037823677063,
"learning_rate": 8.22058306989814e-05,
"loss": 0.6842,
"step": 2975
},
{
"epoch": 0.40879636086072924,
"grad_norm": 0.32901930809020996,
"learning_rate": 8.20302072356867e-05,
"loss": 0.6677,
"step": 3000
},
{
"epoch": 0.4122029972012353,
"grad_norm": 0.34060442447662354,
"learning_rate": 8.185458377239199e-05,
"loss": 0.6728,
"step": 3025
},
{
"epoch": 0.4156096335417414,
"grad_norm": 0.3380946218967438,
"learning_rate": 8.16789603090973e-05,
"loss": 0.7131,
"step": 3050
},
{
"epoch": 0.4190162698822475,
"grad_norm": 0.3136932849884033,
"learning_rate": 8.15033368458026e-05,
"loss": 0.6829,
"step": 3075
},
{
"epoch": 0.4224229062227536,
"grad_norm": 0.2733888328075409,
"learning_rate": 8.132771338250791e-05,
"loss": 0.7026,
"step": 3100
},
{
"epoch": 0.42582954256325967,
"grad_norm": 0.3108291029930115,
"learning_rate": 8.115208991921321e-05,
"loss": 0.689,
"step": 3125
},
{
"epoch": 0.4292361789037657,
"grad_norm": 0.3084828853607178,
"learning_rate": 8.09764664559185e-05,
"loss": 0.6723,
"step": 3150
},
{
"epoch": 0.4326428152442718,
"grad_norm": 0.2837083041667938,
"learning_rate": 8.080084299262381e-05,
"loss": 0.6699,
"step": 3175
},
{
"epoch": 0.43604945158477787,
"grad_norm": 0.3026272654533386,
"learning_rate": 8.062521952932912e-05,
"loss": 0.6771,
"step": 3200
},
{
"epoch": 0.43945608792528396,
"grad_norm": 0.29208138585090637,
"learning_rate": 8.044959606603442e-05,
"loss": 0.6636,
"step": 3225
},
{
"epoch": 0.44286272426579004,
"grad_norm": 0.29942700266838074,
"learning_rate": 8.027397260273973e-05,
"loss": 0.6788,
"step": 3250
},
{
"epoch": 0.44626936060629613,
"grad_norm": 0.262577086687088,
"learning_rate": 8.009834913944503e-05,
"loss": 0.65,
"step": 3275
},
{
"epoch": 0.44967599694680216,
"grad_norm": 0.29627498984336853,
"learning_rate": 7.992272567615034e-05,
"loss": 0.651,
"step": 3300
},
{
"epoch": 0.45308263328730825,
"grad_norm": 0.298234224319458,
"learning_rate": 7.974710221285565e-05,
"loss": 0.6883,
"step": 3325
},
{
"epoch": 0.45648926962781433,
"grad_norm": 0.2771220803260803,
"learning_rate": 7.957147874956095e-05,
"loss": 0.6786,
"step": 3350
},
{
"epoch": 0.4598959059683204,
"grad_norm": 0.29765036702156067,
"learning_rate": 7.939585528626626e-05,
"loss": 0.6835,
"step": 3375
},
{
"epoch": 0.4633025423088265,
"grad_norm": 0.31564295291900635,
"learning_rate": 7.922023182297156e-05,
"loss": 0.6546,
"step": 3400
},
{
"epoch": 0.4667091786493326,
"grad_norm": 0.2750963270664215,
"learning_rate": 7.904460835967685e-05,
"loss": 0.6765,
"step": 3425
},
{
"epoch": 0.4701158149898386,
"grad_norm": 0.28891250491142273,
"learning_rate": 7.886898489638216e-05,
"loss": 0.6719,
"step": 3450
},
{
"epoch": 0.4735224513303447,
"grad_norm": 0.3090450167655945,
"learning_rate": 7.869336143308747e-05,
"loss": 0.6629,
"step": 3475
},
{
"epoch": 0.4769290876708508,
"grad_norm": 0.2829868495464325,
"learning_rate": 7.851773796979277e-05,
"loss": 0.676,
"step": 3500
},
{
"epoch": 0.4803357240113569,
"grad_norm": 0.2897719144821167,
"learning_rate": 7.834211450649808e-05,
"loss": 0.6853,
"step": 3525
},
{
"epoch": 0.48374236035186297,
"grad_norm": 0.31082597374916077,
"learning_rate": 7.816649104320337e-05,
"loss": 0.6818,
"step": 3550
},
{
"epoch": 0.48714899669236905,
"grad_norm": 0.29339250922203064,
"learning_rate": 7.799086757990867e-05,
"loss": 0.673,
"step": 3575
},
{
"epoch": 0.4905556330328751,
"grad_norm": 0.29037582874298096,
"learning_rate": 7.781524411661398e-05,
"loss": 0.6726,
"step": 3600
},
{
"epoch": 0.49396226937338117,
"grad_norm": 0.2868305742740631,
"learning_rate": 7.763962065331929e-05,
"loss": 0.6894,
"step": 3625
},
{
"epoch": 0.49736890571388725,
"grad_norm": 0.28504157066345215,
"learning_rate": 7.746399719002459e-05,
"loss": 0.6479,
"step": 3650
},
{
"epoch": 0.5007755420543933,
"grad_norm": 0.28001776337623596,
"learning_rate": 7.728837372672988e-05,
"loss": 0.6571,
"step": 3675
},
{
"epoch": 0.5041821783948994,
"grad_norm": 0.3145054280757904,
"learning_rate": 7.711275026343519e-05,
"loss": 0.6638,
"step": 3700
},
{
"epoch": 0.5075888147354055,
"grad_norm": 0.3753933012485504,
"learning_rate": 7.693712680014051e-05,
"loss": 0.6617,
"step": 3725
},
{
"epoch": 0.5109954510759116,
"grad_norm": 0.2966824173927307,
"learning_rate": 7.676150333684581e-05,
"loss": 0.6666,
"step": 3750
},
{
"epoch": 0.5144020874164177,
"grad_norm": 0.26844897866249084,
"learning_rate": 7.658587987355112e-05,
"loss": 0.6646,
"step": 3775
},
{
"epoch": 0.5178087237569238,
"grad_norm": 0.3304543197154999,
"learning_rate": 7.641025641025641e-05,
"loss": 0.6704,
"step": 3800
},
{
"epoch": 0.5212153600974297,
"grad_norm": 0.31017178297042847,
"learning_rate": 7.623463294696172e-05,
"loss": 0.6846,
"step": 3825
},
{
"epoch": 0.5246219964379358,
"grad_norm": 0.3583391308784485,
"learning_rate": 7.605900948366702e-05,
"loss": 0.6759,
"step": 3850
},
{
"epoch": 0.5280286327784419,
"grad_norm": 0.28191253542900085,
"learning_rate": 7.588338602037233e-05,
"loss": 0.6369,
"step": 3875
},
{
"epoch": 0.531435269118948,
"grad_norm": 0.27828291058540344,
"learning_rate": 7.570776255707763e-05,
"loss": 0.6818,
"step": 3900
},
{
"epoch": 0.5348419054594541,
"grad_norm": 0.3020898997783661,
"learning_rate": 7.553213909378293e-05,
"loss": 0.6578,
"step": 3925
},
{
"epoch": 0.5382485417999602,
"grad_norm": 0.266043484210968,
"learning_rate": 7.535651563048823e-05,
"loss": 0.6652,
"step": 3950
},
{
"epoch": 0.5416551781404663,
"grad_norm": 0.2994140684604645,
"learning_rate": 7.518089216719354e-05,
"loss": 0.6639,
"step": 3975
},
{
"epoch": 0.5450618144809724,
"grad_norm": 0.29508110880851746,
"learning_rate": 7.500526870389884e-05,
"loss": 0.7025,
"step": 4000
},
{
"epoch": 0.5484684508214784,
"grad_norm": 0.29506492614746094,
"learning_rate": 7.482964524060415e-05,
"loss": 0.6565,
"step": 4025
},
{
"epoch": 0.5518750871619845,
"grad_norm": 0.31011542677879333,
"learning_rate": 7.465402177730946e-05,
"loss": 0.6851,
"step": 4050
},
{
"epoch": 0.5552817235024906,
"grad_norm": 0.3215409219264984,
"learning_rate": 7.447839831401475e-05,
"loss": 0.6559,
"step": 4075
},
{
"epoch": 0.5586883598429967,
"grad_norm": 0.29974356293678284,
"learning_rate": 7.430277485072005e-05,
"loss": 0.6784,
"step": 4100
},
{
"epoch": 0.5620949961835027,
"grad_norm": 0.29619455337524414,
"learning_rate": 7.412715138742536e-05,
"loss": 0.6836,
"step": 4125
},
{
"epoch": 0.5655016325240088,
"grad_norm": 0.3017350733280182,
"learning_rate": 7.395152792413066e-05,
"loss": 0.6533,
"step": 4150
},
{
"epoch": 0.5689082688645148,
"grad_norm": 0.2944609224796295,
"learning_rate": 7.377590446083597e-05,
"loss": 0.6568,
"step": 4175
},
{
"epoch": 0.5723149052050209,
"grad_norm": 0.2667727470397949,
"learning_rate": 7.360028099754128e-05,
"loss": 0.6546,
"step": 4200
},
{
"epoch": 0.575721541545527,
"grad_norm": 0.2934742569923401,
"learning_rate": 7.342465753424658e-05,
"loss": 0.6747,
"step": 4225
},
{
"epoch": 0.5791281778860331,
"grad_norm": 0.3042696416378021,
"learning_rate": 7.324903407095189e-05,
"loss": 0.6939,
"step": 4250
},
{
"epoch": 0.5825348142265392,
"grad_norm": 0.2748386859893799,
"learning_rate": 7.307341060765719e-05,
"loss": 0.6538,
"step": 4275
},
{
"epoch": 0.5859414505670453,
"grad_norm": 0.2834206223487854,
"learning_rate": 7.28977871443625e-05,
"loss": 0.6296,
"step": 4300
},
{
"epoch": 0.5893480869075514,
"grad_norm": 0.31131449341773987,
"learning_rate": 7.272216368106779e-05,
"loss": 0.6773,
"step": 4325
},
{
"epoch": 0.5927547232480574,
"grad_norm": 0.28676480054855347,
"learning_rate": 7.25465402177731e-05,
"loss": 0.6908,
"step": 4350
},
{
"epoch": 0.5961613595885635,
"grad_norm": 0.25846219062805176,
"learning_rate": 7.23709167544784e-05,
"loss": 0.6919,
"step": 4375
},
{
"epoch": 0.5995679959290696,
"grad_norm": 0.3187685012817383,
"learning_rate": 7.219529329118371e-05,
"loss": 0.6816,
"step": 4400
},
{
"epoch": 0.6029746322695757,
"grad_norm": 0.26819366216659546,
"learning_rate": 7.201966982788901e-05,
"loss": 0.6653,
"step": 4425
},
{
"epoch": 0.6063812686100817,
"grad_norm": 0.3077665865421295,
"learning_rate": 7.18440463645943e-05,
"loss": 0.648,
"step": 4450
},
{
"epoch": 0.6097879049505878,
"grad_norm": 0.4746949374675751,
"learning_rate": 7.166842290129961e-05,
"loss": 0.7012,
"step": 4475
},
{
"epoch": 0.6131945412910939,
"grad_norm": 0.2993331849575043,
"learning_rate": 7.149279943800492e-05,
"loss": 0.6779,
"step": 4500
},
{
"epoch": 0.6166011776315999,
"grad_norm": 0.26777181029319763,
"learning_rate": 7.131717597471022e-05,
"loss": 0.6706,
"step": 4525
},
{
"epoch": 0.620007813972106,
"grad_norm": 0.31024643778800964,
"learning_rate": 7.114155251141553e-05,
"loss": 0.6748,
"step": 4550
},
{
"epoch": 0.6234144503126121,
"grad_norm": 0.26060259342193604,
"learning_rate": 7.096592904812083e-05,
"loss": 0.6871,
"step": 4575
},
{
"epoch": 0.6268210866531182,
"grad_norm": 0.2986789643764496,
"learning_rate": 7.079030558482614e-05,
"loss": 0.6638,
"step": 4600
},
{
"epoch": 0.6302277229936243,
"grad_norm": 0.28097572922706604,
"learning_rate": 7.061468212153144e-05,
"loss": 0.661,
"step": 4625
},
{
"epoch": 0.6336343593341304,
"grad_norm": 0.3189026415348053,
"learning_rate": 7.043905865823675e-05,
"loss": 0.6525,
"step": 4650
},
{
"epoch": 0.6370409956746365,
"grad_norm": 0.29876482486724854,
"learning_rate": 7.026343519494206e-05,
"loss": 0.6655,
"step": 4675
},
{
"epoch": 0.6404476320151425,
"grad_norm": 0.30885156989097595,
"learning_rate": 7.008781173164735e-05,
"loss": 0.6797,
"step": 4700
},
{
"epoch": 0.6438542683556486,
"grad_norm": 0.3274490237236023,
"learning_rate": 6.991218826835265e-05,
"loss": 0.6723,
"step": 4725
},
{
"epoch": 0.6472609046961546,
"grad_norm": 0.30561405420303345,
"learning_rate": 6.973656480505796e-05,
"loss": 0.6414,
"step": 4750
},
{
"epoch": 0.6506675410366607,
"grad_norm": 0.3118368089199066,
"learning_rate": 6.956094134176326e-05,
"loss": 0.6786,
"step": 4775
},
{
"epoch": 0.6540741773771668,
"grad_norm": 0.2799939811229706,
"learning_rate": 6.938531787846857e-05,
"loss": 0.6556,
"step": 4800
},
{
"epoch": 0.6574808137176729,
"grad_norm": 0.31232938170433044,
"learning_rate": 6.920969441517388e-05,
"loss": 0.6527,
"step": 4825
},
{
"epoch": 0.660887450058179,
"grad_norm": 0.37503504753112793,
"learning_rate": 6.903407095187917e-05,
"loss": 0.6765,
"step": 4850
},
{
"epoch": 0.664294086398685,
"grad_norm": 0.2833762764930725,
"learning_rate": 6.885844748858447e-05,
"loss": 0.6558,
"step": 4875
},
{
"epoch": 0.6677007227391911,
"grad_norm": 0.27285921573638916,
"learning_rate": 6.868282402528978e-05,
"loss": 0.6717,
"step": 4900
},
{
"epoch": 0.6711073590796972,
"grad_norm": 0.29982614517211914,
"learning_rate": 6.850720056199509e-05,
"loss": 0.6736,
"step": 4925
},
{
"epoch": 0.6745139954202033,
"grad_norm": 0.29421648383140564,
"learning_rate": 6.833157709870039e-05,
"loss": 0.6633,
"step": 4950
},
{
"epoch": 0.6779206317607094,
"grad_norm": 0.30472108721733093,
"learning_rate": 6.815595363540568e-05,
"loss": 0.6418,
"step": 4975
},
{
"epoch": 0.6813272681012155,
"grad_norm": 0.2899680733680725,
"learning_rate": 6.7980330172111e-05,
"loss": 0.681,
"step": 5000
},
{
"epoch": 0.6847339044417216,
"grad_norm": 0.2816316485404968,
"learning_rate": 6.780470670881631e-05,
"loss": 0.6661,
"step": 5025
},
{
"epoch": 0.6881405407822275,
"grad_norm": 0.2587604224681854,
"learning_rate": 6.762908324552161e-05,
"loss": 0.6474,
"step": 5050
},
{
"epoch": 0.6915471771227336,
"grad_norm": 0.36257919669151306,
"learning_rate": 6.745345978222692e-05,
"loss": 0.6449,
"step": 5075
},
{
"epoch": 0.6949538134632397,
"grad_norm": 0.2923133373260498,
"learning_rate": 6.727783631893221e-05,
"loss": 0.6805,
"step": 5100
},
{
"epoch": 0.6983604498037458,
"grad_norm": 0.2798687517642975,
"learning_rate": 6.710221285563752e-05,
"loss": 0.6635,
"step": 5125
},
{
"epoch": 0.7017670861442519,
"grad_norm": 0.2894415259361267,
"learning_rate": 6.692658939234282e-05,
"loss": 0.6748,
"step": 5150
},
{
"epoch": 0.705173722484758,
"grad_norm": 0.31652697920799255,
"learning_rate": 6.675096592904813e-05,
"loss": 0.6434,
"step": 5175
},
{
"epoch": 0.708580358825264,
"grad_norm": 0.28104209899902344,
"learning_rate": 6.657534246575343e-05,
"loss": 0.6693,
"step": 5200
},
{
"epoch": 0.7119869951657701,
"grad_norm": 0.29532864689826965,
"learning_rate": 6.639971900245873e-05,
"loss": 0.665,
"step": 5225
},
{
"epoch": 0.7153936315062762,
"grad_norm": 0.2964029312133789,
"learning_rate": 6.622409553916403e-05,
"loss": 0.6742,
"step": 5250
},
{
"epoch": 0.7188002678467823,
"grad_norm": 0.3458377420902252,
"learning_rate": 6.604847207586934e-05,
"loss": 0.6836,
"step": 5275
},
{
"epoch": 0.7222069041872884,
"grad_norm": 0.30560600757598877,
"learning_rate": 6.587284861257464e-05,
"loss": 0.6883,
"step": 5300
},
{
"epoch": 0.7256135405277945,
"grad_norm": 0.266996294260025,
"learning_rate": 6.569722514927995e-05,
"loss": 0.6674,
"step": 5325
},
{
"epoch": 0.7290201768683005,
"grad_norm": 0.29581552743911743,
"learning_rate": 6.552160168598524e-05,
"loss": 0.6776,
"step": 5350
},
{
"epoch": 0.7324268132088065,
"grad_norm": 0.26997387409210205,
"learning_rate": 6.534597822269055e-05,
"loss": 0.6646,
"step": 5375
},
{
"epoch": 0.7358334495493126,
"grad_norm": 0.3831222355365753,
"learning_rate": 6.517035475939585e-05,
"loss": 0.6626,
"step": 5400
},
{
"epoch": 0.7392400858898187,
"grad_norm": 0.3023781180381775,
"learning_rate": 6.499473129610116e-05,
"loss": 0.672,
"step": 5425
},
{
"epoch": 0.7426467222303248,
"grad_norm": 0.3060709536075592,
"learning_rate": 6.481910783280646e-05,
"loss": 0.68,
"step": 5450
},
{
"epoch": 0.7460533585708309,
"grad_norm": 0.26548224687576294,
"learning_rate": 6.464348436951177e-05,
"loss": 0.6585,
"step": 5475
},
{
"epoch": 0.749459994911337,
"grad_norm": 0.3094407916069031,
"learning_rate": 6.446786090621707e-05,
"loss": 0.6809,
"step": 5500
},
{
"epoch": 0.752866631251843,
"grad_norm": 0.2568422853946686,
"learning_rate": 6.429223744292238e-05,
"loss": 0.6889,
"step": 5525
},
{
"epoch": 0.7562732675923491,
"grad_norm": 0.2860637605190277,
"learning_rate": 6.411661397962769e-05,
"loss": 0.6782,
"step": 5550
},
{
"epoch": 0.7596799039328552,
"grad_norm": 0.3119567036628723,
"learning_rate": 6.394099051633299e-05,
"loss": 0.6481,
"step": 5575
},
{
"epoch": 0.7630865402733613,
"grad_norm": 0.3033282458782196,
"learning_rate": 6.37653670530383e-05,
"loss": 0.6735,
"step": 5600
},
{
"epoch": 0.7664931766138674,
"grad_norm": 0.303755521774292,
"learning_rate": 6.358974358974359e-05,
"loss": 0.6736,
"step": 5625
},
{
"epoch": 0.7698998129543734,
"grad_norm": 0.3597272038459778,
"learning_rate": 6.34141201264489e-05,
"loss": 0.6749,
"step": 5650
},
{
"epoch": 0.7733064492948795,
"grad_norm": 0.26582157611846924,
"learning_rate": 6.32384966631542e-05,
"loss": 0.6646,
"step": 5675
},
{
"epoch": 0.7767130856353855,
"grad_norm": 0.2802767753601074,
"learning_rate": 6.30628731998595e-05,
"loss": 0.6675,
"step": 5700
},
{
"epoch": 0.7801197219758916,
"grad_norm": 0.31844425201416016,
"learning_rate": 6.288724973656481e-05,
"loss": 0.6637,
"step": 5725
},
{
"epoch": 0.7835263583163977,
"grad_norm": 0.31489238142967224,
"learning_rate": 6.27116262732701e-05,
"loss": 0.6627,
"step": 5750
},
{
"epoch": 0.7869329946569038,
"grad_norm": 0.3433346152305603,
"learning_rate": 6.253600280997541e-05,
"loss": 0.6796,
"step": 5775
},
{
"epoch": 0.7903396309974099,
"grad_norm": 0.3126988708972931,
"learning_rate": 6.236037934668072e-05,
"loss": 0.6768,
"step": 5800
},
{
"epoch": 0.793746267337916,
"grad_norm": 0.2817250192165375,
"learning_rate": 6.218475588338602e-05,
"loss": 0.6521,
"step": 5825
},
{
"epoch": 0.7971529036784221,
"grad_norm": 0.30584487318992615,
"learning_rate": 6.200913242009133e-05,
"loss": 0.6357,
"step": 5850
},
{
"epoch": 0.8005595400189282,
"grad_norm": 0.3048030138015747,
"learning_rate": 6.183350895679663e-05,
"loss": 0.6688,
"step": 5875
},
{
"epoch": 0.8039661763594342,
"grad_norm": 0.30793818831443787,
"learning_rate": 6.165788549350194e-05,
"loss": 0.6717,
"step": 5900
},
{
"epoch": 0.8073728126999403,
"grad_norm": 0.2916577458381653,
"learning_rate": 6.148226203020724e-05,
"loss": 0.6432,
"step": 5925
},
{
"epoch": 0.8107794490404464,
"grad_norm": 0.3017030656337738,
"learning_rate": 6.130663856691255e-05,
"loss": 0.6237,
"step": 5950
},
{
"epoch": 0.8141860853809524,
"grad_norm": 0.3230801820755005,
"learning_rate": 6.113101510361785e-05,
"loss": 0.6776,
"step": 5975
},
{
"epoch": 0.8175927217214585,
"grad_norm": 0.3148077130317688,
"learning_rate": 6.095539164032315e-05,
"loss": 0.6602,
"step": 6000
},
{
"epoch": 0.8209993580619646,
"grad_norm": 0.29618728160858154,
"learning_rate": 6.077976817702845e-05,
"loss": 0.6399,
"step": 6025
},
{
"epoch": 0.8244059944024706,
"grad_norm": 0.2636460065841675,
"learning_rate": 6.060414471373376e-05,
"loss": 0.6703,
"step": 6050
},
{
"epoch": 0.8278126307429767,
"grad_norm": 0.29034677147865295,
"learning_rate": 6.0428521250439064e-05,
"loss": 0.6406,
"step": 6075
},
{
"epoch": 0.8312192670834828,
"grad_norm": 0.29782792925834656,
"learning_rate": 6.025289778714437e-05,
"loss": 0.6721,
"step": 6100
},
{
"epoch": 0.8346259034239889,
"grad_norm": 0.29536172747612,
"learning_rate": 6.007727432384966e-05,
"loss": 0.6538,
"step": 6125
},
{
"epoch": 0.838032539764495,
"grad_norm": 0.31480076909065247,
"learning_rate": 5.990165086055497e-05,
"loss": 0.6807,
"step": 6150
},
{
"epoch": 0.8414391761050011,
"grad_norm": 0.28027939796447754,
"learning_rate": 5.972602739726027e-05,
"loss": 0.6602,
"step": 6175
},
{
"epoch": 0.8448458124455072,
"grad_norm": 0.3105228543281555,
"learning_rate": 5.9550403933965585e-05,
"loss": 0.6438,
"step": 6200
},
{
"epoch": 0.8482524487860132,
"grad_norm": 0.3044467568397522,
"learning_rate": 5.937478047067089e-05,
"loss": 0.6474,
"step": 6225
},
{
"epoch": 0.8516590851265193,
"grad_norm": 0.2907201945781708,
"learning_rate": 5.9199157007376196e-05,
"loss": 0.6732,
"step": 6250
},
{
"epoch": 0.8550657214670253,
"grad_norm": 0.3138374984264374,
"learning_rate": 5.902353354408149e-05,
"loss": 0.6629,
"step": 6275
},
{
"epoch": 0.8584723578075314,
"grad_norm": 0.2916734516620636,
"learning_rate": 5.8847910080786794e-05,
"loss": 0.6588,
"step": 6300
},
{
"epoch": 0.8618789941480375,
"grad_norm": 0.2865552604198456,
"learning_rate": 5.86722866174921e-05,
"loss": 0.6541,
"step": 6325
},
{
"epoch": 0.8652856304885436,
"grad_norm": 0.31498757004737854,
"learning_rate": 5.8496663154197405e-05,
"loss": 0.6685,
"step": 6350
},
{
"epoch": 0.8686922668290497,
"grad_norm": 0.30698952078819275,
"learning_rate": 5.832103969090271e-05,
"loss": 0.66,
"step": 6375
},
{
"epoch": 0.8720989031695557,
"grad_norm": 0.30786094069480896,
"learning_rate": 5.814541622760801e-05,
"loss": 0.6641,
"step": 6400
},
{
"epoch": 0.8755055395100618,
"grad_norm": 0.30545827746391296,
"learning_rate": 5.7969792764313316e-05,
"loss": 0.6713,
"step": 6425
},
{
"epoch": 0.8789121758505679,
"grad_norm": 0.2694856524467468,
"learning_rate": 5.779416930101862e-05,
"loss": 0.6711,
"step": 6450
},
{
"epoch": 0.882318812191074,
"grad_norm": 0.3137877881526947,
"learning_rate": 5.761854583772393e-05,
"loss": 0.6032,
"step": 6475
},
{
"epoch": 0.8857254485315801,
"grad_norm": 0.2708178460597992,
"learning_rate": 5.744292237442923e-05,
"loss": 0.6799,
"step": 6500
},
{
"epoch": 0.8891320848720862,
"grad_norm": 0.3112589716911316,
"learning_rate": 5.7267298911134525e-05,
"loss": 0.6839,
"step": 6525
},
{
"epoch": 0.8925387212125923,
"grad_norm": 0.3140416741371155,
"learning_rate": 5.709167544783983e-05,
"loss": 0.6764,
"step": 6550
},
{
"epoch": 0.8959453575530982,
"grad_norm": 0.26221442222595215,
"learning_rate": 5.6916051984545136e-05,
"loss": 0.6619,
"step": 6575
},
{
"epoch": 0.8993519938936043,
"grad_norm": 0.2906458079814911,
"learning_rate": 5.674042852125044e-05,
"loss": 0.6581,
"step": 6600
},
{
"epoch": 0.9027586302341104,
"grad_norm": 0.3028678596019745,
"learning_rate": 5.656480505795575e-05,
"loss": 0.6633,
"step": 6625
},
{
"epoch": 0.9061652665746165,
"grad_norm": 0.31104981899261475,
"learning_rate": 5.6389181594661046e-05,
"loss": 0.6522,
"step": 6650
},
{
"epoch": 0.9095719029151226,
"grad_norm": 0.2633511424064636,
"learning_rate": 5.621355813136635e-05,
"loss": 0.641,
"step": 6675
},
{
"epoch": 0.9129785392556287,
"grad_norm": 0.28382760286331177,
"learning_rate": 5.603793466807166e-05,
"loss": 0.6704,
"step": 6700
},
{
"epoch": 0.9163851755961347,
"grad_norm": 0.26792773604393005,
"learning_rate": 5.586231120477696e-05,
"loss": 0.6321,
"step": 6725
},
{
"epoch": 0.9197918119366408,
"grad_norm": 0.3073452413082123,
"learning_rate": 5.568668774148227e-05,
"loss": 0.6548,
"step": 6750
},
{
"epoch": 0.9231984482771469,
"grad_norm": 0.3464025855064392,
"learning_rate": 5.551106427818756e-05,
"loss": 0.6669,
"step": 6775
},
{
"epoch": 0.926605084617653,
"grad_norm": 0.3169041574001312,
"learning_rate": 5.5335440814892866e-05,
"loss": 0.6739,
"step": 6800
},
{
"epoch": 0.9300117209581591,
"grad_norm": 0.2641547620296478,
"learning_rate": 5.515981735159817e-05,
"loss": 0.6561,
"step": 6825
},
{
"epoch": 0.9334183572986652,
"grad_norm": 0.2953278720378876,
"learning_rate": 5.4984193888303484e-05,
"loss": 0.661,
"step": 6850
},
{
"epoch": 0.9368249936391712,
"grad_norm": 0.3245227336883545,
"learning_rate": 5.480857042500879e-05,
"loss": 0.6709,
"step": 6875
},
{
"epoch": 0.9402316299796772,
"grad_norm": 0.2790832221508026,
"learning_rate": 5.463294696171408e-05,
"loss": 0.6625,
"step": 6900
},
{
"epoch": 0.9436382663201833,
"grad_norm": 0.28198742866516113,
"learning_rate": 5.445732349841939e-05,
"loss": 0.6696,
"step": 6925
},
{
"epoch": 0.9470449026606894,
"grad_norm": 0.27755188941955566,
"learning_rate": 5.4281700035124694e-05,
"loss": 0.6544,
"step": 6950
},
{
"epoch": 0.9504515390011955,
"grad_norm": 0.28444674611091614,
"learning_rate": 5.410607657183e-05,
"loss": 0.6571,
"step": 6975
},
{
"epoch": 0.9538581753417016,
"grad_norm": 0.3238435983657837,
"learning_rate": 5.3930453108535305e-05,
"loss": 0.6391,
"step": 7000
},
{
"epoch": 0.9572648116822077,
"grad_norm": 0.30982422828674316,
"learning_rate": 5.375482964524061e-05,
"loss": 0.6604,
"step": 7025
},
{
"epoch": 0.9606714480227138,
"grad_norm": 0.30013763904571533,
"learning_rate": 5.357920618194591e-05,
"loss": 0.6766,
"step": 7050
},
{
"epoch": 0.9640780843632198,
"grad_norm": 0.33146125078201294,
"learning_rate": 5.3403582718651215e-05,
"loss": 0.6404,
"step": 7075
},
{
"epoch": 0.9674847207037259,
"grad_norm": 0.23431046307086945,
"learning_rate": 5.322795925535652e-05,
"loss": 0.6412,
"step": 7100
},
{
"epoch": 0.970891357044232,
"grad_norm": 0.29649725556373596,
"learning_rate": 5.3052335792061826e-05,
"loss": 0.6462,
"step": 7125
},
{
"epoch": 0.9742979933847381,
"grad_norm": 0.31719136238098145,
"learning_rate": 5.287671232876713e-05,
"loss": 0.6877,
"step": 7150
},
{
"epoch": 0.9777046297252441,
"grad_norm": 0.25989624857902527,
"learning_rate": 5.2701088865472424e-05,
"loss": 0.6416,
"step": 7175
},
{
"epoch": 0.9811112660657502,
"grad_norm": 0.32486599683761597,
"learning_rate": 5.252546540217773e-05,
"loss": 0.6307,
"step": 7200
},
{
"epoch": 0.9845179024062563,
"grad_norm": 0.31052151322364807,
"learning_rate": 5.2349841938883035e-05,
"loss": 0.6686,
"step": 7225
},
{
"epoch": 0.9879245387467623,
"grad_norm": 0.2722817659378052,
"learning_rate": 5.217421847558834e-05,
"loss": 0.6554,
"step": 7250
},
{
"epoch": 0.9913311750872684,
"grad_norm": 0.29433372616767883,
"learning_rate": 5.199859501229365e-05,
"loss": 0.6491,
"step": 7275
},
{
"epoch": 0.9947378114277745,
"grad_norm": 0.2956300973892212,
"learning_rate": 5.1822971548998945e-05,
"loss": 0.6433,
"step": 7300
},
{
"epoch": 0.9981444477682806,
"grad_norm": 0.29263532161712646,
"learning_rate": 5.164734808570425e-05,
"loss": 0.6434,
"step": 7325
},
{
"epoch": 1.0014989199898228,
"grad_norm": 0.2794649600982666,
"learning_rate": 5.147172462240956e-05,
"loss": 0.6228,
"step": 7350
},
{
"epoch": 1.0049055563303289,
"grad_norm": 0.2971467673778534,
"learning_rate": 5.129610115911486e-05,
"loss": 0.5837,
"step": 7375
},
{
"epoch": 1.008312192670835,
"grad_norm": 0.2719724178314209,
"learning_rate": 5.112047769582017e-05,
"loss": 0.5884,
"step": 7400
},
{
"epoch": 1.0117188290113408,
"grad_norm": 0.3151598870754242,
"learning_rate": 5.094485423252546e-05,
"loss": 0.5755,
"step": 7425
},
{
"epoch": 1.015125465351847,
"grad_norm": 0.2950424551963806,
"learning_rate": 5.0769230769230766e-05,
"loss": 0.5849,
"step": 7450
},
{
"epoch": 1.018532101692353,
"grad_norm": 0.2683977782726288,
"learning_rate": 5.059360730593608e-05,
"loss": 0.5864,
"step": 7475
},
{
"epoch": 1.021938738032859,
"grad_norm": 0.32205426692962646,
"learning_rate": 5.0417983842641384e-05,
"loss": 0.5712,
"step": 7500
},
{
"epoch": 1.0253453743733651,
"grad_norm": 0.3280932903289795,
"learning_rate": 5.024236037934669e-05,
"loss": 0.5857,
"step": 7525
},
{
"epoch": 1.0287520107138712,
"grad_norm": 0.2890978753566742,
"learning_rate": 5.006673691605198e-05,
"loss": 0.5899,
"step": 7550
},
{
"epoch": 1.0321586470543773,
"grad_norm": 0.2932690680027008,
"learning_rate": 4.9891113452757294e-05,
"loss": 0.5691,
"step": 7575
},
{
"epoch": 1.0355652833948834,
"grad_norm": 0.3369404673576355,
"learning_rate": 4.971548998946259e-05,
"loss": 0.5946,
"step": 7600
},
{
"epoch": 1.0389719197353895,
"grad_norm": 0.29351669549942017,
"learning_rate": 4.95398665261679e-05,
"loss": 0.5962,
"step": 7625
},
{
"epoch": 1.0423785560758956,
"grad_norm": 0.28703489899635315,
"learning_rate": 4.93642430628732e-05,
"loss": 0.6125,
"step": 7650
},
{
"epoch": 1.0457851924164017,
"grad_norm": 0.30198702216148376,
"learning_rate": 4.91886195995785e-05,
"loss": 0.5828,
"step": 7675
},
{
"epoch": 1.0491918287569078,
"grad_norm": 0.2725096046924591,
"learning_rate": 4.901299613628381e-05,
"loss": 0.5757,
"step": 7700
},
{
"epoch": 1.0525984650974138,
"grad_norm": 0.30557242035865784,
"learning_rate": 4.8837372672989114e-05,
"loss": 0.6002,
"step": 7725
},
{
"epoch": 1.05600510143792,
"grad_norm": 0.30513855814933777,
"learning_rate": 4.866174920969442e-05,
"loss": 0.5882,
"step": 7750
},
{
"epoch": 1.059411737778426,
"grad_norm": 0.28717830777168274,
"learning_rate": 4.848612574639972e-05,
"loss": 0.5846,
"step": 7775
},
{
"epoch": 1.062818374118932,
"grad_norm": 0.3230528235435486,
"learning_rate": 4.8310502283105025e-05,
"loss": 0.5949,
"step": 7800
},
{
"epoch": 1.0662250104594382,
"grad_norm": 0.29600274562835693,
"learning_rate": 4.813487881981033e-05,
"loss": 0.5691,
"step": 7825
},
{
"epoch": 1.0696316467999443,
"grad_norm": 0.27012398838996887,
"learning_rate": 4.795925535651563e-05,
"loss": 0.6006,
"step": 7850
},
{
"epoch": 1.0730382831404504,
"grad_norm": 0.36482536792755127,
"learning_rate": 4.7783631893220935e-05,
"loss": 0.5872,
"step": 7875
},
{
"epoch": 1.0764449194809564,
"grad_norm": 0.3450656235218048,
"learning_rate": 4.760800842992624e-05,
"loss": 0.5777,
"step": 7900
},
{
"epoch": 1.0798515558214625,
"grad_norm": 0.30672258138656616,
"learning_rate": 4.7432384966631546e-05,
"loss": 0.582,
"step": 7925
},
{
"epoch": 1.0832581921619686,
"grad_norm": 0.3871435523033142,
"learning_rate": 4.725676150333685e-05,
"loss": 0.5877,
"step": 7950
},
{
"epoch": 1.0866648285024747,
"grad_norm": 0.3294704258441925,
"learning_rate": 4.708113804004215e-05,
"loss": 0.5606,
"step": 7975
},
{
"epoch": 1.0900714648429806,
"grad_norm": 0.3116719722747803,
"learning_rate": 4.6905514576747456e-05,
"loss": 0.5552,
"step": 8000
},
{
"epoch": 1.0934781011834867,
"grad_norm": 0.35774242877960205,
"learning_rate": 4.672989111345276e-05,
"loss": 0.61,
"step": 8025
},
{
"epoch": 1.0968847375239927,
"grad_norm": 0.3226807117462158,
"learning_rate": 4.655426765015806e-05,
"loss": 0.6,
"step": 8050
},
{
"epoch": 1.1002913738644988,
"grad_norm": 0.2967264652252197,
"learning_rate": 4.6378644186863366e-05,
"loss": 0.5923,
"step": 8075
},
{
"epoch": 1.103698010205005,
"grad_norm": 0.27228492498397827,
"learning_rate": 4.6203020723568665e-05,
"loss": 0.5761,
"step": 8100
},
{
"epoch": 1.107104646545511,
"grad_norm": 0.33212724328041077,
"learning_rate": 4.602739726027398e-05,
"loss": 0.6051,
"step": 8125
},
{
"epoch": 1.110511282886017,
"grad_norm": 0.28719720244407654,
"learning_rate": 4.585177379697928e-05,
"loss": 0.6039,
"step": 8150
},
{
"epoch": 1.1139179192265232,
"grad_norm": 0.32448840141296387,
"learning_rate": 4.567615033368458e-05,
"loss": 0.5913,
"step": 8175
},
{
"epoch": 1.1173245555670293,
"grad_norm": 0.2699980139732361,
"learning_rate": 4.550052687038989e-05,
"loss": 0.606,
"step": 8200
},
{
"epoch": 1.1207311919075353,
"grad_norm": 0.285967081785202,
"learning_rate": 4.5324903407095193e-05,
"loss": 0.5831,
"step": 8225
},
{
"epoch": 1.1241378282480414,
"grad_norm": 0.31957241892814636,
"learning_rate": 4.514927994380049e-05,
"loss": 0.6087,
"step": 8250
},
{
"epoch": 1.1275444645885475,
"grad_norm": 0.3458569049835205,
"learning_rate": 4.49736564805058e-05,
"loss": 0.5916,
"step": 8275
},
{
"epoch": 1.1309511009290536,
"grad_norm": 0.3341424763202667,
"learning_rate": 4.47980330172111e-05,
"loss": 0.5924,
"step": 8300
},
{
"epoch": 1.1343577372695597,
"grad_norm": 0.2719230055809021,
"learning_rate": 4.46224095539164e-05,
"loss": 0.5583,
"step": 8325
},
{
"epoch": 1.1377643736100658,
"grad_norm": 0.3184913098812103,
"learning_rate": 4.444678609062171e-05,
"loss": 0.576,
"step": 8350
},
{
"epoch": 1.1411710099505719,
"grad_norm": 0.31068292260169983,
"learning_rate": 4.4271162627327014e-05,
"loss": 0.6214,
"step": 8375
},
{
"epoch": 1.144577646291078,
"grad_norm": 0.3055650293827057,
"learning_rate": 4.409553916403232e-05,
"loss": 0.5867,
"step": 8400
},
{
"epoch": 1.147984282631584,
"grad_norm": 0.3341415524482727,
"learning_rate": 4.391991570073762e-05,
"loss": 0.573,
"step": 8425
},
{
"epoch": 1.1513909189720901,
"grad_norm": 0.30130231380462646,
"learning_rate": 4.3744292237442924e-05,
"loss": 0.5723,
"step": 8450
},
{
"epoch": 1.1547975553125962,
"grad_norm": 0.3360777199268341,
"learning_rate": 4.356866877414823e-05,
"loss": 0.5857,
"step": 8475
},
{
"epoch": 1.1582041916531023,
"grad_norm": 0.28688421845436096,
"learning_rate": 4.339304531085353e-05,
"loss": 0.5814,
"step": 8500
},
{
"epoch": 1.1616108279936084,
"grad_norm": 0.3305681049823761,
"learning_rate": 4.3217421847558834e-05,
"loss": 0.5829,
"step": 8525
},
{
"epoch": 1.1650174643341145,
"grad_norm": 0.28347936272621155,
"learning_rate": 4.304179838426414e-05,
"loss": 0.5644,
"step": 8550
},
{
"epoch": 1.1684241006746205,
"grad_norm": 0.3450355529785156,
"learning_rate": 4.2866174920969445e-05,
"loss": 0.6092,
"step": 8575
},
{
"epoch": 1.1718307370151266,
"grad_norm": 0.3150685727596283,
"learning_rate": 4.269055145767475e-05,
"loss": 0.5804,
"step": 8600
},
{
"epoch": 1.1752373733556327,
"grad_norm": 0.2739415764808655,
"learning_rate": 4.251492799438005e-05,
"loss": 0.5934,
"step": 8625
},
{
"epoch": 1.1786440096961388,
"grad_norm": 0.28803735971450806,
"learning_rate": 4.2339304531085356e-05,
"loss": 0.5766,
"step": 8650
},
{
"epoch": 1.1820506460366447,
"grad_norm": 0.3043755888938904,
"learning_rate": 4.216368106779066e-05,
"loss": 0.5652,
"step": 8675
},
{
"epoch": 1.1854572823771508,
"grad_norm": 0.3253932297229767,
"learning_rate": 4.198805760449596e-05,
"loss": 0.587,
"step": 8700
},
{
"epoch": 1.1888639187176568,
"grad_norm": 0.332832932472229,
"learning_rate": 4.1812434141201266e-05,
"loss": 0.5877,
"step": 8725
},
{
"epoch": 1.192270555058163,
"grad_norm": 0.30993393063545227,
"learning_rate": 4.163681067790657e-05,
"loss": 0.5862,
"step": 8750
},
{
"epoch": 1.195677191398669,
"grad_norm": 0.3366856276988983,
"learning_rate": 4.146118721461188e-05,
"loss": 0.583,
"step": 8775
},
{
"epoch": 1.199083827739175,
"grad_norm": 0.3211153447628021,
"learning_rate": 4.128556375131718e-05,
"loss": 0.578,
"step": 8800
},
{
"epoch": 1.2024904640796812,
"grad_norm": 0.29055503010749817,
"learning_rate": 4.110994028802248e-05,
"loss": 0.6299,
"step": 8825
},
{
"epoch": 1.2058971004201873,
"grad_norm": 0.32854345440864563,
"learning_rate": 4.093431682472779e-05,
"loss": 0.6044,
"step": 8850
},
{
"epoch": 1.2093037367606934,
"grad_norm": 0.3426538407802582,
"learning_rate": 4.0758693361433086e-05,
"loss": 0.6021,
"step": 8875
},
{
"epoch": 1.2127103731011994,
"grad_norm": 0.32610806822776794,
"learning_rate": 4.058306989813839e-05,
"loss": 0.5936,
"step": 8900
},
{
"epoch": 1.2161170094417055,
"grad_norm": 0.3073614537715912,
"learning_rate": 4.04074464348437e-05,
"loss": 0.5826,
"step": 8925
},
{
"epoch": 1.2195236457822116,
"grad_norm": 0.317548930644989,
"learning_rate": 4.0231822971548996e-05,
"loss": 0.5917,
"step": 8950
},
{
"epoch": 1.2229302821227177,
"grad_norm": 0.3202793300151825,
"learning_rate": 4.00561995082543e-05,
"loss": 0.5757,
"step": 8975
},
{
"epoch": 1.2263369184632238,
"grad_norm": 0.2945701479911804,
"learning_rate": 3.9880576044959614e-05,
"loss": 0.5901,
"step": 9000
},
{
"epoch": 1.2297435548037299,
"grad_norm": 0.3056606650352478,
"learning_rate": 3.970495258166491e-05,
"loss": 0.5979,
"step": 9025
},
{
"epoch": 1.233150191144236,
"grad_norm": 0.3158716559410095,
"learning_rate": 3.952932911837022e-05,
"loss": 0.5945,
"step": 9050
},
{
"epoch": 1.236556827484742,
"grad_norm": 0.3323630094528198,
"learning_rate": 3.935370565507552e-05,
"loss": 0.6056,
"step": 9075
},
{
"epoch": 1.2399634638252481,
"grad_norm": 0.31543681025505066,
"learning_rate": 3.9178082191780823e-05,
"loss": 0.5642,
"step": 9100
},
{
"epoch": 1.2433701001657542,
"grad_norm": 0.33846592903137207,
"learning_rate": 3.900245872848613e-05,
"loss": 0.585,
"step": 9125
},
{
"epoch": 1.2467767365062603,
"grad_norm": 0.2971675992012024,
"learning_rate": 3.882683526519143e-05,
"loss": 0.5953,
"step": 9150
},
{
"epoch": 1.2501833728467662,
"grad_norm": 0.3218803405761719,
"learning_rate": 3.8651211801896734e-05,
"loss": 0.5914,
"step": 9175
},
{
"epoch": 1.2535900091872723,
"grad_norm": 0.3213994801044464,
"learning_rate": 3.847558833860204e-05,
"loss": 0.5822,
"step": 9200
},
{
"epoch": 1.2569966455277783,
"grad_norm": 0.33393600583076477,
"learning_rate": 3.8299964875307345e-05,
"loss": 0.5937,
"step": 9225
},
{
"epoch": 1.2604032818682844,
"grad_norm": 0.27980881929397583,
"learning_rate": 3.812434141201265e-05,
"loss": 0.6056,
"step": 9250
},
{
"epoch": 1.2638099182087905,
"grad_norm": 0.3020496070384979,
"learning_rate": 3.794871794871795e-05,
"loss": 0.6062,
"step": 9275
},
{
"epoch": 1.2672165545492966,
"grad_norm": 0.3256005048751831,
"learning_rate": 3.7773094485423255e-05,
"loss": 0.6021,
"step": 9300
},
{
"epoch": 1.2706231908898027,
"grad_norm": 0.30306681990623474,
"learning_rate": 3.759747102212856e-05,
"loss": 0.608,
"step": 9325
},
{
"epoch": 1.2740298272303088,
"grad_norm": 0.2977856695652008,
"learning_rate": 3.742184755883386e-05,
"loss": 0.5785,
"step": 9350
},
{
"epoch": 1.2774364635708149,
"grad_norm": 0.3653808832168579,
"learning_rate": 3.7246224095539165e-05,
"loss": 0.5907,
"step": 9375
},
{
"epoch": 1.280843099911321,
"grad_norm": 0.3345490097999573,
"learning_rate": 3.707060063224447e-05,
"loss": 0.587,
"step": 9400
},
{
"epoch": 1.284249736251827,
"grad_norm": 0.2982236444950104,
"learning_rate": 3.6894977168949777e-05,
"loss": 0.5547,
"step": 9425
},
{
"epoch": 1.2876563725923331,
"grad_norm": 0.3151262402534485,
"learning_rate": 3.671935370565508e-05,
"loss": 0.5801,
"step": 9450
},
{
"epoch": 1.2910630089328392,
"grad_norm": 0.3142940402030945,
"learning_rate": 3.654373024236038e-05,
"loss": 0.574,
"step": 9475
},
{
"epoch": 1.2944696452733453,
"grad_norm": 0.3357933461666107,
"learning_rate": 3.636810677906569e-05,
"loss": 0.598,
"step": 9500
},
{
"epoch": 1.2978762816138514,
"grad_norm": 0.3080194294452667,
"learning_rate": 3.6192483315770986e-05,
"loss": 0.5895,
"step": 9525
},
{
"epoch": 1.3012829179543575,
"grad_norm": 0.29399389028549194,
"learning_rate": 3.601685985247629e-05,
"loss": 0.5606,
"step": 9550
},
{
"epoch": 1.3046895542948636,
"grad_norm": 0.3045051693916321,
"learning_rate": 3.58412363891816e-05,
"loss": 0.5781,
"step": 9575
},
{
"epoch": 1.3080961906353696,
"grad_norm": 0.3037458062171936,
"learning_rate": 3.5665612925886896e-05,
"loss": 0.5693,
"step": 9600
},
{
"epoch": 1.3115028269758757,
"grad_norm": 0.3291562497615814,
"learning_rate": 3.54899894625922e-05,
"loss": 0.5733,
"step": 9625
},
{
"epoch": 1.3149094633163818,
"grad_norm": 0.33404019474983215,
"learning_rate": 3.531436599929751e-05,
"loss": 0.5834,
"step": 9650
},
{
"epoch": 1.318316099656888,
"grad_norm": 0.3017292320728302,
"learning_rate": 3.513874253600281e-05,
"loss": 0.5724,
"step": 9675
},
{
"epoch": 1.321722735997394,
"grad_norm": 0.31595638394355774,
"learning_rate": 3.496311907270812e-05,
"loss": 0.5901,
"step": 9700
},
{
"epoch": 1.3251293723379,
"grad_norm": 0.29873600602149963,
"learning_rate": 3.478749560941342e-05,
"loss": 0.5758,
"step": 9725
},
{
"epoch": 1.3285360086784062,
"grad_norm": 0.3320747911930084,
"learning_rate": 3.461187214611872e-05,
"loss": 0.5882,
"step": 9750
},
{
"epoch": 1.3319426450189122,
"grad_norm": 0.2872166037559509,
"learning_rate": 3.443624868282403e-05,
"loss": 0.5942,
"step": 9775
},
{
"epoch": 1.3353492813594183,
"grad_norm": 0.2684285342693329,
"learning_rate": 3.426062521952933e-05,
"loss": 0.6047,
"step": 9800
},
{
"epoch": 1.3387559176999244,
"grad_norm": 0.3230988383293152,
"learning_rate": 3.408500175623463e-05,
"loss": 0.5786,
"step": 9825
},
{
"epoch": 1.3421625540404305,
"grad_norm": 0.33714237809181213,
"learning_rate": 3.390937829293994e-05,
"loss": 0.5895,
"step": 9850
},
{
"epoch": 1.3455691903809366,
"grad_norm": 0.29626786708831787,
"learning_rate": 3.3733754829645244e-05,
"loss": 0.5808,
"step": 9875
},
{
"epoch": 1.3489758267214427,
"grad_norm": 0.3248324990272522,
"learning_rate": 3.355813136635055e-05,
"loss": 0.5642,
"step": 9900
},
{
"epoch": 1.3523824630619485,
"grad_norm": 0.32865700125694275,
"learning_rate": 3.338250790305585e-05,
"loss": 0.5807,
"step": 9925
},
{
"epoch": 1.3557890994024546,
"grad_norm": 0.280898779630661,
"learning_rate": 3.3206884439761154e-05,
"loss": 0.6035,
"step": 9950
},
{
"epoch": 1.3591957357429607,
"grad_norm": 0.3238600492477417,
"learning_rate": 3.303126097646645e-05,
"loss": 0.5941,
"step": 9975
},
{
"epoch": 1.3626023720834668,
"grad_norm": 0.3435409367084503,
"learning_rate": 3.285563751317176e-05,
"loss": 0.5831,
"step": 10000
},
{
"epoch": 1.3660090084239729,
"grad_norm": 0.3290906250476837,
"learning_rate": 3.2680014049877065e-05,
"loss": 0.6023,
"step": 10025
},
{
"epoch": 1.369415644764479,
"grad_norm": 0.32303091883659363,
"learning_rate": 3.250439058658237e-05,
"loss": 0.5957,
"step": 10050
},
{
"epoch": 1.372822281104985,
"grad_norm": 0.30665114521980286,
"learning_rate": 3.2328767123287676e-05,
"loss": 0.586,
"step": 10075
},
{
"epoch": 1.3762289174454911,
"grad_norm": 0.32518666982650757,
"learning_rate": 3.215314365999298e-05,
"loss": 0.5829,
"step": 10100
},
{
"epoch": 1.3796355537859972,
"grad_norm": 0.3334820568561554,
"learning_rate": 3.197752019669828e-05,
"loss": 0.5939,
"step": 10125
},
{
"epoch": 1.3830421901265033,
"grad_norm": 0.30552002787590027,
"learning_rate": 3.1801896733403586e-05,
"loss": 0.5826,
"step": 10150
},
{
"epoch": 1.3864488264670094,
"grad_norm": 0.35440927743911743,
"learning_rate": 3.1626273270108885e-05,
"loss": 0.607,
"step": 10175
},
{
"epoch": 1.3898554628075155,
"grad_norm": 0.31508344411849976,
"learning_rate": 3.145064980681419e-05,
"loss": 0.564,
"step": 10200
},
{
"epoch": 1.3932620991480216,
"grad_norm": 0.2713547348976135,
"learning_rate": 3.1275026343519496e-05,
"loss": 0.5798,
"step": 10225
},
{
"epoch": 1.3966687354885277,
"grad_norm": 0.30810999870300293,
"learning_rate": 3.1099402880224795e-05,
"loss": 0.5887,
"step": 10250
},
{
"epoch": 1.4000753718290337,
"grad_norm": 0.29226067662239075,
"learning_rate": 3.092377941693011e-05,
"loss": 0.5992,
"step": 10275
},
{
"epoch": 1.4034820081695398,
"grad_norm": 0.35241231322288513,
"learning_rate": 3.0748155953635406e-05,
"loss": 0.5965,
"step": 10300
},
{
"epoch": 1.406888644510046,
"grad_norm": 0.321756511926651,
"learning_rate": 3.057253249034071e-05,
"loss": 0.6126,
"step": 10325
},
{
"epoch": 1.410295280850552,
"grad_norm": 0.28978297114372253,
"learning_rate": 3.0396909027046018e-05,
"loss": 0.5835,
"step": 10350
},
{
"epoch": 1.413701917191058,
"grad_norm": 0.3767097592353821,
"learning_rate": 3.0221285563751317e-05,
"loss": 0.6188,
"step": 10375
},
{
"epoch": 1.417108553531564,
"grad_norm": 0.2829996347427368,
"learning_rate": 3.0045662100456622e-05,
"loss": 0.6087,
"step": 10400
},
{
"epoch": 1.42051518987207,
"grad_norm": 0.3308793008327484,
"learning_rate": 2.9870038637161928e-05,
"loss": 0.5818,
"step": 10425
},
{
"epoch": 1.4239218262125761,
"grad_norm": 0.3255029618740082,
"learning_rate": 2.969441517386723e-05,
"loss": 0.5788,
"step": 10450
},
{
"epoch": 1.4273284625530822,
"grad_norm": 0.328983873128891,
"learning_rate": 2.9518791710572536e-05,
"loss": 0.6092,
"step": 10475
},
{
"epoch": 1.4307350988935883,
"grad_norm": 0.3405100703239441,
"learning_rate": 2.9343168247277835e-05,
"loss": 0.5924,
"step": 10500
},
{
"epoch": 1.4341417352340944,
"grad_norm": 0.30565881729125977,
"learning_rate": 2.916754478398314e-05,
"loss": 0.5827,
"step": 10525
},
{
"epoch": 1.4375483715746005,
"grad_norm": 0.2973788380622864,
"learning_rate": 2.899192132068845e-05,
"loss": 0.5797,
"step": 10550
},
{
"epoch": 1.4409550079151066,
"grad_norm": 0.316902756690979,
"learning_rate": 2.8816297857393748e-05,
"loss": 0.5763,
"step": 10575
},
{
"epoch": 1.4443616442556126,
"grad_norm": 0.30036553740501404,
"learning_rate": 2.8640674394099054e-05,
"loss": 0.59,
"step": 10600
},
{
"epoch": 1.4477682805961187,
"grad_norm": 0.3541527986526489,
"learning_rate": 2.8465050930804356e-05,
"loss": 0.6028,
"step": 10625
},
{
"epoch": 1.4511749169366248,
"grad_norm": 0.3401188850402832,
"learning_rate": 2.8289427467509662e-05,
"loss": 0.5771,
"step": 10650
},
{
"epoch": 1.454581553277131,
"grad_norm": 0.3095962405204773,
"learning_rate": 2.8113804004214967e-05,
"loss": 0.5766,
"step": 10675
},
{
"epoch": 1.457988189617637,
"grad_norm": 0.3148902952671051,
"learning_rate": 2.7938180540920266e-05,
"loss": 0.5878,
"step": 10700
},
{
"epoch": 1.461394825958143,
"grad_norm": 0.3642762303352356,
"learning_rate": 2.7762557077625572e-05,
"loss": 0.5876,
"step": 10725
},
{
"epoch": 1.4648014622986492,
"grad_norm": 0.3234134018421173,
"learning_rate": 2.7586933614330874e-05,
"loss": 0.5963,
"step": 10750
},
{
"epoch": 1.4682080986391552,
"grad_norm": 0.3760201632976532,
"learning_rate": 2.741131015103618e-05,
"loss": 0.5982,
"step": 10775
},
{
"epoch": 1.4716147349796613,
"grad_norm": 0.3245134651660919,
"learning_rate": 2.7235686687741486e-05,
"loss": 0.5676,
"step": 10800
},
{
"epoch": 1.4750213713201674,
"grad_norm": 0.32639381289482117,
"learning_rate": 2.7060063224446784e-05,
"loss": 0.5614,
"step": 10825
},
{
"epoch": 1.4784280076606735,
"grad_norm": 0.3028101325035095,
"learning_rate": 2.688443976115209e-05,
"loss": 0.5965,
"step": 10850
},
{
"epoch": 1.4818346440011796,
"grad_norm": 0.32848045229911804,
"learning_rate": 2.67088162978574e-05,
"loss": 0.5873,
"step": 10875
},
{
"epoch": 1.4852412803416857,
"grad_norm": 0.30644866824150085,
"learning_rate": 2.6533192834562698e-05,
"loss": 0.5908,
"step": 10900
},
{
"epoch": 1.4886479166821918,
"grad_norm": 0.3460093140602112,
"learning_rate": 2.6357569371268004e-05,
"loss": 0.5897,
"step": 10925
},
{
"epoch": 1.4920545530226978,
"grad_norm": 0.3794984221458435,
"learning_rate": 2.6181945907973306e-05,
"loss": 0.6111,
"step": 10950
},
{
"epoch": 1.495461189363204,
"grad_norm": 0.3346981108188629,
"learning_rate": 2.600632244467861e-05,
"loss": 0.5662,
"step": 10975
},
{
"epoch": 1.49886782570371,
"grad_norm": 0.32983726263046265,
"learning_rate": 2.5830698981383917e-05,
"loss": 0.5525,
"step": 11000
},
{
"epoch": 1.502274462044216,
"grad_norm": 0.3233675956726074,
"learning_rate": 2.5655075518089216e-05,
"loss": 0.5818,
"step": 11025
},
{
"epoch": 1.5056810983847222,
"grad_norm": 0.3246392607688904,
"learning_rate": 2.547945205479452e-05,
"loss": 0.5819,
"step": 11050
},
{
"epoch": 1.5090877347252283,
"grad_norm": 0.31511232256889343,
"learning_rate": 2.5303828591499824e-05,
"loss": 0.5608,
"step": 11075
},
{
"epoch": 1.5124943710657344,
"grad_norm": 0.3445887267589569,
"learning_rate": 2.512820512820513e-05,
"loss": 0.5964,
"step": 11100
},
{
"epoch": 1.5159010074062405,
"grad_norm": 0.29458969831466675,
"learning_rate": 2.4952581664910432e-05,
"loss": 0.562,
"step": 11125
},
{
"epoch": 1.5193076437467465,
"grad_norm": 0.31204789876937866,
"learning_rate": 2.4776958201615737e-05,
"loss": 0.5738,
"step": 11150
},
{
"epoch": 1.5227142800872526,
"grad_norm": 0.2912496030330658,
"learning_rate": 2.460133473832104e-05,
"loss": 0.5669,
"step": 11175
},
{
"epoch": 1.5261209164277587,
"grad_norm": 0.3034071922302246,
"learning_rate": 2.4425711275026345e-05,
"loss": 0.6167,
"step": 11200
},
{
"epoch": 1.5295275527682646,
"grad_norm": 0.3211260735988617,
"learning_rate": 2.4250087811731648e-05,
"loss": 0.5852,
"step": 11225
},
{
"epoch": 1.5329341891087707,
"grad_norm": 0.3125843405723572,
"learning_rate": 2.4074464348436953e-05,
"loss": 0.5985,
"step": 11250
},
{
"epoch": 1.5363408254492767,
"grad_norm": 0.3351622223854065,
"learning_rate": 2.3898840885142256e-05,
"loss": 0.5916,
"step": 11275
},
{
"epoch": 1.5397474617897828,
"grad_norm": 0.29850685596466064,
"learning_rate": 2.372321742184756e-05,
"loss": 0.5848,
"step": 11300
},
{
"epoch": 1.543154098130289,
"grad_norm": 0.30315324664115906,
"learning_rate": 2.3547593958552863e-05,
"loss": 0.5762,
"step": 11325
},
{
"epoch": 1.546560734470795,
"grad_norm": 0.31599122285842896,
"learning_rate": 2.3371970495258166e-05,
"loss": 0.5685,
"step": 11350
},
{
"epoch": 1.549967370811301,
"grad_norm": 0.3213142454624176,
"learning_rate": 2.319634703196347e-05,
"loss": 0.5985,
"step": 11375
},
{
"epoch": 1.5533740071518072,
"grad_norm": 0.3190346360206604,
"learning_rate": 2.3020723568668777e-05,
"loss": 0.5917,
"step": 11400
},
{
"epoch": 1.5567806434923133,
"grad_norm": 0.33793261647224426,
"learning_rate": 2.284510010537408e-05,
"loss": 0.5911,
"step": 11425
},
{
"epoch": 1.5601872798328194,
"grad_norm": 0.3071536421775818,
"learning_rate": 2.266947664207938e-05,
"loss": 0.5668,
"step": 11450
},
{
"epoch": 1.5635939161733254,
"grad_norm": 0.30443212389945984,
"learning_rate": 2.2493853178784687e-05,
"loss": 0.5872,
"step": 11475
},
{
"epoch": 1.5670005525138315,
"grad_norm": 0.2918742895126343,
"learning_rate": 2.231822971548999e-05,
"loss": 0.5678,
"step": 11500
},
{
"epoch": 1.5704071888543374,
"grad_norm": 0.31535714864730835,
"learning_rate": 2.2142606252195295e-05,
"loss": 0.5654,
"step": 11525
},
{
"epoch": 1.5738138251948435,
"grad_norm": 0.308967649936676,
"learning_rate": 2.1966982788900597e-05,
"loss": 0.5807,
"step": 11550
},
{
"epoch": 1.5772204615353496,
"grad_norm": 0.3183216452598572,
"learning_rate": 2.17913593256059e-05,
"loss": 0.5892,
"step": 11575
},
{
"epoch": 1.5806270978758556,
"grad_norm": 0.3481101393699646,
"learning_rate": 2.1615735862311205e-05,
"loss": 0.5777,
"step": 11600
},
{
"epoch": 1.5840337342163617,
"grad_norm": 0.3023226261138916,
"learning_rate": 2.144011239901651e-05,
"loss": 0.6049,
"step": 11625
},
{
"epoch": 1.5874403705568678,
"grad_norm": 0.3095088303089142,
"learning_rate": 2.1264488935721813e-05,
"loss": 0.5618,
"step": 11650
},
{
"epoch": 1.590847006897374,
"grad_norm": 0.3510863482952118,
"learning_rate": 2.1088865472427115e-05,
"loss": 0.5794,
"step": 11675
},
{
"epoch": 1.59425364323788,
"grad_norm": 0.3501506745815277,
"learning_rate": 2.091324200913242e-05,
"loss": 0.5883,
"step": 11700
},
{
"epoch": 1.597660279578386,
"grad_norm": 0.35157808661460876,
"learning_rate": 2.0737618545837727e-05,
"loss": 0.591,
"step": 11725
},
{
"epoch": 1.6010669159188922,
"grad_norm": 0.34667158126831055,
"learning_rate": 2.056199508254303e-05,
"loss": 0.5625,
"step": 11750
},
{
"epoch": 1.6044735522593983,
"grad_norm": 0.36488744616508484,
"learning_rate": 2.038637161924833e-05,
"loss": 0.6236,
"step": 11775
},
{
"epoch": 1.6078801885999043,
"grad_norm": 0.3272364139556885,
"learning_rate": 2.0210748155953637e-05,
"loss": 0.5867,
"step": 11800
},
{
"epoch": 1.6112868249404104,
"grad_norm": 0.3458411395549774,
"learning_rate": 2.0035124692658943e-05,
"loss": 0.5919,
"step": 11825
},
{
"epoch": 1.6146934612809165,
"grad_norm": 0.362530916929245,
"learning_rate": 1.9859501229364245e-05,
"loss": 0.5844,
"step": 11850
},
{
"epoch": 1.6181000976214226,
"grad_norm": 0.3061313331127167,
"learning_rate": 1.9683877766069547e-05,
"loss": 0.578,
"step": 11875
},
{
"epoch": 1.6215067339619287,
"grad_norm": 0.3104388117790222,
"learning_rate": 1.950825430277485e-05,
"loss": 0.5862,
"step": 11900
},
{
"epoch": 1.6249133703024348,
"grad_norm": 0.29367533326148987,
"learning_rate": 1.9332630839480155e-05,
"loss": 0.568,
"step": 11925
},
{
"epoch": 1.6283200066429409,
"grad_norm": 0.3182733952999115,
"learning_rate": 1.915700737618546e-05,
"loss": 0.584,
"step": 11950
},
{
"epoch": 1.631726642983447,
"grad_norm": 0.31644895672798157,
"learning_rate": 1.8981383912890763e-05,
"loss": 0.5666,
"step": 11975
},
{
"epoch": 1.635133279323953,
"grad_norm": 0.297619104385376,
"learning_rate": 1.8805760449596065e-05,
"loss": 0.5908,
"step": 12000
},
{
"epoch": 1.6385399156644591,
"grad_norm": 0.3179234266281128,
"learning_rate": 1.863013698630137e-05,
"loss": 0.589,
"step": 12025
},
{
"epoch": 1.6419465520049652,
"grad_norm": 0.3117496371269226,
"learning_rate": 1.8454513523006676e-05,
"loss": 0.5777,
"step": 12050
},
{
"epoch": 1.6453531883454713,
"grad_norm": 0.3699851930141449,
"learning_rate": 1.827889005971198e-05,
"loss": 0.595,
"step": 12075
},
{
"epoch": 1.6487598246859774,
"grad_norm": 0.30205854773521423,
"learning_rate": 1.810326659641728e-05,
"loss": 0.6009,
"step": 12100
},
{
"epoch": 1.6521664610264835,
"grad_norm": 0.2764529585838318,
"learning_rate": 1.7927643133122583e-05,
"loss": 0.5686,
"step": 12125
},
{
"epoch": 1.6555730973669895,
"grad_norm": 0.306428998708725,
"learning_rate": 1.7752019669827892e-05,
"loss": 0.5955,
"step": 12150
},
{
"epoch": 1.6589797337074956,
"grad_norm": 0.36335381865501404,
"learning_rate": 1.7576396206533195e-05,
"loss": 0.6048,
"step": 12175
},
{
"epoch": 1.6623863700480017,
"grad_norm": 0.2794727087020874,
"learning_rate": 1.7400772743238497e-05,
"loss": 0.5981,
"step": 12200
},
{
"epoch": 1.6657930063885078,
"grad_norm": 0.32972803711891174,
"learning_rate": 1.72251492799438e-05,
"loss": 0.5476,
"step": 12225
},
{
"epoch": 1.6691996427290139,
"grad_norm": 0.31174436211586,
"learning_rate": 1.7049525816649105e-05,
"loss": 0.5732,
"step": 12250
},
{
"epoch": 1.67260627906952,
"grad_norm": 0.3182581067085266,
"learning_rate": 1.687390235335441e-05,
"loss": 0.5736,
"step": 12275
},
{
"epoch": 1.676012915410026,
"grad_norm": 0.3544427752494812,
"learning_rate": 1.6698278890059713e-05,
"loss": 0.6036,
"step": 12300
},
{
"epoch": 1.6794195517505321,
"grad_norm": 0.31909558176994324,
"learning_rate": 1.6522655426765015e-05,
"loss": 0.6003,
"step": 12325
},
{
"epoch": 1.6828261880910382,
"grad_norm": 0.37591150403022766,
"learning_rate": 1.634703196347032e-05,
"loss": 0.5778,
"step": 12350
},
{
"epoch": 1.6862328244315443,
"grad_norm": 0.3523281514644623,
"learning_rate": 1.6171408500175626e-05,
"loss": 0.615,
"step": 12375
},
{
"epoch": 1.6896394607720504,
"grad_norm": 0.3197459578514099,
"learning_rate": 1.599578503688093e-05,
"loss": 0.5955,
"step": 12400
},
{
"epoch": 1.6930460971125565,
"grad_norm": 0.3070302903652191,
"learning_rate": 1.582016157358623e-05,
"loss": 0.5701,
"step": 12425
},
{
"epoch": 1.6964527334530624,
"grad_norm": 0.3082948327064514,
"learning_rate": 1.5644538110291533e-05,
"loss": 0.5852,
"step": 12450
},
{
"epoch": 1.6998593697935684,
"grad_norm": 0.3454219400882721,
"learning_rate": 1.5468914646996842e-05,
"loss": 0.6087,
"step": 12475
},
{
"epoch": 1.7032660061340745,
"grad_norm": 0.3495611548423767,
"learning_rate": 1.5293291183702144e-05,
"loss": 0.6165,
"step": 12500
},
{
"epoch": 1.7066726424745806,
"grad_norm": 0.35227829217910767,
"learning_rate": 1.5117667720407446e-05,
"loss": 0.6058,
"step": 12525
},
{
"epoch": 1.7100792788150867,
"grad_norm": 0.3307322859764099,
"learning_rate": 1.494204425711275e-05,
"loss": 0.613,
"step": 12550
},
{
"epoch": 1.7134859151555928,
"grad_norm": 0.29460999369621277,
"learning_rate": 1.4766420793818056e-05,
"loss": 0.5818,
"step": 12575
},
{
"epoch": 1.7168925514960989,
"grad_norm": 0.2916513979434967,
"learning_rate": 1.4590797330523358e-05,
"loss": 0.6006,
"step": 12600
},
{
"epoch": 1.720299187836605,
"grad_norm": 0.3213001787662506,
"learning_rate": 1.4415173867228662e-05,
"loss": 0.5895,
"step": 12625
},
{
"epoch": 1.723705824177111,
"grad_norm": 0.3271484375,
"learning_rate": 1.4239550403933966e-05,
"loss": 0.5751,
"step": 12650
},
{
"epoch": 1.7271124605176171,
"grad_norm": 0.31554198265075684,
"learning_rate": 1.4063926940639269e-05,
"loss": 0.5619,
"step": 12675
},
{
"epoch": 1.7305190968581232,
"grad_norm": 0.33524471521377563,
"learning_rate": 1.3888303477344574e-05,
"loss": 0.5983,
"step": 12700
},
{
"epoch": 1.7339257331986293,
"grad_norm": 0.2879363000392914,
"learning_rate": 1.3712680014049878e-05,
"loss": 0.5788,
"step": 12725
},
{
"epoch": 1.7373323695391352,
"grad_norm": 0.33548370003700256,
"learning_rate": 1.353705655075518e-05,
"loss": 0.5857,
"step": 12750
},
{
"epoch": 1.7407390058796413,
"grad_norm": 0.31736239790916443,
"learning_rate": 1.3361433087460484e-05,
"loss": 0.5763,
"step": 12775
},
{
"epoch": 1.7441456422201473,
"grad_norm": 0.2964527904987335,
"learning_rate": 1.318580962416579e-05,
"loss": 0.588,
"step": 12800
},
{
"epoch": 1.7475522785606534,
"grad_norm": 0.35111087560653687,
"learning_rate": 1.3010186160871094e-05,
"loss": 0.5727,
"step": 12825
},
{
"epoch": 1.7509589149011595,
"grad_norm": 0.3259890377521515,
"learning_rate": 1.2834562697576396e-05,
"loss": 0.5983,
"step": 12850
},
{
"epoch": 1.7543655512416656,
"grad_norm": 0.3256474435329437,
"learning_rate": 1.26589392342817e-05,
"loss": 0.6077,
"step": 12875
},
{
"epoch": 1.7577721875821717,
"grad_norm": 0.32251888513565063,
"learning_rate": 1.2483315770987004e-05,
"loss": 0.5836,
"step": 12900
},
{
"epoch": 1.7611788239226778,
"grad_norm": 0.3057299256324768,
"learning_rate": 1.230769230769231e-05,
"loss": 0.579,
"step": 12925
},
{
"epoch": 1.7645854602631839,
"grad_norm": 0.3247281610965729,
"learning_rate": 1.2132068844397612e-05,
"loss": 0.5909,
"step": 12950
},
{
"epoch": 1.76799209660369,
"grad_norm": 0.3315178453922272,
"learning_rate": 1.1956445381102916e-05,
"loss": 0.5778,
"step": 12975
},
{
"epoch": 1.771398732944196,
"grad_norm": 0.3456060588359833,
"learning_rate": 1.178082191780822e-05,
"loss": 0.5772,
"step": 13000
},
{
"epoch": 1.7748053692847021,
"grad_norm": 0.2976686358451843,
"learning_rate": 1.1605198454513524e-05,
"loss": 0.5782,
"step": 13025
},
{
"epoch": 1.7782120056252082,
"grad_norm": 0.3840683400630951,
"learning_rate": 1.1429574991218828e-05,
"loss": 0.5909,
"step": 13050
},
{
"epoch": 1.7816186419657143,
"grad_norm": 0.33678537607192993,
"learning_rate": 1.125395152792413e-05,
"loss": 0.6187,
"step": 13075
},
{
"epoch": 1.7850252783062204,
"grad_norm": 0.30530181527137756,
"learning_rate": 1.1078328064629436e-05,
"loss": 0.5983,
"step": 13100
},
{
"epoch": 1.7884319146467265,
"grad_norm": 0.37755846977233887,
"learning_rate": 1.0902704601334738e-05,
"loss": 0.5875,
"step": 13125
},
{
"epoch": 1.7918385509872325,
"grad_norm": 0.3588654100894928,
"learning_rate": 1.0727081138040044e-05,
"loss": 0.5778,
"step": 13150
},
{
"epoch": 1.7952451873277386,
"grad_norm": 0.33472126722335815,
"learning_rate": 1.0551457674745346e-05,
"loss": 0.5903,
"step": 13175
},
{
"epoch": 1.7986518236682447,
"grad_norm": 0.3286832869052887,
"learning_rate": 1.0375834211450652e-05,
"loss": 0.5776,
"step": 13200
},
{
"epoch": 1.8020584600087508,
"grad_norm": 0.32272377610206604,
"learning_rate": 1.0200210748155954e-05,
"loss": 0.6116,
"step": 13225
},
{
"epoch": 1.805465096349257,
"grad_norm": 0.3254378139972687,
"learning_rate": 1.0024587284861258e-05,
"loss": 0.5801,
"step": 13250
},
{
"epoch": 1.808871732689763,
"grad_norm": 0.2955749034881592,
"learning_rate": 9.848963821566562e-06,
"loss": 0.5808,
"step": 13275
},
{
"epoch": 1.812278369030269,
"grad_norm": 0.3526495695114136,
"learning_rate": 9.673340358271866e-06,
"loss": 0.5777,
"step": 13300
},
{
"epoch": 1.8156850053707752,
"grad_norm": 0.30919772386550903,
"learning_rate": 9.49771689497717e-06,
"loss": 0.5912,
"step": 13325
},
{
"epoch": 1.8190916417112812,
"grad_norm": 0.33964163064956665,
"learning_rate": 9.322093431682474e-06,
"loss": 0.5581,
"step": 13350
},
{
"epoch": 1.8224982780517873,
"grad_norm": 0.31757307052612305,
"learning_rate": 9.146469968387778e-06,
"loss": 0.5845,
"step": 13375
},
{
"epoch": 1.8259049143922934,
"grad_norm": 0.3355781137943268,
"learning_rate": 8.970846505093081e-06,
"loss": 0.5958,
"step": 13400
},
{
"epoch": 1.8293115507327995,
"grad_norm": 0.28168126940727234,
"learning_rate": 8.795223041798385e-06,
"loss": 0.5734,
"step": 13425
},
{
"epoch": 1.8327181870733056,
"grad_norm": 0.32637590169906616,
"learning_rate": 8.619599578503688e-06,
"loss": 0.5755,
"step": 13450
},
{
"epoch": 1.8361248234138117,
"grad_norm": 0.3058291971683502,
"learning_rate": 8.443976115208993e-06,
"loss": 0.5756,
"step": 13475
},
{
"epoch": 1.8395314597543178,
"grad_norm": 0.3433995246887207,
"learning_rate": 8.268352651914296e-06,
"loss": 0.5926,
"step": 13500
},
{
"epoch": 1.8429380960948238,
"grad_norm": 0.32179251313209534,
"learning_rate": 8.0927291886196e-06,
"loss": 0.5835,
"step": 13525
},
{
"epoch": 1.84634473243533,
"grad_norm": 0.3159351050853729,
"learning_rate": 7.917105725324904e-06,
"loss": 0.605,
"step": 13550
},
{
"epoch": 1.849751368775836,
"grad_norm": 0.300453782081604,
"learning_rate": 7.741482262030207e-06,
"loss": 0.5771,
"step": 13575
},
{
"epoch": 1.853158005116342,
"grad_norm": 0.29322999715805054,
"learning_rate": 7.565858798735511e-06,
"loss": 0.583,
"step": 13600
},
{
"epoch": 1.8565646414568482,
"grad_norm": 0.3196932375431061,
"learning_rate": 7.3902353354408145e-06,
"loss": 0.5619,
"step": 13625
},
{
"epoch": 1.8599712777973543,
"grad_norm": 0.35355255007743835,
"learning_rate": 7.214611872146119e-06,
"loss": 0.5968,
"step": 13650
},
{
"epoch": 1.8633779141378601,
"grad_norm": 0.29249703884124756,
"learning_rate": 7.0389884088514224e-06,
"loss": 0.59,
"step": 13675
},
{
"epoch": 1.8667845504783662,
"grad_norm": 0.32649850845336914,
"learning_rate": 6.863364945556727e-06,
"loss": 0.592,
"step": 13700
},
{
"epoch": 1.8701911868188723,
"grad_norm": 0.3016286790370941,
"learning_rate": 6.68774148226203e-06,
"loss": 0.5797,
"step": 13725
},
{
"epoch": 1.8735978231593784,
"grad_norm": 0.35082265734672546,
"learning_rate": 6.512118018967335e-06,
"loss": 0.5919,
"step": 13750
},
{
"epoch": 1.8770044594998845,
"grad_norm": 0.35020819306373596,
"learning_rate": 6.336494555672638e-06,
"loss": 0.5811,
"step": 13775
},
{
"epoch": 1.8804110958403906,
"grad_norm": 0.3795909285545349,
"learning_rate": 6.160871092377942e-06,
"loss": 0.5733,
"step": 13800
},
{
"epoch": 1.8838177321808967,
"grad_norm": 0.3251563608646393,
"learning_rate": 5.985247629083246e-06,
"loss": 0.5949,
"step": 13825
},
{
"epoch": 1.8872243685214027,
"grad_norm": 0.2950724959373474,
"learning_rate": 5.80962416578855e-06,
"loss": 0.5824,
"step": 13850
},
{
"epoch": 1.8906310048619088,
"grad_norm": 0.29133903980255127,
"learning_rate": 5.634000702493854e-06,
"loss": 0.5772,
"step": 13875
},
{
"epoch": 1.894037641202415,
"grad_norm": 0.3070613741874695,
"learning_rate": 5.458377239199157e-06,
"loss": 0.5785,
"step": 13900
},
{
"epoch": 1.897444277542921,
"grad_norm": 0.3025205135345459,
"learning_rate": 5.282753775904461e-06,
"loss": 0.6019,
"step": 13925
},
{
"epoch": 1.900850913883427,
"grad_norm": 0.3081129193305969,
"learning_rate": 5.107130312609765e-06,
"loss": 0.5867,
"step": 13950
},
{
"epoch": 1.904257550223933,
"grad_norm": 0.33867502212524414,
"learning_rate": 4.931506849315068e-06,
"loss": 0.5743,
"step": 13975
},
{
"epoch": 1.907664186564439,
"grad_norm": 0.30810752511024475,
"learning_rate": 4.755883386020372e-06,
"loss": 0.5821,
"step": 14000
},
{
"epoch": 1.9110708229049451,
"grad_norm": 0.30959686636924744,
"learning_rate": 4.580259922725676e-06,
"loss": 0.5822,
"step": 14025
},
{
"epoch": 1.9144774592454512,
"grad_norm": 0.35800251364707947,
"learning_rate": 4.40463645943098e-06,
"loss": 0.5697,
"step": 14050
},
{
"epoch": 1.9178840955859573,
"grad_norm": 0.30580955743789673,
"learning_rate": 4.229012996136284e-06,
"loss": 0.578,
"step": 14075
},
{
"epoch": 1.9212907319264634,
"grad_norm": 0.3346704840660095,
"learning_rate": 4.053389532841588e-06,
"loss": 0.5735,
"step": 14100
},
{
"epoch": 1.9246973682669695,
"grad_norm": 0.32596781849861145,
"learning_rate": 3.877766069546892e-06,
"loss": 0.5991,
"step": 14125
},
{
"epoch": 1.9281040046074756,
"grad_norm": 0.3023940920829773,
"learning_rate": 3.702142606252196e-06,
"loss": 0.5689,
"step": 14150
},
{
"epoch": 1.9315106409479816,
"grad_norm": 0.31074970960617065,
"learning_rate": 3.526519142957499e-06,
"loss": 0.5891,
"step": 14175
},
{
"epoch": 1.9349172772884877,
"grad_norm": 0.339542955160141,
"learning_rate": 3.350895679662803e-06,
"loss": 0.5745,
"step": 14200
},
{
"epoch": 1.9383239136289938,
"grad_norm": 0.28495198488235474,
"learning_rate": 3.175272216368107e-06,
"loss": 0.5832,
"step": 14225
},
{
"epoch": 1.9417305499695,
"grad_norm": 0.37374746799468994,
"learning_rate": 2.9996487530734104e-06,
"loss": 0.5965,
"step": 14250
},
{
"epoch": 1.945137186310006,
"grad_norm": 0.30754634737968445,
"learning_rate": 2.8240252897787144e-06,
"loss": 0.5732,
"step": 14275
},
{
"epoch": 1.948543822650512,
"grad_norm": 0.36621904373168945,
"learning_rate": 2.6484018264840183e-06,
"loss": 0.568,
"step": 14300
},
{
"epoch": 1.9519504589910182,
"grad_norm": 0.360920786857605,
"learning_rate": 2.4727783631893223e-06,
"loss": 0.5803,
"step": 14325
},
{
"epoch": 1.9553570953315242,
"grad_norm": 0.36131951212882996,
"learning_rate": 2.2971548998946262e-06,
"loss": 0.6093,
"step": 14350
},
{
"epoch": 1.9587637316720303,
"grad_norm": 0.2969054877758026,
"learning_rate": 2.1215314365999298e-06,
"loss": 0.5818,
"step": 14375
},
{
"epoch": 1.9621703680125364,
"grad_norm": 0.3174475431442261,
"learning_rate": 1.9459079733052337e-06,
"loss": 0.574,
"step": 14400
},
{
"epoch": 1.9655770043530425,
"grad_norm": 0.3457885682582855,
"learning_rate": 1.7702845100105375e-06,
"loss": 0.5844,
"step": 14425
},
{
"epoch": 1.9689836406935486,
"grad_norm": 0.3122316300868988,
"learning_rate": 1.5946610467158414e-06,
"loss": 0.5661,
"step": 14450
},
{
"epoch": 1.9723902770340547,
"grad_norm": 0.3251049518585205,
"learning_rate": 1.4190375834211452e-06,
"loss": 0.5774,
"step": 14475
},
{
"epoch": 1.9757969133745608,
"grad_norm": 0.31165990233421326,
"learning_rate": 1.243414120126449e-06,
"loss": 0.5852,
"step": 14500
},
{
"epoch": 1.9792035497150668,
"grad_norm": 0.3429218530654907,
"learning_rate": 1.0677906568317528e-06,
"loss": 0.5581,
"step": 14525
},
{
"epoch": 1.982610186055573,
"grad_norm": 0.34827032685279846,
"learning_rate": 8.921671935370565e-07,
"loss": 0.5762,
"step": 14550
},
{
"epoch": 1.986016822396079,
"grad_norm": 0.32945892214775085,
"learning_rate": 7.165437302423604e-07,
"loss": 0.5947,
"step": 14575
},
{
"epoch": 1.989423458736585,
"grad_norm": 0.3533202111721039,
"learning_rate": 5.409202669476642e-07,
"loss": 0.5808,
"step": 14600
},
{
"epoch": 1.9928300950770912,
"grad_norm": 0.33338454365730286,
"learning_rate": 3.6529680365296803e-07,
"loss": 0.5559,
"step": 14625
},
{
"epoch": 1.9962367314175973,
"grad_norm": 0.29166388511657715,
"learning_rate": 1.8967334035827188e-07,
"loss": 0.5723,
"step": 14650
},
{
"epoch": 1.9996433677581034,
"grad_norm": 0.3164283335208893,
"learning_rate": 1.4049877063575695e-08,
"loss": 0.5964,
"step": 14675
}
],
"logging_steps": 25,
"max_steps": 14676,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.244907645688297e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}