AusmitM's picture
Upload 31 files
d127049 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 99.5475113122172,
"eval_steps": 20000,
"global_step": 308000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03232062055591468,
"grad_norm": 122.48411560058594,
"learning_rate": 9.900000000000002e-06,
"loss": 5.087,
"step": 100
},
{
"epoch": 0.06464124111182935,
"grad_norm": 87.30136108398438,
"learning_rate": 1.9900000000000003e-05,
"loss": 4.5234,
"step": 200
},
{
"epoch": 0.09696186166774402,
"grad_norm": 14.430279731750488,
"learning_rate": 2.9900000000000002e-05,
"loss": 4.1619,
"step": 300
},
{
"epoch": 0.1292824822236587,
"grad_norm": 7.7585768699646,
"learning_rate": 3.99e-05,
"loss": 4.0647,
"step": 400
},
{
"epoch": 0.16160310277957338,
"grad_norm": 29.10972785949707,
"learning_rate": 4.99e-05,
"loss": 4.0281,
"step": 500
},
{
"epoch": 0.19392372333548805,
"grad_norm": 2.4226176738739014,
"learning_rate": 5.9900000000000006e-05,
"loss": 3.9624,
"step": 600
},
{
"epoch": 0.22624434389140272,
"grad_norm": 6.9928483963012695,
"learning_rate": 6.99e-05,
"loss": 3.9599,
"step": 700
},
{
"epoch": 0.2585649644473174,
"grad_norm": 2.1512234210968018,
"learning_rate": 7.99e-05,
"loss": 3.9121,
"step": 800
},
{
"epoch": 0.2908855850032321,
"grad_norm": 2.635658025741577,
"learning_rate": 8.989999999999999e-05,
"loss": 3.8726,
"step": 900
},
{
"epoch": 0.32320620555914675,
"grad_norm": 2.5363662242889404,
"learning_rate": 9.99e-05,
"loss": 3.8475,
"step": 1000
},
{
"epoch": 0.3555268261150614,
"grad_norm": 2.313490390777588,
"learning_rate": 0.0001099,
"loss": 3.8057,
"step": 1100
},
{
"epoch": 0.3878474466709761,
"grad_norm": 4.086860656738281,
"learning_rate": 0.00011990000000000001,
"loss": 3.7642,
"step": 1200
},
{
"epoch": 0.42016806722689076,
"grad_norm": 2.341607093811035,
"learning_rate": 0.00012989999999999999,
"loss": 3.7752,
"step": 1300
},
{
"epoch": 0.45248868778280543,
"grad_norm": 2.455345630645752,
"learning_rate": 0.0001399,
"loss": 3.759,
"step": 1400
},
{
"epoch": 0.4848093083387201,
"grad_norm": 2.450104236602783,
"learning_rate": 0.0001499,
"loss": 3.7248,
"step": 1500
},
{
"epoch": 0.5171299288946348,
"grad_norm": 15.239123344421387,
"learning_rate": 0.00015989999999999998,
"loss": 3.7233,
"step": 1600
},
{
"epoch": 0.5494505494505495,
"grad_norm": 2.546140670776367,
"learning_rate": 0.0001699,
"loss": 3.6952,
"step": 1700
},
{
"epoch": 0.5817711700064642,
"grad_norm": 2.6291635036468506,
"learning_rate": 0.0001799,
"loss": 3.6845,
"step": 1800
},
{
"epoch": 0.6140917905623788,
"grad_norm": 2.4870760440826416,
"learning_rate": 0.0001899,
"loss": 3.6955,
"step": 1900
},
{
"epoch": 0.6464124111182935,
"grad_norm": 2.939796209335327,
"learning_rate": 0.0001999,
"loss": 3.6431,
"step": 2000
},
{
"epoch": 0.6787330316742082,
"grad_norm": 2.5228323936462402,
"learning_rate": 0.0002099,
"loss": 3.6324,
"step": 2100
},
{
"epoch": 0.7110536522301228,
"grad_norm": 2.6174564361572266,
"learning_rate": 0.0002199,
"loss": 3.6422,
"step": 2200
},
{
"epoch": 0.7433742727860375,
"grad_norm": 2.353490114212036,
"learning_rate": 0.0002299,
"loss": 3.6162,
"step": 2300
},
{
"epoch": 0.7756948933419522,
"grad_norm": 2.3185763359069824,
"learning_rate": 0.0002399,
"loss": 3.6186,
"step": 2400
},
{
"epoch": 0.8080155138978669,
"grad_norm": 3.45831036567688,
"learning_rate": 0.0002499,
"loss": 3.5984,
"step": 2500
},
{
"epoch": 0.8403361344537815,
"grad_norm": 2.644831895828247,
"learning_rate": 0.00025990000000000003,
"loss": 3.5993,
"step": 2600
},
{
"epoch": 0.8726567550096962,
"grad_norm": 2.6962873935699463,
"learning_rate": 0.0002699,
"loss": 3.6063,
"step": 2700
},
{
"epoch": 0.9049773755656109,
"grad_norm": 2.297571897506714,
"learning_rate": 0.0002799,
"loss": 3.6051,
"step": 2800
},
{
"epoch": 0.9372979961215255,
"grad_norm": 2.1946680545806885,
"learning_rate": 0.0002899,
"loss": 3.5727,
"step": 2900
},
{
"epoch": 0.9696186166774402,
"grad_norm": 2.3791117668151855,
"learning_rate": 0.0002999,
"loss": 3.5559,
"step": 3000
},
{
"epoch": 1.0019392372333549,
"grad_norm": 1.4081765413284302,
"learning_rate": 0.0003099,
"loss": 3.5623,
"step": 3100
},
{
"epoch": 1.0342598577892697,
"grad_norm": 1.5577131509780884,
"learning_rate": 0.0003199,
"loss": 3.5032,
"step": 3200
},
{
"epoch": 1.0665804783451842,
"grad_norm": 1.6537963151931763,
"learning_rate": 0.00032990000000000005,
"loss": 3.5059,
"step": 3300
},
{
"epoch": 1.098901098901099,
"grad_norm": 1.1358909606933594,
"learning_rate": 0.00033989999999999997,
"loss": 3.4895,
"step": 3400
},
{
"epoch": 1.1312217194570136,
"grad_norm": 1.8418935537338257,
"learning_rate": 0.0003499,
"loss": 3.5043,
"step": 3500
},
{
"epoch": 1.1635423400129283,
"grad_norm": 1.2883477210998535,
"learning_rate": 0.0003599,
"loss": 3.4928,
"step": 3600
},
{
"epoch": 1.195862960568843,
"grad_norm": 2.2627828121185303,
"learning_rate": 0.0003699,
"loss": 3.484,
"step": 3700
},
{
"epoch": 1.2281835811247577,
"grad_norm": 1.3333110809326172,
"learning_rate": 0.0003799,
"loss": 3.4796,
"step": 3800
},
{
"epoch": 1.2605042016806722,
"grad_norm": 1.2305139303207397,
"learning_rate": 0.00038990000000000004,
"loss": 3.4748,
"step": 3900
},
{
"epoch": 1.292824822236587,
"grad_norm": 1.0505295991897583,
"learning_rate": 0.00039989999999999996,
"loss": 3.4668,
"step": 4000
},
{
"epoch": 1.3251454427925016,
"grad_norm": 1.9814112186431885,
"learning_rate": 0.0004099,
"loss": 3.472,
"step": 4100
},
{
"epoch": 1.3574660633484164,
"grad_norm": 1.0137572288513184,
"learning_rate": 0.0004199,
"loss": 3.476,
"step": 4200
},
{
"epoch": 1.389786683904331,
"grad_norm": 1.4311473369598389,
"learning_rate": 0.0004299,
"loss": 3.4757,
"step": 4300
},
{
"epoch": 1.4221073044602457,
"grad_norm": 1.5800271034240723,
"learning_rate": 0.0004399,
"loss": 3.4591,
"step": 4400
},
{
"epoch": 1.4544279250161603,
"grad_norm": 1.2509582042694092,
"learning_rate": 0.00044990000000000004,
"loss": 3.4495,
"step": 4500
},
{
"epoch": 1.486748545572075,
"grad_norm": 1.3119666576385498,
"learning_rate": 0.0004599,
"loss": 3.4512,
"step": 4600
},
{
"epoch": 1.5190691661279896,
"grad_norm": 1.1141108274459839,
"learning_rate": 0.0004699,
"loss": 3.4655,
"step": 4700
},
{
"epoch": 1.5513897866839044,
"grad_norm": 1.295493721961975,
"learning_rate": 0.0004799,
"loss": 3.4512,
"step": 4800
},
{
"epoch": 1.5837104072398192,
"grad_norm": 0.9966570734977722,
"learning_rate": 0.0004899,
"loss": 3.4436,
"step": 4900
},
{
"epoch": 1.6160310277957337,
"grad_norm": 0.8886629939079285,
"learning_rate": 0.0004999000000000001,
"loss": 3.4193,
"step": 5000
},
{
"epoch": 1.6483516483516483,
"grad_norm": 1.174604892730713,
"learning_rate": 0.0005099,
"loss": 3.4435,
"step": 5100
},
{
"epoch": 1.680672268907563,
"grad_norm": 1.2078009843826294,
"learning_rate": 0.0005199,
"loss": 3.4152,
"step": 5200
},
{
"epoch": 1.7129928894634778,
"grad_norm": 1.044562816619873,
"learning_rate": 0.0005299,
"loss": 3.4225,
"step": 5300
},
{
"epoch": 1.7453135100193924,
"grad_norm": 1.065690517425537,
"learning_rate": 0.0005399000000000001,
"loss": 3.4187,
"step": 5400
},
{
"epoch": 1.777634130575307,
"grad_norm": 0.9627252817153931,
"learning_rate": 0.0005499000000000001,
"loss": 3.4267,
"step": 5500
},
{
"epoch": 1.8099547511312217,
"grad_norm": 0.9242098331451416,
"learning_rate": 0.0005599,
"loss": 3.4098,
"step": 5600
},
{
"epoch": 1.8422753716871365,
"grad_norm": 1.0979822874069214,
"learning_rate": 0.0005698999999999999,
"loss": 3.4007,
"step": 5700
},
{
"epoch": 1.874595992243051,
"grad_norm": 0.9085142016410828,
"learning_rate": 0.0005799,
"loss": 3.3882,
"step": 5800
},
{
"epoch": 1.9069166127989656,
"grad_norm": 1.1204533576965332,
"learning_rate": 0.0005899,
"loss": 3.3966,
"step": 5900
},
{
"epoch": 1.9392372333548804,
"grad_norm": 1.6729990243911743,
"learning_rate": 0.0005999,
"loss": 3.391,
"step": 6000
},
{
"epoch": 1.9715578539107952,
"grad_norm": 1.193272590637207,
"learning_rate": 0.0006099,
"loss": 3.398,
"step": 6100
},
{
"epoch": 2.0038784744667097,
"grad_norm": 0.9911984801292419,
"learning_rate": 0.0006199,
"loss": 3.3991,
"step": 6200
},
{
"epoch": 2.0361990950226243,
"grad_norm": 1.0938817262649536,
"learning_rate": 0.0006299000000000001,
"loss": 3.2833,
"step": 6300
},
{
"epoch": 2.0685197155785393,
"grad_norm": 1.0412477254867554,
"learning_rate": 0.0006399,
"loss": 3.3089,
"step": 6400
},
{
"epoch": 2.100840336134454,
"grad_norm": 1.0674519538879395,
"learning_rate": 0.0006499,
"loss": 3.3082,
"step": 6500
},
{
"epoch": 2.1331609566903684,
"grad_norm": 1.0387150049209595,
"learning_rate": 0.0006599,
"loss": 3.2791,
"step": 6600
},
{
"epoch": 2.165481577246283,
"grad_norm": 1.2339116334915161,
"learning_rate": 0.0006699000000000001,
"loss": 3.3188,
"step": 6700
},
{
"epoch": 2.197802197802198,
"grad_norm": 1.027089238166809,
"learning_rate": 0.0006799,
"loss": 3.3212,
"step": 6800
},
{
"epoch": 2.2301228183581125,
"grad_norm": 0.8910061120986938,
"learning_rate": 0.0006899,
"loss": 3.303,
"step": 6900
},
{
"epoch": 2.262443438914027,
"grad_norm": 1.0685038566589355,
"learning_rate": 0.0006998999999999999,
"loss": 3.3141,
"step": 7000
},
{
"epoch": 2.2947640594699417,
"grad_norm": 1.1801034212112427,
"learning_rate": 0.0007099,
"loss": 3.3029,
"step": 7100
},
{
"epoch": 2.3270846800258567,
"grad_norm": 0.927495539188385,
"learning_rate": 0.0007199,
"loss": 3.3018,
"step": 7200
},
{
"epoch": 2.3594053005817712,
"grad_norm": 0.9997018575668335,
"learning_rate": 0.0007299,
"loss": 3.3107,
"step": 7300
},
{
"epoch": 2.391725921137686,
"grad_norm": 0.9022483825683594,
"learning_rate": 0.0007399,
"loss": 3.3105,
"step": 7400
},
{
"epoch": 2.4240465416936003,
"grad_norm": 1.1087478399276733,
"learning_rate": 0.0007499000000000001,
"loss": 3.3093,
"step": 7500
},
{
"epoch": 2.4563671622495153,
"grad_norm": 0.8586679100990295,
"learning_rate": 0.0007599,
"loss": 3.2884,
"step": 7600
},
{
"epoch": 2.48868778280543,
"grad_norm": 0.9941036105155945,
"learning_rate": 0.0007699,
"loss": 3.286,
"step": 7700
},
{
"epoch": 2.5210084033613445,
"grad_norm": 0.8098440766334534,
"learning_rate": 0.0007799,
"loss": 3.292,
"step": 7800
},
{
"epoch": 2.553329023917259,
"grad_norm": 0.8971424102783203,
"learning_rate": 0.0007899000000000001,
"loss": 3.3003,
"step": 7900
},
{
"epoch": 2.585649644473174,
"grad_norm": 0.7246127724647522,
"learning_rate": 0.0007999000000000001,
"loss": 3.2941,
"step": 8000
},
{
"epoch": 2.6179702650290886,
"grad_norm": 1.125386118888855,
"learning_rate": 0.0008099,
"loss": 3.2923,
"step": 8100
},
{
"epoch": 2.650290885585003,
"grad_norm": 0.8445305228233337,
"learning_rate": 0.0008198999999999999,
"loss": 3.2854,
"step": 8200
},
{
"epoch": 2.682611506140918,
"grad_norm": 0.9022735953330994,
"learning_rate": 0.0008299,
"loss": 3.3128,
"step": 8300
},
{
"epoch": 2.7149321266968327,
"grad_norm": 2.568601608276367,
"learning_rate": 0.0008399,
"loss": 3.2935,
"step": 8400
},
{
"epoch": 2.7472527472527473,
"grad_norm": 0.9319080114364624,
"learning_rate": 0.0008499,
"loss": 3.2836,
"step": 8500
},
{
"epoch": 2.779573367808662,
"grad_norm": 0.9496336579322815,
"learning_rate": 0.0008599,
"loss": 3.2652,
"step": 8600
},
{
"epoch": 2.8118939883645764,
"grad_norm": 0.9994519948959351,
"learning_rate": 0.0008699000000000001,
"loss": 3.2707,
"step": 8700
},
{
"epoch": 2.8442146089204914,
"grad_norm": 0.8992823958396912,
"learning_rate": 0.0008799000000000001,
"loss": 3.2758,
"step": 8800
},
{
"epoch": 2.876535229476406,
"grad_norm": 0.9746232032775879,
"learning_rate": 0.0008899,
"loss": 3.2662,
"step": 8900
},
{
"epoch": 2.9088558500323205,
"grad_norm": 1.0509666204452515,
"learning_rate": 0.0008999,
"loss": 3.287,
"step": 9000
},
{
"epoch": 2.9411764705882355,
"grad_norm": 1.1123683452606201,
"learning_rate": 0.0009099,
"loss": 3.2874,
"step": 9100
},
{
"epoch": 2.97349709114415,
"grad_norm": 0.8263154029846191,
"learning_rate": 0.0009199000000000001,
"loss": 3.2478,
"step": 9200
},
{
"epoch": 3.0058177117000646,
"grad_norm": 0.9960469007492065,
"learning_rate": 0.0009299,
"loss": 3.2641,
"step": 9300
},
{
"epoch": 3.038138332255979,
"grad_norm": 0.9817208051681519,
"learning_rate": 0.0009399,
"loss": 3.1547,
"step": 9400
},
{
"epoch": 3.070458952811894,
"grad_norm": 0.8626974821090698,
"learning_rate": 0.0009498999999999999,
"loss": 3.1857,
"step": 9500
},
{
"epoch": 3.1027795733678087,
"grad_norm": 0.8734513521194458,
"learning_rate": 0.0009599,
"loss": 3.1828,
"step": 9600
},
{
"epoch": 3.1351001939237233,
"grad_norm": 0.9288407564163208,
"learning_rate": 0.0009699,
"loss": 3.1667,
"step": 9700
},
{
"epoch": 3.167420814479638,
"grad_norm": 1.0279629230499268,
"learning_rate": 0.0009799,
"loss": 3.1817,
"step": 9800
},
{
"epoch": 3.199741435035553,
"grad_norm": 0.8655309677124023,
"learning_rate": 0.0009899,
"loss": 3.1927,
"step": 9900
},
{
"epoch": 3.2320620555914674,
"grad_norm": 0.8720046877861023,
"learning_rate": 0.0009999,
"loss": 3.1844,
"step": 10000
},
{
"epoch": 3.264382676147382,
"grad_norm": 0.9178308844566345,
"learning_rate": 0.001,
"loss": 3.1809,
"step": 10100
},
{
"epoch": 3.2967032967032965,
"grad_norm": 1.1305406093597412,
"learning_rate": 0.001,
"loss": 3.1835,
"step": 10200
},
{
"epoch": 3.3290239172592115,
"grad_norm": 1.032468318939209,
"learning_rate": 0.001,
"loss": 3.1599,
"step": 10300
},
{
"epoch": 3.361344537815126,
"grad_norm": 0.9638125896453857,
"learning_rate": 0.001,
"loss": 3.1734,
"step": 10400
},
{
"epoch": 3.3936651583710407,
"grad_norm": 0.8970602750778198,
"learning_rate": 0.001,
"loss": 3.1596,
"step": 10500
},
{
"epoch": 3.425985778926955,
"grad_norm": 0.9424933195114136,
"learning_rate": 0.001,
"loss": 3.1612,
"step": 10600
},
{
"epoch": 3.45830639948287,
"grad_norm": 1.008062720298767,
"learning_rate": 0.001,
"loss": 3.1634,
"step": 10700
},
{
"epoch": 3.490627020038785,
"grad_norm": 0.8778404593467712,
"learning_rate": 0.001,
"loss": 3.1653,
"step": 10800
},
{
"epoch": 3.5229476405946993,
"grad_norm": 1.0320971012115479,
"learning_rate": 0.001,
"loss": 3.1862,
"step": 10900
},
{
"epoch": 3.555268261150614,
"grad_norm": 0.7651787996292114,
"learning_rate": 0.001,
"loss": 3.1588,
"step": 11000
},
{
"epoch": 3.587588881706529,
"grad_norm": 1.0109679698944092,
"learning_rate": 0.001,
"loss": 3.1633,
"step": 11100
},
{
"epoch": 3.6199095022624435,
"grad_norm": 1.1707773208618164,
"learning_rate": 0.001,
"loss": 3.1605,
"step": 11200
},
{
"epoch": 3.652230122818358,
"grad_norm": 0.8497399091720581,
"learning_rate": 0.001,
"loss": 3.1628,
"step": 11300
},
{
"epoch": 3.684550743374273,
"grad_norm": 1.1759374141693115,
"learning_rate": 0.001,
"loss": 3.1498,
"step": 11400
},
{
"epoch": 3.7168713639301876,
"grad_norm": 1.3136417865753174,
"learning_rate": 0.001,
"loss": 3.1711,
"step": 11500
},
{
"epoch": 3.749191984486102,
"grad_norm": 0.8515832424163818,
"learning_rate": 0.001,
"loss": 3.1654,
"step": 11600
},
{
"epoch": 3.7815126050420167,
"grad_norm": 0.9094183444976807,
"learning_rate": 0.001,
"loss": 3.1736,
"step": 11700
},
{
"epoch": 3.8138332255979313,
"grad_norm": 0.7091506719589233,
"learning_rate": 0.001,
"loss": 3.1555,
"step": 11800
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.7586659789085388,
"learning_rate": 0.001,
"loss": 3.1599,
"step": 11900
},
{
"epoch": 3.878474466709761,
"grad_norm": 0.8182700276374817,
"learning_rate": 0.001,
"loss": 3.1678,
"step": 12000
},
{
"epoch": 3.9107950872656754,
"grad_norm": 0.8606002926826477,
"learning_rate": 0.001,
"loss": 3.1539,
"step": 12100
},
{
"epoch": 3.9431157078215904,
"grad_norm": 1.0754919052124023,
"learning_rate": 0.001,
"loss": 3.1526,
"step": 12200
},
{
"epoch": 3.975436328377505,
"grad_norm": 1.1159805059432983,
"learning_rate": 0.001,
"loss": 3.1545,
"step": 12300
},
{
"epoch": 4.0077569489334195,
"grad_norm": 0.8374843597412109,
"learning_rate": 0.001,
"loss": 3.1302,
"step": 12400
},
{
"epoch": 4.040077569489334,
"grad_norm": 0.8675714135169983,
"learning_rate": 0.001,
"loss": 3.0011,
"step": 12500
},
{
"epoch": 4.072398190045249,
"grad_norm": 0.9006522297859192,
"learning_rate": 0.001,
"loss": 3.0227,
"step": 12600
},
{
"epoch": 4.104718810601163,
"grad_norm": 0.8449679613113403,
"learning_rate": 0.001,
"loss": 3.0373,
"step": 12700
},
{
"epoch": 4.137039431157079,
"grad_norm": 1.1275441646575928,
"learning_rate": 0.001,
"loss": 3.029,
"step": 12800
},
{
"epoch": 4.169360051712993,
"grad_norm": 0.9259504079818726,
"learning_rate": 0.001,
"loss": 3.0351,
"step": 12900
},
{
"epoch": 4.201680672268908,
"grad_norm": 1.0209553241729736,
"learning_rate": 0.001,
"loss": 3.0349,
"step": 13000
},
{
"epoch": 4.234001292824822,
"grad_norm": 0.9716935753822327,
"learning_rate": 0.001,
"loss": 3.0446,
"step": 13100
},
{
"epoch": 4.266321913380737,
"grad_norm": 0.8766793608665466,
"learning_rate": 0.001,
"loss": 3.0304,
"step": 13200
},
{
"epoch": 4.298642533936651,
"grad_norm": 0.9728161692619324,
"learning_rate": 0.001,
"loss": 3.0246,
"step": 13300
},
{
"epoch": 4.330963154492566,
"grad_norm": 0.8230871558189392,
"learning_rate": 0.001,
"loss": 3.0515,
"step": 13400
},
{
"epoch": 4.3632837750484805,
"grad_norm": 0.8542620539665222,
"learning_rate": 0.001,
"loss": 3.0577,
"step": 13500
},
{
"epoch": 4.395604395604396,
"grad_norm": 0.7272443175315857,
"learning_rate": 0.001,
"loss": 3.0452,
"step": 13600
},
{
"epoch": 4.4279250161603105,
"grad_norm": 1.0077018737792969,
"learning_rate": 0.001,
"loss": 3.0209,
"step": 13700
},
{
"epoch": 4.460245636716225,
"grad_norm": 1.0092633962631226,
"learning_rate": 0.001,
"loss": 3.05,
"step": 13800
},
{
"epoch": 4.49256625727214,
"grad_norm": 0.7395654916763306,
"learning_rate": 0.001,
"loss": 3.0595,
"step": 13900
},
{
"epoch": 4.524886877828054,
"grad_norm": 0.9956360459327698,
"learning_rate": 0.001,
"loss": 3.0167,
"step": 14000
},
{
"epoch": 4.557207498383969,
"grad_norm": 0.9180870652198792,
"learning_rate": 0.001,
"loss": 3.0449,
"step": 14100
},
{
"epoch": 4.589528118939883,
"grad_norm": 0.9045624732971191,
"learning_rate": 0.001,
"loss": 3.0553,
"step": 14200
},
{
"epoch": 4.621848739495798,
"grad_norm": 0.795102596282959,
"learning_rate": 0.001,
"loss": 3.0544,
"step": 14300
},
{
"epoch": 4.654169360051713,
"grad_norm": 1.0817164182662964,
"learning_rate": 0.001,
"loss": 3.0537,
"step": 14400
},
{
"epoch": 4.686489980607628,
"grad_norm": 0.9697984457015991,
"learning_rate": 0.001,
"loss": 3.0513,
"step": 14500
},
{
"epoch": 4.7188106011635425,
"grad_norm": 0.903374969959259,
"learning_rate": 0.001,
"loss": 3.0283,
"step": 14600
},
{
"epoch": 4.751131221719457,
"grad_norm": 1.154241681098938,
"learning_rate": 0.001,
"loss": 3.0389,
"step": 14700
},
{
"epoch": 4.783451842275372,
"grad_norm": 0.7839650511741638,
"learning_rate": 0.001,
"loss": 3.0233,
"step": 14800
},
{
"epoch": 4.815772462831286,
"grad_norm": 1.3288406133651733,
"learning_rate": 0.001,
"loss": 3.0315,
"step": 14900
},
{
"epoch": 4.848093083387201,
"grad_norm": 0.9391863942146301,
"learning_rate": 0.001,
"loss": 3.0416,
"step": 15000
},
{
"epoch": 4.880413703943116,
"grad_norm": 0.9329890608787537,
"learning_rate": 0.001,
"loss": 3.0265,
"step": 15100
},
{
"epoch": 4.912734324499031,
"grad_norm": 0.9532723426818848,
"learning_rate": 0.001,
"loss": 3.038,
"step": 15200
},
{
"epoch": 4.945054945054945,
"grad_norm": 0.9412986040115356,
"learning_rate": 0.001,
"loss": 3.0333,
"step": 15300
},
{
"epoch": 4.97737556561086,
"grad_norm": 1.205365538597107,
"learning_rate": 0.001,
"loss": 3.0419,
"step": 15400
},
{
"epoch": 5.009696186166774,
"grad_norm": 0.9483180642127991,
"learning_rate": 0.001,
"loss": 3.0027,
"step": 15500
},
{
"epoch": 5.042016806722689,
"grad_norm": 1.1812211275100708,
"learning_rate": 0.001,
"loss": 2.8706,
"step": 15600
},
{
"epoch": 5.0743374272786035,
"grad_norm": 0.8328335881233215,
"learning_rate": 0.001,
"loss": 2.9088,
"step": 15700
},
{
"epoch": 5.106658047834518,
"grad_norm": 0.9921544790267944,
"learning_rate": 0.001,
"loss": 2.9154,
"step": 15800
},
{
"epoch": 5.1389786683904335,
"grad_norm": 0.984326183795929,
"learning_rate": 0.001,
"loss": 2.9198,
"step": 15900
},
{
"epoch": 5.171299288946348,
"grad_norm": 0.9495807886123657,
"learning_rate": 0.001,
"loss": 2.9274,
"step": 16000
},
{
"epoch": 5.203619909502263,
"grad_norm": 1.0462549924850464,
"learning_rate": 0.001,
"loss": 2.9166,
"step": 16100
},
{
"epoch": 5.235940530058177,
"grad_norm": 0.8766588568687439,
"learning_rate": 0.001,
"loss": 2.9068,
"step": 16200
},
{
"epoch": 5.268261150614092,
"grad_norm": 0.7101683616638184,
"learning_rate": 0.001,
"loss": 2.9257,
"step": 16300
},
{
"epoch": 5.300581771170006,
"grad_norm": 0.957855761051178,
"learning_rate": 0.001,
"loss": 2.9323,
"step": 16400
},
{
"epoch": 5.332902391725921,
"grad_norm": 1.2239158153533936,
"learning_rate": 0.001,
"loss": 2.9191,
"step": 16500
},
{
"epoch": 5.365223012281835,
"grad_norm": 0.8347125053405762,
"learning_rate": 0.001,
"loss": 2.9195,
"step": 16600
},
{
"epoch": 5.397543632837751,
"grad_norm": 0.8946273922920227,
"learning_rate": 0.001,
"loss": 2.9198,
"step": 16700
},
{
"epoch": 5.429864253393665,
"grad_norm": 0.8327140212059021,
"learning_rate": 0.001,
"loss": 2.9386,
"step": 16800
},
{
"epoch": 5.46218487394958,
"grad_norm": 0.9591065049171448,
"learning_rate": 0.001,
"loss": 2.9233,
"step": 16900
},
{
"epoch": 5.4945054945054945,
"grad_norm": 0.9572122693061829,
"learning_rate": 0.001,
"loss": 2.9417,
"step": 17000
},
{
"epoch": 5.526826115061409,
"grad_norm": 1.2478572130203247,
"learning_rate": 0.001,
"loss": 2.9227,
"step": 17100
},
{
"epoch": 5.559146735617324,
"grad_norm": 0.7681442499160767,
"learning_rate": 0.001,
"loss": 2.9318,
"step": 17200
},
{
"epoch": 5.591467356173238,
"grad_norm": 0.9959331750869751,
"learning_rate": 0.001,
"loss": 2.9099,
"step": 17300
},
{
"epoch": 5.623787976729153,
"grad_norm": 1.1006550788879395,
"learning_rate": 0.001,
"loss": 2.9279,
"step": 17400
},
{
"epoch": 5.656108597285068,
"grad_norm": 0.9101115465164185,
"learning_rate": 0.001,
"loss": 2.9323,
"step": 17500
},
{
"epoch": 5.688429217840983,
"grad_norm": 1.1711993217468262,
"learning_rate": 0.001,
"loss": 2.9567,
"step": 17600
},
{
"epoch": 5.720749838396897,
"grad_norm": 0.9631717205047607,
"learning_rate": 0.001,
"loss": 2.9452,
"step": 17700
},
{
"epoch": 5.753070458952812,
"grad_norm": 1.028151035308838,
"learning_rate": 0.001,
"loss": 2.9354,
"step": 17800
},
{
"epoch": 5.785391079508726,
"grad_norm": 0.9212917685508728,
"learning_rate": 0.001,
"loss": 2.9225,
"step": 17900
},
{
"epoch": 5.817711700064641,
"grad_norm": 1.1386103630065918,
"learning_rate": 0.001,
"loss": 2.9372,
"step": 18000
},
{
"epoch": 5.850032320620556,
"grad_norm": 0.8606694936752319,
"learning_rate": 0.001,
"loss": 2.9466,
"step": 18100
},
{
"epoch": 5.882352941176471,
"grad_norm": 0.7786224484443665,
"learning_rate": 0.001,
"loss": 2.9463,
"step": 18200
},
{
"epoch": 5.914673561732386,
"grad_norm": 0.8389948010444641,
"learning_rate": 0.001,
"loss": 2.959,
"step": 18300
},
{
"epoch": 5.9469941822883,
"grad_norm": 1.326764464378357,
"learning_rate": 0.001,
"loss": 2.9324,
"step": 18400
},
{
"epoch": 5.979314802844215,
"grad_norm": 1.221278190612793,
"learning_rate": 0.001,
"loss": 2.938,
"step": 18500
},
{
"epoch": 6.011635423400129,
"grad_norm": 0.9087767004966736,
"learning_rate": 0.001,
"loss": 2.8858,
"step": 18600
},
{
"epoch": 6.043956043956044,
"grad_norm": 0.9513673782348633,
"learning_rate": 0.001,
"loss": 2.7961,
"step": 18700
},
{
"epoch": 6.076276664511958,
"grad_norm": 0.9944157004356384,
"learning_rate": 0.001,
"loss": 2.7991,
"step": 18800
},
{
"epoch": 6.108597285067873,
"grad_norm": 1.06227707862854,
"learning_rate": 0.001,
"loss": 2.8068,
"step": 18900
},
{
"epoch": 6.140917905623788,
"grad_norm": 0.8983224034309387,
"learning_rate": 0.001,
"loss": 2.8201,
"step": 19000
},
{
"epoch": 6.173238526179703,
"grad_norm": 0.7782119512557983,
"learning_rate": 0.001,
"loss": 2.8152,
"step": 19100
},
{
"epoch": 6.2055591467356175,
"grad_norm": 0.8689744472503662,
"learning_rate": 0.001,
"loss": 2.8198,
"step": 19200
},
{
"epoch": 6.237879767291532,
"grad_norm": 1.1996628046035767,
"learning_rate": 0.001,
"loss": 2.8325,
"step": 19300
},
{
"epoch": 6.270200387847447,
"grad_norm": 0.844738245010376,
"learning_rate": 0.001,
"loss": 2.8508,
"step": 19400
},
{
"epoch": 6.302521008403361,
"grad_norm": 0.9009729027748108,
"learning_rate": 0.001,
"loss": 2.8238,
"step": 19500
},
{
"epoch": 6.334841628959276,
"grad_norm": 1.2987245321273804,
"learning_rate": 0.001,
"loss": 2.8231,
"step": 19600
},
{
"epoch": 6.36716224951519,
"grad_norm": 1.0361422300338745,
"learning_rate": 0.001,
"loss": 2.8338,
"step": 19700
},
{
"epoch": 6.399482870071106,
"grad_norm": 1.0403621196746826,
"learning_rate": 0.001,
"loss": 2.8508,
"step": 19800
},
{
"epoch": 6.43180349062702,
"grad_norm": 0.9830465912818909,
"learning_rate": 0.001,
"loss": 2.833,
"step": 19900
},
{
"epoch": 6.464124111182935,
"grad_norm": 0.7783878445625305,
"learning_rate": 0.001,
"loss": 2.8237,
"step": 20000
},
{
"epoch": 6.496444731738849,
"grad_norm": 0.8542726635932922,
"learning_rate": 0.001,
"loss": 2.8433,
"step": 20100
},
{
"epoch": 6.528765352294764,
"grad_norm": 0.9122792482376099,
"learning_rate": 0.001,
"loss": 2.8397,
"step": 20200
},
{
"epoch": 6.5610859728506785,
"grad_norm": 1.1444542407989502,
"learning_rate": 0.001,
"loss": 2.8342,
"step": 20300
},
{
"epoch": 6.593406593406593,
"grad_norm": 1.118303894996643,
"learning_rate": 0.001,
"loss": 2.8437,
"step": 20400
},
{
"epoch": 6.625727213962508,
"grad_norm": 1.069883108139038,
"learning_rate": 0.001,
"loss": 2.8436,
"step": 20500
},
{
"epoch": 6.658047834518423,
"grad_norm": 0.9313819408416748,
"learning_rate": 0.001,
"loss": 2.8396,
"step": 20600
},
{
"epoch": 6.690368455074338,
"grad_norm": 1.0891497135162354,
"learning_rate": 0.001,
"loss": 2.8338,
"step": 20700
},
{
"epoch": 6.722689075630252,
"grad_norm": 0.9340724945068359,
"learning_rate": 0.001,
"loss": 2.8487,
"step": 20800
},
{
"epoch": 6.755009696186167,
"grad_norm": 0.8091602325439453,
"learning_rate": 0.001,
"loss": 2.8663,
"step": 20900
},
{
"epoch": 6.787330316742081,
"grad_norm": 0.878212034702301,
"learning_rate": 0.001,
"loss": 2.8463,
"step": 21000
},
{
"epoch": 6.819650937297996,
"grad_norm": 0.7118565440177917,
"learning_rate": 0.001,
"loss": 2.8466,
"step": 21100
},
{
"epoch": 6.85197155785391,
"grad_norm": 0.7867164611816406,
"learning_rate": 0.001,
"loss": 2.8717,
"step": 21200
},
{
"epoch": 6.884292178409826,
"grad_norm": 1.037338137626648,
"learning_rate": 0.001,
"loss": 2.8507,
"step": 21300
},
{
"epoch": 6.91661279896574,
"grad_norm": 1.0783958435058594,
"learning_rate": 0.001,
"loss": 2.8654,
"step": 21400
},
{
"epoch": 6.948933419521655,
"grad_norm": 1.2687408924102783,
"learning_rate": 0.001,
"loss": 2.8545,
"step": 21500
},
{
"epoch": 6.98125404007757,
"grad_norm": 1.236606478691101,
"learning_rate": 0.001,
"loss": 2.8371,
"step": 21600
},
{
"epoch": 7.013574660633484,
"grad_norm": 0.8000703454017639,
"learning_rate": 0.001,
"loss": 2.7781,
"step": 21700
},
{
"epoch": 7.045895281189399,
"grad_norm": 1.101654052734375,
"learning_rate": 0.001,
"loss": 2.7135,
"step": 21800
},
{
"epoch": 7.078215901745313,
"grad_norm": 1.2058213949203491,
"learning_rate": 0.001,
"loss": 2.7033,
"step": 21900
},
{
"epoch": 7.110536522301228,
"grad_norm": 1.1319844722747803,
"learning_rate": 0.001,
"loss": 2.7142,
"step": 22000
},
{
"epoch": 7.142857142857143,
"grad_norm": 1.2110048532485962,
"learning_rate": 0.001,
"loss": 2.7249,
"step": 22100
},
{
"epoch": 7.175177763413058,
"grad_norm": 1.1869632005691528,
"learning_rate": 0.001,
"loss": 2.7229,
"step": 22200
},
{
"epoch": 7.207498383968972,
"grad_norm": 0.8604994416236877,
"learning_rate": 0.001,
"loss": 2.7444,
"step": 22300
},
{
"epoch": 7.239819004524887,
"grad_norm": 1.1467421054840088,
"learning_rate": 0.001,
"loss": 2.7492,
"step": 22400
},
{
"epoch": 7.2721396250808015,
"grad_norm": 0.9088888764381409,
"learning_rate": 0.001,
"loss": 2.7485,
"step": 22500
},
{
"epoch": 7.304460245636716,
"grad_norm": 1.0480924844741821,
"learning_rate": 0.001,
"loss": 2.7441,
"step": 22600
},
{
"epoch": 7.336780866192631,
"grad_norm": 0.8503157496452332,
"learning_rate": 0.001,
"loss": 2.7429,
"step": 22700
},
{
"epoch": 7.369101486748546,
"grad_norm": 0.9091809391975403,
"learning_rate": 0.001,
"loss": 2.765,
"step": 22800
},
{
"epoch": 7.401422107304461,
"grad_norm": 1.3554952144622803,
"learning_rate": 0.001,
"loss": 2.7709,
"step": 22900
},
{
"epoch": 7.433742727860375,
"grad_norm": 1.0642281770706177,
"learning_rate": 0.001,
"loss": 2.7586,
"step": 23000
},
{
"epoch": 7.46606334841629,
"grad_norm": 1.150550127029419,
"learning_rate": 0.001,
"loss": 2.7648,
"step": 23100
},
{
"epoch": 7.498383968972204,
"grad_norm": 0.9647549986839294,
"learning_rate": 0.001,
"loss": 2.7699,
"step": 23200
},
{
"epoch": 7.530704589528119,
"grad_norm": 1.3209148645401,
"learning_rate": 0.001,
"loss": 2.7576,
"step": 23300
},
{
"epoch": 7.563025210084033,
"grad_norm": 0.9192158579826355,
"learning_rate": 0.001,
"loss": 2.7804,
"step": 23400
},
{
"epoch": 7.595345830639948,
"grad_norm": 1.0111809968948364,
"learning_rate": 0.001,
"loss": 2.7506,
"step": 23500
},
{
"epoch": 7.6276664511958625,
"grad_norm": 0.9270694851875305,
"learning_rate": 0.001,
"loss": 2.7643,
"step": 23600
},
{
"epoch": 7.659987071751778,
"grad_norm": 1.0087300539016724,
"learning_rate": 0.001,
"loss": 2.7594,
"step": 23700
},
{
"epoch": 7.6923076923076925,
"grad_norm": 1.1877171993255615,
"learning_rate": 0.001,
"loss": 2.7515,
"step": 23800
},
{
"epoch": 7.724628312863607,
"grad_norm": 1.1692532300949097,
"learning_rate": 0.001,
"loss": 2.7635,
"step": 23900
},
{
"epoch": 7.756948933419522,
"grad_norm": 0.9619380831718445,
"learning_rate": 0.001,
"loss": 2.7688,
"step": 24000
},
{
"epoch": 7.789269553975436,
"grad_norm": 1.0232017040252686,
"learning_rate": 0.001,
"loss": 2.7836,
"step": 24100
},
{
"epoch": 7.821590174531351,
"grad_norm": 1.0013618469238281,
"learning_rate": 0.001,
"loss": 2.7778,
"step": 24200
},
{
"epoch": 7.853910795087265,
"grad_norm": 0.97769695520401,
"learning_rate": 0.001,
"loss": 2.7579,
"step": 24300
},
{
"epoch": 7.886231415643181,
"grad_norm": 0.9150475859642029,
"learning_rate": 0.001,
"loss": 2.7683,
"step": 24400
},
{
"epoch": 7.918552036199095,
"grad_norm": 1.0076943635940552,
"learning_rate": 0.001,
"loss": 2.767,
"step": 24500
},
{
"epoch": 7.95087265675501,
"grad_norm": 1.0352141857147217,
"learning_rate": 0.001,
"loss": 2.8031,
"step": 24600
},
{
"epoch": 7.983193277310924,
"grad_norm": 0.9151076078414917,
"learning_rate": 0.001,
"loss": 2.7802,
"step": 24700
},
{
"epoch": 8.015513897866839,
"grad_norm": 1.3630614280700684,
"learning_rate": 0.001,
"loss": 2.6941,
"step": 24800
},
{
"epoch": 8.047834518422754,
"grad_norm": 1.331531047821045,
"learning_rate": 0.001,
"loss": 2.6443,
"step": 24900
},
{
"epoch": 8.080155138978668,
"grad_norm": 1.6701931953430176,
"learning_rate": 0.001,
"loss": 2.6514,
"step": 25000
},
{
"epoch": 8.112475759534583,
"grad_norm": 1.3195340633392334,
"learning_rate": 0.001,
"loss": 2.6601,
"step": 25100
},
{
"epoch": 8.144796380090497,
"grad_norm": 1.5026133060455322,
"learning_rate": 0.001,
"loss": 2.6744,
"step": 25200
},
{
"epoch": 8.177117000646412,
"grad_norm": 1.1848560571670532,
"learning_rate": 0.001,
"loss": 2.6525,
"step": 25300
},
{
"epoch": 8.209437621202326,
"grad_norm": 1.428328275680542,
"learning_rate": 0.001,
"loss": 2.661,
"step": 25400
},
{
"epoch": 8.241758241758241,
"grad_norm": 1.4003264904022217,
"learning_rate": 0.001,
"loss": 2.6613,
"step": 25500
},
{
"epoch": 8.274078862314157,
"grad_norm": 1.0878551006317139,
"learning_rate": 0.001,
"loss": 2.6452,
"step": 25600
},
{
"epoch": 8.306399482870072,
"grad_norm": 1.2808568477630615,
"learning_rate": 0.001,
"loss": 2.6624,
"step": 25700
},
{
"epoch": 8.338720103425986,
"grad_norm": 1.1686146259307861,
"learning_rate": 0.001,
"loss": 2.6687,
"step": 25800
},
{
"epoch": 8.371040723981901,
"grad_norm": 1.0730260610580444,
"learning_rate": 0.001,
"loss": 2.6768,
"step": 25900
},
{
"epoch": 8.403361344537815,
"grad_norm": 1.2151321172714233,
"learning_rate": 0.001,
"loss": 2.6459,
"step": 26000
},
{
"epoch": 8.43568196509373,
"grad_norm": 1.6539074182510376,
"learning_rate": 0.001,
"loss": 2.6671,
"step": 26100
},
{
"epoch": 8.468002585649645,
"grad_norm": 1.3920577764511108,
"learning_rate": 0.001,
"loss": 2.6961,
"step": 26200
},
{
"epoch": 8.50032320620556,
"grad_norm": 0.968061089515686,
"learning_rate": 0.001,
"loss": 2.6942,
"step": 26300
},
{
"epoch": 8.532643826761474,
"grad_norm": 1.220590353012085,
"learning_rate": 0.001,
"loss": 2.688,
"step": 26400
},
{
"epoch": 8.564964447317388,
"grad_norm": 1.3109445571899414,
"learning_rate": 0.001,
"loss": 2.6982,
"step": 26500
},
{
"epoch": 8.597285067873303,
"grad_norm": 1.6531808376312256,
"learning_rate": 0.001,
"loss": 2.6929,
"step": 26600
},
{
"epoch": 8.629605688429217,
"grad_norm": 1.1348577737808228,
"learning_rate": 0.001,
"loss": 2.7053,
"step": 26700
},
{
"epoch": 8.661926308985132,
"grad_norm": 1.5206853151321411,
"learning_rate": 0.001,
"loss": 2.687,
"step": 26800
},
{
"epoch": 8.694246929541046,
"grad_norm": 1.5895841121673584,
"learning_rate": 0.001,
"loss": 2.7004,
"step": 26900
},
{
"epoch": 8.726567550096961,
"grad_norm": 1.1910431385040283,
"learning_rate": 0.001,
"loss": 2.71,
"step": 27000
},
{
"epoch": 8.758888170652877,
"grad_norm": 1.8874714374542236,
"learning_rate": 0.001,
"loss": 2.6958,
"step": 27100
},
{
"epoch": 8.791208791208792,
"grad_norm": 1.01585054397583,
"learning_rate": 0.001,
"loss": 2.6898,
"step": 27200
},
{
"epoch": 8.823529411764707,
"grad_norm": 1.2293158769607544,
"learning_rate": 0.001,
"loss": 2.7001,
"step": 27300
},
{
"epoch": 8.855850032320621,
"grad_norm": 1.2832460403442383,
"learning_rate": 0.001,
"loss": 2.7213,
"step": 27400
},
{
"epoch": 8.888170652876536,
"grad_norm": 1.0620850324630737,
"learning_rate": 0.001,
"loss": 2.7072,
"step": 27500
},
{
"epoch": 8.92049127343245,
"grad_norm": 1.327416181564331,
"learning_rate": 0.001,
"loss": 2.69,
"step": 27600
},
{
"epoch": 8.952811893988365,
"grad_norm": 0.9163597226142883,
"learning_rate": 0.001,
"loss": 2.7301,
"step": 27700
},
{
"epoch": 8.98513251454428,
"grad_norm": 1.2720648050308228,
"learning_rate": 0.001,
"loss": 2.7143,
"step": 27800
},
{
"epoch": 9.017453135100194,
"grad_norm": 1.5404466390609741,
"learning_rate": 0.001,
"loss": 2.622,
"step": 27900
},
{
"epoch": 9.049773755656108,
"grad_norm": 1.5179469585418701,
"learning_rate": 0.001,
"loss": 2.5708,
"step": 28000
},
{
"epoch": 9.082094376212023,
"grad_norm": 1.2205370664596558,
"learning_rate": 0.001,
"loss": 2.5866,
"step": 28100
},
{
"epoch": 9.114414996767938,
"grad_norm": 1.3983556032180786,
"learning_rate": 0.001,
"loss": 2.5853,
"step": 28200
},
{
"epoch": 9.146735617323852,
"grad_norm": 1.6919760704040527,
"learning_rate": 0.001,
"loss": 2.5896,
"step": 28300
},
{
"epoch": 9.179056237879767,
"grad_norm": 1.3373256921768188,
"learning_rate": 0.001,
"loss": 2.5889,
"step": 28400
},
{
"epoch": 9.211376858435681,
"grad_norm": 1.7062350511550903,
"learning_rate": 0.001,
"loss": 2.5728,
"step": 28500
},
{
"epoch": 9.243697478991596,
"grad_norm": 1.7507829666137695,
"learning_rate": 0.001,
"loss": 2.5811,
"step": 28600
},
{
"epoch": 9.276018099547512,
"grad_norm": 1.2600959539413452,
"learning_rate": 0.001,
"loss": 2.5818,
"step": 28700
},
{
"epoch": 9.308338720103427,
"grad_norm": 1.0494632720947266,
"learning_rate": 0.001,
"loss": 2.6243,
"step": 28800
},
{
"epoch": 9.340659340659341,
"grad_norm": 1.623203158378601,
"learning_rate": 0.001,
"loss": 2.6241,
"step": 28900
},
{
"epoch": 9.372979961215256,
"grad_norm": 1.0560330152511597,
"learning_rate": 0.001,
"loss": 2.6189,
"step": 29000
},
{
"epoch": 9.40530058177117,
"grad_norm": 1.1647675037384033,
"learning_rate": 0.001,
"loss": 2.6266,
"step": 29100
},
{
"epoch": 9.437621202327085,
"grad_norm": 1.3463765382766724,
"learning_rate": 0.001,
"loss": 2.6203,
"step": 29200
},
{
"epoch": 9.469941822883,
"grad_norm": 0.9838733673095703,
"learning_rate": 0.001,
"loss": 2.6273,
"step": 29300
},
{
"epoch": 9.502262443438914,
"grad_norm": 1.7701879739761353,
"learning_rate": 0.001,
"loss": 2.6278,
"step": 29400
},
{
"epoch": 9.534583063994829,
"grad_norm": 0.891035795211792,
"learning_rate": 0.001,
"loss": 2.6208,
"step": 29500
},
{
"epoch": 9.566903684550743,
"grad_norm": 1.5723954439163208,
"learning_rate": 0.001,
"loss": 2.6093,
"step": 29600
},
{
"epoch": 9.599224305106658,
"grad_norm": 1.0477232933044434,
"learning_rate": 0.001,
"loss": 2.6275,
"step": 29700
},
{
"epoch": 9.631544925662572,
"grad_norm": 1.5019673109054565,
"learning_rate": 0.001,
"loss": 2.659,
"step": 29800
},
{
"epoch": 9.663865546218487,
"grad_norm": 1.2560871839523315,
"learning_rate": 0.001,
"loss": 2.6101,
"step": 29900
},
{
"epoch": 9.696186166774401,
"grad_norm": 1.1805680990219116,
"learning_rate": 0.001,
"loss": 2.6174,
"step": 30000
},
{
"epoch": 9.728506787330316,
"grad_norm": 1.0397446155548096,
"learning_rate": 0.001,
"loss": 2.6498,
"step": 30100
},
{
"epoch": 9.760827407886232,
"grad_norm": 1.6623647212982178,
"learning_rate": 0.001,
"loss": 2.6359,
"step": 30200
},
{
"epoch": 9.793148028442147,
"grad_norm": 1.26282799243927,
"learning_rate": 0.001,
"loss": 2.6579,
"step": 30300
},
{
"epoch": 9.825468648998061,
"grad_norm": 1.2911651134490967,
"learning_rate": 0.001,
"loss": 2.6161,
"step": 30400
},
{
"epoch": 9.857789269553976,
"grad_norm": 1.3239177465438843,
"learning_rate": 0.001,
"loss": 2.6292,
"step": 30500
},
{
"epoch": 9.89010989010989,
"grad_norm": 0.9709534049034119,
"learning_rate": 0.001,
"loss": 2.627,
"step": 30600
},
{
"epoch": 9.922430510665805,
"grad_norm": 1.332767367362976,
"learning_rate": 0.001,
"loss": 2.6526,
"step": 30700
},
{
"epoch": 9.95475113122172,
"grad_norm": 1.4657008647918701,
"learning_rate": 0.001,
"loss": 2.6515,
"step": 30800
},
{
"epoch": 9.987071751777634,
"grad_norm": 1.639960765838623,
"learning_rate": 0.001,
"loss": 2.6621,
"step": 30900
},
{
"epoch": 10.019392372333549,
"grad_norm": 1.050683617591858,
"learning_rate": 0.001,
"loss": 2.5767,
"step": 31000
},
{
"epoch": 10.051712992889463,
"grad_norm": 1.1165498495101929,
"learning_rate": 0.001,
"loss": 2.5307,
"step": 31100
},
{
"epoch": 10.084033613445378,
"grad_norm": 1.3107905387878418,
"learning_rate": 0.001,
"loss": 2.5127,
"step": 31200
},
{
"epoch": 10.116354234001292,
"grad_norm": 1.1361440420150757,
"learning_rate": 0.001,
"loss": 2.5248,
"step": 31300
},
{
"epoch": 10.148674854557207,
"grad_norm": 1.4246487617492676,
"learning_rate": 0.001,
"loss": 2.5209,
"step": 31400
},
{
"epoch": 10.180995475113122,
"grad_norm": 0.9404054880142212,
"learning_rate": 0.001,
"loss": 2.5429,
"step": 31500
},
{
"epoch": 10.213316095669036,
"grad_norm": 0.9457871913909912,
"learning_rate": 0.001,
"loss": 2.5414,
"step": 31600
},
{
"epoch": 10.24563671622495,
"grad_norm": 1.1399489641189575,
"learning_rate": 0.001,
"loss": 2.5448,
"step": 31700
},
{
"epoch": 10.277957336780867,
"grad_norm": 1.1223640441894531,
"learning_rate": 0.001,
"loss": 2.5353,
"step": 31800
},
{
"epoch": 10.310277957336782,
"grad_norm": 1.446655511856079,
"learning_rate": 0.001,
"loss": 2.5589,
"step": 31900
},
{
"epoch": 10.342598577892696,
"grad_norm": 0.9820685982704163,
"learning_rate": 0.001,
"loss": 2.5305,
"step": 32000
},
{
"epoch": 10.37491919844861,
"grad_norm": 0.9038324952125549,
"learning_rate": 0.001,
"loss": 2.5389,
"step": 32100
},
{
"epoch": 10.407239819004525,
"grad_norm": 1.184463381767273,
"learning_rate": 0.001,
"loss": 2.5791,
"step": 32200
},
{
"epoch": 10.43956043956044,
"grad_norm": 0.9426921606063843,
"learning_rate": 0.001,
"loss": 2.5612,
"step": 32300
},
{
"epoch": 10.471881060116354,
"grad_norm": 1.2373493909835815,
"learning_rate": 0.001,
"loss": 2.5603,
"step": 32400
},
{
"epoch": 10.504201680672269,
"grad_norm": 1.0498517751693726,
"learning_rate": 0.001,
"loss": 2.5471,
"step": 32500
},
{
"epoch": 10.536522301228183,
"grad_norm": 1.1927897930145264,
"learning_rate": 0.001,
"loss": 2.5663,
"step": 32600
},
{
"epoch": 10.568842921784098,
"grad_norm": 1.0986390113830566,
"learning_rate": 0.001,
"loss": 2.589,
"step": 32700
},
{
"epoch": 10.601163542340013,
"grad_norm": 1.0975826978683472,
"learning_rate": 0.001,
"loss": 2.5746,
"step": 32800
},
{
"epoch": 10.633484162895927,
"grad_norm": 1.4507718086242676,
"learning_rate": 0.001,
"loss": 2.5584,
"step": 32900
},
{
"epoch": 10.665804783451842,
"grad_norm": 0.9978941082954407,
"learning_rate": 0.001,
"loss": 2.5884,
"step": 33000
},
{
"epoch": 10.698125404007756,
"grad_norm": 1.0040268898010254,
"learning_rate": 0.001,
"loss": 2.5802,
"step": 33100
},
{
"epoch": 10.73044602456367,
"grad_norm": 1.1611864566802979,
"learning_rate": 0.001,
"loss": 2.5745,
"step": 33200
},
{
"epoch": 10.762766645119587,
"grad_norm": 0.9444515705108643,
"learning_rate": 0.001,
"loss": 2.5742,
"step": 33300
},
{
"epoch": 10.795087265675502,
"grad_norm": 0.770430326461792,
"learning_rate": 0.001,
"loss": 2.5833,
"step": 33400
},
{
"epoch": 10.827407886231416,
"grad_norm": 1.0619391202926636,
"learning_rate": 0.001,
"loss": 2.5904,
"step": 33500
},
{
"epoch": 10.85972850678733,
"grad_norm": 0.9841474890708923,
"learning_rate": 0.001,
"loss": 2.5825,
"step": 33600
},
{
"epoch": 10.892049127343245,
"grad_norm": 0.8374784588813782,
"learning_rate": 0.001,
"loss": 2.5847,
"step": 33700
},
{
"epoch": 10.92436974789916,
"grad_norm": 1.0929527282714844,
"learning_rate": 0.001,
"loss": 2.5871,
"step": 33800
},
{
"epoch": 10.956690368455074,
"grad_norm": 1.063624382019043,
"learning_rate": 0.001,
"loss": 2.5833,
"step": 33900
},
{
"epoch": 10.989010989010989,
"grad_norm": 0.9693267941474915,
"learning_rate": 0.001,
"loss": 2.5765,
"step": 34000
},
{
"epoch": 11.021331609566904,
"grad_norm": 1.0682049989700317,
"learning_rate": 0.001,
"loss": 2.5111,
"step": 34100
},
{
"epoch": 11.053652230122818,
"grad_norm": 1.2485145330429077,
"learning_rate": 0.001,
"loss": 2.4233,
"step": 34200
},
{
"epoch": 11.085972850678733,
"grad_norm": 1.2661662101745605,
"learning_rate": 0.001,
"loss": 2.4521,
"step": 34300
},
{
"epoch": 11.118293471234647,
"grad_norm": 0.9464433789253235,
"learning_rate": 0.001,
"loss": 2.4781,
"step": 34400
},
{
"epoch": 11.150614091790562,
"grad_norm": 1.4966256618499756,
"learning_rate": 0.001,
"loss": 2.4606,
"step": 34500
},
{
"epoch": 11.182934712346476,
"grad_norm": 0.9548174738883972,
"learning_rate": 0.001,
"loss": 2.474,
"step": 34600
},
{
"epoch": 11.215255332902391,
"grad_norm": 1.2072795629501343,
"learning_rate": 0.001,
"loss": 2.4905,
"step": 34700
},
{
"epoch": 11.247575953458306,
"grad_norm": 1.1185228824615479,
"learning_rate": 0.001,
"loss": 2.4699,
"step": 34800
},
{
"epoch": 11.279896574014222,
"grad_norm": 1.3499680757522583,
"learning_rate": 0.001,
"loss": 2.4886,
"step": 34900
},
{
"epoch": 11.312217194570136,
"grad_norm": 0.9926376938819885,
"learning_rate": 0.001,
"loss": 2.4868,
"step": 35000
},
{
"epoch": 11.344537815126051,
"grad_norm": 1.0508836507797241,
"learning_rate": 0.001,
"loss": 2.4988,
"step": 35100
},
{
"epoch": 11.376858435681966,
"grad_norm": 1.0693625211715698,
"learning_rate": 0.001,
"loss": 2.4944,
"step": 35200
},
{
"epoch": 11.40917905623788,
"grad_norm": 1.2590545415878296,
"learning_rate": 0.001,
"loss": 2.5104,
"step": 35300
},
{
"epoch": 11.441499676793795,
"grad_norm": 1.0293058156967163,
"learning_rate": 0.001,
"loss": 2.5196,
"step": 35400
},
{
"epoch": 11.47382029734971,
"grad_norm": 1.299343466758728,
"learning_rate": 0.001,
"loss": 2.5146,
"step": 35500
},
{
"epoch": 11.506140917905624,
"grad_norm": 1.1421232223510742,
"learning_rate": 0.001,
"loss": 2.5063,
"step": 35600
},
{
"epoch": 11.538461538461538,
"grad_norm": 0.8818128108978271,
"learning_rate": 0.001,
"loss": 2.5242,
"step": 35700
},
{
"epoch": 11.570782159017453,
"grad_norm": 1.0025140047073364,
"learning_rate": 0.001,
"loss": 2.4983,
"step": 35800
},
{
"epoch": 11.603102779573367,
"grad_norm": 0.9715794920921326,
"learning_rate": 0.001,
"loss": 2.5207,
"step": 35900
},
{
"epoch": 11.635423400129282,
"grad_norm": 1.076883316040039,
"learning_rate": 0.001,
"loss": 2.5222,
"step": 36000
},
{
"epoch": 11.667744020685197,
"grad_norm": 1.0112929344177246,
"learning_rate": 0.001,
"loss": 2.5248,
"step": 36100
},
{
"epoch": 11.700064641241111,
"grad_norm": 1.1007529497146606,
"learning_rate": 0.001,
"loss": 2.5448,
"step": 36200
},
{
"epoch": 11.732385261797026,
"grad_norm": 0.9470750093460083,
"learning_rate": 0.001,
"loss": 2.5358,
"step": 36300
},
{
"epoch": 11.764705882352942,
"grad_norm": 1.0336370468139648,
"learning_rate": 0.001,
"loss": 2.5418,
"step": 36400
},
{
"epoch": 11.797026502908857,
"grad_norm": 1.0315172672271729,
"learning_rate": 0.001,
"loss": 2.5124,
"step": 36500
},
{
"epoch": 11.829347123464771,
"grad_norm": 1.18720543384552,
"learning_rate": 0.001,
"loss": 2.5141,
"step": 36600
},
{
"epoch": 11.861667744020686,
"grad_norm": 1.034296989440918,
"learning_rate": 0.001,
"loss": 2.5322,
"step": 36700
},
{
"epoch": 11.8939883645766,
"grad_norm": 1.0513291358947754,
"learning_rate": 0.001,
"loss": 2.5226,
"step": 36800
},
{
"epoch": 11.926308985132515,
"grad_norm": 1.0058174133300781,
"learning_rate": 0.001,
"loss": 2.5452,
"step": 36900
},
{
"epoch": 11.95862960568843,
"grad_norm": 1.342353105545044,
"learning_rate": 0.001,
"loss": 2.546,
"step": 37000
},
{
"epoch": 11.990950226244344,
"grad_norm": 1.0556999444961548,
"learning_rate": 0.001,
"loss": 2.5334,
"step": 37100
},
{
"epoch": 12.023270846800258,
"grad_norm": 1.2010694742202759,
"learning_rate": 0.001,
"loss": 2.4479,
"step": 37200
},
{
"epoch": 12.055591467356173,
"grad_norm": 1.0156108140945435,
"learning_rate": 0.001,
"loss": 2.4137,
"step": 37300
},
{
"epoch": 12.087912087912088,
"grad_norm": 1.216589093208313,
"learning_rate": 0.001,
"loss": 2.3993,
"step": 37400
},
{
"epoch": 12.120232708468002,
"grad_norm": 1.1752727031707764,
"learning_rate": 0.001,
"loss": 2.3963,
"step": 37500
},
{
"epoch": 12.152553329023917,
"grad_norm": 1.212253451347351,
"learning_rate": 0.001,
"loss": 2.4086,
"step": 37600
},
{
"epoch": 12.184873949579831,
"grad_norm": 1.0437322854995728,
"learning_rate": 0.001,
"loss": 2.4183,
"step": 37700
},
{
"epoch": 12.217194570135746,
"grad_norm": 1.2049281597137451,
"learning_rate": 0.001,
"loss": 2.4453,
"step": 37800
},
{
"epoch": 12.24951519069166,
"grad_norm": 1.014501929283142,
"learning_rate": 0.001,
"loss": 2.4162,
"step": 37900
},
{
"epoch": 12.281835811247577,
"grad_norm": 1.2967098951339722,
"learning_rate": 0.001,
"loss": 2.4231,
"step": 38000
},
{
"epoch": 12.314156431803491,
"grad_norm": 1.0043957233428955,
"learning_rate": 0.001,
"loss": 2.4235,
"step": 38100
},
{
"epoch": 12.346477052359406,
"grad_norm": 1.1719709634780884,
"learning_rate": 0.001,
"loss": 2.4608,
"step": 38200
},
{
"epoch": 12.37879767291532,
"grad_norm": 1.6381487846374512,
"learning_rate": 0.001,
"loss": 2.453,
"step": 38300
},
{
"epoch": 12.411118293471235,
"grad_norm": 1.1440620422363281,
"learning_rate": 0.001,
"loss": 2.4596,
"step": 38400
},
{
"epoch": 12.44343891402715,
"grad_norm": 1.389151692390442,
"learning_rate": 0.001,
"loss": 2.4657,
"step": 38500
},
{
"epoch": 12.475759534583064,
"grad_norm": 1.1156370639801025,
"learning_rate": 0.001,
"loss": 2.4662,
"step": 38600
},
{
"epoch": 12.508080155138979,
"grad_norm": 0.8124172687530518,
"learning_rate": 0.001,
"loss": 2.4504,
"step": 38700
},
{
"epoch": 12.540400775694893,
"grad_norm": 1.3011460304260254,
"learning_rate": 0.001,
"loss": 2.4797,
"step": 38800
},
{
"epoch": 12.572721396250808,
"grad_norm": 0.8509902954101562,
"learning_rate": 0.001,
"loss": 2.4685,
"step": 38900
},
{
"epoch": 12.605042016806722,
"grad_norm": 1.1421103477478027,
"learning_rate": 0.001,
"loss": 2.4656,
"step": 39000
},
{
"epoch": 12.637362637362637,
"grad_norm": 1.0977075099945068,
"learning_rate": 0.001,
"loss": 2.4641,
"step": 39100
},
{
"epoch": 12.669683257918551,
"grad_norm": 0.9898512959480286,
"learning_rate": 0.001,
"loss": 2.4688,
"step": 39200
},
{
"epoch": 12.702003878474466,
"grad_norm": 1.0552868843078613,
"learning_rate": 0.001,
"loss": 2.4752,
"step": 39300
},
{
"epoch": 12.73432449903038,
"grad_norm": 0.7937177419662476,
"learning_rate": 0.001,
"loss": 2.4834,
"step": 39400
},
{
"epoch": 12.766645119586297,
"grad_norm": 1.175423264503479,
"learning_rate": 0.001,
"loss": 2.476,
"step": 39500
},
{
"epoch": 12.798965740142211,
"grad_norm": 0.835038959980011,
"learning_rate": 0.001,
"loss": 2.4991,
"step": 39600
},
{
"epoch": 12.831286360698126,
"grad_norm": 1.1652710437774658,
"learning_rate": 0.001,
"loss": 2.4946,
"step": 39700
},
{
"epoch": 12.86360698125404,
"grad_norm": 0.8950003385543823,
"learning_rate": 0.001,
"loss": 2.47,
"step": 39800
},
{
"epoch": 12.895927601809955,
"grad_norm": 1.1447982788085938,
"learning_rate": 0.001,
"loss": 2.4728,
"step": 39900
},
{
"epoch": 12.92824822236587,
"grad_norm": 1.3896085023880005,
"learning_rate": 0.001,
"loss": 2.473,
"step": 40000
},
{
"epoch": 12.960568842921784,
"grad_norm": 1.0610136985778809,
"learning_rate": 0.001,
"loss": 2.4644,
"step": 40100
},
{
"epoch": 12.992889463477699,
"grad_norm": 1.2621580362319946,
"learning_rate": 0.001,
"loss": 2.4881,
"step": 40200
},
{
"epoch": 13.025210084033613,
"grad_norm": 1.0967028141021729,
"learning_rate": 0.001,
"loss": 2.3948,
"step": 40300
},
{
"epoch": 13.057530704589528,
"grad_norm": 0.9754540920257568,
"learning_rate": 0.001,
"loss": 2.362,
"step": 40400
},
{
"epoch": 13.089851325145442,
"grad_norm": 1.023298740386963,
"learning_rate": 0.001,
"loss": 2.3488,
"step": 40500
},
{
"epoch": 13.122171945701357,
"grad_norm": 0.9918268322944641,
"learning_rate": 0.001,
"loss": 2.3396,
"step": 40600
},
{
"epoch": 13.154492566257272,
"grad_norm": 0.865107536315918,
"learning_rate": 0.001,
"loss": 2.3641,
"step": 40700
},
{
"epoch": 13.186813186813186,
"grad_norm": 1.0310488939285278,
"learning_rate": 0.001,
"loss": 2.3617,
"step": 40800
},
{
"epoch": 13.2191338073691,
"grad_norm": 0.8639553189277649,
"learning_rate": 0.001,
"loss": 2.3812,
"step": 40900
},
{
"epoch": 13.251454427925015,
"grad_norm": 1.0706894397735596,
"learning_rate": 0.001,
"loss": 2.3639,
"step": 41000
},
{
"epoch": 13.283775048480932,
"grad_norm": 1.1262027025222778,
"learning_rate": 0.001,
"loss": 2.3943,
"step": 41100
},
{
"epoch": 13.316095669036846,
"grad_norm": 1.2430620193481445,
"learning_rate": 0.001,
"loss": 2.3872,
"step": 41200
},
{
"epoch": 13.34841628959276,
"grad_norm": 0.976929783821106,
"learning_rate": 0.001,
"loss": 2.3958,
"step": 41300
},
{
"epoch": 13.380736910148675,
"grad_norm": 1.1228851079940796,
"learning_rate": 0.001,
"loss": 2.3995,
"step": 41400
},
{
"epoch": 13.41305753070459,
"grad_norm": 1.3405524492263794,
"learning_rate": 0.001,
"loss": 2.3798,
"step": 41500
},
{
"epoch": 13.445378151260504,
"grad_norm": 1.1651748418807983,
"learning_rate": 0.001,
"loss": 2.4097,
"step": 41600
},
{
"epoch": 13.477698771816419,
"grad_norm": 0.9650669693946838,
"learning_rate": 0.001,
"loss": 2.4045,
"step": 41700
},
{
"epoch": 13.510019392372334,
"grad_norm": 1.1848506927490234,
"learning_rate": 0.001,
"loss": 2.4206,
"step": 41800
},
{
"epoch": 13.542340012928248,
"grad_norm": 1.2294769287109375,
"learning_rate": 0.001,
"loss": 2.4142,
"step": 41900
},
{
"epoch": 13.574660633484163,
"grad_norm": 0.9633001685142517,
"learning_rate": 0.001,
"loss": 2.4102,
"step": 42000
},
{
"epoch": 13.606981254040077,
"grad_norm": 1.231916904449463,
"learning_rate": 0.001,
"loss": 2.4432,
"step": 42100
},
{
"epoch": 13.639301874595992,
"grad_norm": 1.1208678483963013,
"learning_rate": 0.001,
"loss": 2.4254,
"step": 42200
},
{
"epoch": 13.671622495151906,
"grad_norm": 1.0216820240020752,
"learning_rate": 0.001,
"loss": 2.4281,
"step": 42300
},
{
"epoch": 13.70394311570782,
"grad_norm": 1.2707115411758423,
"learning_rate": 0.001,
"loss": 2.4229,
"step": 42400
},
{
"epoch": 13.736263736263737,
"grad_norm": 1.4039192199707031,
"learning_rate": 0.001,
"loss": 2.4345,
"step": 42500
},
{
"epoch": 13.768584356819652,
"grad_norm": 0.8980364799499512,
"learning_rate": 0.001,
"loss": 2.4371,
"step": 42600
},
{
"epoch": 13.800904977375566,
"grad_norm": 1.0485930442810059,
"learning_rate": 0.001,
"loss": 2.4274,
"step": 42700
},
{
"epoch": 13.83322559793148,
"grad_norm": 0.94874507188797,
"learning_rate": 0.001,
"loss": 2.4408,
"step": 42800
},
{
"epoch": 13.865546218487395,
"grad_norm": 1.0252995491027832,
"learning_rate": 0.001,
"loss": 2.4494,
"step": 42900
},
{
"epoch": 13.89786683904331,
"grad_norm": 1.4931409358978271,
"learning_rate": 0.001,
"loss": 2.4155,
"step": 43000
},
{
"epoch": 13.930187459599225,
"grad_norm": 1.2104835510253906,
"learning_rate": 0.001,
"loss": 2.4439,
"step": 43100
},
{
"epoch": 13.96250808015514,
"grad_norm": 1.1501274108886719,
"learning_rate": 0.001,
"loss": 2.4477,
"step": 43200
},
{
"epoch": 13.994828700711054,
"grad_norm": 1.2111330032348633,
"learning_rate": 0.001,
"loss": 2.4257,
"step": 43300
},
{
"epoch": 14.027149321266968,
"grad_norm": 1.2563148736953735,
"learning_rate": 0.001,
"loss": 2.3305,
"step": 43400
},
{
"epoch": 14.059469941822883,
"grad_norm": 1.1479803323745728,
"learning_rate": 0.001,
"loss": 2.3233,
"step": 43500
},
{
"epoch": 14.091790562378797,
"grad_norm": 1.4190595149993896,
"learning_rate": 0.001,
"loss": 2.3296,
"step": 43600
},
{
"epoch": 14.124111182934712,
"grad_norm": 1.0025237798690796,
"learning_rate": 0.001,
"loss": 2.3046,
"step": 43700
},
{
"epoch": 14.156431803490626,
"grad_norm": 0.9679207801818848,
"learning_rate": 0.001,
"loss": 2.3314,
"step": 43800
},
{
"epoch": 14.188752424046541,
"grad_norm": 1.1125279664993286,
"learning_rate": 0.001,
"loss": 2.3419,
"step": 43900
},
{
"epoch": 14.221073044602456,
"grad_norm": 0.9743191599845886,
"learning_rate": 0.001,
"loss": 2.3285,
"step": 44000
},
{
"epoch": 14.25339366515837,
"grad_norm": 1.4148427248001099,
"learning_rate": 0.001,
"loss": 2.3147,
"step": 44100
},
{
"epoch": 14.285714285714286,
"grad_norm": 1.012438178062439,
"learning_rate": 0.001,
"loss": 2.325,
"step": 44200
},
{
"epoch": 14.318034906270201,
"grad_norm": 1.4741400480270386,
"learning_rate": 0.001,
"loss": 2.3371,
"step": 44300
},
{
"epoch": 14.350355526826116,
"grad_norm": 1.0528957843780518,
"learning_rate": 0.001,
"loss": 2.347,
"step": 44400
},
{
"epoch": 14.38267614738203,
"grad_norm": 1.1542372703552246,
"learning_rate": 0.001,
"loss": 2.3617,
"step": 44500
},
{
"epoch": 14.414996767937945,
"grad_norm": 1.225242018699646,
"learning_rate": 0.001,
"loss": 2.363,
"step": 44600
},
{
"epoch": 14.44731738849386,
"grad_norm": 1.0331146717071533,
"learning_rate": 0.001,
"loss": 2.3536,
"step": 44700
},
{
"epoch": 14.479638009049774,
"grad_norm": 1.1673269271850586,
"learning_rate": 0.001,
"loss": 2.3626,
"step": 44800
},
{
"epoch": 14.511958629605688,
"grad_norm": 1.0871409177780151,
"learning_rate": 0.001,
"loss": 2.3564,
"step": 44900
},
{
"epoch": 14.544279250161603,
"grad_norm": 1.0477733612060547,
"learning_rate": 0.001,
"loss": 2.361,
"step": 45000
},
{
"epoch": 14.576599870717518,
"grad_norm": 1.1164947748184204,
"learning_rate": 0.001,
"loss": 2.368,
"step": 45100
},
{
"epoch": 14.608920491273432,
"grad_norm": 1.1241072416305542,
"learning_rate": 0.001,
"loss": 2.3684,
"step": 45200
},
{
"epoch": 14.641241111829347,
"grad_norm": 1.1364701986312866,
"learning_rate": 0.001,
"loss": 2.3814,
"step": 45300
},
{
"epoch": 14.673561732385261,
"grad_norm": 1.3487145900726318,
"learning_rate": 0.001,
"loss": 2.3935,
"step": 45400
},
{
"epoch": 14.705882352941176,
"grad_norm": 1.0357153415679932,
"learning_rate": 0.001,
"loss": 2.3942,
"step": 45500
},
{
"epoch": 14.738202973497092,
"grad_norm": 1.5036461353302002,
"learning_rate": 0.001,
"loss": 2.4049,
"step": 45600
},
{
"epoch": 14.770523594053007,
"grad_norm": 1.0585894584655762,
"learning_rate": 0.001,
"loss": 2.3785,
"step": 45700
},
{
"epoch": 14.802844214608921,
"grad_norm": 0.9382624626159668,
"learning_rate": 0.001,
"loss": 2.3854,
"step": 45800
},
{
"epoch": 14.835164835164836,
"grad_norm": 1.0960943698883057,
"learning_rate": 0.001,
"loss": 2.3735,
"step": 45900
},
{
"epoch": 14.86748545572075,
"grad_norm": 1.1341259479522705,
"learning_rate": 0.001,
"loss": 2.3893,
"step": 46000
},
{
"epoch": 14.899806076276665,
"grad_norm": 0.8915718197822571,
"learning_rate": 0.001,
"loss": 2.3789,
"step": 46100
},
{
"epoch": 14.93212669683258,
"grad_norm": 1.1936798095703125,
"learning_rate": 0.001,
"loss": 2.4101,
"step": 46200
},
{
"epoch": 14.964447317388494,
"grad_norm": 1.2099779844284058,
"learning_rate": 0.001,
"loss": 2.4176,
"step": 46300
},
{
"epoch": 14.996767937944409,
"grad_norm": 0.8536863327026367,
"learning_rate": 0.001,
"loss": 2.3897,
"step": 46400
},
{
"epoch": 15.029088558500323,
"grad_norm": 1.2651400566101074,
"learning_rate": 0.001,
"loss": 2.2709,
"step": 46500
},
{
"epoch": 15.061409179056238,
"grad_norm": 1.1207914352416992,
"learning_rate": 0.001,
"loss": 2.2619,
"step": 46600
},
{
"epoch": 15.093729799612152,
"grad_norm": 1.0109786987304688,
"learning_rate": 0.001,
"loss": 2.275,
"step": 46700
},
{
"epoch": 15.126050420168067,
"grad_norm": 1.6071540117263794,
"learning_rate": 0.001,
"loss": 2.2764,
"step": 46800
},
{
"epoch": 15.158371040723981,
"grad_norm": 1.2585033178329468,
"learning_rate": 0.001,
"loss": 2.291,
"step": 46900
},
{
"epoch": 15.190691661279896,
"grad_norm": 1.313089370727539,
"learning_rate": 0.001,
"loss": 2.277,
"step": 47000
},
{
"epoch": 15.22301228183581,
"grad_norm": 0.8985034227371216,
"learning_rate": 0.001,
"loss": 2.2836,
"step": 47100
},
{
"epoch": 15.255332902391725,
"grad_norm": 1.3318949937820435,
"learning_rate": 0.001,
"loss": 2.2753,
"step": 47200
},
{
"epoch": 15.287653522947641,
"grad_norm": 1.6827569007873535,
"learning_rate": 0.001,
"loss": 2.2958,
"step": 47300
},
{
"epoch": 15.319974143503556,
"grad_norm": 1.5016133785247803,
"learning_rate": 0.001,
"loss": 2.3056,
"step": 47400
},
{
"epoch": 15.35229476405947,
"grad_norm": 1.3888620138168335,
"learning_rate": 0.001,
"loss": 2.3095,
"step": 47500
},
{
"epoch": 15.384615384615385,
"grad_norm": 0.9856404662132263,
"learning_rate": 0.001,
"loss": 2.294,
"step": 47600
},
{
"epoch": 15.4169360051713,
"grad_norm": 1.1952743530273438,
"learning_rate": 0.001,
"loss": 2.3156,
"step": 47700
},
{
"epoch": 15.449256625727214,
"grad_norm": 1.3369444608688354,
"learning_rate": 0.001,
"loss": 2.3049,
"step": 47800
},
{
"epoch": 15.481577246283129,
"grad_norm": 1.258984923362732,
"learning_rate": 0.001,
"loss": 2.3135,
"step": 47900
},
{
"epoch": 15.513897866839043,
"grad_norm": 1.1995731592178345,
"learning_rate": 0.001,
"loss": 2.3567,
"step": 48000
},
{
"epoch": 15.546218487394958,
"grad_norm": 1.1886205673217773,
"learning_rate": 0.001,
"loss": 2.3331,
"step": 48100
},
{
"epoch": 15.578539107950872,
"grad_norm": 1.1722344160079956,
"learning_rate": 0.001,
"loss": 2.3292,
"step": 48200
},
{
"epoch": 15.610859728506787,
"grad_norm": 1.4253953695297241,
"learning_rate": 0.001,
"loss": 2.3449,
"step": 48300
},
{
"epoch": 15.643180349062701,
"grad_norm": 1.1338748931884766,
"learning_rate": 0.001,
"loss": 2.3234,
"step": 48400
},
{
"epoch": 15.675500969618616,
"grad_norm": 1.2208466529846191,
"learning_rate": 0.001,
"loss": 2.3402,
"step": 48500
},
{
"epoch": 15.70782159017453,
"grad_norm": 1.2015690803527832,
"learning_rate": 0.001,
"loss": 2.3361,
"step": 48600
},
{
"epoch": 15.740142210730447,
"grad_norm": 1.0186476707458496,
"learning_rate": 0.001,
"loss": 2.3455,
"step": 48700
},
{
"epoch": 15.772462831286362,
"grad_norm": 1.4595075845718384,
"learning_rate": 0.001,
"loss": 2.3425,
"step": 48800
},
{
"epoch": 15.804783451842276,
"grad_norm": 1.0360519886016846,
"learning_rate": 0.001,
"loss": 2.3361,
"step": 48900
},
{
"epoch": 15.83710407239819,
"grad_norm": 0.9627425074577332,
"learning_rate": 0.001,
"loss": 2.3359,
"step": 49000
},
{
"epoch": 15.869424692954105,
"grad_norm": 0.9317019581794739,
"learning_rate": 0.001,
"loss": 2.3536,
"step": 49100
},
{
"epoch": 15.90174531351002,
"grad_norm": 1.2822422981262207,
"learning_rate": 0.001,
"loss": 2.3559,
"step": 49200
},
{
"epoch": 15.934065934065934,
"grad_norm": 1.5487724542617798,
"learning_rate": 0.001,
"loss": 2.3575,
"step": 49300
},
{
"epoch": 15.966386554621849,
"grad_norm": 1.4528721570968628,
"learning_rate": 0.001,
"loss": 2.351,
"step": 49400
},
{
"epoch": 15.998707175177763,
"grad_norm": 1.8106807470321655,
"learning_rate": 0.001,
"loss": 2.3607,
"step": 49500
},
{
"epoch": 16.031027795733678,
"grad_norm": 1.4708492755889893,
"learning_rate": 0.001,
"loss": 2.2137,
"step": 49600
},
{
"epoch": 16.063348416289593,
"grad_norm": 1.2355926036834717,
"learning_rate": 0.001,
"loss": 2.2269,
"step": 49700
},
{
"epoch": 16.095669036845507,
"grad_norm": 2.1963531970977783,
"learning_rate": 0.001,
"loss": 2.2293,
"step": 49800
},
{
"epoch": 16.12798965740142,
"grad_norm": 1.46670663356781,
"learning_rate": 0.001,
"loss": 2.2242,
"step": 49900
},
{
"epoch": 16.160310277957336,
"grad_norm": 1.289772629737854,
"learning_rate": 0.001,
"loss": 2.2503,
"step": 50000
},
{
"epoch": 16.19263089851325,
"grad_norm": 1.4664719104766846,
"learning_rate": 0.001,
"loss": 2.2517,
"step": 50100
},
{
"epoch": 16.224951519069165,
"grad_norm": 1.0733598470687866,
"learning_rate": 0.001,
"loss": 2.2682,
"step": 50200
},
{
"epoch": 16.25727213962508,
"grad_norm": 1.5189355611801147,
"learning_rate": 0.001,
"loss": 2.2561,
"step": 50300
},
{
"epoch": 16.289592760180994,
"grad_norm": 1.3452093601226807,
"learning_rate": 0.001,
"loss": 2.2513,
"step": 50400
},
{
"epoch": 16.32191338073691,
"grad_norm": 1.3040879964828491,
"learning_rate": 0.001,
"loss": 2.249,
"step": 50500
},
{
"epoch": 16.354234001292824,
"grad_norm": 1.6627320051193237,
"learning_rate": 0.001,
"loss": 2.2694,
"step": 50600
},
{
"epoch": 16.386554621848738,
"grad_norm": 1.5199756622314453,
"learning_rate": 0.001,
"loss": 2.2729,
"step": 50700
},
{
"epoch": 16.418875242404653,
"grad_norm": 1.412865161895752,
"learning_rate": 0.001,
"loss": 2.2636,
"step": 50800
},
{
"epoch": 16.451195862960567,
"grad_norm": 1.1884099245071411,
"learning_rate": 0.001,
"loss": 2.281,
"step": 50900
},
{
"epoch": 16.483516483516482,
"grad_norm": 1.5446977615356445,
"learning_rate": 0.001,
"loss": 2.2719,
"step": 51000
},
{
"epoch": 16.5158371040724,
"grad_norm": 1.6802927255630493,
"learning_rate": 0.001,
"loss": 2.2849,
"step": 51100
},
{
"epoch": 16.548157724628314,
"grad_norm": 1.2737250328063965,
"learning_rate": 0.001,
"loss": 2.2851,
"step": 51200
},
{
"epoch": 16.58047834518423,
"grad_norm": 1.3500823974609375,
"learning_rate": 0.001,
"loss": 2.3008,
"step": 51300
},
{
"epoch": 16.612798965740144,
"grad_norm": 1.1918673515319824,
"learning_rate": 0.001,
"loss": 2.2985,
"step": 51400
},
{
"epoch": 16.645119586296058,
"grad_norm": 1.1218078136444092,
"learning_rate": 0.001,
"loss": 2.2982,
"step": 51500
},
{
"epoch": 16.677440206851973,
"grad_norm": 1.5831869840621948,
"learning_rate": 0.001,
"loss": 2.3034,
"step": 51600
},
{
"epoch": 16.709760827407887,
"grad_norm": 1.3978112936019897,
"learning_rate": 0.001,
"loss": 2.3151,
"step": 51700
},
{
"epoch": 16.742081447963802,
"grad_norm": 1.4054855108261108,
"learning_rate": 0.001,
"loss": 2.286,
"step": 51800
},
{
"epoch": 16.774402068519716,
"grad_norm": 0.9958290457725525,
"learning_rate": 0.001,
"loss": 2.3094,
"step": 51900
},
{
"epoch": 16.80672268907563,
"grad_norm": 1.3509937524795532,
"learning_rate": 0.001,
"loss": 2.2887,
"step": 52000
},
{
"epoch": 16.839043309631545,
"grad_norm": 2.143278121948242,
"learning_rate": 0.001,
"loss": 2.3244,
"step": 52100
},
{
"epoch": 16.87136393018746,
"grad_norm": 1.5843268632888794,
"learning_rate": 0.001,
"loss": 2.3059,
"step": 52200
},
{
"epoch": 16.903684550743375,
"grad_norm": 1.4028642177581787,
"learning_rate": 0.001,
"loss": 2.3053,
"step": 52300
},
{
"epoch": 16.93600517129929,
"grad_norm": 1.5138201713562012,
"learning_rate": 0.001,
"loss": 2.3194,
"step": 52400
},
{
"epoch": 16.968325791855204,
"grad_norm": 1.6168174743652344,
"learning_rate": 0.001,
"loss": 2.3039,
"step": 52500
},
{
"epoch": 17.00064641241112,
"grad_norm": 1.1482666730880737,
"learning_rate": 0.001,
"loss": 2.302,
"step": 52600
},
{
"epoch": 17.032967032967033,
"grad_norm": 1.4270600080490112,
"learning_rate": 0.001,
"loss": 2.1655,
"step": 52700
},
{
"epoch": 17.065287653522947,
"grad_norm": 1.8423911333084106,
"learning_rate": 0.001,
"loss": 2.1833,
"step": 52800
},
{
"epoch": 17.097608274078862,
"grad_norm": 1.6184693574905396,
"learning_rate": 0.001,
"loss": 2.1812,
"step": 52900
},
{
"epoch": 17.129928894634777,
"grad_norm": 1.0999650955200195,
"learning_rate": 0.001,
"loss": 2.194,
"step": 53000
},
{
"epoch": 17.16224951519069,
"grad_norm": 1.5708781480789185,
"learning_rate": 0.001,
"loss": 2.1813,
"step": 53100
},
{
"epoch": 17.194570135746606,
"grad_norm": 1.5949249267578125,
"learning_rate": 0.001,
"loss": 2.1901,
"step": 53200
},
{
"epoch": 17.22689075630252,
"grad_norm": 1.490331768989563,
"learning_rate": 0.001,
"loss": 2.2041,
"step": 53300
},
{
"epoch": 17.259211376858435,
"grad_norm": 1.5599461793899536,
"learning_rate": 0.001,
"loss": 2.2346,
"step": 53400
},
{
"epoch": 17.29153199741435,
"grad_norm": 1.9876192808151245,
"learning_rate": 0.001,
"loss": 2.2065,
"step": 53500
},
{
"epoch": 17.323852617970264,
"grad_norm": 1.8799535036087036,
"learning_rate": 0.001,
"loss": 2.241,
"step": 53600
},
{
"epoch": 17.35617323852618,
"grad_norm": 2.051346778869629,
"learning_rate": 0.001,
"loss": 2.2141,
"step": 53700
},
{
"epoch": 17.388493859082093,
"grad_norm": 2.10467529296875,
"learning_rate": 0.001,
"loss": 2.2266,
"step": 53800
},
{
"epoch": 17.420814479638008,
"grad_norm": 1.1607954502105713,
"learning_rate": 0.001,
"loss": 2.2431,
"step": 53900
},
{
"epoch": 17.453135100193922,
"grad_norm": 1.8151345252990723,
"learning_rate": 0.001,
"loss": 2.2645,
"step": 54000
},
{
"epoch": 17.485455720749837,
"grad_norm": 1.368027925491333,
"learning_rate": 0.001,
"loss": 2.2507,
"step": 54100
},
{
"epoch": 17.517776341305755,
"grad_norm": 1.8371237516403198,
"learning_rate": 0.001,
"loss": 2.2506,
"step": 54200
},
{
"epoch": 17.55009696186167,
"grad_norm": 1.2181835174560547,
"learning_rate": 0.001,
"loss": 2.2249,
"step": 54300
},
{
"epoch": 17.582417582417584,
"grad_norm": 1.3113195896148682,
"learning_rate": 0.001,
"loss": 2.233,
"step": 54400
},
{
"epoch": 17.6147382029735,
"grad_norm": 1.2849704027175903,
"learning_rate": 0.001,
"loss": 2.2755,
"step": 54500
},
{
"epoch": 17.647058823529413,
"grad_norm": 1.2958779335021973,
"learning_rate": 0.001,
"loss": 2.2668,
"step": 54600
},
{
"epoch": 17.679379444085328,
"grad_norm": 1.5940552949905396,
"learning_rate": 0.001,
"loss": 2.2648,
"step": 54700
},
{
"epoch": 17.711700064641242,
"grad_norm": 1.6866202354431152,
"learning_rate": 0.001,
"loss": 2.2646,
"step": 54800
},
{
"epoch": 17.744020685197157,
"grad_norm": 1.7330199480056763,
"learning_rate": 0.001,
"loss": 2.2786,
"step": 54900
},
{
"epoch": 17.77634130575307,
"grad_norm": 1.4443870782852173,
"learning_rate": 0.001,
"loss": 2.2595,
"step": 55000
},
{
"epoch": 17.808661926308986,
"grad_norm": 1.4749608039855957,
"learning_rate": 0.001,
"loss": 2.2711,
"step": 55100
},
{
"epoch": 17.8409825468649,
"grad_norm": 1.5407294034957886,
"learning_rate": 0.001,
"loss": 2.263,
"step": 55200
},
{
"epoch": 17.873303167420815,
"grad_norm": 1.521584153175354,
"learning_rate": 0.001,
"loss": 2.2822,
"step": 55300
},
{
"epoch": 17.90562378797673,
"grad_norm": 1.7927467823028564,
"learning_rate": 0.001,
"loss": 2.273,
"step": 55400
},
{
"epoch": 17.937944408532644,
"grad_norm": 1.3927431106567383,
"learning_rate": 0.001,
"loss": 2.2797,
"step": 55500
},
{
"epoch": 17.97026502908856,
"grad_norm": 2.010420799255371,
"learning_rate": 0.001,
"loss": 2.2923,
"step": 55600
},
{
"epoch": 18.002585649644473,
"grad_norm": 1.3878021240234375,
"learning_rate": 0.001,
"loss": 2.2861,
"step": 55700
},
{
"epoch": 18.034906270200388,
"grad_norm": 1.6571372747421265,
"learning_rate": 0.001,
"loss": 2.1246,
"step": 55800
},
{
"epoch": 18.067226890756302,
"grad_norm": 1.3559300899505615,
"learning_rate": 0.001,
"loss": 2.1514,
"step": 55900
},
{
"epoch": 18.099547511312217,
"grad_norm": 1.1670371294021606,
"learning_rate": 0.001,
"loss": 2.1436,
"step": 56000
},
{
"epoch": 18.13186813186813,
"grad_norm": 1.331639051437378,
"learning_rate": 0.001,
"loss": 2.1682,
"step": 56100
},
{
"epoch": 18.164188752424046,
"grad_norm": 1.2827179431915283,
"learning_rate": 0.001,
"loss": 2.1676,
"step": 56200
},
{
"epoch": 18.19650937297996,
"grad_norm": 1.556694746017456,
"learning_rate": 0.001,
"loss": 2.1648,
"step": 56300
},
{
"epoch": 18.228829993535875,
"grad_norm": 1.428536057472229,
"learning_rate": 0.001,
"loss": 2.1962,
"step": 56400
},
{
"epoch": 18.26115061409179,
"grad_norm": 1.1887167692184448,
"learning_rate": 0.001,
"loss": 2.1626,
"step": 56500
},
{
"epoch": 18.293471234647704,
"grad_norm": 1.2648141384124756,
"learning_rate": 0.001,
"loss": 2.1923,
"step": 56600
},
{
"epoch": 18.32579185520362,
"grad_norm": 1.1101640462875366,
"learning_rate": 0.001,
"loss": 2.217,
"step": 56700
},
{
"epoch": 18.358112475759533,
"grad_norm": 0.9787229895591736,
"learning_rate": 0.001,
"loss": 2.1767,
"step": 56800
},
{
"epoch": 18.390433096315448,
"grad_norm": 1.4049594402313232,
"learning_rate": 0.001,
"loss": 2.1986,
"step": 56900
},
{
"epoch": 18.422753716871362,
"grad_norm": 1.415755033493042,
"learning_rate": 0.001,
"loss": 2.2025,
"step": 57000
},
{
"epoch": 18.455074337427277,
"grad_norm": 1.1880539655685425,
"learning_rate": 0.001,
"loss": 2.2029,
"step": 57100
},
{
"epoch": 18.48739495798319,
"grad_norm": 1.154875636100769,
"learning_rate": 0.001,
"loss": 2.2117,
"step": 57200
},
{
"epoch": 18.51971557853911,
"grad_norm": 0.9201914668083191,
"learning_rate": 0.001,
"loss": 2.2086,
"step": 57300
},
{
"epoch": 18.552036199095024,
"grad_norm": 1.2845484018325806,
"learning_rate": 0.001,
"loss": 2.2137,
"step": 57400
},
{
"epoch": 18.58435681965094,
"grad_norm": 1.026379108428955,
"learning_rate": 0.001,
"loss": 2.2051,
"step": 57500
},
{
"epoch": 18.616677440206853,
"grad_norm": 0.9726206064224243,
"learning_rate": 0.001,
"loss": 2.2064,
"step": 57600
},
{
"epoch": 18.648998060762768,
"grad_norm": 1.5916303396224976,
"learning_rate": 0.001,
"loss": 2.216,
"step": 57700
},
{
"epoch": 18.681318681318682,
"grad_norm": 1.2654246091842651,
"learning_rate": 0.001,
"loss": 2.2234,
"step": 57800
},
{
"epoch": 18.713639301874597,
"grad_norm": 1.6823699474334717,
"learning_rate": 0.001,
"loss": 2.2388,
"step": 57900
},
{
"epoch": 18.74595992243051,
"grad_norm": 1.0970423221588135,
"learning_rate": 0.001,
"loss": 2.2332,
"step": 58000
},
{
"epoch": 18.778280542986426,
"grad_norm": 1.136433720588684,
"learning_rate": 0.001,
"loss": 2.2384,
"step": 58100
},
{
"epoch": 18.81060116354234,
"grad_norm": 1.2828449010849,
"learning_rate": 0.001,
"loss": 2.2142,
"step": 58200
},
{
"epoch": 18.842921784098255,
"grad_norm": 1.010236144065857,
"learning_rate": 0.001,
"loss": 2.2349,
"step": 58300
},
{
"epoch": 18.87524240465417,
"grad_norm": 1.3149129152297974,
"learning_rate": 0.001,
"loss": 2.2471,
"step": 58400
},
{
"epoch": 18.907563025210084,
"grad_norm": 1.5493013858795166,
"learning_rate": 0.001,
"loss": 2.2713,
"step": 58500
},
{
"epoch": 18.939883645766,
"grad_norm": 1.3114402294158936,
"learning_rate": 0.001,
"loss": 2.2392,
"step": 58600
},
{
"epoch": 18.972204266321913,
"grad_norm": 0.9853785037994385,
"learning_rate": 0.001,
"loss": 2.2405,
"step": 58700
},
{
"epoch": 19.004524886877828,
"grad_norm": 1.2043988704681396,
"learning_rate": 0.001,
"loss": 2.2477,
"step": 58800
},
{
"epoch": 19.036845507433743,
"grad_norm": 1.6636931896209717,
"learning_rate": 0.001,
"loss": 2.0882,
"step": 58900
},
{
"epoch": 19.069166127989657,
"grad_norm": 0.8756181597709656,
"learning_rate": 0.001,
"loss": 2.0999,
"step": 59000
},
{
"epoch": 19.10148674854557,
"grad_norm": 1.1063123941421509,
"learning_rate": 0.001,
"loss": 2.1199,
"step": 59100
},
{
"epoch": 19.133807369101486,
"grad_norm": 1.2338377237319946,
"learning_rate": 0.001,
"loss": 2.1267,
"step": 59200
},
{
"epoch": 19.1661279896574,
"grad_norm": 1.44084632396698,
"learning_rate": 0.001,
"loss": 2.1371,
"step": 59300
},
{
"epoch": 19.198448610213315,
"grad_norm": 1.1796813011169434,
"learning_rate": 0.001,
"loss": 2.1499,
"step": 59400
},
{
"epoch": 19.23076923076923,
"grad_norm": 1.1995784044265747,
"learning_rate": 0.001,
"loss": 2.1438,
"step": 59500
},
{
"epoch": 19.263089851325145,
"grad_norm": 0.9502636790275574,
"learning_rate": 0.001,
"loss": 2.1585,
"step": 59600
},
{
"epoch": 19.29541047188106,
"grad_norm": 0.9679586291313171,
"learning_rate": 0.001,
"loss": 2.1544,
"step": 59700
},
{
"epoch": 19.327731092436974,
"grad_norm": 1.2242112159729004,
"learning_rate": 0.001,
"loss": 2.1633,
"step": 59800
},
{
"epoch": 19.360051712992888,
"grad_norm": 1.2602921724319458,
"learning_rate": 0.001,
"loss": 2.157,
"step": 59900
},
{
"epoch": 19.392372333548803,
"grad_norm": 1.55171799659729,
"learning_rate": 0.001,
"loss": 2.1749,
"step": 60000
},
{
"epoch": 19.424692954104717,
"grad_norm": 1.2272818088531494,
"learning_rate": 0.001,
"loss": 2.1719,
"step": 60100
},
{
"epoch": 19.457013574660632,
"grad_norm": 1.1695352792739868,
"learning_rate": 0.001,
"loss": 2.1562,
"step": 60200
},
{
"epoch": 19.489334195216546,
"grad_norm": 1.1918383836746216,
"learning_rate": 0.001,
"loss": 2.1781,
"step": 60300
},
{
"epoch": 19.521654815772465,
"grad_norm": 1.1158689260482788,
"learning_rate": 0.001,
"loss": 2.175,
"step": 60400
},
{
"epoch": 19.55397543632838,
"grad_norm": 1.2081654071807861,
"learning_rate": 0.001,
"loss": 2.1811,
"step": 60500
},
{
"epoch": 19.586296056884294,
"grad_norm": 1.0976276397705078,
"learning_rate": 0.001,
"loss": 2.18,
"step": 60600
},
{
"epoch": 19.618616677440208,
"grad_norm": 1.0541235208511353,
"learning_rate": 0.001,
"loss": 2.1821,
"step": 60700
},
{
"epoch": 19.650937297996123,
"grad_norm": 1.1526302099227905,
"learning_rate": 0.001,
"loss": 2.1947,
"step": 60800
},
{
"epoch": 19.683257918552037,
"grad_norm": 1.3280059099197388,
"learning_rate": 0.001,
"loss": 2.1674,
"step": 60900
},
{
"epoch": 19.715578539107952,
"grad_norm": 1.2853764295578003,
"learning_rate": 0.001,
"loss": 2.1935,
"step": 61000
},
{
"epoch": 19.747899159663866,
"grad_norm": 1.0324232578277588,
"learning_rate": 0.001,
"loss": 2.2013,
"step": 61100
},
{
"epoch": 19.78021978021978,
"grad_norm": 1.3006898164749146,
"learning_rate": 0.001,
"loss": 2.2024,
"step": 61200
},
{
"epoch": 19.812540400775696,
"grad_norm": 1.1228374242782593,
"learning_rate": 0.001,
"loss": 2.193,
"step": 61300
},
{
"epoch": 19.84486102133161,
"grad_norm": 1.2684869766235352,
"learning_rate": 0.001,
"loss": 2.2083,
"step": 61400
},
{
"epoch": 19.877181641887525,
"grad_norm": 0.8672001361846924,
"learning_rate": 0.001,
"loss": 2.2101,
"step": 61500
},
{
"epoch": 19.90950226244344,
"grad_norm": 1.0683246850967407,
"learning_rate": 0.001,
"loss": 2.2139,
"step": 61600
},
{
"epoch": 19.941822882999354,
"grad_norm": 1.1728013753890991,
"learning_rate": 0.001,
"loss": 2.1988,
"step": 61700
},
{
"epoch": 19.97414350355527,
"grad_norm": 1.2360259294509888,
"learning_rate": 0.001,
"loss": 2.206,
"step": 61800
},
{
"epoch": 20.006464124111183,
"grad_norm": 1.1226767301559448,
"learning_rate": 0.001,
"loss": 2.1834,
"step": 61900
},
{
"epoch": 20.038784744667097,
"grad_norm": 1.2612074613571167,
"learning_rate": 0.001,
"loss": 2.0784,
"step": 62000
},
{
"epoch": 20.071105365223012,
"grad_norm": 1.2865725755691528,
"learning_rate": 0.001,
"loss": 2.0913,
"step": 62100
},
{
"epoch": 20.103425985778927,
"grad_norm": 1.2894771099090576,
"learning_rate": 0.001,
"loss": 2.0799,
"step": 62200
},
{
"epoch": 20.13574660633484,
"grad_norm": 1.3527390956878662,
"learning_rate": 0.001,
"loss": 2.0959,
"step": 62300
},
{
"epoch": 20.168067226890756,
"grad_norm": 1.1928151845932007,
"learning_rate": 0.001,
"loss": 2.103,
"step": 62400
},
{
"epoch": 20.20038784744667,
"grad_norm": 1.0168992280960083,
"learning_rate": 0.001,
"loss": 2.0941,
"step": 62500
},
{
"epoch": 20.232708468002585,
"grad_norm": 1.0409342050552368,
"learning_rate": 0.001,
"loss": 2.1038,
"step": 62600
},
{
"epoch": 20.2650290885585,
"grad_norm": 1.2547892332077026,
"learning_rate": 0.001,
"loss": 2.0966,
"step": 62700
},
{
"epoch": 20.297349709114414,
"grad_norm": 1.2054287195205688,
"learning_rate": 0.001,
"loss": 2.1298,
"step": 62800
},
{
"epoch": 20.32967032967033,
"grad_norm": 1.0080070495605469,
"learning_rate": 0.001,
"loss": 2.1166,
"step": 62900
},
{
"epoch": 20.361990950226243,
"grad_norm": 1.103588342666626,
"learning_rate": 0.001,
"loss": 2.1441,
"step": 63000
},
{
"epoch": 20.394311570782158,
"grad_norm": 1.1245242357254028,
"learning_rate": 0.001,
"loss": 2.1216,
"step": 63100
},
{
"epoch": 20.426632191338072,
"grad_norm": 1.236649513244629,
"learning_rate": 0.001,
"loss": 2.1274,
"step": 63200
},
{
"epoch": 20.458952811893987,
"grad_norm": 1.1738601922988892,
"learning_rate": 0.001,
"loss": 2.1229,
"step": 63300
},
{
"epoch": 20.4912734324499,
"grad_norm": 1.0897395610809326,
"learning_rate": 0.001,
"loss": 2.1431,
"step": 63400
},
{
"epoch": 20.52359405300582,
"grad_norm": 1.4749375581741333,
"learning_rate": 0.001,
"loss": 2.1278,
"step": 63500
},
{
"epoch": 20.555914673561734,
"grad_norm": 1.4643535614013672,
"learning_rate": 0.001,
"loss": 2.1512,
"step": 63600
},
{
"epoch": 20.58823529411765,
"grad_norm": 0.9917904734611511,
"learning_rate": 0.001,
"loss": 2.1503,
"step": 63700
},
{
"epoch": 20.620555914673563,
"grad_norm": 1.2945317029953003,
"learning_rate": 0.001,
"loss": 2.1454,
"step": 63800
},
{
"epoch": 20.652876535229478,
"grad_norm": 1.1126689910888672,
"learning_rate": 0.001,
"loss": 2.1353,
"step": 63900
},
{
"epoch": 20.685197155785392,
"grad_norm": 1.2969932556152344,
"learning_rate": 0.001,
"loss": 2.1496,
"step": 64000
},
{
"epoch": 20.717517776341307,
"grad_norm": 1.3333299160003662,
"learning_rate": 0.001,
"loss": 2.1623,
"step": 64100
},
{
"epoch": 20.74983839689722,
"grad_norm": 1.0493760108947754,
"learning_rate": 0.001,
"loss": 2.1668,
"step": 64200
},
{
"epoch": 20.782159017453136,
"grad_norm": 1.0862622261047363,
"learning_rate": 0.001,
"loss": 2.1579,
"step": 64300
},
{
"epoch": 20.81447963800905,
"grad_norm": 1.3162381649017334,
"learning_rate": 0.001,
"loss": 2.1769,
"step": 64400
},
{
"epoch": 20.846800258564965,
"grad_norm": 1.7097973823547363,
"learning_rate": 0.001,
"loss": 2.1642,
"step": 64500
},
{
"epoch": 20.87912087912088,
"grad_norm": 1.0230882167816162,
"learning_rate": 0.001,
"loss": 2.1939,
"step": 64600
},
{
"epoch": 20.911441499676794,
"grad_norm": 0.8909283876419067,
"learning_rate": 0.001,
"loss": 2.1795,
"step": 64700
},
{
"epoch": 20.94376212023271,
"grad_norm": 1.1475738286972046,
"learning_rate": 0.001,
"loss": 2.1833,
"step": 64800
},
{
"epoch": 20.976082740788623,
"grad_norm": 1.3730794191360474,
"learning_rate": 0.001,
"loss": 2.1868,
"step": 64900
},
{
"epoch": 21.008403361344538,
"grad_norm": 1.518537163734436,
"learning_rate": 0.001,
"loss": 2.1312,
"step": 65000
},
{
"epoch": 21.040723981900452,
"grad_norm": 1.2940282821655273,
"learning_rate": 0.001,
"loss": 2.0509,
"step": 65100
},
{
"epoch": 21.073044602456367,
"grad_norm": 1.860131859779358,
"learning_rate": 0.001,
"loss": 2.0563,
"step": 65200
},
{
"epoch": 21.10536522301228,
"grad_norm": 1.4131841659545898,
"learning_rate": 0.001,
"loss": 2.0643,
"step": 65300
},
{
"epoch": 21.137685843568196,
"grad_norm": 1.429236888885498,
"learning_rate": 0.001,
"loss": 2.0731,
"step": 65400
},
{
"epoch": 21.17000646412411,
"grad_norm": 1.1104481220245361,
"learning_rate": 0.001,
"loss": 2.062,
"step": 65500
},
{
"epoch": 21.202327084680025,
"grad_norm": 1.18645179271698,
"learning_rate": 0.001,
"loss": 2.0773,
"step": 65600
},
{
"epoch": 21.23464770523594,
"grad_norm": 1.384711742401123,
"learning_rate": 0.001,
"loss": 2.0734,
"step": 65700
},
{
"epoch": 21.266968325791854,
"grad_norm": 1.1969033479690552,
"learning_rate": 0.001,
"loss": 2.0751,
"step": 65800
},
{
"epoch": 21.29928894634777,
"grad_norm": 1.297878384590149,
"learning_rate": 0.001,
"loss": 2.0942,
"step": 65900
},
{
"epoch": 21.331609566903683,
"grad_norm": 1.4161620140075684,
"learning_rate": 0.001,
"loss": 2.1058,
"step": 66000
},
{
"epoch": 21.363930187459598,
"grad_norm": 1.1466299295425415,
"learning_rate": 0.001,
"loss": 2.0875,
"step": 66100
},
{
"epoch": 21.396250808015512,
"grad_norm": 1.0590342283248901,
"learning_rate": 0.001,
"loss": 2.0883,
"step": 66200
},
{
"epoch": 21.428571428571427,
"grad_norm": 1.4948805570602417,
"learning_rate": 0.001,
"loss": 2.1005,
"step": 66300
},
{
"epoch": 21.46089204912734,
"grad_norm": 1.6177111864089966,
"learning_rate": 0.001,
"loss": 2.1093,
"step": 66400
},
{
"epoch": 21.49321266968326,
"grad_norm": 1.3227946758270264,
"learning_rate": 0.001,
"loss": 2.0811,
"step": 66500
},
{
"epoch": 21.525533290239174,
"grad_norm": 1.3259391784667969,
"learning_rate": 0.001,
"loss": 2.0773,
"step": 66600
},
{
"epoch": 21.55785391079509,
"grad_norm": 1.352561354637146,
"learning_rate": 0.001,
"loss": 2.111,
"step": 66700
},
{
"epoch": 21.590174531351003,
"grad_norm": 1.1337064504623413,
"learning_rate": 0.001,
"loss": 2.1252,
"step": 66800
},
{
"epoch": 21.622495151906918,
"grad_norm": 1.2298682928085327,
"learning_rate": 0.001,
"loss": 2.1198,
"step": 66900
},
{
"epoch": 21.654815772462833,
"grad_norm": 1.2822353839874268,
"learning_rate": 0.001,
"loss": 2.1258,
"step": 67000
},
{
"epoch": 21.687136393018747,
"grad_norm": 0.9597915410995483,
"learning_rate": 0.001,
"loss": 2.1283,
"step": 67100
},
{
"epoch": 21.71945701357466,
"grad_norm": 0.9812031388282776,
"learning_rate": 0.001,
"loss": 2.1439,
"step": 67200
},
{
"epoch": 21.751777634130576,
"grad_norm": 1.0487492084503174,
"learning_rate": 0.001,
"loss": 2.1357,
"step": 67300
},
{
"epoch": 21.78409825468649,
"grad_norm": 1.1485785245895386,
"learning_rate": 0.001,
"loss": 2.1601,
"step": 67400
},
{
"epoch": 21.816418875242405,
"grad_norm": 0.9800519347190857,
"learning_rate": 0.001,
"loss": 2.1565,
"step": 67500
},
{
"epoch": 21.84873949579832,
"grad_norm": 1.2664602994918823,
"learning_rate": 0.001,
"loss": 2.1442,
"step": 67600
},
{
"epoch": 21.881060116354234,
"grad_norm": 1.3896493911743164,
"learning_rate": 0.001,
"loss": 2.1419,
"step": 67700
},
{
"epoch": 21.91338073691015,
"grad_norm": 1.5186612606048584,
"learning_rate": 0.001,
"loss": 2.1331,
"step": 67800
},
{
"epoch": 21.945701357466064,
"grad_norm": 1.1328755617141724,
"learning_rate": 0.001,
"loss": 2.1423,
"step": 67900
},
{
"epoch": 21.978021978021978,
"grad_norm": 1.7095478773117065,
"learning_rate": 0.001,
"loss": 2.144,
"step": 68000
},
{
"epoch": 22.010342598577893,
"grad_norm": 1.1905454397201538,
"learning_rate": 0.001,
"loss": 2.0849,
"step": 68100
},
{
"epoch": 22.042663219133807,
"grad_norm": 1.151082992553711,
"learning_rate": 0.001,
"loss": 2.0151,
"step": 68200
},
{
"epoch": 22.07498383968972,
"grad_norm": 1.414963960647583,
"learning_rate": 0.001,
"loss": 2.0046,
"step": 68300
},
{
"epoch": 22.107304460245636,
"grad_norm": 1.0465073585510254,
"learning_rate": 0.001,
"loss": 2.0297,
"step": 68400
},
{
"epoch": 22.13962508080155,
"grad_norm": 1.1696205139160156,
"learning_rate": 0.001,
"loss": 2.0333,
"step": 68500
},
{
"epoch": 22.171945701357465,
"grad_norm": 1.165142297744751,
"learning_rate": 0.001,
"loss": 2.0432,
"step": 68600
},
{
"epoch": 22.20426632191338,
"grad_norm": 1.5192077159881592,
"learning_rate": 0.001,
"loss": 2.0496,
"step": 68700
},
{
"epoch": 22.236586942469295,
"grad_norm": 1.7254102230072021,
"learning_rate": 0.001,
"loss": 2.04,
"step": 68800
},
{
"epoch": 22.26890756302521,
"grad_norm": 1.0662051439285278,
"learning_rate": 0.001,
"loss": 2.0316,
"step": 68900
},
{
"epoch": 22.301228183581124,
"grad_norm": 1.3857476711273193,
"learning_rate": 0.001,
"loss": 2.0576,
"step": 69000
},
{
"epoch": 22.33354880413704,
"grad_norm": 1.332242727279663,
"learning_rate": 0.001,
"loss": 2.0349,
"step": 69100
},
{
"epoch": 22.365869424692953,
"grad_norm": 1.3924357891082764,
"learning_rate": 0.001,
"loss": 2.0848,
"step": 69200
},
{
"epoch": 22.398190045248867,
"grad_norm": 1.1637581586837769,
"learning_rate": 0.001,
"loss": 2.0862,
"step": 69300
},
{
"epoch": 22.430510665804782,
"grad_norm": 1.474599003791809,
"learning_rate": 0.001,
"loss": 2.0771,
"step": 69400
},
{
"epoch": 22.462831286360696,
"grad_norm": 1.2520018815994263,
"learning_rate": 0.001,
"loss": 2.0771,
"step": 69500
},
{
"epoch": 22.49515190691661,
"grad_norm": 1.3370542526245117,
"learning_rate": 0.001,
"loss": 2.0893,
"step": 69600
},
{
"epoch": 22.52747252747253,
"grad_norm": 0.9184688925743103,
"learning_rate": 0.001,
"loss": 2.0766,
"step": 69700
},
{
"epoch": 22.559793148028444,
"grad_norm": 1.0229378938674927,
"learning_rate": 0.001,
"loss": 2.0814,
"step": 69800
},
{
"epoch": 22.59211376858436,
"grad_norm": 1.6382076740264893,
"learning_rate": 0.001,
"loss": 2.0853,
"step": 69900
},
{
"epoch": 22.624434389140273,
"grad_norm": 1.4604605436325073,
"learning_rate": 0.001,
"loss": 2.094,
"step": 70000
},
{
"epoch": 22.656755009696187,
"grad_norm": 0.9782185554504395,
"learning_rate": 0.001,
"loss": 2.0849,
"step": 70100
},
{
"epoch": 22.689075630252102,
"grad_norm": 1.2251577377319336,
"learning_rate": 0.001,
"loss": 2.1201,
"step": 70200
},
{
"epoch": 22.721396250808017,
"grad_norm": 0.9563422799110413,
"learning_rate": 0.001,
"loss": 2.105,
"step": 70300
},
{
"epoch": 22.75371687136393,
"grad_norm": 1.2224578857421875,
"learning_rate": 0.001,
"loss": 2.1025,
"step": 70400
},
{
"epoch": 22.786037491919846,
"grad_norm": 1.3068078756332397,
"learning_rate": 0.001,
"loss": 2.1108,
"step": 70500
},
{
"epoch": 22.81835811247576,
"grad_norm": 1.2106717824935913,
"learning_rate": 0.001,
"loss": 2.1068,
"step": 70600
},
{
"epoch": 22.850678733031675,
"grad_norm": 1.1505300998687744,
"learning_rate": 0.001,
"loss": 2.0991,
"step": 70700
},
{
"epoch": 22.88299935358759,
"grad_norm": 1.1176759004592896,
"learning_rate": 0.001,
"loss": 2.1206,
"step": 70800
},
{
"epoch": 22.915319974143504,
"grad_norm": 1.4180165529251099,
"learning_rate": 0.001,
"loss": 2.108,
"step": 70900
},
{
"epoch": 22.94764059469942,
"grad_norm": 1.2236053943634033,
"learning_rate": 0.001,
"loss": 2.1049,
"step": 71000
},
{
"epoch": 22.979961215255333,
"grad_norm": 1.3975863456726074,
"learning_rate": 0.001,
"loss": 2.1208,
"step": 71100
},
{
"epoch": 23.012281835811248,
"grad_norm": 1.246389627456665,
"learning_rate": 0.001,
"loss": 2.0397,
"step": 71200
},
{
"epoch": 23.044602456367162,
"grad_norm": 1.5850071907043457,
"learning_rate": 0.001,
"loss": 1.9887,
"step": 71300
},
{
"epoch": 23.076923076923077,
"grad_norm": 1.1955056190490723,
"learning_rate": 0.001,
"loss": 1.9772,
"step": 71400
},
{
"epoch": 23.10924369747899,
"grad_norm": 1.3734443187713623,
"learning_rate": 0.001,
"loss": 1.9873,
"step": 71500
},
{
"epoch": 23.141564318034906,
"grad_norm": 1.61884343624115,
"learning_rate": 0.001,
"loss": 2.0123,
"step": 71600
},
{
"epoch": 23.17388493859082,
"grad_norm": 1.601148009300232,
"learning_rate": 0.001,
"loss": 2.0035,
"step": 71700
},
{
"epoch": 23.206205559146735,
"grad_norm": 1.3028818368911743,
"learning_rate": 0.001,
"loss": 2.0261,
"step": 71800
},
{
"epoch": 23.23852617970265,
"grad_norm": 1.3658092021942139,
"learning_rate": 0.001,
"loss": 2.0145,
"step": 71900
},
{
"epoch": 23.270846800258564,
"grad_norm": 1.2859296798706055,
"learning_rate": 0.001,
"loss": 2.0193,
"step": 72000
},
{
"epoch": 23.30316742081448,
"grad_norm": 1.5372207164764404,
"learning_rate": 0.001,
"loss": 2.0275,
"step": 72100
},
{
"epoch": 23.335488041370393,
"grad_norm": 1.2374789714813232,
"learning_rate": 0.001,
"loss": 2.0197,
"step": 72200
},
{
"epoch": 23.367808661926308,
"grad_norm": 1.213862657546997,
"learning_rate": 0.001,
"loss": 2.0317,
"step": 72300
},
{
"epoch": 23.400129282482222,
"grad_norm": 1.0549333095550537,
"learning_rate": 0.001,
"loss": 2.0409,
"step": 72400
},
{
"epoch": 23.432449903038137,
"grad_norm": 1.0349382162094116,
"learning_rate": 0.001,
"loss": 2.0287,
"step": 72500
},
{
"epoch": 23.46477052359405,
"grad_norm": 1.197283387184143,
"learning_rate": 0.001,
"loss": 2.0392,
"step": 72600
},
{
"epoch": 23.49709114414997,
"grad_norm": 1.5242730379104614,
"learning_rate": 0.001,
"loss": 2.048,
"step": 72700
},
{
"epoch": 23.529411764705884,
"grad_norm": 1.352638602256775,
"learning_rate": 0.001,
"loss": 2.0576,
"step": 72800
},
{
"epoch": 23.5617323852618,
"grad_norm": 0.8858992457389832,
"learning_rate": 0.001,
"loss": 2.0455,
"step": 72900
},
{
"epoch": 23.594053005817713,
"grad_norm": 1.298511028289795,
"learning_rate": 0.001,
"loss": 2.0653,
"step": 73000
},
{
"epoch": 23.626373626373628,
"grad_norm": 1.6291836500167847,
"learning_rate": 0.001,
"loss": 2.0624,
"step": 73100
},
{
"epoch": 23.658694246929542,
"grad_norm": 1.0975252389907837,
"learning_rate": 0.001,
"loss": 2.0855,
"step": 73200
},
{
"epoch": 23.691014867485457,
"grad_norm": 1.221786618232727,
"learning_rate": 0.001,
"loss": 2.0684,
"step": 73300
},
{
"epoch": 23.72333548804137,
"grad_norm": 1.5622029304504395,
"learning_rate": 0.001,
"loss": 2.0895,
"step": 73400
},
{
"epoch": 23.755656108597286,
"grad_norm": 1.9987596273422241,
"learning_rate": 0.001,
"loss": 2.083,
"step": 73500
},
{
"epoch": 23.7879767291532,
"grad_norm": 1.3780443668365479,
"learning_rate": 0.001,
"loss": 2.1051,
"step": 73600
},
{
"epoch": 23.820297349709115,
"grad_norm": 1.5061843395233154,
"learning_rate": 0.001,
"loss": 2.0651,
"step": 73700
},
{
"epoch": 23.85261797026503,
"grad_norm": 1.7464600801467896,
"learning_rate": 0.001,
"loss": 2.0765,
"step": 73800
},
{
"epoch": 23.884938590820944,
"grad_norm": 1.4121919870376587,
"learning_rate": 0.001,
"loss": 2.0719,
"step": 73900
},
{
"epoch": 23.91725921137686,
"grad_norm": 1.759332299232483,
"learning_rate": 0.001,
"loss": 2.0994,
"step": 74000
},
{
"epoch": 23.949579831932773,
"grad_norm": 1.6104328632354736,
"learning_rate": 0.001,
"loss": 2.0959,
"step": 74100
},
{
"epoch": 23.981900452488688,
"grad_norm": 1.543933391571045,
"learning_rate": 0.001,
"loss": 2.0865,
"step": 74200
},
{
"epoch": 24.014221073044602,
"grad_norm": 1.7527161836624146,
"learning_rate": 0.001,
"loss": 2.0309,
"step": 74300
},
{
"epoch": 24.046541693600517,
"grad_norm": 1.540932297706604,
"learning_rate": 0.001,
"loss": 1.9614,
"step": 74400
},
{
"epoch": 24.07886231415643,
"grad_norm": 1.3225823640823364,
"learning_rate": 0.001,
"loss": 1.9619,
"step": 74500
},
{
"epoch": 24.111182934712346,
"grad_norm": 1.4268288612365723,
"learning_rate": 0.001,
"loss": 1.9617,
"step": 74600
},
{
"epoch": 24.14350355526826,
"grad_norm": 1.436618685722351,
"learning_rate": 0.001,
"loss": 1.9802,
"step": 74700
},
{
"epoch": 24.175824175824175,
"grad_norm": 2.0793025493621826,
"learning_rate": 0.001,
"loss": 2.0074,
"step": 74800
},
{
"epoch": 24.20814479638009,
"grad_norm": 1.332802176475525,
"learning_rate": 0.001,
"loss": 1.9643,
"step": 74900
},
{
"epoch": 24.240465416936004,
"grad_norm": 1.3457491397857666,
"learning_rate": 0.001,
"loss": 1.99,
"step": 75000
},
{
"epoch": 24.27278603749192,
"grad_norm": 1.360344409942627,
"learning_rate": 0.001,
"loss": 2.0099,
"step": 75100
},
{
"epoch": 24.305106658047833,
"grad_norm": 1.846103549003601,
"learning_rate": 0.001,
"loss": 2.0058,
"step": 75200
},
{
"epoch": 24.337427278603748,
"grad_norm": 1.547523856163025,
"learning_rate": 0.001,
"loss": 2.0142,
"step": 75300
},
{
"epoch": 24.369747899159663,
"grad_norm": 1.4123824834823608,
"learning_rate": 0.001,
"loss": 1.9988,
"step": 75400
},
{
"epoch": 24.402068519715577,
"grad_norm": 1.568945050239563,
"learning_rate": 0.001,
"loss": 1.9981,
"step": 75500
},
{
"epoch": 24.43438914027149,
"grad_norm": 1.4623258113861084,
"learning_rate": 0.001,
"loss": 2.0148,
"step": 75600
},
{
"epoch": 24.466709760827406,
"grad_norm": 1.5970401763916016,
"learning_rate": 0.001,
"loss": 1.9942,
"step": 75700
},
{
"epoch": 24.49903038138332,
"grad_norm": 1.8230788707733154,
"learning_rate": 0.001,
"loss": 2.0233,
"step": 75800
},
{
"epoch": 24.53135100193924,
"grad_norm": 1.3407267332077026,
"learning_rate": 0.001,
"loss": 2.0037,
"step": 75900
},
{
"epoch": 24.563671622495153,
"grad_norm": 1.5265939235687256,
"learning_rate": 0.001,
"loss": 2.0392,
"step": 76000
},
{
"epoch": 24.595992243051068,
"grad_norm": 1.3966178894042969,
"learning_rate": 0.001,
"loss": 2.0244,
"step": 76100
},
{
"epoch": 24.628312863606983,
"grad_norm": 1.5279021263122559,
"learning_rate": 0.001,
"loss": 2.0293,
"step": 76200
},
{
"epoch": 24.660633484162897,
"grad_norm": 1.4958832263946533,
"learning_rate": 0.001,
"loss": 2.0439,
"step": 76300
},
{
"epoch": 24.69295410471881,
"grad_norm": 1.5250979661941528,
"learning_rate": 0.001,
"loss": 2.0475,
"step": 76400
},
{
"epoch": 24.725274725274726,
"grad_norm": 1.5153076648712158,
"learning_rate": 0.001,
"loss": 2.0397,
"step": 76500
},
{
"epoch": 24.75759534583064,
"grad_norm": 1.655045747756958,
"learning_rate": 0.001,
"loss": 2.0539,
"step": 76600
},
{
"epoch": 24.789915966386555,
"grad_norm": 1.5840027332305908,
"learning_rate": 0.001,
"loss": 2.0401,
"step": 76700
},
{
"epoch": 24.82223658694247,
"grad_norm": 1.7858777046203613,
"learning_rate": 0.001,
"loss": 2.0484,
"step": 76800
},
{
"epoch": 24.854557207498384,
"grad_norm": 1.2337720394134521,
"learning_rate": 0.001,
"loss": 2.0456,
"step": 76900
},
{
"epoch": 24.8868778280543,
"grad_norm": 1.4665465354919434,
"learning_rate": 0.001,
"loss": 2.086,
"step": 77000
},
{
"epoch": 24.919198448610214,
"grad_norm": 1.840309977531433,
"learning_rate": 0.001,
"loss": 2.0634,
"step": 77100
},
{
"epoch": 24.951519069166128,
"grad_norm": 1.268938660621643,
"learning_rate": 0.001,
"loss": 2.0702,
"step": 77200
},
{
"epoch": 24.983839689722043,
"grad_norm": 1.476412057876587,
"learning_rate": 0.001,
"loss": 2.066,
"step": 77300
},
{
"epoch": 25.016160310277957,
"grad_norm": 2.2744739055633545,
"learning_rate": 0.001,
"loss": 1.9527,
"step": 77400
},
{
"epoch": 25.048480930833872,
"grad_norm": 2.6017308235168457,
"learning_rate": 0.001,
"loss": 1.9749,
"step": 77500
},
{
"epoch": 25.080801551389786,
"grad_norm": 2.104499101638794,
"learning_rate": 0.001,
"loss": 1.9307,
"step": 77600
},
{
"epoch": 25.1131221719457,
"grad_norm": 2.4056954383850098,
"learning_rate": 0.001,
"loss": 1.9341,
"step": 77700
},
{
"epoch": 25.145442792501616,
"grad_norm": 2.033249855041504,
"learning_rate": 0.001,
"loss": 1.9518,
"step": 77800
},
{
"epoch": 25.17776341305753,
"grad_norm": 1.6417148113250732,
"learning_rate": 0.001,
"loss": 1.9826,
"step": 77900
},
{
"epoch": 25.210084033613445,
"grad_norm": 1.9145283699035645,
"learning_rate": 0.001,
"loss": 1.9765,
"step": 78000
},
{
"epoch": 25.24240465416936,
"grad_norm": 1.8713525533676147,
"learning_rate": 0.001,
"loss": 1.964,
"step": 78100
},
{
"epoch": 25.274725274725274,
"grad_norm": 1.9054548740386963,
"learning_rate": 0.001,
"loss": 1.9553,
"step": 78200
},
{
"epoch": 25.30704589528119,
"grad_norm": 1.673055648803711,
"learning_rate": 0.001,
"loss": 1.9904,
"step": 78300
},
{
"epoch": 25.339366515837103,
"grad_norm": 2.0248286724090576,
"learning_rate": 0.001,
"loss": 1.9886,
"step": 78400
},
{
"epoch": 25.371687136393017,
"grad_norm": 1.8936235904693604,
"learning_rate": 0.001,
"loss": 1.9842,
"step": 78500
},
{
"epoch": 25.404007756948932,
"grad_norm": 2.2731919288635254,
"learning_rate": 0.001,
"loss": 1.9873,
"step": 78600
},
{
"epoch": 25.436328377504847,
"grad_norm": 1.883629560470581,
"learning_rate": 0.001,
"loss": 1.9706,
"step": 78700
},
{
"epoch": 25.46864899806076,
"grad_norm": 1.7781099081039429,
"learning_rate": 0.001,
"loss": 1.9913,
"step": 78800
},
{
"epoch": 25.50096961861668,
"grad_norm": 1.5534332990646362,
"learning_rate": 0.001,
"loss": 1.9678,
"step": 78900
},
{
"epoch": 25.533290239172594,
"grad_norm": 1.9587346315383911,
"learning_rate": 0.001,
"loss": 2.0279,
"step": 79000
},
{
"epoch": 25.56561085972851,
"grad_norm": 1.8949452638626099,
"learning_rate": 0.001,
"loss": 1.9936,
"step": 79100
},
{
"epoch": 25.597931480284423,
"grad_norm": 1.916925311088562,
"learning_rate": 0.001,
"loss": 2.0152,
"step": 79200
},
{
"epoch": 25.630252100840337,
"grad_norm": 2.057035446166992,
"learning_rate": 0.001,
"loss": 1.9873,
"step": 79300
},
{
"epoch": 25.662572721396252,
"grad_norm": 2.0889744758605957,
"learning_rate": 0.001,
"loss": 2.0126,
"step": 79400
},
{
"epoch": 25.694893341952167,
"grad_norm": 2.0734827518463135,
"learning_rate": 0.001,
"loss": 2.0183,
"step": 79500
},
{
"epoch": 25.72721396250808,
"grad_norm": 1.9695285558700562,
"learning_rate": 0.001,
"loss": 2.0178,
"step": 79600
},
{
"epoch": 25.759534583063996,
"grad_norm": 1.844465970993042,
"learning_rate": 0.001,
"loss": 2.0045,
"step": 79700
},
{
"epoch": 25.79185520361991,
"grad_norm": 2.1013882160186768,
"learning_rate": 0.001,
"loss": 2.0245,
"step": 79800
},
{
"epoch": 25.824175824175825,
"grad_norm": 2.3713462352752686,
"learning_rate": 0.001,
"loss": 2.0091,
"step": 79900
},
{
"epoch": 25.85649644473174,
"grad_norm": 1.981223464012146,
"learning_rate": 0.001,
"loss": 2.0356,
"step": 80000
},
{
"epoch": 25.888817065287654,
"grad_norm": 2.089397668838501,
"learning_rate": 0.001,
"loss": 2.0271,
"step": 80100
},
{
"epoch": 25.92113768584357,
"grad_norm": 2.173301935195923,
"learning_rate": 0.001,
"loss": 2.0314,
"step": 80200
},
{
"epoch": 25.953458306399483,
"grad_norm": 1.814848780632019,
"learning_rate": 0.001,
"loss": 2.0307,
"step": 80300
},
{
"epoch": 25.985778926955398,
"grad_norm": 2.236610174179077,
"learning_rate": 0.001,
"loss": 2.0317,
"step": 80400
},
{
"epoch": 26.018099547511312,
"grad_norm": 1.870568037033081,
"learning_rate": 0.001,
"loss": 1.9713,
"step": 80500
},
{
"epoch": 26.050420168067227,
"grad_norm": 1.495255947113037,
"learning_rate": 0.001,
"loss": 1.9004,
"step": 80600
},
{
"epoch": 26.08274078862314,
"grad_norm": 1.3151029348373413,
"learning_rate": 0.001,
"loss": 1.9098,
"step": 80700
},
{
"epoch": 26.115061409179056,
"grad_norm": 1.3545511960983276,
"learning_rate": 0.001,
"loss": 1.9286,
"step": 80800
},
{
"epoch": 26.14738202973497,
"grad_norm": 1.536292552947998,
"learning_rate": 0.001,
"loss": 1.9249,
"step": 80900
},
{
"epoch": 26.179702650290885,
"grad_norm": 1.2461029291152954,
"learning_rate": 0.001,
"loss": 1.9315,
"step": 81000
},
{
"epoch": 26.2120232708468,
"grad_norm": 1.3342663049697876,
"learning_rate": 0.001,
"loss": 1.9299,
"step": 81100
},
{
"epoch": 26.244343891402714,
"grad_norm": 1.148798942565918,
"learning_rate": 0.001,
"loss": 1.9422,
"step": 81200
},
{
"epoch": 26.27666451195863,
"grad_norm": 1.2964235544204712,
"learning_rate": 0.001,
"loss": 1.9585,
"step": 81300
},
{
"epoch": 26.308985132514543,
"grad_norm": 1.3100039958953857,
"learning_rate": 0.001,
"loss": 1.9527,
"step": 81400
},
{
"epoch": 26.341305753070458,
"grad_norm": 1.6557549238204956,
"learning_rate": 0.001,
"loss": 1.9658,
"step": 81500
},
{
"epoch": 26.373626373626372,
"grad_norm": 1.9575376510620117,
"learning_rate": 0.001,
"loss": 1.9577,
"step": 81600
},
{
"epoch": 26.405946994182287,
"grad_norm": 1.3891215324401855,
"learning_rate": 0.001,
"loss": 1.9592,
"step": 81700
},
{
"epoch": 26.4382676147382,
"grad_norm": 1.4893115758895874,
"learning_rate": 0.001,
"loss": 1.9441,
"step": 81800
},
{
"epoch": 26.470588235294116,
"grad_norm": 1.5343430042266846,
"learning_rate": 0.001,
"loss": 1.9507,
"step": 81900
},
{
"epoch": 26.50290885585003,
"grad_norm": 1.5884416103363037,
"learning_rate": 0.001,
"loss": 1.9655,
"step": 82000
},
{
"epoch": 26.53522947640595,
"grad_norm": 1.4402801990509033,
"learning_rate": 0.001,
"loss": 1.9629,
"step": 82100
},
{
"epoch": 26.567550096961863,
"grad_norm": 1.8175873756408691,
"learning_rate": 0.001,
"loss": 1.9686,
"step": 82200
},
{
"epoch": 26.599870717517778,
"grad_norm": 1.0564377307891846,
"learning_rate": 0.001,
"loss": 1.977,
"step": 82300
},
{
"epoch": 26.632191338073692,
"grad_norm": 1.5843056440353394,
"learning_rate": 0.001,
"loss": 1.9762,
"step": 82400
},
{
"epoch": 26.664511958629607,
"grad_norm": 1.2666752338409424,
"learning_rate": 0.001,
"loss": 1.9911,
"step": 82500
},
{
"epoch": 26.69683257918552,
"grad_norm": 1.8625491857528687,
"learning_rate": 0.001,
"loss": 1.9856,
"step": 82600
},
{
"epoch": 26.729153199741436,
"grad_norm": 1.5447437763214111,
"learning_rate": 0.001,
"loss": 1.9967,
"step": 82700
},
{
"epoch": 26.76147382029735,
"grad_norm": 1.376990795135498,
"learning_rate": 0.001,
"loss": 1.9884,
"step": 82800
},
{
"epoch": 26.793794440853265,
"grad_norm": 1.804949164390564,
"learning_rate": 0.001,
"loss": 2.005,
"step": 82900
},
{
"epoch": 26.82611506140918,
"grad_norm": 1.2310376167297363,
"learning_rate": 0.001,
"loss": 2.0221,
"step": 83000
},
{
"epoch": 26.858435681965094,
"grad_norm": 1.287807583808899,
"learning_rate": 0.001,
"loss": 2.0364,
"step": 83100
},
{
"epoch": 26.89075630252101,
"grad_norm": 1.3025513887405396,
"learning_rate": 0.001,
"loss": 2.0196,
"step": 83200
},
{
"epoch": 26.923076923076923,
"grad_norm": 1.4322696924209595,
"learning_rate": 0.001,
"loss": 1.9999,
"step": 83300
},
{
"epoch": 26.955397543632838,
"grad_norm": 1.3139070272445679,
"learning_rate": 0.001,
"loss": 2.015,
"step": 83400
},
{
"epoch": 26.987718164188752,
"grad_norm": 1.5377211570739746,
"learning_rate": 0.001,
"loss": 2.0096,
"step": 83500
},
{
"epoch": 27.020038784744667,
"grad_norm": 1.1557625532150269,
"learning_rate": 0.001,
"loss": 1.9483,
"step": 83600
},
{
"epoch": 27.05235940530058,
"grad_norm": 1.446763277053833,
"learning_rate": 0.001,
"loss": 1.8683,
"step": 83700
},
{
"epoch": 27.084680025856496,
"grad_norm": 1.820147156715393,
"learning_rate": 0.001,
"loss": 1.8726,
"step": 83800
},
{
"epoch": 27.11700064641241,
"grad_norm": 1.3174505233764648,
"learning_rate": 0.001,
"loss": 1.9013,
"step": 83900
},
{
"epoch": 27.149321266968325,
"grad_norm": 1.4255985021591187,
"learning_rate": 0.001,
"loss": 1.9021,
"step": 84000
},
{
"epoch": 27.18164188752424,
"grad_norm": 1.037857174873352,
"learning_rate": 0.001,
"loss": 1.8889,
"step": 84100
},
{
"epoch": 27.213962508080154,
"grad_norm": 1.2322443723678589,
"learning_rate": 0.001,
"loss": 1.919,
"step": 84200
},
{
"epoch": 27.24628312863607,
"grad_norm": 1.1215150356292725,
"learning_rate": 0.001,
"loss": 1.8952,
"step": 84300
},
{
"epoch": 27.278603749191983,
"grad_norm": 1.5352097749710083,
"learning_rate": 0.001,
"loss": 1.9379,
"step": 84400
},
{
"epoch": 27.310924369747898,
"grad_norm": 1.6588249206542969,
"learning_rate": 0.001,
"loss": 1.9141,
"step": 84500
},
{
"epoch": 27.343244990303813,
"grad_norm": 1.0232058763504028,
"learning_rate": 0.001,
"loss": 1.9484,
"step": 84600
},
{
"epoch": 27.375565610859727,
"grad_norm": 1.1663068532943726,
"learning_rate": 0.001,
"loss": 1.9368,
"step": 84700
},
{
"epoch": 27.40788623141564,
"grad_norm": 1.4564636945724487,
"learning_rate": 0.001,
"loss": 1.9377,
"step": 84800
},
{
"epoch": 27.440206851971556,
"grad_norm": 1.569733738899231,
"learning_rate": 0.001,
"loss": 1.9459,
"step": 84900
},
{
"epoch": 27.47252747252747,
"grad_norm": 1.4780761003494263,
"learning_rate": 0.001,
"loss": 1.9596,
"step": 85000
},
{
"epoch": 27.50484809308339,
"grad_norm": 1.6503818035125732,
"learning_rate": 0.001,
"loss": 1.9514,
"step": 85100
},
{
"epoch": 27.537168713639304,
"grad_norm": 1.4085288047790527,
"learning_rate": 0.001,
"loss": 1.9569,
"step": 85200
},
{
"epoch": 27.569489334195218,
"grad_norm": 1.341328501701355,
"learning_rate": 0.001,
"loss": 1.9578,
"step": 85300
},
{
"epoch": 27.601809954751133,
"grad_norm": 1.232920527458191,
"learning_rate": 0.001,
"loss": 1.9485,
"step": 85400
},
{
"epoch": 27.634130575307047,
"grad_norm": 1.274560809135437,
"learning_rate": 0.001,
"loss": 1.9575,
"step": 85500
},
{
"epoch": 27.66645119586296,
"grad_norm": 1.2799341678619385,
"learning_rate": 0.001,
"loss": 1.9717,
"step": 85600
},
{
"epoch": 27.698771816418876,
"grad_norm": 1.42722749710083,
"learning_rate": 0.001,
"loss": 1.9646,
"step": 85700
},
{
"epoch": 27.73109243697479,
"grad_norm": 1.578813076019287,
"learning_rate": 0.001,
"loss": 1.965,
"step": 85800
},
{
"epoch": 27.763413057530705,
"grad_norm": 1.4803584814071655,
"learning_rate": 0.001,
"loss": 1.9622,
"step": 85900
},
{
"epoch": 27.79573367808662,
"grad_norm": 1.6228522062301636,
"learning_rate": 0.001,
"loss": 1.9464,
"step": 86000
},
{
"epoch": 27.828054298642535,
"grad_norm": 1.2919968366622925,
"learning_rate": 0.001,
"loss": 1.9683,
"step": 86100
},
{
"epoch": 27.86037491919845,
"grad_norm": 1.221200942993164,
"learning_rate": 0.001,
"loss": 1.9735,
"step": 86200
},
{
"epoch": 27.892695539754364,
"grad_norm": 1.462760090827942,
"learning_rate": 0.001,
"loss": 1.9756,
"step": 86300
},
{
"epoch": 27.92501616031028,
"grad_norm": 0.960984468460083,
"learning_rate": 0.001,
"loss": 1.9819,
"step": 86400
},
{
"epoch": 27.957336780866193,
"grad_norm": 1.4694859981536865,
"learning_rate": 0.001,
"loss": 1.9928,
"step": 86500
},
{
"epoch": 27.989657401422107,
"grad_norm": 1.3951972723007202,
"learning_rate": 0.001,
"loss": 1.9907,
"step": 86600
},
{
"epoch": 28.021978021978022,
"grad_norm": 1.2100982666015625,
"learning_rate": 0.001,
"loss": 1.9312,
"step": 86700
},
{
"epoch": 28.054298642533936,
"grad_norm": 1.3661807775497437,
"learning_rate": 0.001,
"loss": 1.8497,
"step": 86800
},
{
"epoch": 28.08661926308985,
"grad_norm": 1.0485259294509888,
"learning_rate": 0.001,
"loss": 1.8621,
"step": 86900
},
{
"epoch": 28.118939883645766,
"grad_norm": 1.33681321144104,
"learning_rate": 0.001,
"loss": 1.8819,
"step": 87000
},
{
"epoch": 28.15126050420168,
"grad_norm": 1.643082857131958,
"learning_rate": 0.001,
"loss": 1.8681,
"step": 87100
},
{
"epoch": 28.183581124757595,
"grad_norm": 1.5186346769332886,
"learning_rate": 0.001,
"loss": 1.8834,
"step": 87200
},
{
"epoch": 28.21590174531351,
"grad_norm": 0.9780005812644958,
"learning_rate": 0.001,
"loss": 1.8804,
"step": 87300
},
{
"epoch": 28.248222365869424,
"grad_norm": 1.3443195819854736,
"learning_rate": 0.001,
"loss": 1.8871,
"step": 87400
},
{
"epoch": 28.28054298642534,
"grad_norm": 1.2283108234405518,
"learning_rate": 0.001,
"loss": 1.9005,
"step": 87500
},
{
"epoch": 28.312863606981253,
"grad_norm": 1.2878503799438477,
"learning_rate": 0.001,
"loss": 1.9133,
"step": 87600
},
{
"epoch": 28.345184227537167,
"grad_norm": 1.4921066761016846,
"learning_rate": 0.001,
"loss": 1.8821,
"step": 87700
},
{
"epoch": 28.377504848093082,
"grad_norm": 1.0940135717391968,
"learning_rate": 0.001,
"loss": 1.9035,
"step": 87800
},
{
"epoch": 28.409825468648997,
"grad_norm": 1.5041978359222412,
"learning_rate": 0.001,
"loss": 1.9182,
"step": 87900
},
{
"epoch": 28.44214608920491,
"grad_norm": 1.3685961961746216,
"learning_rate": 0.001,
"loss": 1.9104,
"step": 88000
},
{
"epoch": 28.474466709760826,
"grad_norm": 0.9715288281440735,
"learning_rate": 0.001,
"loss": 1.8933,
"step": 88100
},
{
"epoch": 28.50678733031674,
"grad_norm": 1.1276092529296875,
"learning_rate": 0.001,
"loss": 1.9218,
"step": 88200
},
{
"epoch": 28.53910795087266,
"grad_norm": 1.2015578746795654,
"learning_rate": 0.001,
"loss": 1.9177,
"step": 88300
},
{
"epoch": 28.571428571428573,
"grad_norm": 1.405713438987732,
"learning_rate": 0.001,
"loss": 1.9161,
"step": 88400
},
{
"epoch": 28.603749191984488,
"grad_norm": 1.2195446491241455,
"learning_rate": 0.001,
"loss": 1.9379,
"step": 88500
},
{
"epoch": 28.636069812540402,
"grad_norm": 1.1534994840621948,
"learning_rate": 0.001,
"loss": 1.9198,
"step": 88600
},
{
"epoch": 28.668390433096317,
"grad_norm": 1.2401294708251953,
"learning_rate": 0.001,
"loss": 1.9546,
"step": 88700
},
{
"epoch": 28.70071105365223,
"grad_norm": 1.0914874076843262,
"learning_rate": 0.001,
"loss": 1.9235,
"step": 88800
},
{
"epoch": 28.733031674208146,
"grad_norm": 1.068054437637329,
"learning_rate": 0.001,
"loss": 1.9317,
"step": 88900
},
{
"epoch": 28.76535229476406,
"grad_norm": 1.1873078346252441,
"learning_rate": 0.001,
"loss": 1.9632,
"step": 89000
},
{
"epoch": 28.797672915319975,
"grad_norm": 1.529497504234314,
"learning_rate": 0.001,
"loss": 1.9459,
"step": 89100
},
{
"epoch": 28.82999353587589,
"grad_norm": 1.1832150220870972,
"learning_rate": 0.001,
"loss": 1.9613,
"step": 89200
},
{
"epoch": 28.862314156431804,
"grad_norm": 1.134078860282898,
"learning_rate": 0.001,
"loss": 1.9669,
"step": 89300
},
{
"epoch": 28.89463477698772,
"grad_norm": 1.0518510341644287,
"learning_rate": 0.001,
"loss": 1.9613,
"step": 89400
},
{
"epoch": 28.926955397543633,
"grad_norm": 1.4630978107452393,
"learning_rate": 0.001,
"loss": 1.9742,
"step": 89500
},
{
"epoch": 28.959276018099548,
"grad_norm": 1.2870162725448608,
"learning_rate": 0.001,
"loss": 1.9635,
"step": 89600
},
{
"epoch": 28.991596638655462,
"grad_norm": 1.2716412544250488,
"learning_rate": 0.001,
"loss": 1.9711,
"step": 89700
},
{
"epoch": 29.023917259211377,
"grad_norm": 1.6266603469848633,
"learning_rate": 0.001,
"loss": 1.8602,
"step": 89800
},
{
"epoch": 29.05623787976729,
"grad_norm": 1.565943956375122,
"learning_rate": 0.001,
"loss": 1.8294,
"step": 89900
},
{
"epoch": 29.088558500323206,
"grad_norm": 1.1085901260375977,
"learning_rate": 0.001,
"loss": 1.8302,
"step": 90000
},
{
"epoch": 29.12087912087912,
"grad_norm": 1.1395728588104248,
"learning_rate": 0.001,
"loss": 1.8454,
"step": 90100
},
{
"epoch": 29.153199741435035,
"grad_norm": 1.566756248474121,
"learning_rate": 0.001,
"loss": 1.8563,
"step": 90200
},
{
"epoch": 29.18552036199095,
"grad_norm": 0.9699623584747314,
"learning_rate": 0.001,
"loss": 1.8623,
"step": 90300
},
{
"epoch": 29.217840982546864,
"grad_norm": 1.4705579280853271,
"learning_rate": 0.001,
"loss": 1.8527,
"step": 90400
},
{
"epoch": 29.25016160310278,
"grad_norm": 1.355370044708252,
"learning_rate": 0.001,
"loss": 1.893,
"step": 90500
},
{
"epoch": 29.282482223658693,
"grad_norm": 1.3909432888031006,
"learning_rate": 0.001,
"loss": 1.8812,
"step": 90600
},
{
"epoch": 29.314802844214608,
"grad_norm": 1.1816900968551636,
"learning_rate": 0.001,
"loss": 1.8655,
"step": 90700
},
{
"epoch": 29.347123464770522,
"grad_norm": 1.5452797412872314,
"learning_rate": 0.001,
"loss": 1.877,
"step": 90800
},
{
"epoch": 29.379444085326437,
"grad_norm": 1.3184372186660767,
"learning_rate": 0.001,
"loss": 1.8757,
"step": 90900
},
{
"epoch": 29.41176470588235,
"grad_norm": 1.2689801454544067,
"learning_rate": 0.001,
"loss": 1.8703,
"step": 91000
},
{
"epoch": 29.444085326438266,
"grad_norm": 1.005915641784668,
"learning_rate": 0.001,
"loss": 1.8925,
"step": 91100
},
{
"epoch": 29.47640594699418,
"grad_norm": 1.5966804027557373,
"learning_rate": 0.001,
"loss": 1.8971,
"step": 91200
},
{
"epoch": 29.5087265675501,
"grad_norm": 1.195678949356079,
"learning_rate": 0.001,
"loss": 1.911,
"step": 91300
},
{
"epoch": 29.541047188106013,
"grad_norm": 1.8636404275894165,
"learning_rate": 0.001,
"loss": 1.9171,
"step": 91400
},
{
"epoch": 29.573367808661928,
"grad_norm": 1.2436506748199463,
"learning_rate": 0.001,
"loss": 1.8951,
"step": 91500
},
{
"epoch": 29.605688429217842,
"grad_norm": 1.3966453075408936,
"learning_rate": 0.001,
"loss": 1.9076,
"step": 91600
},
{
"epoch": 29.638009049773757,
"grad_norm": 1.2091529369354248,
"learning_rate": 0.001,
"loss": 1.9215,
"step": 91700
},
{
"epoch": 29.67032967032967,
"grad_norm": 1.365408182144165,
"learning_rate": 0.001,
"loss": 1.9221,
"step": 91800
},
{
"epoch": 29.702650290885586,
"grad_norm": 1.1305558681488037,
"learning_rate": 0.001,
"loss": 1.9288,
"step": 91900
},
{
"epoch": 29.7349709114415,
"grad_norm": 1.3631125688552856,
"learning_rate": 0.001,
"loss": 1.9105,
"step": 92000
},
{
"epoch": 29.767291531997415,
"grad_norm": 1.4561924934387207,
"learning_rate": 0.001,
"loss": 1.9325,
"step": 92100
},
{
"epoch": 29.79961215255333,
"grad_norm": 1.1379411220550537,
"learning_rate": 0.001,
"loss": 1.9197,
"step": 92200
},
{
"epoch": 29.831932773109244,
"grad_norm": 1.0893300771713257,
"learning_rate": 0.001,
"loss": 1.9179,
"step": 92300
},
{
"epoch": 29.86425339366516,
"grad_norm": 1.2213011980056763,
"learning_rate": 0.001,
"loss": 1.942,
"step": 92400
},
{
"epoch": 29.896574014221073,
"grad_norm": 1.494968056678772,
"learning_rate": 0.001,
"loss": 1.9272,
"step": 92500
},
{
"epoch": 29.928894634776988,
"grad_norm": 1.1858004331588745,
"learning_rate": 0.001,
"loss": 1.9369,
"step": 92600
},
{
"epoch": 29.961215255332903,
"grad_norm": 1.262840747833252,
"learning_rate": 0.001,
"loss": 1.9503,
"step": 92700
},
{
"epoch": 29.993535875888817,
"grad_norm": 1.2398407459259033,
"learning_rate": 0.001,
"loss": 1.9521,
"step": 92800
},
{
"epoch": 30.02585649644473,
"grad_norm": 1.3323872089385986,
"learning_rate": 0.001,
"loss": 1.8326,
"step": 92900
},
{
"epoch": 30.058177117000646,
"grad_norm": 1.5650134086608887,
"learning_rate": 0.001,
"loss": 1.8103,
"step": 93000
},
{
"epoch": 30.09049773755656,
"grad_norm": 1.6720815896987915,
"learning_rate": 0.001,
"loss": 1.8213,
"step": 93100
},
{
"epoch": 30.122818358112475,
"grad_norm": 1.2393896579742432,
"learning_rate": 0.001,
"loss": 1.8266,
"step": 93200
},
{
"epoch": 30.15513897866839,
"grad_norm": 1.468858242034912,
"learning_rate": 0.001,
"loss": 1.8387,
"step": 93300
},
{
"epoch": 30.187459599224304,
"grad_norm": 1.7441996335983276,
"learning_rate": 0.001,
"loss": 1.8402,
"step": 93400
},
{
"epoch": 30.21978021978022,
"grad_norm": 1.106845498085022,
"learning_rate": 0.001,
"loss": 1.8426,
"step": 93500
},
{
"epoch": 30.252100840336134,
"grad_norm": 1.4239120483398438,
"learning_rate": 0.001,
"loss": 1.8457,
"step": 93600
},
{
"epoch": 30.284421460892048,
"grad_norm": 1.1072075366973877,
"learning_rate": 0.001,
"loss": 1.8712,
"step": 93700
},
{
"epoch": 30.316742081447963,
"grad_norm": 1.4010498523712158,
"learning_rate": 0.001,
"loss": 1.8449,
"step": 93800
},
{
"epoch": 30.349062702003877,
"grad_norm": 1.7070467472076416,
"learning_rate": 0.001,
"loss": 1.8591,
"step": 93900
},
{
"epoch": 30.381383322559792,
"grad_norm": 1.3719133138656616,
"learning_rate": 0.001,
"loss": 1.8659,
"step": 94000
},
{
"epoch": 30.413703943115706,
"grad_norm": 1.2224758863449097,
"learning_rate": 0.001,
"loss": 1.8523,
"step": 94100
},
{
"epoch": 30.44602456367162,
"grad_norm": 1.2274152040481567,
"learning_rate": 0.001,
"loss": 1.8777,
"step": 94200
},
{
"epoch": 30.478345184227535,
"grad_norm": 1.23996901512146,
"learning_rate": 0.001,
"loss": 1.8823,
"step": 94300
},
{
"epoch": 30.51066580478345,
"grad_norm": 1.41676926612854,
"learning_rate": 0.001,
"loss": 1.8862,
"step": 94400
},
{
"epoch": 30.542986425339368,
"grad_norm": 1.3740698099136353,
"learning_rate": 0.001,
"loss": 1.8776,
"step": 94500
},
{
"epoch": 30.575307045895283,
"grad_norm": 1.242758870124817,
"learning_rate": 0.001,
"loss": 1.8893,
"step": 94600
},
{
"epoch": 30.607627666451197,
"grad_norm": 1.2667688131332397,
"learning_rate": 0.001,
"loss": 1.8815,
"step": 94700
},
{
"epoch": 30.639948287007112,
"grad_norm": 1.2830685377120972,
"learning_rate": 0.001,
"loss": 1.8888,
"step": 94800
},
{
"epoch": 30.672268907563026,
"grad_norm": 1.6480231285095215,
"learning_rate": 0.001,
"loss": 1.9095,
"step": 94900
},
{
"epoch": 30.70458952811894,
"grad_norm": 1.4405111074447632,
"learning_rate": 0.001,
"loss": 1.8965,
"step": 95000
},
{
"epoch": 30.736910148674855,
"grad_norm": 1.1821390390396118,
"learning_rate": 0.001,
"loss": 1.8876,
"step": 95100
},
{
"epoch": 30.76923076923077,
"grad_norm": 1.060267686843872,
"learning_rate": 0.001,
"loss": 1.9112,
"step": 95200
},
{
"epoch": 30.801551389786685,
"grad_norm": 1.1657590866088867,
"learning_rate": 0.001,
"loss": 1.8993,
"step": 95300
},
{
"epoch": 30.8338720103426,
"grad_norm": 1.247584581375122,
"learning_rate": 0.001,
"loss": 1.8987,
"step": 95400
},
{
"epoch": 30.866192630898514,
"grad_norm": 1.1204445362091064,
"learning_rate": 0.001,
"loss": 1.9222,
"step": 95500
},
{
"epoch": 30.89851325145443,
"grad_norm": 1.2892838716506958,
"learning_rate": 0.001,
"loss": 1.8837,
"step": 95600
},
{
"epoch": 30.930833872010343,
"grad_norm": 1.3222150802612305,
"learning_rate": 0.001,
"loss": 1.9183,
"step": 95700
},
{
"epoch": 30.963154492566257,
"grad_norm": 1.1754868030548096,
"learning_rate": 0.001,
"loss": 1.8873,
"step": 95800
},
{
"epoch": 30.995475113122172,
"grad_norm": 2.0036423206329346,
"learning_rate": 0.001,
"loss": 1.8993,
"step": 95900
},
{
"epoch": 31.027795733678087,
"grad_norm": 1.5189635753631592,
"learning_rate": 0.001,
"loss": 1.8267,
"step": 96000
},
{
"epoch": 31.060116354234,
"grad_norm": 1.5205469131469727,
"learning_rate": 0.001,
"loss": 1.7728,
"step": 96100
},
{
"epoch": 31.092436974789916,
"grad_norm": 1.5693349838256836,
"learning_rate": 0.001,
"loss": 1.7906,
"step": 96200
},
{
"epoch": 31.12475759534583,
"grad_norm": 1.1951963901519775,
"learning_rate": 0.001,
"loss": 1.7979,
"step": 96300
},
{
"epoch": 31.157078215901745,
"grad_norm": 1.4093998670578003,
"learning_rate": 0.001,
"loss": 1.807,
"step": 96400
},
{
"epoch": 31.18939883645766,
"grad_norm": 1.1882504224777222,
"learning_rate": 0.001,
"loss": 1.8112,
"step": 96500
},
{
"epoch": 31.221719457013574,
"grad_norm": 1.4225773811340332,
"learning_rate": 0.001,
"loss": 1.808,
"step": 96600
},
{
"epoch": 31.25404007756949,
"grad_norm": 1.5231945514678955,
"learning_rate": 0.001,
"loss": 1.817,
"step": 96700
},
{
"epoch": 31.286360698125403,
"grad_norm": 1.1047606468200684,
"learning_rate": 0.001,
"loss": 1.8177,
"step": 96800
},
{
"epoch": 31.318681318681318,
"grad_norm": 1.7534669637680054,
"learning_rate": 0.001,
"loss": 1.8447,
"step": 96900
},
{
"epoch": 31.351001939237232,
"grad_norm": 1.1873900890350342,
"learning_rate": 0.001,
"loss": 1.8534,
"step": 97000
},
{
"epoch": 31.383322559793147,
"grad_norm": 1.5391671657562256,
"learning_rate": 0.001,
"loss": 1.8465,
"step": 97100
},
{
"epoch": 31.41564318034906,
"grad_norm": 1.2355151176452637,
"learning_rate": 0.001,
"loss": 1.8367,
"step": 97200
},
{
"epoch": 31.447963800904976,
"grad_norm": 1.5391733646392822,
"learning_rate": 0.001,
"loss": 1.8533,
"step": 97300
},
{
"epoch": 31.48028442146089,
"grad_norm": 1.5837594270706177,
"learning_rate": 0.001,
"loss": 1.852,
"step": 97400
},
{
"epoch": 31.51260504201681,
"grad_norm": 1.621614694595337,
"learning_rate": 0.001,
"loss": 1.8666,
"step": 97500
},
{
"epoch": 31.544925662572723,
"grad_norm": 1.3326274156570435,
"learning_rate": 0.001,
"loss": 1.8625,
"step": 97600
},
{
"epoch": 31.577246283128638,
"grad_norm": 1.7983978986740112,
"learning_rate": 0.001,
"loss": 1.855,
"step": 97700
},
{
"epoch": 31.609566903684552,
"grad_norm": 1.7443538904190063,
"learning_rate": 0.001,
"loss": 1.8759,
"step": 97800
},
{
"epoch": 31.641887524240467,
"grad_norm": 1.2091596126556396,
"learning_rate": 0.001,
"loss": 1.8732,
"step": 97900
},
{
"epoch": 31.67420814479638,
"grad_norm": 1.17062509059906,
"learning_rate": 0.001,
"loss": 1.8758,
"step": 98000
},
{
"epoch": 31.706528765352296,
"grad_norm": 1.6161521673202515,
"learning_rate": 0.001,
"loss": 1.8711,
"step": 98100
},
{
"epoch": 31.73884938590821,
"grad_norm": 1.6055936813354492,
"learning_rate": 0.001,
"loss": 1.8694,
"step": 98200
},
{
"epoch": 31.771170006464125,
"grad_norm": 1.7089295387268066,
"learning_rate": 0.001,
"loss": 1.8783,
"step": 98300
},
{
"epoch": 31.80349062702004,
"grad_norm": 1.6026934385299683,
"learning_rate": 0.001,
"loss": 1.8962,
"step": 98400
},
{
"epoch": 31.835811247575954,
"grad_norm": 1.5359739065170288,
"learning_rate": 0.001,
"loss": 1.8792,
"step": 98500
},
{
"epoch": 31.86813186813187,
"grad_norm": 1.1889594793319702,
"learning_rate": 0.001,
"loss": 1.9086,
"step": 98600
},
{
"epoch": 31.900452488687783,
"grad_norm": 1.362362265586853,
"learning_rate": 0.001,
"loss": 1.8811,
"step": 98700
},
{
"epoch": 31.932773109243698,
"grad_norm": 1.6680059432983398,
"learning_rate": 0.001,
"loss": 1.8937,
"step": 98800
},
{
"epoch": 31.965093729799612,
"grad_norm": 1.2521178722381592,
"learning_rate": 0.001,
"loss": 1.8801,
"step": 98900
},
{
"epoch": 31.997414350355527,
"grad_norm": 1.616936206817627,
"learning_rate": 0.001,
"loss": 1.8914,
"step": 99000
},
{
"epoch": 32.02973497091144,
"grad_norm": 1.6018805503845215,
"learning_rate": 0.001,
"loss": 1.7961,
"step": 99100
},
{
"epoch": 32.062055591467356,
"grad_norm": 1.572320818901062,
"learning_rate": 0.001,
"loss": 1.7659,
"step": 99200
},
{
"epoch": 32.09437621202327,
"grad_norm": 1.987259030342102,
"learning_rate": 0.001,
"loss": 1.7972,
"step": 99300
},
{
"epoch": 32.126696832579185,
"grad_norm": 1.6123111248016357,
"learning_rate": 0.001,
"loss": 1.7724,
"step": 99400
},
{
"epoch": 32.1590174531351,
"grad_norm": 1.2040059566497803,
"learning_rate": 0.001,
"loss": 1.7894,
"step": 99500
},
{
"epoch": 32.191338073691014,
"grad_norm": 1.4743309020996094,
"learning_rate": 0.001,
"loss": 1.8069,
"step": 99600
},
{
"epoch": 32.22365869424693,
"grad_norm": 1.7867310047149658,
"learning_rate": 0.001,
"loss": 1.8028,
"step": 99700
},
{
"epoch": 32.25597931480284,
"grad_norm": 1.8878384828567505,
"learning_rate": 0.001,
"loss": 1.8005,
"step": 99800
},
{
"epoch": 32.28829993535876,
"grad_norm": 1.6064954996109009,
"learning_rate": 0.001,
"loss": 1.8001,
"step": 99900
},
{
"epoch": 32.32062055591467,
"grad_norm": 1.521106243133545,
"learning_rate": 0.001,
"loss": 1.7954,
"step": 100000
},
{
"epoch": 32.35294117647059,
"grad_norm": 1.6338764429092407,
"learning_rate": 0.001,
"loss": 1.8231,
"step": 100100
},
{
"epoch": 32.3852617970265,
"grad_norm": 1.635077714920044,
"learning_rate": 0.001,
"loss": 1.8095,
"step": 100200
},
{
"epoch": 32.417582417582416,
"grad_norm": 1.2968257665634155,
"learning_rate": 0.001,
"loss": 1.8286,
"step": 100300
},
{
"epoch": 32.44990303813833,
"grad_norm": 1.4315623044967651,
"learning_rate": 0.001,
"loss": 1.8202,
"step": 100400
},
{
"epoch": 32.482223658694245,
"grad_norm": 1.7625807523727417,
"learning_rate": 0.001,
"loss": 1.8428,
"step": 100500
},
{
"epoch": 32.51454427925016,
"grad_norm": 1.3693516254425049,
"learning_rate": 0.001,
"loss": 1.8412,
"step": 100600
},
{
"epoch": 32.546864899806074,
"grad_norm": 1.441098690032959,
"learning_rate": 0.001,
"loss": 1.8156,
"step": 100700
},
{
"epoch": 32.57918552036199,
"grad_norm": 1.4699690341949463,
"learning_rate": 0.001,
"loss": 1.8586,
"step": 100800
},
{
"epoch": 32.6115061409179,
"grad_norm": 2.550903558731079,
"learning_rate": 0.001,
"loss": 1.8482,
"step": 100900
},
{
"epoch": 32.64382676147382,
"grad_norm": 1.6919152736663818,
"learning_rate": 0.001,
"loss": 1.8602,
"step": 101000
},
{
"epoch": 32.67614738202973,
"grad_norm": 1.5435961484909058,
"learning_rate": 0.001,
"loss": 1.8381,
"step": 101100
},
{
"epoch": 32.70846800258565,
"grad_norm": 1.6188323497772217,
"learning_rate": 0.001,
"loss": 1.8566,
"step": 101200
},
{
"epoch": 32.74078862314156,
"grad_norm": 1.7124555110931396,
"learning_rate": 0.001,
"loss": 1.8561,
"step": 101300
},
{
"epoch": 32.773109243697476,
"grad_norm": 1.8261075019836426,
"learning_rate": 0.001,
"loss": 1.8466,
"step": 101400
},
{
"epoch": 32.80542986425339,
"grad_norm": 1.6595410108566284,
"learning_rate": 0.001,
"loss": 1.858,
"step": 101500
},
{
"epoch": 32.837750484809305,
"grad_norm": 1.3070226907730103,
"learning_rate": 0.001,
"loss": 1.8703,
"step": 101600
},
{
"epoch": 32.87007110536522,
"grad_norm": 1.761932373046875,
"learning_rate": 0.001,
"loss": 1.8731,
"step": 101700
},
{
"epoch": 32.902391725921134,
"grad_norm": 1.431853175163269,
"learning_rate": 0.001,
"loss": 1.8585,
"step": 101800
},
{
"epoch": 32.93471234647705,
"grad_norm": 1.4691671133041382,
"learning_rate": 0.001,
"loss": 1.8466,
"step": 101900
},
{
"epoch": 32.967032967032964,
"grad_norm": 1.5675320625305176,
"learning_rate": 0.001,
"loss": 1.8724,
"step": 102000
},
{
"epoch": 32.999353587588885,
"grad_norm": 1.831527829170227,
"learning_rate": 0.001,
"loss": 1.8737,
"step": 102100
},
{
"epoch": 33.0316742081448,
"grad_norm": 1.768676519393921,
"learning_rate": 0.001,
"loss": 1.7363,
"step": 102200
},
{
"epoch": 33.063994828700714,
"grad_norm": 1.7403342723846436,
"learning_rate": 0.001,
"loss": 1.7487,
"step": 102300
},
{
"epoch": 33.09631544925663,
"grad_norm": 1.5318952798843384,
"learning_rate": 0.001,
"loss": 1.765,
"step": 102400
},
{
"epoch": 33.12863606981254,
"grad_norm": 1.767539143562317,
"learning_rate": 0.001,
"loss": 1.7783,
"step": 102500
},
{
"epoch": 33.16095669036846,
"grad_norm": 2.2086029052734375,
"learning_rate": 0.001,
"loss": 1.7807,
"step": 102600
},
{
"epoch": 33.19327731092437,
"grad_norm": 2.058333396911621,
"learning_rate": 0.001,
"loss": 1.7926,
"step": 102700
},
{
"epoch": 33.22559793148029,
"grad_norm": 1.7875993251800537,
"learning_rate": 0.001,
"loss": 1.7916,
"step": 102800
},
{
"epoch": 33.2579185520362,
"grad_norm": 2.276167869567871,
"learning_rate": 0.001,
"loss": 1.7611,
"step": 102900
},
{
"epoch": 33.290239172592116,
"grad_norm": 1.4356229305267334,
"learning_rate": 0.001,
"loss": 1.7932,
"step": 103000
},
{
"epoch": 33.32255979314803,
"grad_norm": 2.1756107807159424,
"learning_rate": 0.001,
"loss": 1.7796,
"step": 103100
},
{
"epoch": 33.354880413703945,
"grad_norm": 2.1314845085144043,
"learning_rate": 0.001,
"loss": 1.7952,
"step": 103200
},
{
"epoch": 33.38720103425986,
"grad_norm": 1.6774673461914062,
"learning_rate": 0.001,
"loss": 1.8078,
"step": 103300
},
{
"epoch": 33.419521654815775,
"grad_norm": 1.5583240985870361,
"learning_rate": 0.001,
"loss": 1.7953,
"step": 103400
},
{
"epoch": 33.45184227537169,
"grad_norm": 2.216525077819824,
"learning_rate": 0.001,
"loss": 1.7887,
"step": 103500
},
{
"epoch": 33.484162895927604,
"grad_norm": 1.5822532176971436,
"learning_rate": 0.001,
"loss": 1.7857,
"step": 103600
},
{
"epoch": 33.51648351648352,
"grad_norm": 2.059272527694702,
"learning_rate": 0.001,
"loss": 1.8239,
"step": 103700
},
{
"epoch": 33.54880413703943,
"grad_norm": 2.067791223526001,
"learning_rate": 0.001,
"loss": 1.8228,
"step": 103800
},
{
"epoch": 33.58112475759535,
"grad_norm": 1.7426116466522217,
"learning_rate": 0.001,
"loss": 1.8187,
"step": 103900
},
{
"epoch": 33.61344537815126,
"grad_norm": 2.092301845550537,
"learning_rate": 0.001,
"loss": 1.8198,
"step": 104000
},
{
"epoch": 33.645765998707176,
"grad_norm": 1.4148643016815186,
"learning_rate": 0.001,
"loss": 1.8446,
"step": 104100
},
{
"epoch": 33.67808661926309,
"grad_norm": 2.115337610244751,
"learning_rate": 0.001,
"loss": 1.8341,
"step": 104200
},
{
"epoch": 33.710407239819006,
"grad_norm": 1.8922119140625,
"learning_rate": 0.001,
"loss": 1.8314,
"step": 104300
},
{
"epoch": 33.74272786037492,
"grad_norm": 2.526000499725342,
"learning_rate": 0.001,
"loss": 1.8322,
"step": 104400
},
{
"epoch": 33.775048480930835,
"grad_norm": 1.95648992061615,
"learning_rate": 0.001,
"loss": 1.8507,
"step": 104500
},
{
"epoch": 33.80736910148675,
"grad_norm": 2.0690479278564453,
"learning_rate": 0.001,
"loss": 1.8452,
"step": 104600
},
{
"epoch": 33.839689722042664,
"grad_norm": 1.7053970098495483,
"learning_rate": 0.001,
"loss": 1.8557,
"step": 104700
},
{
"epoch": 33.87201034259858,
"grad_norm": 1.5044736862182617,
"learning_rate": 0.001,
"loss": 1.8416,
"step": 104800
},
{
"epoch": 33.90433096315449,
"grad_norm": 1.7018338441848755,
"learning_rate": 0.001,
"loss": 1.8698,
"step": 104900
},
{
"epoch": 33.93665158371041,
"grad_norm": 1.5839446783065796,
"learning_rate": 0.001,
"loss": 1.8678,
"step": 105000
},
{
"epoch": 33.96897220426632,
"grad_norm": 1.8129023313522339,
"learning_rate": 0.001,
"loss": 1.8497,
"step": 105100
},
{
"epoch": 34.00129282482224,
"grad_norm": 2.0042531490325928,
"learning_rate": 0.001,
"loss": 1.8455,
"step": 105200
},
{
"epoch": 34.03361344537815,
"grad_norm": 2.1890511512756348,
"learning_rate": 0.001,
"loss": 1.7252,
"step": 105300
},
{
"epoch": 34.065934065934066,
"grad_norm": 1.5610796213150024,
"learning_rate": 0.001,
"loss": 1.7255,
"step": 105400
},
{
"epoch": 34.09825468648998,
"grad_norm": 1.540030837059021,
"learning_rate": 0.001,
"loss": 1.7401,
"step": 105500
},
{
"epoch": 34.130575307045895,
"grad_norm": 1.5150138139724731,
"learning_rate": 0.001,
"loss": 1.7475,
"step": 105600
},
{
"epoch": 34.16289592760181,
"grad_norm": 1.734230875968933,
"learning_rate": 0.001,
"loss": 1.7407,
"step": 105700
},
{
"epoch": 34.195216548157724,
"grad_norm": 1.4981082677841187,
"learning_rate": 0.001,
"loss": 1.7366,
"step": 105800
},
{
"epoch": 34.22753716871364,
"grad_norm": 1.9089019298553467,
"learning_rate": 0.001,
"loss": 1.756,
"step": 105900
},
{
"epoch": 34.25985778926955,
"grad_norm": 1.513126254081726,
"learning_rate": 0.001,
"loss": 1.772,
"step": 106000
},
{
"epoch": 34.29217840982547,
"grad_norm": 1.825900912284851,
"learning_rate": 0.001,
"loss": 1.7671,
"step": 106100
},
{
"epoch": 34.32449903038138,
"grad_norm": 1.3419309854507446,
"learning_rate": 0.001,
"loss": 1.7717,
"step": 106200
},
{
"epoch": 34.3568196509373,
"grad_norm": 1.7643647193908691,
"learning_rate": 0.001,
"loss": 1.7825,
"step": 106300
},
{
"epoch": 34.38914027149321,
"grad_norm": 1.6718119382858276,
"learning_rate": 0.001,
"loss": 1.7848,
"step": 106400
},
{
"epoch": 34.421460892049126,
"grad_norm": 1.4287859201431274,
"learning_rate": 0.001,
"loss": 1.7878,
"step": 106500
},
{
"epoch": 34.45378151260504,
"grad_norm": 1.3736450672149658,
"learning_rate": 0.001,
"loss": 1.7913,
"step": 106600
},
{
"epoch": 34.486102133160955,
"grad_norm": 1.4067167043685913,
"learning_rate": 0.001,
"loss": 1.7939,
"step": 106700
},
{
"epoch": 34.51842275371687,
"grad_norm": 1.4027693271636963,
"learning_rate": 0.001,
"loss": 1.7663,
"step": 106800
},
{
"epoch": 34.550743374272784,
"grad_norm": 1.3773787021636963,
"learning_rate": 0.001,
"loss": 1.8117,
"step": 106900
},
{
"epoch": 34.5830639948287,
"grad_norm": 1.795181155204773,
"learning_rate": 0.001,
"loss": 1.7823,
"step": 107000
},
{
"epoch": 34.61538461538461,
"grad_norm": 1.54495108127594,
"learning_rate": 0.001,
"loss": 1.7914,
"step": 107100
},
{
"epoch": 34.64770523594053,
"grad_norm": 2.048011064529419,
"learning_rate": 0.001,
"loss": 1.8232,
"step": 107200
},
{
"epoch": 34.68002585649644,
"grad_norm": 1.1294292211532593,
"learning_rate": 0.001,
"loss": 1.8143,
"step": 107300
},
{
"epoch": 34.71234647705236,
"grad_norm": 1.6068726778030396,
"learning_rate": 0.001,
"loss": 1.8199,
"step": 107400
},
{
"epoch": 34.74466709760827,
"grad_norm": 1.8478591442108154,
"learning_rate": 0.001,
"loss": 1.8298,
"step": 107500
},
{
"epoch": 34.776987718164186,
"grad_norm": 1.4019745588302612,
"learning_rate": 0.001,
"loss": 1.8242,
"step": 107600
},
{
"epoch": 34.8093083387201,
"grad_norm": 1.3042278289794922,
"learning_rate": 0.001,
"loss": 1.8359,
"step": 107700
},
{
"epoch": 34.841628959276015,
"grad_norm": 1.7335820198059082,
"learning_rate": 0.001,
"loss": 1.8355,
"step": 107800
},
{
"epoch": 34.87394957983193,
"grad_norm": 1.7621225118637085,
"learning_rate": 0.001,
"loss": 1.8295,
"step": 107900
},
{
"epoch": 34.906270200387844,
"grad_norm": 1.2129218578338623,
"learning_rate": 0.001,
"loss": 1.839,
"step": 108000
},
{
"epoch": 34.93859082094376,
"grad_norm": 1.3511784076690674,
"learning_rate": 0.001,
"loss": 1.8281,
"step": 108100
},
{
"epoch": 34.97091144149967,
"grad_norm": 1.543796181678772,
"learning_rate": 0.001,
"loss": 1.8349,
"step": 108200
},
{
"epoch": 35.003232062055595,
"grad_norm": 1.3133898973464966,
"learning_rate": 0.001,
"loss": 1.8232,
"step": 108300
},
{
"epoch": 35.03555268261151,
"grad_norm": 1.309926986694336,
"learning_rate": 0.001,
"loss": 1.71,
"step": 108400
},
{
"epoch": 35.067873303167424,
"grad_norm": 1.7792454957962036,
"learning_rate": 0.001,
"loss": 1.7366,
"step": 108500
},
{
"epoch": 35.10019392372334,
"grad_norm": 1.2482664585113525,
"learning_rate": 0.001,
"loss": 1.7138,
"step": 108600
},
{
"epoch": 35.13251454427925,
"grad_norm": 1.5493508577346802,
"learning_rate": 0.001,
"loss": 1.7139,
"step": 108700
},
{
"epoch": 35.16483516483517,
"grad_norm": 1.4078187942504883,
"learning_rate": 0.001,
"loss": 1.7421,
"step": 108800
},
{
"epoch": 35.19715578539108,
"grad_norm": 1.512757420539856,
"learning_rate": 0.001,
"loss": 1.753,
"step": 108900
},
{
"epoch": 35.229476405947,
"grad_norm": 1.444258451461792,
"learning_rate": 0.001,
"loss": 1.7136,
"step": 109000
},
{
"epoch": 35.26179702650291,
"grad_norm": 1.8412730693817139,
"learning_rate": 0.001,
"loss": 1.7397,
"step": 109100
},
{
"epoch": 35.294117647058826,
"grad_norm": 1.1742063760757446,
"learning_rate": 0.001,
"loss": 1.7359,
"step": 109200
},
{
"epoch": 35.32643826761474,
"grad_norm": 1.5857884883880615,
"learning_rate": 0.001,
"loss": 1.773,
"step": 109300
},
{
"epoch": 35.358758888170655,
"grad_norm": 1.8618083000183105,
"learning_rate": 0.001,
"loss": 1.775,
"step": 109400
},
{
"epoch": 35.39107950872657,
"grad_norm": 1.5522814989089966,
"learning_rate": 0.001,
"loss": 1.7807,
"step": 109500
},
{
"epoch": 35.423400129282484,
"grad_norm": 1.4634019136428833,
"learning_rate": 0.001,
"loss": 1.7788,
"step": 109600
},
{
"epoch": 35.4557207498384,
"grad_norm": 1.684096097946167,
"learning_rate": 0.001,
"loss": 1.7812,
"step": 109700
},
{
"epoch": 35.48804137039431,
"grad_norm": 1.911356806755066,
"learning_rate": 0.001,
"loss": 1.7775,
"step": 109800
},
{
"epoch": 35.52036199095023,
"grad_norm": 1.1707732677459717,
"learning_rate": 0.001,
"loss": 1.7715,
"step": 109900
},
{
"epoch": 35.55268261150614,
"grad_norm": 1.108163595199585,
"learning_rate": 0.001,
"loss": 1.7673,
"step": 110000
},
{
"epoch": 35.58500323206206,
"grad_norm": 1.791603684425354,
"learning_rate": 0.001,
"loss": 1.7828,
"step": 110100
},
{
"epoch": 35.61732385261797,
"grad_norm": 1.3673819303512573,
"learning_rate": 0.001,
"loss": 1.7974,
"step": 110200
},
{
"epoch": 35.649644473173886,
"grad_norm": 1.5180928707122803,
"learning_rate": 0.001,
"loss": 1.7823,
"step": 110300
},
{
"epoch": 35.6819650937298,
"grad_norm": 1.694899082183838,
"learning_rate": 0.001,
"loss": 1.7895,
"step": 110400
},
{
"epoch": 35.714285714285715,
"grad_norm": 1.395687222480774,
"learning_rate": 0.001,
"loss": 1.7905,
"step": 110500
},
{
"epoch": 35.74660633484163,
"grad_norm": 1.430392861366272,
"learning_rate": 0.001,
"loss": 1.8125,
"step": 110600
},
{
"epoch": 35.778926955397544,
"grad_norm": 1.394409418106079,
"learning_rate": 0.001,
"loss": 1.7847,
"step": 110700
},
{
"epoch": 35.81124757595346,
"grad_norm": 2.1065635681152344,
"learning_rate": 0.001,
"loss": 1.8037,
"step": 110800
},
{
"epoch": 35.84356819650937,
"grad_norm": 1.502328634262085,
"learning_rate": 0.001,
"loss": 1.8033,
"step": 110900
},
{
"epoch": 35.87588881706529,
"grad_norm": 1.1700212955474854,
"learning_rate": 0.001,
"loss": 1.8106,
"step": 111000
},
{
"epoch": 35.9082094376212,
"grad_norm": 1.6413511037826538,
"learning_rate": 0.001,
"loss": 1.8146,
"step": 111100
},
{
"epoch": 35.94053005817712,
"grad_norm": 1.7164719104766846,
"learning_rate": 0.001,
"loss": 1.8177,
"step": 111200
},
{
"epoch": 35.97285067873303,
"grad_norm": 1.4148567914962769,
"learning_rate": 0.001,
"loss": 1.805,
"step": 111300
},
{
"epoch": 36.005171299288946,
"grad_norm": 1.369500994682312,
"learning_rate": 0.001,
"loss": 1.7962,
"step": 111400
},
{
"epoch": 36.03749191984486,
"grad_norm": 1.2702453136444092,
"learning_rate": 0.001,
"loss": 1.6798,
"step": 111500
},
{
"epoch": 36.069812540400775,
"grad_norm": 1.3947583436965942,
"learning_rate": 0.001,
"loss": 1.7028,
"step": 111600
},
{
"epoch": 36.10213316095669,
"grad_norm": 1.4396507740020752,
"learning_rate": 0.001,
"loss": 1.7236,
"step": 111700
},
{
"epoch": 36.134453781512605,
"grad_norm": 1.1698938608169556,
"learning_rate": 0.001,
"loss": 1.6912,
"step": 111800
},
{
"epoch": 36.16677440206852,
"grad_norm": 1.5718566179275513,
"learning_rate": 0.001,
"loss": 1.7184,
"step": 111900
},
{
"epoch": 36.199095022624434,
"grad_norm": 1.6437325477600098,
"learning_rate": 0.001,
"loss": 1.7065,
"step": 112000
},
{
"epoch": 36.23141564318035,
"grad_norm": 1.4717180728912354,
"learning_rate": 0.001,
"loss": 1.709,
"step": 112100
},
{
"epoch": 36.26373626373626,
"grad_norm": 1.5429532527923584,
"learning_rate": 0.001,
"loss": 1.7359,
"step": 112200
},
{
"epoch": 36.29605688429218,
"grad_norm": 1.5812757015228271,
"learning_rate": 0.001,
"loss": 1.7223,
"step": 112300
},
{
"epoch": 36.32837750484809,
"grad_norm": 1.7356760501861572,
"learning_rate": 0.001,
"loss": 1.7252,
"step": 112400
},
{
"epoch": 36.36069812540401,
"grad_norm": 1.375969409942627,
"learning_rate": 0.001,
"loss": 1.7433,
"step": 112500
},
{
"epoch": 36.39301874595992,
"grad_norm": 1.131496787071228,
"learning_rate": 0.001,
"loss": 1.7571,
"step": 112600
},
{
"epoch": 36.425339366515836,
"grad_norm": 1.422635793685913,
"learning_rate": 0.001,
"loss": 1.7496,
"step": 112700
},
{
"epoch": 36.45765998707175,
"grad_norm": 1.3067173957824707,
"learning_rate": 0.001,
"loss": 1.7501,
"step": 112800
},
{
"epoch": 36.489980607627665,
"grad_norm": 1.524130940437317,
"learning_rate": 0.001,
"loss": 1.7565,
"step": 112900
},
{
"epoch": 36.52230122818358,
"grad_norm": 1.2107754945755005,
"learning_rate": 0.001,
"loss": 1.7719,
"step": 113000
},
{
"epoch": 36.554621848739494,
"grad_norm": 1.3486363887786865,
"learning_rate": 0.001,
"loss": 1.7725,
"step": 113100
},
{
"epoch": 36.58694246929541,
"grad_norm": 1.1192071437835693,
"learning_rate": 0.001,
"loss": 1.7699,
"step": 113200
},
{
"epoch": 36.61926308985132,
"grad_norm": 1.3475347757339478,
"learning_rate": 0.001,
"loss": 1.7544,
"step": 113300
},
{
"epoch": 36.65158371040724,
"grad_norm": 1.3668153285980225,
"learning_rate": 0.001,
"loss": 1.7797,
"step": 113400
},
{
"epoch": 36.68390433096315,
"grad_norm": 1.4461634159088135,
"learning_rate": 0.001,
"loss": 1.7845,
"step": 113500
},
{
"epoch": 36.71622495151907,
"grad_norm": 1.4558709859848022,
"learning_rate": 0.001,
"loss": 1.7933,
"step": 113600
},
{
"epoch": 36.74854557207498,
"grad_norm": 1.2568235397338867,
"learning_rate": 0.001,
"loss": 1.7781,
"step": 113700
},
{
"epoch": 36.780866192630896,
"grad_norm": 1.2808581590652466,
"learning_rate": 0.001,
"loss": 1.7947,
"step": 113800
},
{
"epoch": 36.81318681318681,
"grad_norm": 1.3239548206329346,
"learning_rate": 0.001,
"loss": 1.787,
"step": 113900
},
{
"epoch": 36.845507433742725,
"grad_norm": 1.446629285812378,
"learning_rate": 0.001,
"loss": 1.7931,
"step": 114000
},
{
"epoch": 36.87782805429864,
"grad_norm": 1.5585201978683472,
"learning_rate": 0.001,
"loss": 1.7733,
"step": 114100
},
{
"epoch": 36.910148674854554,
"grad_norm": 1.1408222913742065,
"learning_rate": 0.001,
"loss": 1.807,
"step": 114200
},
{
"epoch": 36.94246929541047,
"grad_norm": 1.409403681755066,
"learning_rate": 0.001,
"loss": 1.7972,
"step": 114300
},
{
"epoch": 36.97478991596638,
"grad_norm": 1.6675604581832886,
"learning_rate": 0.001,
"loss": 1.7797,
"step": 114400
},
{
"epoch": 37.007110536522305,
"grad_norm": 1.211574673652649,
"learning_rate": 0.001,
"loss": 1.7839,
"step": 114500
},
{
"epoch": 37.03943115707822,
"grad_norm": 1.6282230615615845,
"learning_rate": 0.001,
"loss": 1.6775,
"step": 114600
},
{
"epoch": 37.071751777634134,
"grad_norm": 1.6066936254501343,
"learning_rate": 0.001,
"loss": 1.668,
"step": 114700
},
{
"epoch": 37.10407239819005,
"grad_norm": 1.384175419807434,
"learning_rate": 0.001,
"loss": 1.6721,
"step": 114800
},
{
"epoch": 37.13639301874596,
"grad_norm": 1.2160314321517944,
"learning_rate": 0.001,
"loss": 1.6847,
"step": 114900
},
{
"epoch": 37.16871363930188,
"grad_norm": 1.1695581674575806,
"learning_rate": 0.001,
"loss": 1.6923,
"step": 115000
},
{
"epoch": 37.20103425985779,
"grad_norm": 1.4204388856887817,
"learning_rate": 0.001,
"loss": 1.6962,
"step": 115100
},
{
"epoch": 37.23335488041371,
"grad_norm": 1.1568971872329712,
"learning_rate": 0.001,
"loss": 1.709,
"step": 115200
},
{
"epoch": 37.26567550096962,
"grad_norm": 1.467090129852295,
"learning_rate": 0.001,
"loss": 1.7089,
"step": 115300
},
{
"epoch": 37.297996121525536,
"grad_norm": 1.3767518997192383,
"learning_rate": 0.001,
"loss": 1.7442,
"step": 115400
},
{
"epoch": 37.33031674208145,
"grad_norm": 1.2017751932144165,
"learning_rate": 0.001,
"loss": 1.7259,
"step": 115500
},
{
"epoch": 37.362637362637365,
"grad_norm": 1.2458261251449585,
"learning_rate": 0.001,
"loss": 1.7122,
"step": 115600
},
{
"epoch": 37.39495798319328,
"grad_norm": 1.1405456066131592,
"learning_rate": 0.001,
"loss": 1.7276,
"step": 115700
},
{
"epoch": 37.427278603749194,
"grad_norm": 1.1807092428207397,
"learning_rate": 0.001,
"loss": 1.7349,
"step": 115800
},
{
"epoch": 37.45959922430511,
"grad_norm": 1.2478501796722412,
"learning_rate": 0.001,
"loss": 1.7379,
"step": 115900
},
{
"epoch": 37.49191984486102,
"grad_norm": 1.7146801948547363,
"learning_rate": 0.001,
"loss": 1.7503,
"step": 116000
},
{
"epoch": 37.52424046541694,
"grad_norm": 1.3230942487716675,
"learning_rate": 0.001,
"loss": 1.7545,
"step": 116100
},
{
"epoch": 37.55656108597285,
"grad_norm": 1.4975101947784424,
"learning_rate": 0.001,
"loss": 1.7507,
"step": 116200
},
{
"epoch": 37.58888170652877,
"grad_norm": 1.3942639827728271,
"learning_rate": 0.001,
"loss": 1.7369,
"step": 116300
},
{
"epoch": 37.62120232708468,
"grad_norm": 1.3690227270126343,
"learning_rate": 0.001,
"loss": 1.7387,
"step": 116400
},
{
"epoch": 37.653522947640596,
"grad_norm": 1.3362698554992676,
"learning_rate": 0.001,
"loss": 1.7416,
"step": 116500
},
{
"epoch": 37.68584356819651,
"grad_norm": 1.4864155054092407,
"learning_rate": 0.001,
"loss": 1.7471,
"step": 116600
},
{
"epoch": 37.718164188752425,
"grad_norm": 1.328788161277771,
"learning_rate": 0.001,
"loss": 1.7511,
"step": 116700
},
{
"epoch": 37.75048480930834,
"grad_norm": 1.7826188802719116,
"learning_rate": 0.001,
"loss": 1.7506,
"step": 116800
},
{
"epoch": 37.782805429864254,
"grad_norm": 1.579100489616394,
"learning_rate": 0.001,
"loss": 1.7819,
"step": 116900
},
{
"epoch": 37.81512605042017,
"grad_norm": 1.4960875511169434,
"learning_rate": 0.001,
"loss": 1.7827,
"step": 117000
},
{
"epoch": 37.84744667097608,
"grad_norm": 1.3491441011428833,
"learning_rate": 0.001,
"loss": 1.7732,
"step": 117100
},
{
"epoch": 37.879767291532,
"grad_norm": 1.6198680400848389,
"learning_rate": 0.001,
"loss": 1.7596,
"step": 117200
},
{
"epoch": 37.91208791208791,
"grad_norm": 1.4920895099639893,
"learning_rate": 0.001,
"loss": 1.7824,
"step": 117300
},
{
"epoch": 37.94440853264383,
"grad_norm": 1.17668879032135,
"learning_rate": 0.001,
"loss": 1.7766,
"step": 117400
},
{
"epoch": 37.97672915319974,
"grad_norm": 1.381245493888855,
"learning_rate": 0.001,
"loss": 1.7859,
"step": 117500
},
{
"epoch": 38.009049773755656,
"grad_norm": 1.8120436668395996,
"learning_rate": 0.001,
"loss": 1.7456,
"step": 117600
},
{
"epoch": 38.04137039431157,
"grad_norm": 1.2770622968673706,
"learning_rate": 0.001,
"loss": 1.6473,
"step": 117700
},
{
"epoch": 38.073691014867485,
"grad_norm": 1.495666265487671,
"learning_rate": 0.001,
"loss": 1.6665,
"step": 117800
},
{
"epoch": 38.1060116354234,
"grad_norm": 1.3319461345672607,
"learning_rate": 0.001,
"loss": 1.6721,
"step": 117900
},
{
"epoch": 38.138332255979314,
"grad_norm": 1.181028127670288,
"learning_rate": 0.001,
"loss": 1.6755,
"step": 118000
},
{
"epoch": 38.17065287653523,
"grad_norm": 1.7435935735702515,
"learning_rate": 0.001,
"loss": 1.6728,
"step": 118100
},
{
"epoch": 38.20297349709114,
"grad_norm": 1.2086992263793945,
"learning_rate": 0.001,
"loss": 1.6858,
"step": 118200
},
{
"epoch": 38.23529411764706,
"grad_norm": 1.749821662902832,
"learning_rate": 0.001,
"loss": 1.692,
"step": 118300
},
{
"epoch": 38.26761473820297,
"grad_norm": 1.2589175701141357,
"learning_rate": 0.001,
"loss": 1.6806,
"step": 118400
},
{
"epoch": 38.29993535875889,
"grad_norm": 1.6938502788543701,
"learning_rate": 0.001,
"loss": 1.7078,
"step": 118500
},
{
"epoch": 38.3322559793148,
"grad_norm": 1.2834974527359009,
"learning_rate": 0.001,
"loss": 1.7057,
"step": 118600
},
{
"epoch": 38.364576599870716,
"grad_norm": 1.4331868886947632,
"learning_rate": 0.001,
"loss": 1.7216,
"step": 118700
},
{
"epoch": 38.39689722042663,
"grad_norm": 1.5875442028045654,
"learning_rate": 0.001,
"loss": 1.718,
"step": 118800
},
{
"epoch": 38.429217840982545,
"grad_norm": 1.463834285736084,
"learning_rate": 0.001,
"loss": 1.7014,
"step": 118900
},
{
"epoch": 38.46153846153846,
"grad_norm": 1.7719365358352661,
"learning_rate": 0.001,
"loss": 1.7055,
"step": 119000
},
{
"epoch": 38.493859082094374,
"grad_norm": 1.751819372177124,
"learning_rate": 0.001,
"loss": 1.7061,
"step": 119100
},
{
"epoch": 38.52617970265029,
"grad_norm": 1.264541506767273,
"learning_rate": 0.001,
"loss": 1.7394,
"step": 119200
},
{
"epoch": 38.558500323206204,
"grad_norm": 1.691881537437439,
"learning_rate": 0.001,
"loss": 1.7198,
"step": 119300
},
{
"epoch": 38.59082094376212,
"grad_norm": 1.7032718658447266,
"learning_rate": 0.001,
"loss": 1.736,
"step": 119400
},
{
"epoch": 38.62314156431803,
"grad_norm": 1.6821051836013794,
"learning_rate": 0.001,
"loss": 1.7352,
"step": 119500
},
{
"epoch": 38.65546218487395,
"grad_norm": 1.8984935283660889,
"learning_rate": 0.001,
"loss": 1.7463,
"step": 119600
},
{
"epoch": 38.68778280542986,
"grad_norm": 1.8338229656219482,
"learning_rate": 0.001,
"loss": 1.7445,
"step": 119700
},
{
"epoch": 38.720103425985776,
"grad_norm": 1.3673371076583862,
"learning_rate": 0.001,
"loss": 1.7467,
"step": 119800
},
{
"epoch": 38.75242404654169,
"grad_norm": 1.401082992553711,
"learning_rate": 0.001,
"loss": 1.7207,
"step": 119900
},
{
"epoch": 38.784744667097605,
"grad_norm": 1.3542741537094116,
"learning_rate": 0.001,
"loss": 1.7632,
"step": 120000
},
{
"epoch": 38.81706528765352,
"grad_norm": 1.6251832246780396,
"learning_rate": 0.001,
"loss": 1.751,
"step": 120100
},
{
"epoch": 38.849385908209435,
"grad_norm": 1.4141253232955933,
"learning_rate": 0.001,
"loss": 1.7618,
"step": 120200
},
{
"epoch": 38.88170652876535,
"grad_norm": 1.387152075767517,
"learning_rate": 0.001,
"loss": 1.7421,
"step": 120300
},
{
"epoch": 38.914027149321264,
"grad_norm": 2.0019166469573975,
"learning_rate": 0.001,
"loss": 1.7439,
"step": 120400
},
{
"epoch": 38.94634776987718,
"grad_norm": 1.5084810256958008,
"learning_rate": 0.001,
"loss": 1.7447,
"step": 120500
},
{
"epoch": 38.97866839043309,
"grad_norm": 1.3848720788955688,
"learning_rate": 0.001,
"loss": 1.7649,
"step": 120600
},
{
"epoch": 39.010989010989015,
"grad_norm": 1.547417402267456,
"learning_rate": 0.001,
"loss": 1.7236,
"step": 120700
},
{
"epoch": 39.04330963154493,
"grad_norm": 1.2557010650634766,
"learning_rate": 0.001,
"loss": 1.6258,
"step": 120800
},
{
"epoch": 39.075630252100844,
"grad_norm": 1.5394682884216309,
"learning_rate": 0.001,
"loss": 1.6437,
"step": 120900
},
{
"epoch": 39.10795087265676,
"grad_norm": 1.2757493257522583,
"learning_rate": 0.001,
"loss": 1.6441,
"step": 121000
},
{
"epoch": 39.14027149321267,
"grad_norm": 1.468744158744812,
"learning_rate": 0.001,
"loss": 1.6495,
"step": 121100
},
{
"epoch": 39.17259211376859,
"grad_norm": 1.1792622804641724,
"learning_rate": 0.001,
"loss": 1.6726,
"step": 121200
},
{
"epoch": 39.2049127343245,
"grad_norm": 1.330483078956604,
"learning_rate": 0.001,
"loss": 1.6983,
"step": 121300
},
{
"epoch": 39.237233354880416,
"grad_norm": 1.6947458982467651,
"learning_rate": 0.001,
"loss": 1.6798,
"step": 121400
},
{
"epoch": 39.26955397543633,
"grad_norm": 1.8113237619400024,
"learning_rate": 0.001,
"loss": 1.6815,
"step": 121500
},
{
"epoch": 39.301874595992246,
"grad_norm": 1.3059000968933105,
"learning_rate": 0.001,
"loss": 1.68,
"step": 121600
},
{
"epoch": 39.33419521654816,
"grad_norm": 2.113922119140625,
"learning_rate": 0.001,
"loss": 1.6893,
"step": 121700
},
{
"epoch": 39.366515837104075,
"grad_norm": 1.4525998830795288,
"learning_rate": 0.001,
"loss": 1.6622,
"step": 121800
},
{
"epoch": 39.39883645765999,
"grad_norm": 1.6511586904525757,
"learning_rate": 0.001,
"loss": 1.6923,
"step": 121900
},
{
"epoch": 39.431157078215904,
"grad_norm": 1.427304744720459,
"learning_rate": 0.001,
"loss": 1.6679,
"step": 122000
},
{
"epoch": 39.46347769877182,
"grad_norm": 1.3397293090820312,
"learning_rate": 0.001,
"loss": 1.7052,
"step": 122100
},
{
"epoch": 39.49579831932773,
"grad_norm": 1.3981614112854004,
"learning_rate": 0.001,
"loss": 1.697,
"step": 122200
},
{
"epoch": 39.52811893988365,
"grad_norm": 1.4614068269729614,
"learning_rate": 0.001,
"loss": 1.7255,
"step": 122300
},
{
"epoch": 39.56043956043956,
"grad_norm": 1.6252069473266602,
"learning_rate": 0.001,
"loss": 1.7105,
"step": 122400
},
{
"epoch": 39.59276018099548,
"grad_norm": 1.5491560697555542,
"learning_rate": 0.001,
"loss": 1.7287,
"step": 122500
},
{
"epoch": 39.62508080155139,
"grad_norm": 1.5289865732192993,
"learning_rate": 0.001,
"loss": 1.7312,
"step": 122600
},
{
"epoch": 39.657401422107306,
"grad_norm": 1.3649866580963135,
"learning_rate": 0.001,
"loss": 1.7001,
"step": 122700
},
{
"epoch": 39.68972204266322,
"grad_norm": 1.8227938413619995,
"learning_rate": 0.001,
"loss": 1.7325,
"step": 122800
},
{
"epoch": 39.722042663219135,
"grad_norm": 1.4305684566497803,
"learning_rate": 0.001,
"loss": 1.7297,
"step": 122900
},
{
"epoch": 39.75436328377505,
"grad_norm": 2.0663459300994873,
"learning_rate": 0.001,
"loss": 1.7248,
"step": 123000
},
{
"epoch": 39.786683904330964,
"grad_norm": 1.6483802795410156,
"learning_rate": 0.001,
"loss": 1.7143,
"step": 123100
},
{
"epoch": 39.81900452488688,
"grad_norm": 1.3471261262893677,
"learning_rate": 0.001,
"loss": 1.7331,
"step": 123200
},
{
"epoch": 39.85132514544279,
"grad_norm": 1.501815676689148,
"learning_rate": 0.001,
"loss": 1.7242,
"step": 123300
},
{
"epoch": 39.88364576599871,
"grad_norm": 2.0373356342315674,
"learning_rate": 0.001,
"loss": 1.7363,
"step": 123400
},
{
"epoch": 39.91596638655462,
"grad_norm": 1.8736695051193237,
"learning_rate": 0.001,
"loss": 1.727,
"step": 123500
},
{
"epoch": 39.94828700711054,
"grad_norm": 1.5386576652526855,
"learning_rate": 0.001,
"loss": 1.7602,
"step": 123600
},
{
"epoch": 39.98060762766645,
"grad_norm": 1.287049412727356,
"learning_rate": 0.001,
"loss": 1.7529,
"step": 123700
},
{
"epoch": 40.012928248222366,
"grad_norm": 1.3691580295562744,
"learning_rate": 0.001,
"loss": 1.6845,
"step": 123800
},
{
"epoch": 40.04524886877828,
"grad_norm": 1.5578200817108154,
"learning_rate": 0.001,
"loss": 1.624,
"step": 123900
},
{
"epoch": 40.077569489334195,
"grad_norm": 1.394068956375122,
"learning_rate": 0.001,
"loss": 1.6299,
"step": 124000
},
{
"epoch": 40.10989010989011,
"grad_norm": 1.5257561206817627,
"learning_rate": 0.001,
"loss": 1.6362,
"step": 124100
},
{
"epoch": 40.142210730446024,
"grad_norm": 1.6746106147766113,
"learning_rate": 0.001,
"loss": 1.6353,
"step": 124200
},
{
"epoch": 40.17453135100194,
"grad_norm": 1.3863023519515991,
"learning_rate": 0.001,
"loss": 1.6562,
"step": 124300
},
{
"epoch": 40.20685197155785,
"grad_norm": 1.9362484216690063,
"learning_rate": 0.001,
"loss": 1.6472,
"step": 124400
},
{
"epoch": 40.23917259211377,
"grad_norm": 1.3526794910430908,
"learning_rate": 0.001,
"loss": 1.6398,
"step": 124500
},
{
"epoch": 40.27149321266968,
"grad_norm": 1.2201699018478394,
"learning_rate": 0.001,
"loss": 1.6627,
"step": 124600
},
{
"epoch": 40.3038138332256,
"grad_norm": 1.4147565364837646,
"learning_rate": 0.001,
"loss": 1.653,
"step": 124700
},
{
"epoch": 40.33613445378151,
"grad_norm": 1.7582532167434692,
"learning_rate": 0.001,
"loss": 1.6675,
"step": 124800
},
{
"epoch": 40.368455074337426,
"grad_norm": 1.7984539270401,
"learning_rate": 0.001,
"loss": 1.6676,
"step": 124900
},
{
"epoch": 40.40077569489334,
"grad_norm": 1.4224402904510498,
"learning_rate": 0.001,
"loss": 1.6743,
"step": 125000
},
{
"epoch": 40.433096315449255,
"grad_norm": 1.6389150619506836,
"learning_rate": 0.001,
"loss": 1.6881,
"step": 125100
},
{
"epoch": 40.46541693600517,
"grad_norm": 1.8313899040222168,
"learning_rate": 0.001,
"loss": 1.6708,
"step": 125200
},
{
"epoch": 40.497737556561084,
"grad_norm": 2.3224472999572754,
"learning_rate": 0.001,
"loss": 1.6873,
"step": 125300
},
{
"epoch": 40.530058177117,
"grad_norm": 1.3956483602523804,
"learning_rate": 0.001,
"loss": 1.6993,
"step": 125400
},
{
"epoch": 40.56237879767291,
"grad_norm": 1.9109262228012085,
"learning_rate": 0.001,
"loss": 1.6995,
"step": 125500
},
{
"epoch": 40.59469941822883,
"grad_norm": 1.7382196187973022,
"learning_rate": 0.001,
"loss": 1.6979,
"step": 125600
},
{
"epoch": 40.62702003878474,
"grad_norm": 1.6517763137817383,
"learning_rate": 0.001,
"loss": 1.6993,
"step": 125700
},
{
"epoch": 40.65934065934066,
"grad_norm": 1.7835556268692017,
"learning_rate": 0.001,
"loss": 1.6925,
"step": 125800
},
{
"epoch": 40.69166127989657,
"grad_norm": 2.0330710411071777,
"learning_rate": 0.001,
"loss": 1.7008,
"step": 125900
},
{
"epoch": 40.723981900452486,
"grad_norm": 1.426033854484558,
"learning_rate": 0.001,
"loss": 1.7268,
"step": 126000
},
{
"epoch": 40.7563025210084,
"grad_norm": 1.7070313692092896,
"learning_rate": 0.001,
"loss": 1.7111,
"step": 126100
},
{
"epoch": 40.788623141564315,
"grad_norm": 2.436343193054199,
"learning_rate": 0.001,
"loss": 1.7108,
"step": 126200
},
{
"epoch": 40.82094376212023,
"grad_norm": 1.3666476011276245,
"learning_rate": 0.001,
"loss": 1.7239,
"step": 126300
},
{
"epoch": 40.853264382676144,
"grad_norm": 1.7341768741607666,
"learning_rate": 0.001,
"loss": 1.7275,
"step": 126400
},
{
"epoch": 40.88558500323206,
"grad_norm": 1.452822208404541,
"learning_rate": 0.001,
"loss": 1.7243,
"step": 126500
},
{
"epoch": 40.91790562378797,
"grad_norm": 1.5342082977294922,
"learning_rate": 0.001,
"loss": 1.7287,
"step": 126600
},
{
"epoch": 40.95022624434389,
"grad_norm": 1.5138624906539917,
"learning_rate": 0.001,
"loss": 1.7283,
"step": 126700
},
{
"epoch": 40.9825468648998,
"grad_norm": 1.7864232063293457,
"learning_rate": 0.001,
"loss": 1.7306,
"step": 126800
},
{
"epoch": 41.014867485455724,
"grad_norm": 1.4858065843582153,
"learning_rate": 0.001,
"loss": 1.6671,
"step": 126900
},
{
"epoch": 41.04718810601164,
"grad_norm": 2.3251335620880127,
"learning_rate": 0.001,
"loss": 1.6027,
"step": 127000
},
{
"epoch": 41.07950872656755,
"grad_norm": 2.122349262237549,
"learning_rate": 0.001,
"loss": 1.6185,
"step": 127100
},
{
"epoch": 41.11182934712347,
"grad_norm": 1.4440858364105225,
"learning_rate": 0.001,
"loss": 1.6143,
"step": 127200
},
{
"epoch": 41.14414996767938,
"grad_norm": 1.6865544319152832,
"learning_rate": 0.001,
"loss": 1.6151,
"step": 127300
},
{
"epoch": 41.1764705882353,
"grad_norm": 1.3845254182815552,
"learning_rate": 0.001,
"loss": 1.6216,
"step": 127400
},
{
"epoch": 41.20879120879121,
"grad_norm": 1.5724061727523804,
"learning_rate": 0.001,
"loss": 1.6334,
"step": 127500
},
{
"epoch": 41.241111829347126,
"grad_norm": 1.6744935512542725,
"learning_rate": 0.001,
"loss": 1.6441,
"step": 127600
},
{
"epoch": 41.27343244990304,
"grad_norm": 1.6147488355636597,
"learning_rate": 0.001,
"loss": 1.6485,
"step": 127700
},
{
"epoch": 41.305753070458955,
"grad_norm": 1.930953025817871,
"learning_rate": 0.001,
"loss": 1.6538,
"step": 127800
},
{
"epoch": 41.33807369101487,
"grad_norm": 1.6779670715332031,
"learning_rate": 0.001,
"loss": 1.6641,
"step": 127900
},
{
"epoch": 41.370394311570784,
"grad_norm": 2.094191074371338,
"learning_rate": 0.001,
"loss": 1.638,
"step": 128000
},
{
"epoch": 41.4027149321267,
"grad_norm": 1.5966159105300903,
"learning_rate": 0.001,
"loss": 1.6603,
"step": 128100
},
{
"epoch": 41.43503555268261,
"grad_norm": 1.9661147594451904,
"learning_rate": 0.001,
"loss": 1.6588,
"step": 128200
},
{
"epoch": 41.46735617323853,
"grad_norm": 2.126356363296509,
"learning_rate": 0.001,
"loss": 1.6594,
"step": 128300
},
{
"epoch": 41.49967679379444,
"grad_norm": 1.5801960229873657,
"learning_rate": 0.001,
"loss": 1.6669,
"step": 128400
},
{
"epoch": 41.53199741435036,
"grad_norm": 1.6362019777297974,
"learning_rate": 0.001,
"loss": 1.6799,
"step": 128500
},
{
"epoch": 41.56431803490627,
"grad_norm": 1.570550799369812,
"learning_rate": 0.001,
"loss": 1.6684,
"step": 128600
},
{
"epoch": 41.596638655462186,
"grad_norm": 1.7719006538391113,
"learning_rate": 0.001,
"loss": 1.6921,
"step": 128700
},
{
"epoch": 41.6289592760181,
"grad_norm": 1.702435851097107,
"learning_rate": 0.001,
"loss": 1.6774,
"step": 128800
},
{
"epoch": 41.661279896574015,
"grad_norm": 1.8023364543914795,
"learning_rate": 0.001,
"loss": 1.71,
"step": 128900
},
{
"epoch": 41.69360051712993,
"grad_norm": 1.429628849029541,
"learning_rate": 0.001,
"loss": 1.6754,
"step": 129000
},
{
"epoch": 41.725921137685845,
"grad_norm": 1.6636650562286377,
"learning_rate": 0.001,
"loss": 1.6908,
"step": 129100
},
{
"epoch": 41.75824175824176,
"grad_norm": 1.9453206062316895,
"learning_rate": 0.001,
"loss": 1.7009,
"step": 129200
},
{
"epoch": 41.790562378797674,
"grad_norm": 1.7370038032531738,
"learning_rate": 0.001,
"loss": 1.7091,
"step": 129300
},
{
"epoch": 41.82288299935359,
"grad_norm": 1.8082547187805176,
"learning_rate": 0.001,
"loss": 1.7062,
"step": 129400
},
{
"epoch": 41.8552036199095,
"grad_norm": 1.7985368967056274,
"learning_rate": 0.001,
"loss": 1.6996,
"step": 129500
},
{
"epoch": 41.88752424046542,
"grad_norm": 1.752468466758728,
"learning_rate": 0.001,
"loss": 1.7066,
"step": 129600
},
{
"epoch": 41.91984486102133,
"grad_norm": 1.7616372108459473,
"learning_rate": 0.001,
"loss": 1.6999,
"step": 129700
},
{
"epoch": 41.95216548157725,
"grad_norm": 1.5394361019134521,
"learning_rate": 0.001,
"loss": 1.7093,
"step": 129800
},
{
"epoch": 41.98448610213316,
"grad_norm": 1.9267899990081787,
"learning_rate": 0.001,
"loss": 1.731,
"step": 129900
},
{
"epoch": 42.016806722689076,
"grad_norm": 1.9508917331695557,
"learning_rate": 0.001,
"loss": 1.6271,
"step": 130000
},
{
"epoch": 42.04912734324499,
"grad_norm": 1.8610914945602417,
"learning_rate": 0.001,
"loss": 1.5836,
"step": 130100
},
{
"epoch": 42.081447963800905,
"grad_norm": 1.9641599655151367,
"learning_rate": 0.001,
"loss": 1.6183,
"step": 130200
},
{
"epoch": 42.11376858435682,
"grad_norm": 1.5812021493911743,
"learning_rate": 0.001,
"loss": 1.5993,
"step": 130300
},
{
"epoch": 42.146089204912734,
"grad_norm": 2.2505440711975098,
"learning_rate": 0.001,
"loss": 1.6031,
"step": 130400
},
{
"epoch": 42.17840982546865,
"grad_norm": 1.7632458209991455,
"learning_rate": 0.001,
"loss": 1.6272,
"step": 130500
},
{
"epoch": 42.21073044602456,
"grad_norm": 1.9542481899261475,
"learning_rate": 0.001,
"loss": 1.6128,
"step": 130600
},
{
"epoch": 42.24305106658048,
"grad_norm": 2.813657283782959,
"learning_rate": 0.001,
"loss": 1.6198,
"step": 130700
},
{
"epoch": 42.27537168713639,
"grad_norm": 1.9514821767807007,
"learning_rate": 0.001,
"loss": 1.6152,
"step": 130800
},
{
"epoch": 42.30769230769231,
"grad_norm": 2.014662981033325,
"learning_rate": 0.001,
"loss": 1.6251,
"step": 130900
},
{
"epoch": 42.34001292824822,
"grad_norm": 2.202014446258545,
"learning_rate": 0.001,
"loss": 1.6405,
"step": 131000
},
{
"epoch": 42.372333548804136,
"grad_norm": 2.675868034362793,
"learning_rate": 0.001,
"loss": 1.6475,
"step": 131100
},
{
"epoch": 42.40465416936005,
"grad_norm": 2.102691888809204,
"learning_rate": 0.001,
"loss": 1.6568,
"step": 131200
},
{
"epoch": 42.436974789915965,
"grad_norm": 2.1783900260925293,
"learning_rate": 0.001,
"loss": 1.637,
"step": 131300
},
{
"epoch": 42.46929541047188,
"grad_norm": 1.476436734199524,
"learning_rate": 0.001,
"loss": 1.6476,
"step": 131400
},
{
"epoch": 42.501616031027794,
"grad_norm": 1.7832789421081543,
"learning_rate": 0.001,
"loss": 1.6625,
"step": 131500
},
{
"epoch": 42.53393665158371,
"grad_norm": 2.12839674949646,
"learning_rate": 0.001,
"loss": 1.6465,
"step": 131600
},
{
"epoch": 42.56625727213962,
"grad_norm": 1.5003304481506348,
"learning_rate": 0.001,
"loss": 1.6794,
"step": 131700
},
{
"epoch": 42.59857789269554,
"grad_norm": 1.5726544857025146,
"learning_rate": 0.001,
"loss": 1.6767,
"step": 131800
},
{
"epoch": 42.63089851325145,
"grad_norm": 2.2940423488616943,
"learning_rate": 0.001,
"loss": 1.6638,
"step": 131900
},
{
"epoch": 42.66321913380737,
"grad_norm": 1.992763638496399,
"learning_rate": 0.001,
"loss": 1.6713,
"step": 132000
},
{
"epoch": 42.69553975436328,
"grad_norm": 2.10316801071167,
"learning_rate": 0.001,
"loss": 1.6719,
"step": 132100
},
{
"epoch": 42.727860374919196,
"grad_norm": 1.1966867446899414,
"learning_rate": 0.001,
"loss": 1.6687,
"step": 132200
},
{
"epoch": 42.76018099547511,
"grad_norm": 1.6396232843399048,
"learning_rate": 0.001,
"loss": 1.6927,
"step": 132300
},
{
"epoch": 42.792501616031025,
"grad_norm": 2.2217977046966553,
"learning_rate": 0.001,
"loss": 1.6911,
"step": 132400
},
{
"epoch": 42.82482223658694,
"grad_norm": 1.3404631614685059,
"learning_rate": 0.001,
"loss": 1.6951,
"step": 132500
},
{
"epoch": 42.857142857142854,
"grad_norm": 2.155172348022461,
"learning_rate": 0.001,
"loss": 1.7034,
"step": 132600
},
{
"epoch": 42.88946347769877,
"grad_norm": 1.7725328207015991,
"learning_rate": 0.001,
"loss": 1.6975,
"step": 132700
},
{
"epoch": 42.92178409825468,
"grad_norm": 1.6837446689605713,
"learning_rate": 0.001,
"loss": 1.7219,
"step": 132800
},
{
"epoch": 42.9541047188106,
"grad_norm": 1.77871835231781,
"learning_rate": 0.001,
"loss": 1.7004,
"step": 132900
},
{
"epoch": 42.98642533936652,
"grad_norm": 1.9019044637680054,
"learning_rate": 0.001,
"loss": 1.7157,
"step": 133000
},
{
"epoch": 43.018745959922434,
"grad_norm": 1.3710943460464478,
"learning_rate": 0.001,
"loss": 1.6493,
"step": 133100
},
{
"epoch": 43.05106658047835,
"grad_norm": 1.582306146621704,
"learning_rate": 0.001,
"loss": 1.5689,
"step": 133200
},
{
"epoch": 43.08338720103426,
"grad_norm": 1.378099799156189,
"learning_rate": 0.001,
"loss": 1.5964,
"step": 133300
},
{
"epoch": 43.11570782159018,
"grad_norm": 1.4372540712356567,
"learning_rate": 0.001,
"loss": 1.5997,
"step": 133400
},
{
"epoch": 43.14802844214609,
"grad_norm": 1.6468878984451294,
"learning_rate": 0.001,
"loss": 1.5879,
"step": 133500
},
{
"epoch": 43.18034906270201,
"grad_norm": 1.480918049812317,
"learning_rate": 0.001,
"loss": 1.5902,
"step": 133600
},
{
"epoch": 43.21266968325792,
"grad_norm": 1.351524829864502,
"learning_rate": 0.001,
"loss": 1.6,
"step": 133700
},
{
"epoch": 43.244990303813836,
"grad_norm": 1.486530065536499,
"learning_rate": 0.001,
"loss": 1.5967,
"step": 133800
},
{
"epoch": 43.27731092436975,
"grad_norm": 1.9295352697372437,
"learning_rate": 0.001,
"loss": 1.6096,
"step": 133900
},
{
"epoch": 43.309631544925665,
"grad_norm": 1.7292808294296265,
"learning_rate": 0.001,
"loss": 1.6064,
"step": 134000
},
{
"epoch": 43.34195216548158,
"grad_norm": 1.8573253154754639,
"learning_rate": 0.001,
"loss": 1.6132,
"step": 134100
},
{
"epoch": 43.374272786037494,
"grad_norm": 1.320823311805725,
"learning_rate": 0.001,
"loss": 1.6472,
"step": 134200
},
{
"epoch": 43.40659340659341,
"grad_norm": 1.955378532409668,
"learning_rate": 0.001,
"loss": 1.6251,
"step": 134300
},
{
"epoch": 43.43891402714932,
"grad_norm": 1.4546319246292114,
"learning_rate": 0.001,
"loss": 1.6083,
"step": 134400
},
{
"epoch": 43.47123464770524,
"grad_norm": 2.06282639503479,
"learning_rate": 0.001,
"loss": 1.6473,
"step": 134500
},
{
"epoch": 43.50355526826115,
"grad_norm": 1.3789846897125244,
"learning_rate": 0.001,
"loss": 1.64,
"step": 134600
},
{
"epoch": 43.53587588881707,
"grad_norm": 1.3253904581069946,
"learning_rate": 0.001,
"loss": 1.643,
"step": 134700
},
{
"epoch": 43.56819650937298,
"grad_norm": 1.5497514009475708,
"learning_rate": 0.001,
"loss": 1.6411,
"step": 134800
},
{
"epoch": 43.600517129928896,
"grad_norm": 1.4692950248718262,
"learning_rate": 0.001,
"loss": 1.6484,
"step": 134900
},
{
"epoch": 43.63283775048481,
"grad_norm": 1.3008520603179932,
"learning_rate": 0.001,
"loss": 1.6697,
"step": 135000
},
{
"epoch": 43.665158371040725,
"grad_norm": 1.8452352285385132,
"learning_rate": 0.001,
"loss": 1.6622,
"step": 135100
},
{
"epoch": 43.69747899159664,
"grad_norm": 1.386189341545105,
"learning_rate": 0.001,
"loss": 1.6648,
"step": 135200
},
{
"epoch": 43.729799612152554,
"grad_norm": 1.5888075828552246,
"learning_rate": 0.001,
"loss": 1.6699,
"step": 135300
},
{
"epoch": 43.76212023270847,
"grad_norm": 1.8917977809906006,
"learning_rate": 0.001,
"loss": 1.6639,
"step": 135400
},
{
"epoch": 43.79444085326438,
"grad_norm": 1.619673252105713,
"learning_rate": 0.001,
"loss": 1.67,
"step": 135500
},
{
"epoch": 43.8267614738203,
"grad_norm": 1.4577540159225464,
"learning_rate": 0.001,
"loss": 1.6833,
"step": 135600
},
{
"epoch": 43.85908209437621,
"grad_norm": 1.8329697847366333,
"learning_rate": 0.001,
"loss": 1.6777,
"step": 135700
},
{
"epoch": 43.89140271493213,
"grad_norm": 1.7369554042816162,
"learning_rate": 0.001,
"loss": 1.6815,
"step": 135800
},
{
"epoch": 43.92372333548804,
"grad_norm": 1.4946892261505127,
"learning_rate": 0.001,
"loss": 1.6964,
"step": 135900
},
{
"epoch": 43.956043956043956,
"grad_norm": 1.434802532196045,
"learning_rate": 0.001,
"loss": 1.695,
"step": 136000
},
{
"epoch": 43.98836457659987,
"grad_norm": 2.0379035472869873,
"learning_rate": 0.001,
"loss": 1.699,
"step": 136100
},
{
"epoch": 44.020685197155785,
"grad_norm": 1.2588499784469604,
"learning_rate": 0.001,
"loss": 1.6163,
"step": 136200
},
{
"epoch": 44.0530058177117,
"grad_norm": 1.443988561630249,
"learning_rate": 0.001,
"loss": 1.5757,
"step": 136300
},
{
"epoch": 44.085326438267614,
"grad_norm": 1.7067530155181885,
"learning_rate": 0.001,
"loss": 1.5694,
"step": 136400
},
{
"epoch": 44.11764705882353,
"grad_norm": 1.8109402656555176,
"learning_rate": 0.001,
"loss": 1.5745,
"step": 136500
},
{
"epoch": 44.14996767937944,
"grad_norm": 1.415626883506775,
"learning_rate": 0.001,
"loss": 1.5756,
"step": 136600
},
{
"epoch": 44.18228829993536,
"grad_norm": 1.2374643087387085,
"learning_rate": 0.001,
"loss": 1.5818,
"step": 136700
},
{
"epoch": 44.21460892049127,
"grad_norm": 1.1942877769470215,
"learning_rate": 0.001,
"loss": 1.6047,
"step": 136800
},
{
"epoch": 44.24692954104719,
"grad_norm": 1.377187967300415,
"learning_rate": 0.001,
"loss": 1.5992,
"step": 136900
},
{
"epoch": 44.2792501616031,
"grad_norm": 1.3861560821533203,
"learning_rate": 0.001,
"loss": 1.6016,
"step": 137000
},
{
"epoch": 44.311570782159016,
"grad_norm": 1.5110067129135132,
"learning_rate": 0.001,
"loss": 1.612,
"step": 137100
},
{
"epoch": 44.34389140271493,
"grad_norm": 1.3622841835021973,
"learning_rate": 0.001,
"loss": 1.6092,
"step": 137200
},
{
"epoch": 44.376212023270845,
"grad_norm": 1.599679708480835,
"learning_rate": 0.001,
"loss": 1.6228,
"step": 137300
},
{
"epoch": 44.40853264382676,
"grad_norm": 1.0782546997070312,
"learning_rate": 0.001,
"loss": 1.6112,
"step": 137400
},
{
"epoch": 44.440853264382675,
"grad_norm": 1.423081398010254,
"learning_rate": 0.001,
"loss": 1.6277,
"step": 137500
},
{
"epoch": 44.47317388493859,
"grad_norm": 1.9224975109100342,
"learning_rate": 0.001,
"loss": 1.6241,
"step": 137600
},
{
"epoch": 44.505494505494504,
"grad_norm": 1.9734402894973755,
"learning_rate": 0.001,
"loss": 1.6014,
"step": 137700
},
{
"epoch": 44.53781512605042,
"grad_norm": 1.341483235359192,
"learning_rate": 0.001,
"loss": 1.6249,
"step": 137800
},
{
"epoch": 44.57013574660633,
"grad_norm": 1.2303094863891602,
"learning_rate": 0.001,
"loss": 1.6322,
"step": 137900
},
{
"epoch": 44.60245636716225,
"grad_norm": 1.8994379043579102,
"learning_rate": 0.001,
"loss": 1.65,
"step": 138000
},
{
"epoch": 44.63477698771816,
"grad_norm": 1.8510922193527222,
"learning_rate": 0.001,
"loss": 1.6313,
"step": 138100
},
{
"epoch": 44.66709760827408,
"grad_norm": 1.5857359170913696,
"learning_rate": 0.001,
"loss": 1.6423,
"step": 138200
},
{
"epoch": 44.69941822882999,
"grad_norm": 1.4746818542480469,
"learning_rate": 0.001,
"loss": 1.6534,
"step": 138300
},
{
"epoch": 44.731738849385906,
"grad_norm": 1.5729411840438843,
"learning_rate": 0.001,
"loss": 1.6496,
"step": 138400
},
{
"epoch": 44.76405946994182,
"grad_norm": 1.5360504388809204,
"learning_rate": 0.001,
"loss": 1.6552,
"step": 138500
},
{
"epoch": 44.796380090497735,
"grad_norm": 1.364283800125122,
"learning_rate": 0.001,
"loss": 1.6658,
"step": 138600
},
{
"epoch": 44.82870071105365,
"grad_norm": 1.3980382680892944,
"learning_rate": 0.001,
"loss": 1.6547,
"step": 138700
},
{
"epoch": 44.861021331609564,
"grad_norm": 1.4113725423812866,
"learning_rate": 0.001,
"loss": 1.6739,
"step": 138800
},
{
"epoch": 44.89334195216548,
"grad_norm": 1.829128623008728,
"learning_rate": 0.001,
"loss": 1.6694,
"step": 138900
},
{
"epoch": 44.92566257272139,
"grad_norm": 1.3401802778244019,
"learning_rate": 0.001,
"loss": 1.6741,
"step": 139000
},
{
"epoch": 44.95798319327731,
"grad_norm": 1.678005337715149,
"learning_rate": 0.001,
"loss": 1.6568,
"step": 139100
},
{
"epoch": 44.99030381383322,
"grad_norm": 1.1385924816131592,
"learning_rate": 0.001,
"loss": 1.6321,
"step": 139200
},
{
"epoch": 45.022624434389144,
"grad_norm": 1.8373512029647827,
"learning_rate": 0.001,
"loss": 1.6038,
"step": 139300
},
{
"epoch": 45.05494505494506,
"grad_norm": 1.5865776538848877,
"learning_rate": 0.001,
"loss": 1.5343,
"step": 139400
},
{
"epoch": 45.08726567550097,
"grad_norm": 1.5001276731491089,
"learning_rate": 0.001,
"loss": 1.5549,
"step": 139500
},
{
"epoch": 45.11958629605689,
"grad_norm": 1.2797919511795044,
"learning_rate": 0.001,
"loss": 1.5443,
"step": 139600
},
{
"epoch": 45.1519069166128,
"grad_norm": 1.455429196357727,
"learning_rate": 0.001,
"loss": 1.5619,
"step": 139700
},
{
"epoch": 45.18422753716872,
"grad_norm": 1.4949995279312134,
"learning_rate": 0.001,
"loss": 1.5724,
"step": 139800
},
{
"epoch": 45.21654815772463,
"grad_norm": 1.2570242881774902,
"learning_rate": 0.001,
"loss": 1.5687,
"step": 139900
},
{
"epoch": 45.248868778280546,
"grad_norm": 1.2480852603912354,
"learning_rate": 0.001,
"loss": 1.5789,
"step": 140000
},
{
"epoch": 45.28118939883646,
"grad_norm": 1.5493923425674438,
"learning_rate": 0.001,
"loss": 1.5761,
"step": 140100
},
{
"epoch": 45.313510019392375,
"grad_norm": 1.478066086769104,
"learning_rate": 0.001,
"loss": 1.5899,
"step": 140200
},
{
"epoch": 45.34583063994829,
"grad_norm": 1.2514175176620483,
"learning_rate": 0.001,
"loss": 1.5733,
"step": 140300
},
{
"epoch": 45.378151260504204,
"grad_norm": 1.3240177631378174,
"learning_rate": 0.001,
"loss": 1.6065,
"step": 140400
},
{
"epoch": 45.41047188106012,
"grad_norm": 1.09498929977417,
"learning_rate": 0.001,
"loss": 1.6061,
"step": 140500
},
{
"epoch": 45.44279250161603,
"grad_norm": 1.4801580905914307,
"learning_rate": 0.001,
"loss": 1.6009,
"step": 140600
},
{
"epoch": 45.47511312217195,
"grad_norm": 1.4095863103866577,
"learning_rate": 0.001,
"loss": 1.606,
"step": 140700
},
{
"epoch": 45.50743374272786,
"grad_norm": 1.3383103609085083,
"learning_rate": 0.001,
"loss": 1.5947,
"step": 140800
},
{
"epoch": 45.53975436328378,
"grad_norm": 1.3287218809127808,
"learning_rate": 0.001,
"loss": 1.6218,
"step": 140900
},
{
"epoch": 45.57207498383969,
"grad_norm": 1.1489287614822388,
"learning_rate": 0.001,
"loss": 1.6112,
"step": 141000
},
{
"epoch": 45.604395604395606,
"grad_norm": 1.3265559673309326,
"learning_rate": 0.001,
"loss": 1.6198,
"step": 141100
},
{
"epoch": 45.63671622495152,
"grad_norm": 1.4024713039398193,
"learning_rate": 0.001,
"loss": 1.6408,
"step": 141200
},
{
"epoch": 45.669036845507435,
"grad_norm": 1.2570629119873047,
"learning_rate": 0.001,
"loss": 1.6241,
"step": 141300
},
{
"epoch": 45.70135746606335,
"grad_norm": 1.234506607055664,
"learning_rate": 0.001,
"loss": 1.6337,
"step": 141400
},
{
"epoch": 45.733678086619264,
"grad_norm": 1.523465871810913,
"learning_rate": 0.001,
"loss": 1.6442,
"step": 141500
},
{
"epoch": 45.76599870717518,
"grad_norm": 1.3609179258346558,
"learning_rate": 0.001,
"loss": 1.6457,
"step": 141600
},
{
"epoch": 45.79831932773109,
"grad_norm": 2.108837604522705,
"learning_rate": 0.001,
"loss": 1.6568,
"step": 141700
},
{
"epoch": 45.83063994828701,
"grad_norm": 1.3636060953140259,
"learning_rate": 0.001,
"loss": 1.6442,
"step": 141800
},
{
"epoch": 45.86296056884292,
"grad_norm": 1.3937108516693115,
"learning_rate": 0.001,
"loss": 1.6542,
"step": 141900
},
{
"epoch": 45.89528118939884,
"grad_norm": 1.1740082502365112,
"learning_rate": 0.001,
"loss": 1.6613,
"step": 142000
},
{
"epoch": 45.92760180995475,
"grad_norm": 1.1534056663513184,
"learning_rate": 0.001,
"loss": 1.6553,
"step": 142100
},
{
"epoch": 45.959922430510666,
"grad_norm": 1.6877473592758179,
"learning_rate": 0.001,
"loss": 1.6476,
"step": 142200
},
{
"epoch": 45.99224305106658,
"grad_norm": 1.163017988204956,
"learning_rate": 0.001,
"loss": 1.6809,
"step": 142300
},
{
"epoch": 46.024563671622495,
"grad_norm": 1.32700777053833,
"learning_rate": 0.001,
"loss": 1.5725,
"step": 142400
},
{
"epoch": 46.05688429217841,
"grad_norm": 1.7165447473526,
"learning_rate": 0.001,
"loss": 1.5315,
"step": 142500
},
{
"epoch": 46.089204912734324,
"grad_norm": 1.582828402519226,
"learning_rate": 0.001,
"loss": 1.542,
"step": 142600
},
{
"epoch": 46.12152553329024,
"grad_norm": 1.316243290901184,
"learning_rate": 0.001,
"loss": 1.5475,
"step": 142700
},
{
"epoch": 46.15384615384615,
"grad_norm": 1.4395389556884766,
"learning_rate": 0.001,
"loss": 1.5609,
"step": 142800
},
{
"epoch": 46.18616677440207,
"grad_norm": 1.7110323905944824,
"learning_rate": 0.001,
"loss": 1.5623,
"step": 142900
},
{
"epoch": 46.21848739495798,
"grad_norm": 1.58608078956604,
"learning_rate": 0.001,
"loss": 1.5753,
"step": 143000
},
{
"epoch": 46.2508080155139,
"grad_norm": 1.436051845550537,
"learning_rate": 0.001,
"loss": 1.5688,
"step": 143100
},
{
"epoch": 46.28312863606981,
"grad_norm": 1.8955553770065308,
"learning_rate": 0.001,
"loss": 1.5732,
"step": 143200
},
{
"epoch": 46.315449256625726,
"grad_norm": 1.6417391300201416,
"learning_rate": 0.001,
"loss": 1.5687,
"step": 143300
},
{
"epoch": 46.34776987718164,
"grad_norm": 1.4715759754180908,
"learning_rate": 0.001,
"loss": 1.568,
"step": 143400
},
{
"epoch": 46.380090497737555,
"grad_norm": 1.410852074623108,
"learning_rate": 0.001,
"loss": 1.5836,
"step": 143500
},
{
"epoch": 46.41241111829347,
"grad_norm": 1.6234686374664307,
"learning_rate": 0.001,
"loss": 1.5823,
"step": 143600
},
{
"epoch": 46.444731738849384,
"grad_norm": 1.5069587230682373,
"learning_rate": 0.001,
"loss": 1.5963,
"step": 143700
},
{
"epoch": 46.4770523594053,
"grad_norm": 1.4912326335906982,
"learning_rate": 0.001,
"loss": 1.6026,
"step": 143800
},
{
"epoch": 46.50937297996121,
"grad_norm": 2.0091121196746826,
"learning_rate": 0.001,
"loss": 1.5956,
"step": 143900
},
{
"epoch": 46.54169360051713,
"grad_norm": 1.2358951568603516,
"learning_rate": 0.001,
"loss": 1.5907,
"step": 144000
},
{
"epoch": 46.57401422107304,
"grad_norm": 1.4463825225830078,
"learning_rate": 0.001,
"loss": 1.5941,
"step": 144100
},
{
"epoch": 46.60633484162896,
"grad_norm": 1.6849217414855957,
"learning_rate": 0.001,
"loss": 1.6197,
"step": 144200
},
{
"epoch": 46.63865546218487,
"grad_norm": 1.4761693477630615,
"learning_rate": 0.001,
"loss": 1.6111,
"step": 144300
},
{
"epoch": 46.670976082740786,
"grad_norm": 1.9993269443511963,
"learning_rate": 0.001,
"loss": 1.6298,
"step": 144400
},
{
"epoch": 46.7032967032967,
"grad_norm": 1.3961095809936523,
"learning_rate": 0.001,
"loss": 1.6271,
"step": 144500
},
{
"epoch": 46.735617323852615,
"grad_norm": 1.3440773487091064,
"learning_rate": 0.001,
"loss": 1.6165,
"step": 144600
},
{
"epoch": 46.76793794440853,
"grad_norm": 1.7641288042068481,
"learning_rate": 0.001,
"loss": 1.6196,
"step": 144700
},
{
"epoch": 46.800258564964444,
"grad_norm": 2.0289649963378906,
"learning_rate": 0.001,
"loss": 1.6436,
"step": 144800
},
{
"epoch": 46.83257918552036,
"grad_norm": 1.651640772819519,
"learning_rate": 0.001,
"loss": 1.628,
"step": 144900
},
{
"epoch": 46.864899806076274,
"grad_norm": 1.3994982242584229,
"learning_rate": 0.001,
"loss": 1.6434,
"step": 145000
},
{
"epoch": 46.89722042663219,
"grad_norm": 1.3219499588012695,
"learning_rate": 0.001,
"loss": 1.6448,
"step": 145100
},
{
"epoch": 46.9295410471881,
"grad_norm": 1.4943182468414307,
"learning_rate": 0.001,
"loss": 1.6185,
"step": 145200
},
{
"epoch": 46.96186166774402,
"grad_norm": 1.338044285774231,
"learning_rate": 0.001,
"loss": 1.6472,
"step": 145300
},
{
"epoch": 46.99418228829994,
"grad_norm": 1.2849805355072021,
"learning_rate": 0.001,
"loss": 1.6408,
"step": 145400
},
{
"epoch": 47.02650290885585,
"grad_norm": 1.4934114217758179,
"learning_rate": 0.001,
"loss": 1.549,
"step": 145500
},
{
"epoch": 47.05882352941177,
"grad_norm": 1.7186708450317383,
"learning_rate": 0.001,
"loss": 1.5166,
"step": 145600
},
{
"epoch": 47.09114414996768,
"grad_norm": 1.237968921661377,
"learning_rate": 0.001,
"loss": 1.5266,
"step": 145700
},
{
"epoch": 47.1234647705236,
"grad_norm": 1.4882655143737793,
"learning_rate": 0.001,
"loss": 1.5339,
"step": 145800
},
{
"epoch": 47.15578539107951,
"grad_norm": 1.618738055229187,
"learning_rate": 0.001,
"loss": 1.5317,
"step": 145900
},
{
"epoch": 47.188106011635426,
"grad_norm": 2.116640090942383,
"learning_rate": 0.001,
"loss": 1.5316,
"step": 146000
},
{
"epoch": 47.22042663219134,
"grad_norm": 1.3013315200805664,
"learning_rate": 0.001,
"loss": 1.5584,
"step": 146100
},
{
"epoch": 47.252747252747255,
"grad_norm": 2.401732921600342,
"learning_rate": 0.001,
"loss": 1.5343,
"step": 146200
},
{
"epoch": 47.28506787330317,
"grad_norm": 1.1736658811569214,
"learning_rate": 0.001,
"loss": 1.5728,
"step": 146300
},
{
"epoch": 47.317388493859085,
"grad_norm": 1.4418681859970093,
"learning_rate": 0.001,
"loss": 1.5624,
"step": 146400
},
{
"epoch": 47.349709114415,
"grad_norm": 1.4937100410461426,
"learning_rate": 0.001,
"loss": 1.5712,
"step": 146500
},
{
"epoch": 47.382029734970914,
"grad_norm": 1.6570427417755127,
"learning_rate": 0.001,
"loss": 1.5789,
"step": 146600
},
{
"epoch": 47.41435035552683,
"grad_norm": 1.9760403633117676,
"learning_rate": 0.001,
"loss": 1.5811,
"step": 146700
},
{
"epoch": 47.44667097608274,
"grad_norm": 1.6386915445327759,
"learning_rate": 0.001,
"loss": 1.5614,
"step": 146800
},
{
"epoch": 47.47899159663866,
"grad_norm": 1.7690401077270508,
"learning_rate": 0.001,
"loss": 1.563,
"step": 146900
},
{
"epoch": 47.51131221719457,
"grad_norm": 1.3809525966644287,
"learning_rate": 0.001,
"loss": 1.5941,
"step": 147000
},
{
"epoch": 47.543632837750486,
"grad_norm": 1.2929129600524902,
"learning_rate": 0.001,
"loss": 1.6022,
"step": 147100
},
{
"epoch": 47.5759534583064,
"grad_norm": 1.6390870809555054,
"learning_rate": 0.001,
"loss": 1.5974,
"step": 147200
},
{
"epoch": 47.608274078862316,
"grad_norm": 1.5908600091934204,
"learning_rate": 0.001,
"loss": 1.605,
"step": 147300
},
{
"epoch": 47.64059469941823,
"grad_norm": 1.6766973733901978,
"learning_rate": 0.001,
"loss": 1.6013,
"step": 147400
},
{
"epoch": 47.672915319974145,
"grad_norm": 1.483106017112732,
"learning_rate": 0.001,
"loss": 1.6074,
"step": 147500
},
{
"epoch": 47.70523594053006,
"grad_norm": 1.9718176126480103,
"learning_rate": 0.001,
"loss": 1.5976,
"step": 147600
},
{
"epoch": 47.737556561085974,
"grad_norm": 1.3451755046844482,
"learning_rate": 0.001,
"loss": 1.607,
"step": 147700
},
{
"epoch": 47.76987718164189,
"grad_norm": 1.6685712337493896,
"learning_rate": 0.001,
"loss": 1.6157,
"step": 147800
},
{
"epoch": 47.8021978021978,
"grad_norm": 1.4344922304153442,
"learning_rate": 0.001,
"loss": 1.6238,
"step": 147900
},
{
"epoch": 47.83451842275372,
"grad_norm": 1.1716681718826294,
"learning_rate": 0.001,
"loss": 1.613,
"step": 148000
},
{
"epoch": 47.86683904330963,
"grad_norm": 1.5936914682388306,
"learning_rate": 0.001,
"loss": 1.622,
"step": 148100
},
{
"epoch": 47.89915966386555,
"grad_norm": 1.8645175695419312,
"learning_rate": 0.001,
"loss": 1.6352,
"step": 148200
},
{
"epoch": 47.93148028442146,
"grad_norm": 1.4023220539093018,
"learning_rate": 0.001,
"loss": 1.6131,
"step": 148300
},
{
"epoch": 47.963800904977376,
"grad_norm": 1.3344358205795288,
"learning_rate": 0.001,
"loss": 1.6347,
"step": 148400
},
{
"epoch": 47.99612152553329,
"grad_norm": 1.6620981693267822,
"learning_rate": 0.001,
"loss": 1.606,
"step": 148500
},
{
"epoch": 48.028442146089205,
"grad_norm": 1.550585389137268,
"learning_rate": 0.001,
"loss": 1.5141,
"step": 148600
},
{
"epoch": 48.06076276664512,
"grad_norm": 2.354085922241211,
"learning_rate": 0.001,
"loss": 1.5095,
"step": 148700
},
{
"epoch": 48.093083387201034,
"grad_norm": 1.6882182359695435,
"learning_rate": 0.001,
"loss": 1.5126,
"step": 148800
},
{
"epoch": 48.12540400775695,
"grad_norm": 1.8086196184158325,
"learning_rate": 0.001,
"loss": 1.5342,
"step": 148900
},
{
"epoch": 48.15772462831286,
"grad_norm": 1.2464903593063354,
"learning_rate": 0.001,
"loss": 1.5145,
"step": 149000
},
{
"epoch": 48.19004524886878,
"grad_norm": 1.5906083583831787,
"learning_rate": 0.001,
"loss": 1.5339,
"step": 149100
},
{
"epoch": 48.22236586942469,
"grad_norm": 1.7611628770828247,
"learning_rate": 0.001,
"loss": 1.5241,
"step": 149200
},
{
"epoch": 48.25468648998061,
"grad_norm": 1.8777153491973877,
"learning_rate": 0.001,
"loss": 1.5343,
"step": 149300
},
{
"epoch": 48.28700711053652,
"grad_norm": 1.7436455488204956,
"learning_rate": 0.001,
"loss": 1.5538,
"step": 149400
},
{
"epoch": 48.319327731092436,
"grad_norm": 1.5883677005767822,
"learning_rate": 0.001,
"loss": 1.5295,
"step": 149500
},
{
"epoch": 48.35164835164835,
"grad_norm": 1.8835762739181519,
"learning_rate": 0.001,
"loss": 1.5582,
"step": 149600
},
{
"epoch": 48.383968972204265,
"grad_norm": 1.5518923997879028,
"learning_rate": 0.001,
"loss": 1.5665,
"step": 149700
},
{
"epoch": 48.41628959276018,
"grad_norm": 1.4821540117263794,
"learning_rate": 0.001,
"loss": 1.5448,
"step": 149800
},
{
"epoch": 48.448610213316094,
"grad_norm": 1.5595805644989014,
"learning_rate": 0.001,
"loss": 1.577,
"step": 149900
},
{
"epoch": 48.48093083387201,
"grad_norm": 1.2543803453445435,
"learning_rate": 0.001,
"loss": 1.5539,
"step": 150000
},
{
"epoch": 48.51325145442792,
"grad_norm": 1.782151699066162,
"learning_rate": 0.001,
"loss": 1.5728,
"step": 150100
},
{
"epoch": 48.54557207498384,
"grad_norm": 1.6886430978775024,
"learning_rate": 0.001,
"loss": 1.5822,
"step": 150200
},
{
"epoch": 48.57789269553975,
"grad_norm": 1.2113796472549438,
"learning_rate": 0.001,
"loss": 1.5753,
"step": 150300
},
{
"epoch": 48.61021331609567,
"grad_norm": 1.291333794593811,
"learning_rate": 0.001,
"loss": 1.5937,
"step": 150400
},
{
"epoch": 48.64253393665158,
"grad_norm": 2.103907823562622,
"learning_rate": 0.001,
"loss": 1.5899,
"step": 150500
},
{
"epoch": 48.674854557207496,
"grad_norm": 1.4811424016952515,
"learning_rate": 0.001,
"loss": 1.5865,
"step": 150600
},
{
"epoch": 48.70717517776341,
"grad_norm": 1.5195327997207642,
"learning_rate": 0.001,
"loss": 1.598,
"step": 150700
},
{
"epoch": 48.739495798319325,
"grad_norm": 1.237692952156067,
"learning_rate": 0.001,
"loss": 1.5986,
"step": 150800
},
{
"epoch": 48.77181641887524,
"grad_norm": 1.5451624393463135,
"learning_rate": 0.001,
"loss": 1.6045,
"step": 150900
},
{
"epoch": 48.804137039431154,
"grad_norm": 1.2580842971801758,
"learning_rate": 0.001,
"loss": 1.6003,
"step": 151000
},
{
"epoch": 48.83645765998707,
"grad_norm": 1.950588345527649,
"learning_rate": 0.001,
"loss": 1.6034,
"step": 151100
},
{
"epoch": 48.86877828054298,
"grad_norm": 2.6268444061279297,
"learning_rate": 0.001,
"loss": 1.618,
"step": 151200
},
{
"epoch": 48.9010989010989,
"grad_norm": 1.5038946866989136,
"learning_rate": 0.001,
"loss": 1.6149,
"step": 151300
},
{
"epoch": 48.93341952165481,
"grad_norm": 1.7941875457763672,
"learning_rate": 0.001,
"loss": 1.6236,
"step": 151400
},
{
"epoch": 48.96574014221073,
"grad_norm": 1.6113241910934448,
"learning_rate": 0.001,
"loss": 1.6138,
"step": 151500
},
{
"epoch": 48.99806076276664,
"grad_norm": 2.102522134780884,
"learning_rate": 0.001,
"loss": 1.5985,
"step": 151600
},
{
"epoch": 49.03038138332256,
"grad_norm": 1.8292262554168701,
"learning_rate": 0.001,
"loss": 1.4972,
"step": 151700
},
{
"epoch": 49.06270200387848,
"grad_norm": 1.5576742887496948,
"learning_rate": 0.001,
"loss": 1.5018,
"step": 151800
},
{
"epoch": 49.09502262443439,
"grad_norm": 1.8177223205566406,
"learning_rate": 0.001,
"loss": 1.5011,
"step": 151900
},
{
"epoch": 49.12734324499031,
"grad_norm": 1.7465415000915527,
"learning_rate": 0.001,
"loss": 1.5105,
"step": 152000
},
{
"epoch": 49.15966386554622,
"grad_norm": 1.664346694946289,
"learning_rate": 0.001,
"loss": 1.5247,
"step": 152100
},
{
"epoch": 49.191984486102136,
"grad_norm": 1.8005129098892212,
"learning_rate": 0.001,
"loss": 1.5064,
"step": 152200
},
{
"epoch": 49.22430510665805,
"grad_norm": 1.9255499839782715,
"learning_rate": 0.001,
"loss": 1.5209,
"step": 152300
},
{
"epoch": 49.256625727213965,
"grad_norm": 1.71510648727417,
"learning_rate": 0.001,
"loss": 1.5359,
"step": 152400
},
{
"epoch": 49.28894634776988,
"grad_norm": 1.9309266805648804,
"learning_rate": 0.001,
"loss": 1.5303,
"step": 152500
},
{
"epoch": 49.321266968325794,
"grad_norm": 1.8289529085159302,
"learning_rate": 0.001,
"loss": 1.5395,
"step": 152600
},
{
"epoch": 49.35358758888171,
"grad_norm": 1.4219285249710083,
"learning_rate": 0.001,
"loss": 1.5633,
"step": 152700
},
{
"epoch": 49.38590820943762,
"grad_norm": 1.34114408493042,
"learning_rate": 0.001,
"loss": 1.5322,
"step": 152800
},
{
"epoch": 49.41822882999354,
"grad_norm": 2.261221408843994,
"learning_rate": 0.001,
"loss": 1.55,
"step": 152900
},
{
"epoch": 49.45054945054945,
"grad_norm": 1.7596498727798462,
"learning_rate": 0.001,
"loss": 1.5583,
"step": 153000
},
{
"epoch": 49.48287007110537,
"grad_norm": 2.1398353576660156,
"learning_rate": 0.001,
"loss": 1.5551,
"step": 153100
},
{
"epoch": 49.51519069166128,
"grad_norm": 1.5906637907028198,
"learning_rate": 0.001,
"loss": 1.5674,
"step": 153200
},
{
"epoch": 49.547511312217196,
"grad_norm": 1.7475155591964722,
"learning_rate": 0.001,
"loss": 1.5704,
"step": 153300
},
{
"epoch": 49.57983193277311,
"grad_norm": 1.7171398401260376,
"learning_rate": 0.001,
"loss": 1.5796,
"step": 153400
},
{
"epoch": 49.612152553329025,
"grad_norm": 1.4808603525161743,
"learning_rate": 0.001,
"loss": 1.5742,
"step": 153500
},
{
"epoch": 49.64447317388494,
"grad_norm": 2.099364995956421,
"learning_rate": 0.001,
"loss": 1.5767,
"step": 153600
},
{
"epoch": 49.676793794440854,
"grad_norm": 1.7997703552246094,
"learning_rate": 0.001,
"loss": 1.5549,
"step": 153700
},
{
"epoch": 49.70911441499677,
"grad_norm": 1.9364619255065918,
"learning_rate": 0.001,
"loss": 1.5791,
"step": 153800
},
{
"epoch": 49.74143503555268,
"grad_norm": 1.8328566551208496,
"learning_rate": 0.001,
"loss": 1.5916,
"step": 153900
},
{
"epoch": 49.7737556561086,
"grad_norm": 1.5092872381210327,
"learning_rate": 0.001,
"loss": 1.5775,
"step": 154000
},
{
"epoch": 49.80607627666451,
"grad_norm": 1.8696229457855225,
"learning_rate": 0.001,
"loss": 1.574,
"step": 154100
},
{
"epoch": 49.83839689722043,
"grad_norm": 1.8740304708480835,
"learning_rate": 0.001,
"loss": 1.5895,
"step": 154200
},
{
"epoch": 49.87071751777634,
"grad_norm": 1.6615220308303833,
"learning_rate": 0.001,
"loss": 1.596,
"step": 154300
},
{
"epoch": 49.903038138332256,
"grad_norm": 1.6122123003005981,
"learning_rate": 0.001,
"loss": 1.5956,
"step": 154400
},
{
"epoch": 49.93535875888817,
"grad_norm": 1.8043103218078613,
"learning_rate": 0.001,
"loss": 1.5898,
"step": 154500
},
{
"epoch": 49.967679379444085,
"grad_norm": 1.4336738586425781,
"learning_rate": 0.001,
"loss": 1.6143,
"step": 154600
},
{
"epoch": 50.0,
"grad_norm": 2.6835567951202393,
"learning_rate": 0.001,
"loss": 1.5645,
"step": 154700
},
{
"epoch": 50.032320620555915,
"grad_norm": 1.7635300159454346,
"learning_rate": 0.001,
"loss": 1.4718,
"step": 154800
},
{
"epoch": 50.06464124111183,
"grad_norm": 2.1710870265960693,
"learning_rate": 0.001,
"loss": 1.4931,
"step": 154900
},
{
"epoch": 50.096961861667744,
"grad_norm": 1.8564825057983398,
"learning_rate": 0.001,
"loss": 1.5021,
"step": 155000
},
{
"epoch": 50.12928248222366,
"grad_norm": 2.041884422302246,
"learning_rate": 0.001,
"loss": 1.51,
"step": 155100
},
{
"epoch": 50.16160310277957,
"grad_norm": 1.9597643613815308,
"learning_rate": 0.001,
"loss": 1.5013,
"step": 155200
},
{
"epoch": 50.19392372333549,
"grad_norm": 2.2390921115875244,
"learning_rate": 0.001,
"loss": 1.5206,
"step": 155300
},
{
"epoch": 50.2262443438914,
"grad_norm": 2.3989622592926025,
"learning_rate": 0.001,
"loss": 1.5063,
"step": 155400
},
{
"epoch": 50.25856496444732,
"grad_norm": 2.2737696170806885,
"learning_rate": 0.001,
"loss": 1.5145,
"step": 155500
},
{
"epoch": 50.29088558500323,
"grad_norm": 1.896588683128357,
"learning_rate": 0.001,
"loss": 1.5299,
"step": 155600
},
{
"epoch": 50.323206205559146,
"grad_norm": 3.11960506439209,
"learning_rate": 0.001,
"loss": 1.5229,
"step": 155700
},
{
"epoch": 50.35552682611506,
"grad_norm": 2.0307984352111816,
"learning_rate": 0.001,
"loss": 1.5363,
"step": 155800
},
{
"epoch": 50.387847446670975,
"grad_norm": 1.7273118495941162,
"learning_rate": 0.001,
"loss": 1.536,
"step": 155900
},
{
"epoch": 50.42016806722689,
"grad_norm": 2.4120867252349854,
"learning_rate": 0.001,
"loss": 1.5523,
"step": 156000
},
{
"epoch": 50.452488687782804,
"grad_norm": 1.8841583728790283,
"learning_rate": 0.001,
"loss": 1.535,
"step": 156100
},
{
"epoch": 50.48480930833872,
"grad_norm": 2.2487380504608154,
"learning_rate": 0.001,
"loss": 1.5375,
"step": 156200
},
{
"epoch": 50.51712992889463,
"grad_norm": 2.0784618854522705,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 156300
},
{
"epoch": 50.54945054945055,
"grad_norm": 2.4231741428375244,
"learning_rate": 0.001,
"loss": 1.5446,
"step": 156400
},
{
"epoch": 50.58177117000646,
"grad_norm": 2.341740131378174,
"learning_rate": 0.001,
"loss": 1.5486,
"step": 156500
},
{
"epoch": 50.61409179056238,
"grad_norm": 1.7603386640548706,
"learning_rate": 0.001,
"loss": 1.5682,
"step": 156600
},
{
"epoch": 50.64641241111829,
"grad_norm": 2.165457248687744,
"learning_rate": 0.001,
"loss": 1.5494,
"step": 156700
},
{
"epoch": 50.678733031674206,
"grad_norm": 1.895553708076477,
"learning_rate": 0.001,
"loss": 1.5577,
"step": 156800
},
{
"epoch": 50.71105365223012,
"grad_norm": 2.9701766967773438,
"learning_rate": 0.001,
"loss": 1.5556,
"step": 156900
},
{
"epoch": 50.743374272786035,
"grad_norm": 1.9630506038665771,
"learning_rate": 0.001,
"loss": 1.5856,
"step": 157000
},
{
"epoch": 50.77569489334195,
"grad_norm": 2.3503477573394775,
"learning_rate": 0.001,
"loss": 1.561,
"step": 157100
},
{
"epoch": 50.808015513897864,
"grad_norm": 2.1852824687957764,
"learning_rate": 0.001,
"loss": 1.5687,
"step": 157200
},
{
"epoch": 50.84033613445378,
"grad_norm": 1.8669474124908447,
"learning_rate": 0.001,
"loss": 1.5816,
"step": 157300
},
{
"epoch": 50.87265675500969,
"grad_norm": 1.9535120725631714,
"learning_rate": 0.001,
"loss": 1.5901,
"step": 157400
},
{
"epoch": 50.90497737556561,
"grad_norm": 2.5935044288635254,
"learning_rate": 0.001,
"loss": 1.5801,
"step": 157500
},
{
"epoch": 50.93729799612152,
"grad_norm": 1.8464397192001343,
"learning_rate": 0.001,
"loss": 1.5987,
"step": 157600
},
{
"epoch": 50.96961861667744,
"grad_norm": 2.1178231239318848,
"learning_rate": 0.001,
"loss": 1.5834,
"step": 157700
},
{
"epoch": 51.00193923723336,
"grad_norm": 1.3759008646011353,
"learning_rate": 0.001,
"loss": 1.6042,
"step": 157800
},
{
"epoch": 51.03425985778927,
"grad_norm": 1.1606357097625732,
"learning_rate": 0.001,
"loss": 1.4657,
"step": 157900
},
{
"epoch": 51.06658047834519,
"grad_norm": 1.3527188301086426,
"learning_rate": 0.001,
"loss": 1.4664,
"step": 158000
},
{
"epoch": 51.0989010989011,
"grad_norm": 1.8263195753097534,
"learning_rate": 0.001,
"loss": 1.485,
"step": 158100
},
{
"epoch": 51.13122171945702,
"grad_norm": 2.2132692337036133,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 158200
},
{
"epoch": 51.16354234001293,
"grad_norm": 1.8447751998901367,
"learning_rate": 0.001,
"loss": 1.4897,
"step": 158300
},
{
"epoch": 51.195862960568846,
"grad_norm": 1.335484266281128,
"learning_rate": 0.001,
"loss": 1.4808,
"step": 158400
},
{
"epoch": 51.22818358112476,
"grad_norm": 1.5198925733566284,
"learning_rate": 0.001,
"loss": 1.5121,
"step": 158500
},
{
"epoch": 51.260504201680675,
"grad_norm": 1.7248674631118774,
"learning_rate": 0.001,
"loss": 1.5077,
"step": 158600
},
{
"epoch": 51.29282482223659,
"grad_norm": 1.5273159742355347,
"learning_rate": 0.001,
"loss": 1.5231,
"step": 158700
},
{
"epoch": 51.325145442792504,
"grad_norm": 2.1961324214935303,
"learning_rate": 0.001,
"loss": 1.5029,
"step": 158800
},
{
"epoch": 51.35746606334842,
"grad_norm": 1.5844553709030151,
"learning_rate": 0.001,
"loss": 1.5136,
"step": 158900
},
{
"epoch": 51.38978668390433,
"grad_norm": 1.2531520128250122,
"learning_rate": 0.001,
"loss": 1.5301,
"step": 159000
},
{
"epoch": 51.42210730446025,
"grad_norm": 1.56679368019104,
"learning_rate": 0.001,
"loss": 1.5119,
"step": 159100
},
{
"epoch": 51.45442792501616,
"grad_norm": 1.5963170528411865,
"learning_rate": 0.001,
"loss": 1.5456,
"step": 159200
},
{
"epoch": 51.48674854557208,
"grad_norm": 1.8117672204971313,
"learning_rate": 0.001,
"loss": 1.536,
"step": 159300
},
{
"epoch": 51.51906916612799,
"grad_norm": 1.429930567741394,
"learning_rate": 0.001,
"loss": 1.5189,
"step": 159400
},
{
"epoch": 51.551389786683906,
"grad_norm": 2.074106216430664,
"learning_rate": 0.001,
"loss": 1.5547,
"step": 159500
},
{
"epoch": 51.58371040723982,
"grad_norm": 1.420963168144226,
"learning_rate": 0.001,
"loss": 1.5466,
"step": 159600
},
{
"epoch": 51.616031027795735,
"grad_norm": 1.265044093132019,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 159700
},
{
"epoch": 51.64835164835165,
"grad_norm": 1.8755286931991577,
"learning_rate": 0.001,
"loss": 1.5581,
"step": 159800
},
{
"epoch": 51.680672268907564,
"grad_norm": 1.8416346311569214,
"learning_rate": 0.001,
"loss": 1.5525,
"step": 159900
},
{
"epoch": 51.71299288946348,
"grad_norm": 1.696976900100708,
"learning_rate": 0.001,
"loss": 1.5643,
"step": 160000
},
{
"epoch": 51.74531351001939,
"grad_norm": 1.5795621871948242,
"learning_rate": 0.001,
"loss": 1.561,
"step": 160100
},
{
"epoch": 51.77763413057531,
"grad_norm": 1.3417747020721436,
"learning_rate": 0.001,
"loss": 1.5681,
"step": 160200
},
{
"epoch": 51.80995475113122,
"grad_norm": 1.4509358406066895,
"learning_rate": 0.001,
"loss": 1.5706,
"step": 160300
},
{
"epoch": 51.84227537168714,
"grad_norm": 1.5150309801101685,
"learning_rate": 0.001,
"loss": 1.561,
"step": 160400
},
{
"epoch": 51.87459599224305,
"grad_norm": 1.656905174255371,
"learning_rate": 0.001,
"loss": 1.57,
"step": 160500
},
{
"epoch": 51.906916612798966,
"grad_norm": 1.7837170362472534,
"learning_rate": 0.001,
"loss": 1.5764,
"step": 160600
},
{
"epoch": 51.93923723335488,
"grad_norm": 1.3933733701705933,
"learning_rate": 0.001,
"loss": 1.5688,
"step": 160700
},
{
"epoch": 51.971557853910795,
"grad_norm": 1.3882452249526978,
"learning_rate": 0.001,
"loss": 1.5767,
"step": 160800
},
{
"epoch": 52.00387847446671,
"grad_norm": 1.3166024684906006,
"learning_rate": 0.001,
"loss": 1.5777,
"step": 160900
},
{
"epoch": 52.036199095022624,
"grad_norm": 1.8539063930511475,
"learning_rate": 0.001,
"loss": 1.4721,
"step": 161000
},
{
"epoch": 52.06851971557854,
"grad_norm": 1.6465736627578735,
"learning_rate": 0.001,
"loss": 1.4729,
"step": 161100
},
{
"epoch": 52.10084033613445,
"grad_norm": 1.6113590002059937,
"learning_rate": 0.001,
"loss": 1.4844,
"step": 161200
},
{
"epoch": 52.13316095669037,
"grad_norm": 1.2973045110702515,
"learning_rate": 0.001,
"loss": 1.4766,
"step": 161300
},
{
"epoch": 52.16548157724628,
"grad_norm": 1.8479392528533936,
"learning_rate": 0.001,
"loss": 1.4733,
"step": 161400
},
{
"epoch": 52.1978021978022,
"grad_norm": 2.0937154293060303,
"learning_rate": 0.001,
"loss": 1.4808,
"step": 161500
},
{
"epoch": 52.23012281835811,
"grad_norm": 1.845787525177002,
"learning_rate": 0.001,
"loss": 1.4955,
"step": 161600
},
{
"epoch": 52.262443438914026,
"grad_norm": 1.343607783317566,
"learning_rate": 0.001,
"loss": 1.4973,
"step": 161700
},
{
"epoch": 52.29476405946994,
"grad_norm": 1.5876832008361816,
"learning_rate": 0.001,
"loss": 1.4992,
"step": 161800
},
{
"epoch": 52.327084680025855,
"grad_norm": 1.8784267902374268,
"learning_rate": 0.001,
"loss": 1.4917,
"step": 161900
},
{
"epoch": 52.35940530058177,
"grad_norm": 1.451918601989746,
"learning_rate": 0.001,
"loss": 1.5066,
"step": 162000
},
{
"epoch": 52.391725921137684,
"grad_norm": 1.790123701095581,
"learning_rate": 0.001,
"loss": 1.5211,
"step": 162100
},
{
"epoch": 52.4240465416936,
"grad_norm": 1.447482943534851,
"learning_rate": 0.001,
"loss": 1.521,
"step": 162200
},
{
"epoch": 52.456367162249514,
"grad_norm": 1.4804128408432007,
"learning_rate": 0.001,
"loss": 1.5163,
"step": 162300
},
{
"epoch": 52.48868778280543,
"grad_norm": 1.4562474489212036,
"learning_rate": 0.001,
"loss": 1.5279,
"step": 162400
},
{
"epoch": 52.52100840336134,
"grad_norm": 2.2251856327056885,
"learning_rate": 0.001,
"loss": 1.5263,
"step": 162500
},
{
"epoch": 52.55332902391726,
"grad_norm": 1.5162007808685303,
"learning_rate": 0.001,
"loss": 1.5191,
"step": 162600
},
{
"epoch": 52.58564964447317,
"grad_norm": 1.2665784358978271,
"learning_rate": 0.001,
"loss": 1.5238,
"step": 162700
},
{
"epoch": 52.617970265029086,
"grad_norm": 1.884289264678955,
"learning_rate": 0.001,
"loss": 1.5426,
"step": 162800
},
{
"epoch": 52.650290885585,
"grad_norm": 1.6486436128616333,
"learning_rate": 0.001,
"loss": 1.5336,
"step": 162900
},
{
"epoch": 52.682611506140915,
"grad_norm": 1.4439828395843506,
"learning_rate": 0.001,
"loss": 1.5398,
"step": 163000
},
{
"epoch": 52.71493212669683,
"grad_norm": 1.2459328174591064,
"learning_rate": 0.001,
"loss": 1.5498,
"step": 163100
},
{
"epoch": 52.747252747252745,
"grad_norm": 1.3897027969360352,
"learning_rate": 0.001,
"loss": 1.5223,
"step": 163200
},
{
"epoch": 52.77957336780866,
"grad_norm": 1.3424687385559082,
"learning_rate": 0.001,
"loss": 1.529,
"step": 163300
},
{
"epoch": 52.811893988364574,
"grad_norm": 1.7006032466888428,
"learning_rate": 0.001,
"loss": 1.5353,
"step": 163400
},
{
"epoch": 52.84421460892049,
"grad_norm": 1.9055922031402588,
"learning_rate": 0.001,
"loss": 1.5481,
"step": 163500
},
{
"epoch": 52.8765352294764,
"grad_norm": 1.2229498624801636,
"learning_rate": 0.001,
"loss": 1.5771,
"step": 163600
},
{
"epoch": 52.90885585003232,
"grad_norm": 1.1328636407852173,
"learning_rate": 0.001,
"loss": 1.5554,
"step": 163700
},
{
"epoch": 52.94117647058823,
"grad_norm": 1.1441473960876465,
"learning_rate": 0.001,
"loss": 1.5656,
"step": 163800
},
{
"epoch": 52.97349709114415,
"grad_norm": 1.0366441011428833,
"learning_rate": 0.001,
"loss": 1.5629,
"step": 163900
},
{
"epoch": 53.00581771170007,
"grad_norm": 1.6409794092178345,
"learning_rate": 0.001,
"loss": 1.5575,
"step": 164000
},
{
"epoch": 53.03813833225598,
"grad_norm": 1.6734297275543213,
"learning_rate": 0.001,
"loss": 1.46,
"step": 164100
},
{
"epoch": 53.0704589528119,
"grad_norm": 1.6403851509094238,
"learning_rate": 0.001,
"loss": 1.4606,
"step": 164200
},
{
"epoch": 53.10277957336781,
"grad_norm": 1.1308412551879883,
"learning_rate": 0.001,
"loss": 1.4333,
"step": 164300
},
{
"epoch": 53.135100193923726,
"grad_norm": 1.4857252836227417,
"learning_rate": 0.001,
"loss": 1.4513,
"step": 164400
},
{
"epoch": 53.16742081447964,
"grad_norm": 1.3973990678787231,
"learning_rate": 0.001,
"loss": 1.4633,
"step": 164500
},
{
"epoch": 53.199741435035556,
"grad_norm": 1.6644614934921265,
"learning_rate": 0.001,
"loss": 1.4753,
"step": 164600
},
{
"epoch": 53.23206205559147,
"grad_norm": 1.3670461177825928,
"learning_rate": 0.001,
"loss": 1.4784,
"step": 164700
},
{
"epoch": 53.264382676147385,
"grad_norm": 1.745236873626709,
"learning_rate": 0.001,
"loss": 1.4781,
"step": 164800
},
{
"epoch": 53.2967032967033,
"grad_norm": 1.3107473850250244,
"learning_rate": 0.001,
"loss": 1.4922,
"step": 164900
},
{
"epoch": 53.329023917259214,
"grad_norm": 1.3133132457733154,
"learning_rate": 0.001,
"loss": 1.4846,
"step": 165000
},
{
"epoch": 53.36134453781513,
"grad_norm": 1.6218743324279785,
"learning_rate": 0.001,
"loss": 1.4851,
"step": 165100
},
{
"epoch": 53.39366515837104,
"grad_norm": 1.9341623783111572,
"learning_rate": 0.001,
"loss": 1.5117,
"step": 165200
},
{
"epoch": 53.42598577892696,
"grad_norm": 2.1856162548065186,
"learning_rate": 0.001,
"loss": 1.4999,
"step": 165300
},
{
"epoch": 53.45830639948287,
"grad_norm": 1.6803065538406372,
"learning_rate": 0.001,
"loss": 1.5102,
"step": 165400
},
{
"epoch": 53.49062702003879,
"grad_norm": 1.2509381771087646,
"learning_rate": 0.001,
"loss": 1.5022,
"step": 165500
},
{
"epoch": 53.5229476405947,
"grad_norm": 1.7496399879455566,
"learning_rate": 0.001,
"loss": 1.5154,
"step": 165600
},
{
"epoch": 53.555268261150616,
"grad_norm": 1.5907295942306519,
"learning_rate": 0.001,
"loss": 1.5101,
"step": 165700
},
{
"epoch": 53.58758888170653,
"grad_norm": 1.764585256576538,
"learning_rate": 0.001,
"loss": 1.4961,
"step": 165800
},
{
"epoch": 53.619909502262445,
"grad_norm": 1.676477074623108,
"learning_rate": 0.001,
"loss": 1.5318,
"step": 165900
},
{
"epoch": 53.65223012281836,
"grad_norm": 1.4604623317718506,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 166000
},
{
"epoch": 53.684550743374274,
"grad_norm": 1.1267753839492798,
"learning_rate": 0.001,
"loss": 1.5256,
"step": 166100
},
{
"epoch": 53.71687136393019,
"grad_norm": 1.5615739822387695,
"learning_rate": 0.001,
"loss": 1.5222,
"step": 166200
},
{
"epoch": 53.7491919844861,
"grad_norm": 1.6060326099395752,
"learning_rate": 0.001,
"loss": 1.5438,
"step": 166300
},
{
"epoch": 53.78151260504202,
"grad_norm": 1.4747138023376465,
"learning_rate": 0.001,
"loss": 1.5469,
"step": 166400
},
{
"epoch": 53.81383322559793,
"grad_norm": 1.2557308673858643,
"learning_rate": 0.001,
"loss": 1.5408,
"step": 166500
},
{
"epoch": 53.84615384615385,
"grad_norm": 1.4080893993377686,
"learning_rate": 0.001,
"loss": 1.5441,
"step": 166600
},
{
"epoch": 53.87847446670976,
"grad_norm": 1.555981159210205,
"learning_rate": 0.001,
"loss": 1.5635,
"step": 166700
},
{
"epoch": 53.910795087265676,
"grad_norm": 1.2787331342697144,
"learning_rate": 0.001,
"loss": 1.5212,
"step": 166800
},
{
"epoch": 53.94311570782159,
"grad_norm": 1.1718072891235352,
"learning_rate": 0.001,
"loss": 1.5554,
"step": 166900
},
{
"epoch": 53.975436328377505,
"grad_norm": 1.5635219812393188,
"learning_rate": 0.001,
"loss": 1.5665,
"step": 167000
},
{
"epoch": 54.00775694893342,
"grad_norm": 1.2549020051956177,
"learning_rate": 0.001,
"loss": 1.5393,
"step": 167100
},
{
"epoch": 54.040077569489334,
"grad_norm": 1.4312152862548828,
"learning_rate": 0.001,
"loss": 1.43,
"step": 167200
},
{
"epoch": 54.07239819004525,
"grad_norm": 1.757952332496643,
"learning_rate": 0.001,
"loss": 1.439,
"step": 167300
},
{
"epoch": 54.10471881060116,
"grad_norm": 1.4769724607467651,
"learning_rate": 0.001,
"loss": 1.4521,
"step": 167400
},
{
"epoch": 54.13703943115708,
"grad_norm": 1.9982845783233643,
"learning_rate": 0.001,
"loss": 1.4369,
"step": 167500
},
{
"epoch": 54.16936005171299,
"grad_norm": 1.7478052377700806,
"learning_rate": 0.001,
"loss": 1.4519,
"step": 167600
},
{
"epoch": 54.20168067226891,
"grad_norm": 1.38131844997406,
"learning_rate": 0.001,
"loss": 1.4713,
"step": 167700
},
{
"epoch": 54.23400129282482,
"grad_norm": 1.747186303138733,
"learning_rate": 0.001,
"loss": 1.4481,
"step": 167800
},
{
"epoch": 54.266321913380736,
"grad_norm": 1.3846583366394043,
"learning_rate": 0.001,
"loss": 1.4738,
"step": 167900
},
{
"epoch": 54.29864253393665,
"grad_norm": 1.2487694025039673,
"learning_rate": 0.001,
"loss": 1.5031,
"step": 168000
},
{
"epoch": 54.330963154492565,
"grad_norm": 1.5017812252044678,
"learning_rate": 0.001,
"loss": 1.4877,
"step": 168100
},
{
"epoch": 54.36328377504848,
"grad_norm": 1.3167939186096191,
"learning_rate": 0.001,
"loss": 1.4894,
"step": 168200
},
{
"epoch": 54.395604395604394,
"grad_norm": 1.418789267539978,
"learning_rate": 0.001,
"loss": 1.4917,
"step": 168300
},
{
"epoch": 54.42792501616031,
"grad_norm": 1.5393143892288208,
"learning_rate": 0.001,
"loss": 1.4844,
"step": 168400
},
{
"epoch": 54.46024563671622,
"grad_norm": 1.560189127922058,
"learning_rate": 0.001,
"loss": 1.4849,
"step": 168500
},
{
"epoch": 54.49256625727214,
"grad_norm": 1.949209213256836,
"learning_rate": 0.001,
"loss": 1.5023,
"step": 168600
},
{
"epoch": 54.52488687782805,
"grad_norm": 1.287278175354004,
"learning_rate": 0.001,
"loss": 1.5138,
"step": 168700
},
{
"epoch": 54.55720749838397,
"grad_norm": 1.3032569885253906,
"learning_rate": 0.001,
"loss": 1.4824,
"step": 168800
},
{
"epoch": 54.58952811893988,
"grad_norm": 1.5704611539840698,
"learning_rate": 0.001,
"loss": 1.4988,
"step": 168900
},
{
"epoch": 54.621848739495796,
"grad_norm": 1.5014654397964478,
"learning_rate": 0.001,
"loss": 1.5112,
"step": 169000
},
{
"epoch": 54.65416936005171,
"grad_norm": 1.3570882081985474,
"learning_rate": 0.001,
"loss": 1.5177,
"step": 169100
},
{
"epoch": 54.686489980607625,
"grad_norm": 1.2611489295959473,
"learning_rate": 0.001,
"loss": 1.525,
"step": 169200
},
{
"epoch": 54.71881060116354,
"grad_norm": 1.2757374048233032,
"learning_rate": 0.001,
"loss": 1.5313,
"step": 169300
},
{
"epoch": 54.751131221719454,
"grad_norm": 1.571677327156067,
"learning_rate": 0.001,
"loss": 1.5313,
"step": 169400
},
{
"epoch": 54.78345184227537,
"grad_norm": 1.2518600225448608,
"learning_rate": 0.001,
"loss": 1.5307,
"step": 169500
},
{
"epoch": 54.81577246283128,
"grad_norm": 1.4769446849822998,
"learning_rate": 0.001,
"loss": 1.5224,
"step": 169600
},
{
"epoch": 54.8480930833872,
"grad_norm": 1.2939919233322144,
"learning_rate": 0.001,
"loss": 1.5528,
"step": 169700
},
{
"epoch": 54.88041370394311,
"grad_norm": 1.5968825817108154,
"learning_rate": 0.001,
"loss": 1.5279,
"step": 169800
},
{
"epoch": 54.91273432449903,
"grad_norm": 1.4943368434906006,
"learning_rate": 0.001,
"loss": 1.5162,
"step": 169900
},
{
"epoch": 54.94505494505494,
"grad_norm": 1.4535393714904785,
"learning_rate": 0.001,
"loss": 1.5206,
"step": 170000
},
{
"epoch": 54.977375565610856,
"grad_norm": 1.3544466495513916,
"learning_rate": 0.001,
"loss": 1.5482,
"step": 170100
},
{
"epoch": 55.00969618616678,
"grad_norm": 1.2927088737487793,
"learning_rate": 0.001,
"loss": 1.4979,
"step": 170200
},
{
"epoch": 55.04201680672269,
"grad_norm": 1.5981316566467285,
"learning_rate": 0.001,
"loss": 1.4178,
"step": 170300
},
{
"epoch": 55.07433742727861,
"grad_norm": 1.2203686237335205,
"learning_rate": 0.001,
"loss": 1.4376,
"step": 170400
},
{
"epoch": 55.10665804783452,
"grad_norm": 1.7450850009918213,
"learning_rate": 0.001,
"loss": 1.4392,
"step": 170500
},
{
"epoch": 55.138978668390436,
"grad_norm": 1.4528957605361938,
"learning_rate": 0.001,
"loss": 1.4457,
"step": 170600
},
{
"epoch": 55.17129928894635,
"grad_norm": 1.6601139307022095,
"learning_rate": 0.001,
"loss": 1.4468,
"step": 170700
},
{
"epoch": 55.203619909502265,
"grad_norm": 1.4350769519805908,
"learning_rate": 0.001,
"loss": 1.4602,
"step": 170800
},
{
"epoch": 55.23594053005818,
"grad_norm": 1.992567777633667,
"learning_rate": 0.001,
"loss": 1.4495,
"step": 170900
},
{
"epoch": 55.268261150614094,
"grad_norm": 1.6500948667526245,
"learning_rate": 0.001,
"loss": 1.4729,
"step": 171000
},
{
"epoch": 55.30058177117001,
"grad_norm": 1.863330364227295,
"learning_rate": 0.001,
"loss": 1.4435,
"step": 171100
},
{
"epoch": 55.33290239172592,
"grad_norm": 1.3665803670883179,
"learning_rate": 0.001,
"loss": 1.4696,
"step": 171200
},
{
"epoch": 55.36522301228184,
"grad_norm": 1.5130021572113037,
"learning_rate": 0.001,
"loss": 1.4687,
"step": 171300
},
{
"epoch": 55.39754363283775,
"grad_norm": 1.8200185298919678,
"learning_rate": 0.001,
"loss": 1.4711,
"step": 171400
},
{
"epoch": 55.42986425339367,
"grad_norm": 1.1587629318237305,
"learning_rate": 0.001,
"loss": 1.4784,
"step": 171500
},
{
"epoch": 55.46218487394958,
"grad_norm": 1.3702373504638672,
"learning_rate": 0.001,
"loss": 1.4809,
"step": 171600
},
{
"epoch": 55.494505494505496,
"grad_norm": 1.5741314888000488,
"learning_rate": 0.001,
"loss": 1.4934,
"step": 171700
},
{
"epoch": 55.52682611506141,
"grad_norm": 1.3632556200027466,
"learning_rate": 0.001,
"loss": 1.4945,
"step": 171800
},
{
"epoch": 55.559146735617325,
"grad_norm": 1.5309301614761353,
"learning_rate": 0.001,
"loss": 1.489,
"step": 171900
},
{
"epoch": 55.59146735617324,
"grad_norm": 1.448372721672058,
"learning_rate": 0.001,
"loss": 1.4978,
"step": 172000
},
{
"epoch": 55.623787976729155,
"grad_norm": 1.7779254913330078,
"learning_rate": 0.001,
"loss": 1.5016,
"step": 172100
},
{
"epoch": 55.65610859728507,
"grad_norm": 2.014350652694702,
"learning_rate": 0.001,
"loss": 1.4941,
"step": 172200
},
{
"epoch": 55.688429217840984,
"grad_norm": 1.3058133125305176,
"learning_rate": 0.001,
"loss": 1.5198,
"step": 172300
},
{
"epoch": 55.7207498383969,
"grad_norm": 1.4762943983078003,
"learning_rate": 0.001,
"loss": 1.4958,
"step": 172400
},
{
"epoch": 55.75307045895281,
"grad_norm": 1.649828314781189,
"learning_rate": 0.001,
"loss": 1.5166,
"step": 172500
},
{
"epoch": 55.78539107950873,
"grad_norm": 2.230229616165161,
"learning_rate": 0.001,
"loss": 1.4897,
"step": 172600
},
{
"epoch": 55.81771170006464,
"grad_norm": 1.9107069969177246,
"learning_rate": 0.001,
"loss": 1.5423,
"step": 172700
},
{
"epoch": 55.85003232062056,
"grad_norm": 1.5066547393798828,
"learning_rate": 0.001,
"loss": 1.5171,
"step": 172800
},
{
"epoch": 55.88235294117647,
"grad_norm": 1.2946594953536987,
"learning_rate": 0.001,
"loss": 1.5258,
"step": 172900
},
{
"epoch": 55.914673561732386,
"grad_norm": 1.9029631614685059,
"learning_rate": 0.001,
"loss": 1.5255,
"step": 173000
},
{
"epoch": 55.9469941822883,
"grad_norm": 1.4923566579818726,
"learning_rate": 0.001,
"loss": 1.5301,
"step": 173100
},
{
"epoch": 55.979314802844215,
"grad_norm": 1.3440992832183838,
"learning_rate": 0.001,
"loss": 1.5311,
"step": 173200
},
{
"epoch": 56.01163542340013,
"grad_norm": 1.6094200611114502,
"learning_rate": 0.001,
"loss": 1.486,
"step": 173300
},
{
"epoch": 56.043956043956044,
"grad_norm": 1.6978893280029297,
"learning_rate": 0.001,
"loss": 1.4141,
"step": 173400
},
{
"epoch": 56.07627666451196,
"grad_norm": 1.399122714996338,
"learning_rate": 0.001,
"loss": 1.4154,
"step": 173500
},
{
"epoch": 56.10859728506787,
"grad_norm": 1.4120837450027466,
"learning_rate": 0.001,
"loss": 1.4248,
"step": 173600
},
{
"epoch": 56.14091790562379,
"grad_norm": 1.6494534015655518,
"learning_rate": 0.001,
"loss": 1.4429,
"step": 173700
},
{
"epoch": 56.1732385261797,
"grad_norm": 1.2671856880187988,
"learning_rate": 0.001,
"loss": 1.425,
"step": 173800
},
{
"epoch": 56.20555914673562,
"grad_norm": 1.5630028247833252,
"learning_rate": 0.001,
"loss": 1.4297,
"step": 173900
},
{
"epoch": 56.23787976729153,
"grad_norm": 1.9019289016723633,
"learning_rate": 0.001,
"loss": 1.4784,
"step": 174000
},
{
"epoch": 56.270200387847446,
"grad_norm": 1.5625338554382324,
"learning_rate": 0.001,
"loss": 1.464,
"step": 174100
},
{
"epoch": 56.30252100840336,
"grad_norm": 1.8091297149658203,
"learning_rate": 0.001,
"loss": 1.4641,
"step": 174200
},
{
"epoch": 56.334841628959275,
"grad_norm": 1.6739461421966553,
"learning_rate": 0.001,
"loss": 1.4426,
"step": 174300
},
{
"epoch": 56.36716224951519,
"grad_norm": 1.7141815423965454,
"learning_rate": 0.001,
"loss": 1.4348,
"step": 174400
},
{
"epoch": 56.399482870071104,
"grad_norm": 1.2906224727630615,
"learning_rate": 0.001,
"loss": 1.4667,
"step": 174500
},
{
"epoch": 56.43180349062702,
"grad_norm": 1.4699000120162964,
"learning_rate": 0.001,
"loss": 1.4707,
"step": 174600
},
{
"epoch": 56.46412411118293,
"grad_norm": 1.8251467943191528,
"learning_rate": 0.001,
"loss": 1.4793,
"step": 174700
},
{
"epoch": 56.49644473173885,
"grad_norm": 1.5201164484024048,
"learning_rate": 0.001,
"loss": 1.4912,
"step": 174800
},
{
"epoch": 56.52876535229476,
"grad_norm": 1.3517299890518188,
"learning_rate": 0.001,
"loss": 1.488,
"step": 174900
},
{
"epoch": 56.56108597285068,
"grad_norm": 1.4030410051345825,
"learning_rate": 0.001,
"loss": 1.4762,
"step": 175000
},
{
"epoch": 56.59340659340659,
"grad_norm": 1.5813250541687012,
"learning_rate": 0.001,
"loss": 1.4818,
"step": 175100
},
{
"epoch": 56.625727213962506,
"grad_norm": 1.785956859588623,
"learning_rate": 0.001,
"loss": 1.4726,
"step": 175200
},
{
"epoch": 56.65804783451842,
"grad_norm": 1.752875566482544,
"learning_rate": 0.001,
"loss": 1.4817,
"step": 175300
},
{
"epoch": 56.690368455074335,
"grad_norm": 1.5739023685455322,
"learning_rate": 0.001,
"loss": 1.4835,
"step": 175400
},
{
"epoch": 56.72268907563025,
"grad_norm": 1.4878536462783813,
"learning_rate": 0.001,
"loss": 1.4908,
"step": 175500
},
{
"epoch": 56.755009696186164,
"grad_norm": 1.5592995882034302,
"learning_rate": 0.001,
"loss": 1.5006,
"step": 175600
},
{
"epoch": 56.78733031674208,
"grad_norm": 1.7241151332855225,
"learning_rate": 0.001,
"loss": 1.4934,
"step": 175700
},
{
"epoch": 56.81965093729799,
"grad_norm": 1.491036295890808,
"learning_rate": 0.001,
"loss": 1.5126,
"step": 175800
},
{
"epoch": 56.85197155785391,
"grad_norm": 1.9235769510269165,
"learning_rate": 0.001,
"loss": 1.5234,
"step": 175900
},
{
"epoch": 56.88429217840982,
"grad_norm": 1.616104245185852,
"learning_rate": 0.001,
"loss": 1.5355,
"step": 176000
},
{
"epoch": 56.91661279896574,
"grad_norm": 1.8652502298355103,
"learning_rate": 0.001,
"loss": 1.5193,
"step": 176100
},
{
"epoch": 56.94893341952165,
"grad_norm": 1.5958667993545532,
"learning_rate": 0.001,
"loss": 1.5206,
"step": 176200
},
{
"epoch": 56.981254040077566,
"grad_norm": 1.3792390823364258,
"learning_rate": 0.001,
"loss": 1.5136,
"step": 176300
},
{
"epoch": 57.01357466063349,
"grad_norm": 1.4925225973129272,
"learning_rate": 0.001,
"loss": 1.4569,
"step": 176400
},
{
"epoch": 57.0458952811894,
"grad_norm": 1.5995019674301147,
"learning_rate": 0.001,
"loss": 1.414,
"step": 176500
},
{
"epoch": 57.07821590174532,
"grad_norm": 1.842695951461792,
"learning_rate": 0.001,
"loss": 1.4157,
"step": 176600
},
{
"epoch": 57.11053652230123,
"grad_norm": 2.0299296379089355,
"learning_rate": 0.001,
"loss": 1.4066,
"step": 176700
},
{
"epoch": 57.142857142857146,
"grad_norm": 1.4141572713851929,
"learning_rate": 0.001,
"loss": 1.4208,
"step": 176800
},
{
"epoch": 57.17517776341306,
"grad_norm": 1.6466530561447144,
"learning_rate": 0.001,
"loss": 1.4173,
"step": 176900
},
{
"epoch": 57.207498383968975,
"grad_norm": 1.8495573997497559,
"learning_rate": 0.001,
"loss": 1.4368,
"step": 177000
},
{
"epoch": 57.23981900452489,
"grad_norm": 1.8739982843399048,
"learning_rate": 0.001,
"loss": 1.4374,
"step": 177100
},
{
"epoch": 57.272139625080804,
"grad_norm": 1.5880706310272217,
"learning_rate": 0.001,
"loss": 1.4305,
"step": 177200
},
{
"epoch": 57.30446024563672,
"grad_norm": 2.1057231426239014,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 177300
},
{
"epoch": 57.33678086619263,
"grad_norm": 1.5921692848205566,
"learning_rate": 0.001,
"loss": 1.4301,
"step": 177400
},
{
"epoch": 57.36910148674855,
"grad_norm": 1.5669559240341187,
"learning_rate": 0.001,
"loss": 1.4547,
"step": 177500
},
{
"epoch": 57.40142210730446,
"grad_norm": 1.8174595832824707,
"learning_rate": 0.001,
"loss": 1.4568,
"step": 177600
},
{
"epoch": 57.43374272786038,
"grad_norm": 1.5995457172393799,
"learning_rate": 0.001,
"loss": 1.4631,
"step": 177700
},
{
"epoch": 57.46606334841629,
"grad_norm": 1.9321708679199219,
"learning_rate": 0.001,
"loss": 1.4577,
"step": 177800
},
{
"epoch": 57.498383968972206,
"grad_norm": 1.4424487352371216,
"learning_rate": 0.001,
"loss": 1.4615,
"step": 177900
},
{
"epoch": 57.53070458952812,
"grad_norm": 2.30383038520813,
"learning_rate": 0.001,
"loss": 1.4617,
"step": 178000
},
{
"epoch": 57.563025210084035,
"grad_norm": 1.7871912717819214,
"learning_rate": 0.001,
"loss": 1.4794,
"step": 178100
},
{
"epoch": 57.59534583063995,
"grad_norm": 1.3103487491607666,
"learning_rate": 0.001,
"loss": 1.4688,
"step": 178200
},
{
"epoch": 57.627666451195864,
"grad_norm": 1.822178602218628,
"learning_rate": 0.001,
"loss": 1.4942,
"step": 178300
},
{
"epoch": 57.65998707175178,
"grad_norm": 1.6625324487686157,
"learning_rate": 0.001,
"loss": 1.4696,
"step": 178400
},
{
"epoch": 57.69230769230769,
"grad_norm": 1.7712842226028442,
"learning_rate": 0.001,
"loss": 1.4974,
"step": 178500
},
{
"epoch": 57.72462831286361,
"grad_norm": 1.611838698387146,
"learning_rate": 0.001,
"loss": 1.5015,
"step": 178600
},
{
"epoch": 57.75694893341952,
"grad_norm": 1.9365757703781128,
"learning_rate": 0.001,
"loss": 1.4802,
"step": 178700
},
{
"epoch": 57.78926955397544,
"grad_norm": 1.673500657081604,
"learning_rate": 0.001,
"loss": 1.5027,
"step": 178800
},
{
"epoch": 57.82159017453135,
"grad_norm": 1.4796686172485352,
"learning_rate": 0.001,
"loss": 1.5001,
"step": 178900
},
{
"epoch": 57.853910795087266,
"grad_norm": 1.7874492406845093,
"learning_rate": 0.001,
"loss": 1.4892,
"step": 179000
},
{
"epoch": 57.88623141564318,
"grad_norm": 1.7060503959655762,
"learning_rate": 0.001,
"loss": 1.5039,
"step": 179100
},
{
"epoch": 57.918552036199095,
"grad_norm": 1.9481016397476196,
"learning_rate": 0.001,
"loss": 1.5026,
"step": 179200
},
{
"epoch": 57.95087265675501,
"grad_norm": 1.857441782951355,
"learning_rate": 0.001,
"loss": 1.5044,
"step": 179300
},
{
"epoch": 57.983193277310924,
"grad_norm": 1.7272157669067383,
"learning_rate": 0.001,
"loss": 1.5088,
"step": 179400
},
{
"epoch": 58.01551389786684,
"grad_norm": 1.9517948627471924,
"learning_rate": 0.001,
"loss": 1.4247,
"step": 179500
},
{
"epoch": 58.04783451842275,
"grad_norm": 1.648769736289978,
"learning_rate": 0.001,
"loss": 1.3887,
"step": 179600
},
{
"epoch": 58.08015513897867,
"grad_norm": 2.3552958965301514,
"learning_rate": 0.001,
"loss": 1.3846,
"step": 179700
},
{
"epoch": 58.11247575953458,
"grad_norm": 1.8240126371383667,
"learning_rate": 0.001,
"loss": 1.3953,
"step": 179800
},
{
"epoch": 58.1447963800905,
"grad_norm": 2.1630699634552,
"learning_rate": 0.001,
"loss": 1.4126,
"step": 179900
},
{
"epoch": 58.17711700064641,
"grad_norm": 1.6678168773651123,
"learning_rate": 0.001,
"loss": 1.4289,
"step": 180000
},
{
"epoch": 58.209437621202326,
"grad_norm": 1.5961699485778809,
"learning_rate": 0.001,
"loss": 1.4081,
"step": 180100
},
{
"epoch": 58.24175824175824,
"grad_norm": 2.3336539268493652,
"learning_rate": 0.001,
"loss": 1.4226,
"step": 180200
},
{
"epoch": 58.274078862314155,
"grad_norm": 2.197392463684082,
"learning_rate": 0.001,
"loss": 1.438,
"step": 180300
},
{
"epoch": 58.30639948287007,
"grad_norm": 2.0521631240844727,
"learning_rate": 0.001,
"loss": 1.4269,
"step": 180400
},
{
"epoch": 58.338720103425985,
"grad_norm": 1.9566603899002075,
"learning_rate": 0.001,
"loss": 1.4343,
"step": 180500
},
{
"epoch": 58.3710407239819,
"grad_norm": 2.1111741065979004,
"learning_rate": 0.001,
"loss": 1.4378,
"step": 180600
},
{
"epoch": 58.403361344537814,
"grad_norm": 2.358926296234131,
"learning_rate": 0.001,
"loss": 1.4458,
"step": 180700
},
{
"epoch": 58.43568196509373,
"grad_norm": 1.9776195287704468,
"learning_rate": 0.001,
"loss": 1.4596,
"step": 180800
},
{
"epoch": 58.46800258564964,
"grad_norm": 1.6591037511825562,
"learning_rate": 0.001,
"loss": 1.4484,
"step": 180900
},
{
"epoch": 58.50032320620556,
"grad_norm": 2.1159605979919434,
"learning_rate": 0.001,
"loss": 1.4584,
"step": 181000
},
{
"epoch": 58.53264382676147,
"grad_norm": 1.8147565126419067,
"learning_rate": 0.001,
"loss": 1.4609,
"step": 181100
},
{
"epoch": 58.56496444731739,
"grad_norm": 2.349036455154419,
"learning_rate": 0.001,
"loss": 1.4592,
"step": 181200
},
{
"epoch": 58.5972850678733,
"grad_norm": 1.9788084030151367,
"learning_rate": 0.001,
"loss": 1.4589,
"step": 181300
},
{
"epoch": 58.629605688429216,
"grad_norm": 1.711326003074646,
"learning_rate": 0.001,
"loss": 1.4564,
"step": 181400
},
{
"epoch": 58.66192630898513,
"grad_norm": 1.824481725692749,
"learning_rate": 0.001,
"loss": 1.4739,
"step": 181500
},
{
"epoch": 58.694246929541045,
"grad_norm": 1.8846087455749512,
"learning_rate": 0.001,
"loss": 1.4962,
"step": 181600
},
{
"epoch": 58.72656755009696,
"grad_norm": 1.3911494016647339,
"learning_rate": 0.001,
"loss": 1.4869,
"step": 181700
},
{
"epoch": 58.758888170652874,
"grad_norm": 2.168950080871582,
"learning_rate": 0.001,
"loss": 1.4737,
"step": 181800
},
{
"epoch": 58.79120879120879,
"grad_norm": 2.162898063659668,
"learning_rate": 0.001,
"loss": 1.4896,
"step": 181900
},
{
"epoch": 58.8235294117647,
"grad_norm": 1.8316357135772705,
"learning_rate": 0.001,
"loss": 1.4872,
"step": 182000
},
{
"epoch": 58.85585003232062,
"grad_norm": 1.8830862045288086,
"learning_rate": 0.001,
"loss": 1.4822,
"step": 182100
},
{
"epoch": 58.88817065287653,
"grad_norm": 1.4162304401397705,
"learning_rate": 0.001,
"loss": 1.4798,
"step": 182200
},
{
"epoch": 58.92049127343245,
"grad_norm": 2.3811752796173096,
"learning_rate": 0.001,
"loss": 1.5042,
"step": 182300
},
{
"epoch": 58.95281189398836,
"grad_norm": 2.115408182144165,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 182400
},
{
"epoch": 58.985132514544276,
"grad_norm": 1.5884422063827515,
"learning_rate": 0.001,
"loss": 1.4966,
"step": 182500
},
{
"epoch": 59.0174531351002,
"grad_norm": 1.5762377977371216,
"learning_rate": 0.001,
"loss": 1.4325,
"step": 182600
},
{
"epoch": 59.04977375565611,
"grad_norm": 1.9034700393676758,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 182700
},
{
"epoch": 59.08209437621203,
"grad_norm": 1.8262866735458374,
"learning_rate": 0.001,
"loss": 1.4013,
"step": 182800
},
{
"epoch": 59.11441499676794,
"grad_norm": 1.5713839530944824,
"learning_rate": 0.001,
"loss": 1.3897,
"step": 182900
},
{
"epoch": 59.146735617323856,
"grad_norm": 2.6167213916778564,
"learning_rate": 0.001,
"loss": 1.4069,
"step": 183000
},
{
"epoch": 59.17905623787977,
"grad_norm": 1.6956098079681396,
"learning_rate": 0.001,
"loss": 1.4126,
"step": 183100
},
{
"epoch": 59.211376858435685,
"grad_norm": 1.5501548051834106,
"learning_rate": 0.001,
"loss": 1.4041,
"step": 183200
},
{
"epoch": 59.2436974789916,
"grad_norm": 1.632919430732727,
"learning_rate": 0.001,
"loss": 1.4135,
"step": 183300
},
{
"epoch": 59.276018099547514,
"grad_norm": 1.7773116827011108,
"learning_rate": 0.001,
"loss": 1.4199,
"step": 183400
},
{
"epoch": 59.30833872010343,
"grad_norm": 1.6854921579360962,
"learning_rate": 0.001,
"loss": 1.4134,
"step": 183500
},
{
"epoch": 59.34065934065934,
"grad_norm": 1.7805248498916626,
"learning_rate": 0.001,
"loss": 1.4246,
"step": 183600
},
{
"epoch": 59.37297996121526,
"grad_norm": 1.943892002105713,
"learning_rate": 0.001,
"loss": 1.4288,
"step": 183700
},
{
"epoch": 59.40530058177117,
"grad_norm": 1.6803494691848755,
"learning_rate": 0.001,
"loss": 1.4369,
"step": 183800
},
{
"epoch": 59.43762120232709,
"grad_norm": 2.133312940597534,
"learning_rate": 0.001,
"loss": 1.452,
"step": 183900
},
{
"epoch": 59.469941822883,
"grad_norm": 1.6594949960708618,
"learning_rate": 0.001,
"loss": 1.4436,
"step": 184000
},
{
"epoch": 59.502262443438916,
"grad_norm": 1.515779733657837,
"learning_rate": 0.001,
"loss": 1.4516,
"step": 184100
},
{
"epoch": 59.53458306399483,
"grad_norm": 1.8933998346328735,
"learning_rate": 0.001,
"loss": 1.444,
"step": 184200
},
{
"epoch": 59.566903684550745,
"grad_norm": 1.962734580039978,
"learning_rate": 0.001,
"loss": 1.4462,
"step": 184300
},
{
"epoch": 59.59922430510666,
"grad_norm": 1.573004126548767,
"learning_rate": 0.001,
"loss": 1.4622,
"step": 184400
},
{
"epoch": 59.631544925662574,
"grad_norm": 2.281193256378174,
"learning_rate": 0.001,
"loss": 1.4556,
"step": 184500
},
{
"epoch": 59.66386554621849,
"grad_norm": 1.6149003505706787,
"learning_rate": 0.001,
"loss": 1.4537,
"step": 184600
},
{
"epoch": 59.6961861667744,
"grad_norm": 1.7026011943817139,
"learning_rate": 0.001,
"loss": 1.4564,
"step": 184700
},
{
"epoch": 59.72850678733032,
"grad_norm": 1.9733768701553345,
"learning_rate": 0.001,
"loss": 1.4769,
"step": 184800
},
{
"epoch": 59.76082740788623,
"grad_norm": 1.584242343902588,
"learning_rate": 0.001,
"loss": 1.4786,
"step": 184900
},
{
"epoch": 59.79314802844215,
"grad_norm": 1.8707184791564941,
"learning_rate": 0.001,
"loss": 1.4709,
"step": 185000
},
{
"epoch": 59.82546864899806,
"grad_norm": 2.813775062561035,
"learning_rate": 0.001,
"loss": 1.4704,
"step": 185100
},
{
"epoch": 59.857789269553976,
"grad_norm": 1.6593128442764282,
"learning_rate": 0.001,
"loss": 1.4652,
"step": 185200
},
{
"epoch": 59.89010989010989,
"grad_norm": 2.4981613159179688,
"learning_rate": 0.001,
"loss": 1.4661,
"step": 185300
},
{
"epoch": 59.922430510665805,
"grad_norm": 1.8080812692642212,
"learning_rate": 0.001,
"loss": 1.4756,
"step": 185400
},
{
"epoch": 59.95475113122172,
"grad_norm": 1.5147720575332642,
"learning_rate": 0.001,
"loss": 1.4866,
"step": 185500
},
{
"epoch": 59.987071751777634,
"grad_norm": 2.0482633113861084,
"learning_rate": 0.001,
"loss": 1.4873,
"step": 185600
},
{
"epoch": 60.01939237233355,
"grad_norm": 1.222092628479004,
"learning_rate": 0.001,
"loss": 1.4183,
"step": 185700
},
{
"epoch": 60.05171299288946,
"grad_norm": 1.309107780456543,
"learning_rate": 0.001,
"loss": 1.3756,
"step": 185800
},
{
"epoch": 60.08403361344538,
"grad_norm": 1.9363985061645508,
"learning_rate": 0.001,
"loss": 1.3947,
"step": 185900
},
{
"epoch": 60.11635423400129,
"grad_norm": 1.837631106376648,
"learning_rate": 0.001,
"loss": 1.3782,
"step": 186000
},
{
"epoch": 60.14867485455721,
"grad_norm": 1.3715769052505493,
"learning_rate": 0.001,
"loss": 1.3752,
"step": 186100
},
{
"epoch": 60.18099547511312,
"grad_norm": 1.2659372091293335,
"learning_rate": 0.001,
"loss": 1.3928,
"step": 186200
},
{
"epoch": 60.213316095669036,
"grad_norm": 1.2593166828155518,
"learning_rate": 0.001,
"loss": 1.3924,
"step": 186300
},
{
"epoch": 60.24563671622495,
"grad_norm": 1.8180540800094604,
"learning_rate": 0.001,
"loss": 1.3932,
"step": 186400
},
{
"epoch": 60.277957336780865,
"grad_norm": 2.077974557876587,
"learning_rate": 0.001,
"loss": 1.4231,
"step": 186500
},
{
"epoch": 60.31027795733678,
"grad_norm": 1.648301124572754,
"learning_rate": 0.001,
"loss": 1.4192,
"step": 186600
},
{
"epoch": 60.342598577892694,
"grad_norm": 1.756912350654602,
"learning_rate": 0.001,
"loss": 1.4235,
"step": 186700
},
{
"epoch": 60.37491919844861,
"grad_norm": 1.4692977666854858,
"learning_rate": 0.001,
"loss": 1.4013,
"step": 186800
},
{
"epoch": 60.40723981900452,
"grad_norm": 1.488180160522461,
"learning_rate": 0.001,
"loss": 1.4257,
"step": 186900
},
{
"epoch": 60.43956043956044,
"grad_norm": 1.2721911668777466,
"learning_rate": 0.001,
"loss": 1.4242,
"step": 187000
},
{
"epoch": 60.47188106011635,
"grad_norm": 1.282834529876709,
"learning_rate": 0.001,
"loss": 1.4192,
"step": 187100
},
{
"epoch": 60.50420168067227,
"grad_norm": 1.254542589187622,
"learning_rate": 0.001,
"loss": 1.4391,
"step": 187200
},
{
"epoch": 60.53652230122818,
"grad_norm": 1.5919004678726196,
"learning_rate": 0.001,
"loss": 1.4214,
"step": 187300
},
{
"epoch": 60.568842921784096,
"grad_norm": 1.6546337604522705,
"learning_rate": 0.001,
"loss": 1.4659,
"step": 187400
},
{
"epoch": 60.60116354234001,
"grad_norm": 1.2467652559280396,
"learning_rate": 0.001,
"loss": 1.4507,
"step": 187500
},
{
"epoch": 60.633484162895925,
"grad_norm": 1.228142499923706,
"learning_rate": 0.001,
"loss": 1.4256,
"step": 187600
},
{
"epoch": 60.66580478345184,
"grad_norm": 1.2312934398651123,
"learning_rate": 0.001,
"loss": 1.4597,
"step": 187700
},
{
"epoch": 60.698125404007754,
"grad_norm": 1.362423062324524,
"learning_rate": 0.001,
"loss": 1.456,
"step": 187800
},
{
"epoch": 60.73044602456367,
"grad_norm": 1.8674683570861816,
"learning_rate": 0.001,
"loss": 1.4538,
"step": 187900
},
{
"epoch": 60.762766645119584,
"grad_norm": 1.8995897769927979,
"learning_rate": 0.001,
"loss": 1.4637,
"step": 188000
},
{
"epoch": 60.7950872656755,
"grad_norm": 1.3285086154937744,
"learning_rate": 0.001,
"loss": 1.4647,
"step": 188100
},
{
"epoch": 60.82740788623141,
"grad_norm": 1.7458430528640747,
"learning_rate": 0.001,
"loss": 1.4563,
"step": 188200
},
{
"epoch": 60.85972850678733,
"grad_norm": 1.4835847616195679,
"learning_rate": 0.001,
"loss": 1.4792,
"step": 188300
},
{
"epoch": 60.89204912734324,
"grad_norm": 2.0184285640716553,
"learning_rate": 0.001,
"loss": 1.471,
"step": 188400
},
{
"epoch": 60.924369747899156,
"grad_norm": 2.2625019550323486,
"learning_rate": 0.001,
"loss": 1.4764,
"step": 188500
},
{
"epoch": 60.95669036845507,
"grad_norm": 1.3040214776992798,
"learning_rate": 0.001,
"loss": 1.4861,
"step": 188600
},
{
"epoch": 60.98901098901099,
"grad_norm": 2.382171392440796,
"learning_rate": 0.001,
"loss": 1.4735,
"step": 188700
},
{
"epoch": 61.02133160956691,
"grad_norm": 1.2037104368209839,
"learning_rate": 0.001,
"loss": 1.4076,
"step": 188800
},
{
"epoch": 61.05365223012282,
"grad_norm": 1.1736313104629517,
"learning_rate": 0.001,
"loss": 1.369,
"step": 188900
},
{
"epoch": 61.085972850678736,
"grad_norm": 1.3122565746307373,
"learning_rate": 0.001,
"loss": 1.369,
"step": 189000
},
{
"epoch": 61.11829347123465,
"grad_norm": 1.4312031269073486,
"learning_rate": 0.001,
"loss": 1.3899,
"step": 189100
},
{
"epoch": 61.150614091790565,
"grad_norm": 1.6407980918884277,
"learning_rate": 0.001,
"loss": 1.3582,
"step": 189200
},
{
"epoch": 61.18293471234648,
"grad_norm": 1.3378984928131104,
"learning_rate": 0.001,
"loss": 1.3773,
"step": 189300
},
{
"epoch": 61.215255332902395,
"grad_norm": 1.4513931274414062,
"learning_rate": 0.001,
"loss": 1.3769,
"step": 189400
},
{
"epoch": 61.24757595345831,
"grad_norm": 1.5741044282913208,
"learning_rate": 0.001,
"loss": 1.3863,
"step": 189500
},
{
"epoch": 61.279896574014224,
"grad_norm": 2.1121723651885986,
"learning_rate": 0.001,
"loss": 1.3781,
"step": 189600
},
{
"epoch": 61.31221719457014,
"grad_norm": 1.7127248048782349,
"learning_rate": 0.001,
"loss": 1.4187,
"step": 189700
},
{
"epoch": 61.34453781512605,
"grad_norm": 1.1498795747756958,
"learning_rate": 0.001,
"loss": 1.4092,
"step": 189800
},
{
"epoch": 61.37685843568197,
"grad_norm": 1.63968825340271,
"learning_rate": 0.001,
"loss": 1.3977,
"step": 189900
},
{
"epoch": 61.40917905623788,
"grad_norm": 1.6286283731460571,
"learning_rate": 0.001,
"loss": 1.426,
"step": 190000
},
{
"epoch": 61.441499676793796,
"grad_norm": 1.6149277687072754,
"learning_rate": 0.001,
"loss": 1.4122,
"step": 190100
},
{
"epoch": 61.47382029734971,
"grad_norm": 1.7483794689178467,
"learning_rate": 0.001,
"loss": 1.4236,
"step": 190200
},
{
"epoch": 61.506140917905626,
"grad_norm": 1.2920759916305542,
"learning_rate": 0.001,
"loss": 1.4363,
"step": 190300
},
{
"epoch": 61.53846153846154,
"grad_norm": 1.121927261352539,
"learning_rate": 0.001,
"loss": 1.4203,
"step": 190400
},
{
"epoch": 61.570782159017455,
"grad_norm": 1.5375155210494995,
"learning_rate": 0.001,
"loss": 1.4306,
"step": 190500
},
{
"epoch": 61.60310277957337,
"grad_norm": 1.3647403717041016,
"learning_rate": 0.001,
"loss": 1.4306,
"step": 190600
},
{
"epoch": 61.635423400129284,
"grad_norm": 1.3525989055633545,
"learning_rate": 0.001,
"loss": 1.4301,
"step": 190700
},
{
"epoch": 61.6677440206852,
"grad_norm": 1.311113715171814,
"learning_rate": 0.001,
"loss": 1.4498,
"step": 190800
},
{
"epoch": 61.70006464124111,
"grad_norm": 1.514700174331665,
"learning_rate": 0.001,
"loss": 1.436,
"step": 190900
},
{
"epoch": 61.73238526179703,
"grad_norm": 1.467130422592163,
"learning_rate": 0.001,
"loss": 1.4369,
"step": 191000
},
{
"epoch": 61.76470588235294,
"grad_norm": 1.9764527082443237,
"learning_rate": 0.001,
"loss": 1.4512,
"step": 191100
},
{
"epoch": 61.79702650290886,
"grad_norm": 1.0968976020812988,
"learning_rate": 0.001,
"loss": 1.4429,
"step": 191200
},
{
"epoch": 61.82934712346477,
"grad_norm": 1.7983734607696533,
"learning_rate": 0.001,
"loss": 1.4687,
"step": 191300
},
{
"epoch": 61.861667744020686,
"grad_norm": 1.818927526473999,
"learning_rate": 0.001,
"loss": 1.4679,
"step": 191400
},
{
"epoch": 61.8939883645766,
"grad_norm": 1.3303331136703491,
"learning_rate": 0.001,
"loss": 1.4805,
"step": 191500
},
{
"epoch": 61.926308985132515,
"grad_norm": 1.4884530305862427,
"learning_rate": 0.001,
"loss": 1.4622,
"step": 191600
},
{
"epoch": 61.95862960568843,
"grad_norm": 1.3443821668624878,
"learning_rate": 0.001,
"loss": 1.4598,
"step": 191700
},
{
"epoch": 61.990950226244344,
"grad_norm": 1.5385671854019165,
"learning_rate": 0.001,
"loss": 1.4846,
"step": 191800
},
{
"epoch": 62.02327084680026,
"grad_norm": 1.9123036861419678,
"learning_rate": 0.001,
"loss": 1.4038,
"step": 191900
},
{
"epoch": 62.05559146735617,
"grad_norm": 1.686424970626831,
"learning_rate": 0.001,
"loss": 1.3439,
"step": 192000
},
{
"epoch": 62.08791208791209,
"grad_norm": 1.4864493608474731,
"learning_rate": 0.001,
"loss": 1.3631,
"step": 192100
},
{
"epoch": 62.120232708468,
"grad_norm": 1.57008957862854,
"learning_rate": 0.001,
"loss": 1.3621,
"step": 192200
},
{
"epoch": 62.15255332902392,
"grad_norm": 1.1772888898849487,
"learning_rate": 0.001,
"loss": 1.3641,
"step": 192300
},
{
"epoch": 62.18487394957983,
"grad_norm": 1.5883642435073853,
"learning_rate": 0.001,
"loss": 1.3795,
"step": 192400
},
{
"epoch": 62.217194570135746,
"grad_norm": 1.4293068647384644,
"learning_rate": 0.001,
"loss": 1.38,
"step": 192500
},
{
"epoch": 62.24951519069166,
"grad_norm": 1.596917986869812,
"learning_rate": 0.001,
"loss": 1.388,
"step": 192600
},
{
"epoch": 62.281835811247575,
"grad_norm": 1.4482003450393677,
"learning_rate": 0.001,
"loss": 1.3861,
"step": 192700
},
{
"epoch": 62.31415643180349,
"grad_norm": 1.8990685939788818,
"learning_rate": 0.001,
"loss": 1.3957,
"step": 192800
},
{
"epoch": 62.346477052359404,
"grad_norm": 1.6335766315460205,
"learning_rate": 0.001,
"loss": 1.3829,
"step": 192900
},
{
"epoch": 62.37879767291532,
"grad_norm": 1.9832218885421753,
"learning_rate": 0.001,
"loss": 1.4018,
"step": 193000
},
{
"epoch": 62.41111829347123,
"grad_norm": 1.4106504917144775,
"learning_rate": 0.001,
"loss": 1.4156,
"step": 193100
},
{
"epoch": 62.44343891402715,
"grad_norm": 1.4505540132522583,
"learning_rate": 0.001,
"loss": 1.3913,
"step": 193200
},
{
"epoch": 62.47575953458306,
"grad_norm": 1.5402984619140625,
"learning_rate": 0.001,
"loss": 1.4008,
"step": 193300
},
{
"epoch": 62.50808015513898,
"grad_norm": 1.4702680110931396,
"learning_rate": 0.001,
"loss": 1.4007,
"step": 193400
},
{
"epoch": 62.54040077569489,
"grad_norm": 1.7161102294921875,
"learning_rate": 0.001,
"loss": 1.4311,
"step": 193500
},
{
"epoch": 62.572721396250806,
"grad_norm": 1.1644012928009033,
"learning_rate": 0.001,
"loss": 1.4264,
"step": 193600
},
{
"epoch": 62.60504201680672,
"grad_norm": 1.5992450714111328,
"learning_rate": 0.001,
"loss": 1.4156,
"step": 193700
},
{
"epoch": 62.637362637362635,
"grad_norm": 1.2773569822311401,
"learning_rate": 0.001,
"loss": 1.4371,
"step": 193800
},
{
"epoch": 62.66968325791855,
"grad_norm": 1.7517127990722656,
"learning_rate": 0.001,
"loss": 1.4243,
"step": 193900
},
{
"epoch": 62.702003878474464,
"grad_norm": 1.5832035541534424,
"learning_rate": 0.001,
"loss": 1.4346,
"step": 194000
},
{
"epoch": 62.73432449903038,
"grad_norm": 1.5175061225891113,
"learning_rate": 0.001,
"loss": 1.4373,
"step": 194100
},
{
"epoch": 62.76664511958629,
"grad_norm": 1.4983898401260376,
"learning_rate": 0.001,
"loss": 1.4354,
"step": 194200
},
{
"epoch": 62.79896574014221,
"grad_norm": 1.923129916191101,
"learning_rate": 0.001,
"loss": 1.4652,
"step": 194300
},
{
"epoch": 62.83128636069812,
"grad_norm": 1.5332492589950562,
"learning_rate": 0.001,
"loss": 1.4446,
"step": 194400
},
{
"epoch": 62.86360698125404,
"grad_norm": 1.3865859508514404,
"learning_rate": 0.001,
"loss": 1.4332,
"step": 194500
},
{
"epoch": 62.89592760180995,
"grad_norm": 1.3664981126785278,
"learning_rate": 0.001,
"loss": 1.4324,
"step": 194600
},
{
"epoch": 62.928248222365866,
"grad_norm": 1.507145881652832,
"learning_rate": 0.001,
"loss": 1.477,
"step": 194700
},
{
"epoch": 62.96056884292178,
"grad_norm": 1.1725382804870605,
"learning_rate": 0.001,
"loss": 1.4616,
"step": 194800
},
{
"epoch": 62.992889463477695,
"grad_norm": 1.6565682888031006,
"learning_rate": 0.001,
"loss": 1.4521,
"step": 194900
},
{
"epoch": 63.02521008403362,
"grad_norm": 1.4437625408172607,
"learning_rate": 0.001,
"loss": 1.3588,
"step": 195000
},
{
"epoch": 63.05753070458953,
"grad_norm": 1.420046329498291,
"learning_rate": 0.001,
"loss": 1.3529,
"step": 195100
},
{
"epoch": 63.089851325145446,
"grad_norm": 1.674965262413025,
"learning_rate": 0.001,
"loss": 1.3499,
"step": 195200
},
{
"epoch": 63.12217194570136,
"grad_norm": 1.4057883024215698,
"learning_rate": 0.001,
"loss": 1.3682,
"step": 195300
},
{
"epoch": 63.154492566257275,
"grad_norm": 1.432799220085144,
"learning_rate": 0.001,
"loss": 1.3541,
"step": 195400
},
{
"epoch": 63.18681318681319,
"grad_norm": 1.248239517211914,
"learning_rate": 0.001,
"loss": 1.3637,
"step": 195500
},
{
"epoch": 63.219133807369104,
"grad_norm": 1.353018045425415,
"learning_rate": 0.001,
"loss": 1.3654,
"step": 195600
},
{
"epoch": 63.25145442792502,
"grad_norm": 1.305019497871399,
"learning_rate": 0.001,
"loss": 1.3749,
"step": 195700
},
{
"epoch": 63.28377504848093,
"grad_norm": 1.4432069063186646,
"learning_rate": 0.001,
"loss": 1.3658,
"step": 195800
},
{
"epoch": 63.31609566903685,
"grad_norm": 1.7704466581344604,
"learning_rate": 0.001,
"loss": 1.3877,
"step": 195900
},
{
"epoch": 63.34841628959276,
"grad_norm": 1.480355978012085,
"learning_rate": 0.001,
"loss": 1.3974,
"step": 196000
},
{
"epoch": 63.38073691014868,
"grad_norm": 1.5801160335540771,
"learning_rate": 0.001,
"loss": 1.3832,
"step": 196100
},
{
"epoch": 63.41305753070459,
"grad_norm": 1.4862293004989624,
"learning_rate": 0.001,
"loss": 1.4066,
"step": 196200
},
{
"epoch": 63.445378151260506,
"grad_norm": 1.883362889289856,
"learning_rate": 0.001,
"loss": 1.3955,
"step": 196300
},
{
"epoch": 63.47769877181642,
"grad_norm": 1.3884245157241821,
"learning_rate": 0.001,
"loss": 1.3998,
"step": 196400
},
{
"epoch": 63.510019392372335,
"grad_norm": 1.5210504531860352,
"learning_rate": 0.001,
"loss": 1.4087,
"step": 196500
},
{
"epoch": 63.54234001292825,
"grad_norm": 1.1657111644744873,
"learning_rate": 0.001,
"loss": 1.3979,
"step": 196600
},
{
"epoch": 63.574660633484164,
"grad_norm": 1.3850380182266235,
"learning_rate": 0.001,
"loss": 1.3997,
"step": 196700
},
{
"epoch": 63.60698125404008,
"grad_norm": 1.7572051286697388,
"learning_rate": 0.001,
"loss": 1.3992,
"step": 196800
},
{
"epoch": 63.63930187459599,
"grad_norm": 1.5824946165084839,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 196900
},
{
"epoch": 63.67162249515191,
"grad_norm": 1.6165339946746826,
"learning_rate": 0.001,
"loss": 1.4299,
"step": 197000
},
{
"epoch": 63.70394311570782,
"grad_norm": 1.657211422920227,
"learning_rate": 0.001,
"loss": 1.4234,
"step": 197100
},
{
"epoch": 63.73626373626374,
"grad_norm": 2.030299186706543,
"learning_rate": 0.001,
"loss": 1.4291,
"step": 197200
},
{
"epoch": 63.76858435681965,
"grad_norm": 1.864864706993103,
"learning_rate": 0.001,
"loss": 1.4228,
"step": 197300
},
{
"epoch": 63.800904977375566,
"grad_norm": 1.3653168678283691,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 197400
},
{
"epoch": 63.83322559793148,
"grad_norm": 1.5868542194366455,
"learning_rate": 0.001,
"loss": 1.4279,
"step": 197500
},
{
"epoch": 63.865546218487395,
"grad_norm": 1.6072509288787842,
"learning_rate": 0.001,
"loss": 1.463,
"step": 197600
},
{
"epoch": 63.89786683904331,
"grad_norm": 1.2431442737579346,
"learning_rate": 0.001,
"loss": 1.4426,
"step": 197700
},
{
"epoch": 63.930187459599225,
"grad_norm": 1.8268156051635742,
"learning_rate": 0.001,
"loss": 1.4428,
"step": 197800
},
{
"epoch": 63.96250808015514,
"grad_norm": 1.5013717412948608,
"learning_rate": 0.001,
"loss": 1.4635,
"step": 197900
},
{
"epoch": 63.994828700711054,
"grad_norm": 1.5922578573226929,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 198000
},
{
"epoch": 64.02714932126698,
"grad_norm": 1.7445586919784546,
"learning_rate": 0.001,
"loss": 1.3447,
"step": 198100
},
{
"epoch": 64.05946994182288,
"grad_norm": 1.9003431797027588,
"learning_rate": 0.001,
"loss": 1.3404,
"step": 198200
},
{
"epoch": 64.0917905623788,
"grad_norm": 1.6859298944473267,
"learning_rate": 0.001,
"loss": 1.3413,
"step": 198300
},
{
"epoch": 64.12411118293471,
"grad_norm": 1.992749571800232,
"learning_rate": 0.001,
"loss": 1.3571,
"step": 198400
},
{
"epoch": 64.15643180349063,
"grad_norm": 1.800107717514038,
"learning_rate": 0.001,
"loss": 1.3495,
"step": 198500
},
{
"epoch": 64.18875242404654,
"grad_norm": 1.5717955827713013,
"learning_rate": 0.001,
"loss": 1.3485,
"step": 198600
},
{
"epoch": 64.22107304460246,
"grad_norm": 1.43135404586792,
"learning_rate": 0.001,
"loss": 1.373,
"step": 198700
},
{
"epoch": 64.25339366515837,
"grad_norm": 1.5628668069839478,
"learning_rate": 0.001,
"loss": 1.3637,
"step": 198800
},
{
"epoch": 64.28571428571429,
"grad_norm": 1.3284069299697876,
"learning_rate": 0.001,
"loss": 1.3696,
"step": 198900
},
{
"epoch": 64.3180349062702,
"grad_norm": 1.7991653680801392,
"learning_rate": 0.001,
"loss": 1.3685,
"step": 199000
},
{
"epoch": 64.35035552682612,
"grad_norm": 1.3946183919906616,
"learning_rate": 0.001,
"loss": 1.3773,
"step": 199100
},
{
"epoch": 64.38267614738203,
"grad_norm": 1.5314682722091675,
"learning_rate": 0.001,
"loss": 1.3858,
"step": 199200
},
{
"epoch": 64.41499676793795,
"grad_norm": 1.5692931413650513,
"learning_rate": 0.001,
"loss": 1.3734,
"step": 199300
},
{
"epoch": 64.44731738849386,
"grad_norm": 2.0414280891418457,
"learning_rate": 0.001,
"loss": 1.3784,
"step": 199400
},
{
"epoch": 64.47963800904978,
"grad_norm": 1.4214924573898315,
"learning_rate": 0.001,
"loss": 1.3956,
"step": 199500
},
{
"epoch": 64.51195862960569,
"grad_norm": 1.3940602540969849,
"learning_rate": 0.001,
"loss": 1.406,
"step": 199600
},
{
"epoch": 64.54427925016161,
"grad_norm": 2.5642523765563965,
"learning_rate": 0.001,
"loss": 1.4053,
"step": 199700
},
{
"epoch": 64.57659987071752,
"grad_norm": 1.6160247325897217,
"learning_rate": 0.001,
"loss": 1.4018,
"step": 199800
},
{
"epoch": 64.60892049127344,
"grad_norm": 1.409174919128418,
"learning_rate": 0.001,
"loss": 1.4122,
"step": 199900
},
{
"epoch": 64.64124111182934,
"grad_norm": 1.6104872226715088,
"learning_rate": 0.001,
"loss": 1.4068,
"step": 200000
},
{
"epoch": 64.67356173238527,
"grad_norm": 1.4003249406814575,
"learning_rate": 0.001,
"loss": 1.4067,
"step": 200100
},
{
"epoch": 64.70588235294117,
"grad_norm": 1.6747041940689087,
"learning_rate": 0.001,
"loss": 1.4152,
"step": 200200
},
{
"epoch": 64.7382029734971,
"grad_norm": 2.0438196659088135,
"learning_rate": 0.001,
"loss": 1.4197,
"step": 200300
},
{
"epoch": 64.770523594053,
"grad_norm": 2.127326726913452,
"learning_rate": 0.001,
"loss": 1.4225,
"step": 200400
},
{
"epoch": 64.80284421460892,
"grad_norm": 1.4178484678268433,
"learning_rate": 0.001,
"loss": 1.4386,
"step": 200500
},
{
"epoch": 64.83516483516483,
"grad_norm": 1.8293603658676147,
"learning_rate": 0.001,
"loss": 1.4268,
"step": 200600
},
{
"epoch": 64.86748545572075,
"grad_norm": 2.1161253452301025,
"learning_rate": 0.001,
"loss": 1.4351,
"step": 200700
},
{
"epoch": 64.89980607627666,
"grad_norm": 1.7479701042175293,
"learning_rate": 0.001,
"loss": 1.4261,
"step": 200800
},
{
"epoch": 64.93212669683258,
"grad_norm": 1.317136526107788,
"learning_rate": 0.001,
"loss": 1.4205,
"step": 200900
},
{
"epoch": 64.96444731738849,
"grad_norm": 1.641223669052124,
"learning_rate": 0.001,
"loss": 1.4455,
"step": 201000
},
{
"epoch": 64.99676793794441,
"grad_norm": 1.5180416107177734,
"learning_rate": 0.001,
"loss": 1.4195,
"step": 201100
},
{
"epoch": 65.02908855850032,
"grad_norm": 1.7227802276611328,
"learning_rate": 0.001,
"loss": 1.3384,
"step": 201200
},
{
"epoch": 65.06140917905624,
"grad_norm": 1.5802651643753052,
"learning_rate": 0.001,
"loss": 1.3211,
"step": 201300
},
{
"epoch": 65.09372979961215,
"grad_norm": 1.1960089206695557,
"learning_rate": 0.001,
"loss": 1.3314,
"step": 201400
},
{
"epoch": 65.12605042016807,
"grad_norm": 1.4870516061782837,
"learning_rate": 0.001,
"loss": 1.3412,
"step": 201500
},
{
"epoch": 65.15837104072398,
"grad_norm": 1.4223541021347046,
"learning_rate": 0.001,
"loss": 1.3444,
"step": 201600
},
{
"epoch": 65.1906916612799,
"grad_norm": 1.4809181690216064,
"learning_rate": 0.001,
"loss": 1.3426,
"step": 201700
},
{
"epoch": 65.2230122818358,
"grad_norm": 2.217803478240967,
"learning_rate": 0.001,
"loss": 1.3414,
"step": 201800
},
{
"epoch": 65.25533290239173,
"grad_norm": 1.7147786617279053,
"learning_rate": 0.001,
"loss": 1.3636,
"step": 201900
},
{
"epoch": 65.28765352294764,
"grad_norm": 1.6113975048065186,
"learning_rate": 0.001,
"loss": 1.3743,
"step": 202000
},
{
"epoch": 65.31997414350356,
"grad_norm": 2.024049997329712,
"learning_rate": 0.001,
"loss": 1.3468,
"step": 202100
},
{
"epoch": 65.35229476405947,
"grad_norm": 1.9836534261703491,
"learning_rate": 0.001,
"loss": 1.371,
"step": 202200
},
{
"epoch": 65.38461538461539,
"grad_norm": 1.769474744796753,
"learning_rate": 0.001,
"loss": 1.3667,
"step": 202300
},
{
"epoch": 65.4169360051713,
"grad_norm": 1.8485040664672852,
"learning_rate": 0.001,
"loss": 1.3754,
"step": 202400
},
{
"epoch": 65.44925662572722,
"grad_norm": 1.2611091136932373,
"learning_rate": 0.001,
"loss": 1.3839,
"step": 202500
},
{
"epoch": 65.48157724628312,
"grad_norm": 1.3341859579086304,
"learning_rate": 0.001,
"loss": 1.3773,
"step": 202600
},
{
"epoch": 65.51389786683905,
"grad_norm": 1.6250958442687988,
"learning_rate": 0.001,
"loss": 1.384,
"step": 202700
},
{
"epoch": 65.54621848739495,
"grad_norm": 1.9036827087402344,
"learning_rate": 0.001,
"loss": 1.3946,
"step": 202800
},
{
"epoch": 65.57853910795087,
"grad_norm": 1.4791826009750366,
"learning_rate": 0.001,
"loss": 1.3823,
"step": 202900
},
{
"epoch": 65.61085972850678,
"grad_norm": 2.300971031188965,
"learning_rate": 0.001,
"loss": 1.3839,
"step": 203000
},
{
"epoch": 65.6431803490627,
"grad_norm": 1.2696810960769653,
"learning_rate": 0.001,
"loss": 1.4193,
"step": 203100
},
{
"epoch": 65.67550096961861,
"grad_norm": 1.4126332998275757,
"learning_rate": 0.001,
"loss": 1.4136,
"step": 203200
},
{
"epoch": 65.70782159017453,
"grad_norm": 1.5814741849899292,
"learning_rate": 0.001,
"loss": 1.4094,
"step": 203300
},
{
"epoch": 65.74014221073044,
"grad_norm": 1.7776198387145996,
"learning_rate": 0.001,
"loss": 1.4044,
"step": 203400
},
{
"epoch": 65.77246283128636,
"grad_norm": 1.5066275596618652,
"learning_rate": 0.001,
"loss": 1.4129,
"step": 203500
},
{
"epoch": 65.80478345184227,
"grad_norm": 2.751493453979492,
"learning_rate": 0.001,
"loss": 1.4055,
"step": 203600
},
{
"epoch": 65.83710407239819,
"grad_norm": 1.5740598440170288,
"learning_rate": 0.001,
"loss": 1.4154,
"step": 203700
},
{
"epoch": 65.8694246929541,
"grad_norm": 1.510005235671997,
"learning_rate": 0.001,
"loss": 1.4266,
"step": 203800
},
{
"epoch": 65.90174531351002,
"grad_norm": 2.236680746078491,
"learning_rate": 0.001,
"loss": 1.4052,
"step": 203900
},
{
"epoch": 65.93406593406593,
"grad_norm": 1.301505208015442,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 204000
},
{
"epoch": 65.96638655462185,
"grad_norm": 2.374619722366333,
"learning_rate": 0.001,
"loss": 1.4352,
"step": 204100
},
{
"epoch": 65.99870717517777,
"grad_norm": 1.7071536779403687,
"learning_rate": 0.001,
"loss": 1.4323,
"step": 204200
},
{
"epoch": 66.03102779573368,
"grad_norm": 2.193371295928955,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 204300
},
{
"epoch": 66.0633484162896,
"grad_norm": 1.6853256225585938,
"learning_rate": 0.001,
"loss": 1.3084,
"step": 204400
},
{
"epoch": 66.0956690368455,
"grad_norm": 1.849713683128357,
"learning_rate": 0.001,
"loss": 1.3098,
"step": 204500
},
{
"epoch": 66.12798965740143,
"grad_norm": 1.6212743520736694,
"learning_rate": 0.001,
"loss": 1.3458,
"step": 204600
},
{
"epoch": 66.16031027795734,
"grad_norm": 2.065988779067993,
"learning_rate": 0.001,
"loss": 1.3158,
"step": 204700
},
{
"epoch": 66.19263089851326,
"grad_norm": 1.4905744791030884,
"learning_rate": 0.001,
"loss": 1.3421,
"step": 204800
},
{
"epoch": 66.22495151906917,
"grad_norm": 1.9883614778518677,
"learning_rate": 0.001,
"loss": 1.3606,
"step": 204900
},
{
"epoch": 66.25727213962509,
"grad_norm": 1.4957332611083984,
"learning_rate": 0.001,
"loss": 1.3492,
"step": 205000
},
{
"epoch": 66.289592760181,
"grad_norm": 1.8609826564788818,
"learning_rate": 0.001,
"loss": 1.3277,
"step": 205100
},
{
"epoch": 66.32191338073692,
"grad_norm": 1.843746304512024,
"learning_rate": 0.001,
"loss": 1.352,
"step": 205200
},
{
"epoch": 66.35423400129282,
"grad_norm": 2.3093807697296143,
"learning_rate": 0.001,
"loss": 1.3516,
"step": 205300
},
{
"epoch": 66.38655462184875,
"grad_norm": 1.9718279838562012,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 205400
},
{
"epoch": 66.41887524240465,
"grad_norm": 1.7455960512161255,
"learning_rate": 0.001,
"loss": 1.3772,
"step": 205500
},
{
"epoch": 66.45119586296057,
"grad_norm": 1.5213228464126587,
"learning_rate": 0.001,
"loss": 1.3873,
"step": 205600
},
{
"epoch": 66.48351648351648,
"grad_norm": 1.6287715435028076,
"learning_rate": 0.001,
"loss": 1.3853,
"step": 205700
},
{
"epoch": 66.5158371040724,
"grad_norm": 1.6267790794372559,
"learning_rate": 0.001,
"loss": 1.3773,
"step": 205800
},
{
"epoch": 66.54815772462831,
"grad_norm": 2.083371877670288,
"learning_rate": 0.001,
"loss": 1.369,
"step": 205900
},
{
"epoch": 66.58047834518423,
"grad_norm": 2.5718905925750732,
"learning_rate": 0.001,
"loss": 1.3731,
"step": 206000
},
{
"epoch": 66.61279896574014,
"grad_norm": 2.2809901237487793,
"learning_rate": 0.001,
"loss": 1.3964,
"step": 206100
},
{
"epoch": 66.64511958629606,
"grad_norm": 1.6251013278961182,
"learning_rate": 0.001,
"loss": 1.3975,
"step": 206200
},
{
"epoch": 66.67744020685197,
"grad_norm": 1.7564525604248047,
"learning_rate": 0.001,
"loss": 1.3877,
"step": 206300
},
{
"epoch": 66.70976082740789,
"grad_norm": 1.748744010925293,
"learning_rate": 0.001,
"loss": 1.3938,
"step": 206400
},
{
"epoch": 66.7420814479638,
"grad_norm": 2.089381217956543,
"learning_rate": 0.001,
"loss": 1.4197,
"step": 206500
},
{
"epoch": 66.77440206851972,
"grad_norm": 2.1797640323638916,
"learning_rate": 0.001,
"loss": 1.3991,
"step": 206600
},
{
"epoch": 66.80672268907563,
"grad_norm": 1.92826509475708,
"learning_rate": 0.001,
"loss": 1.4103,
"step": 206700
},
{
"epoch": 66.83904330963155,
"grad_norm": 1.3981081247329712,
"learning_rate": 0.001,
"loss": 1.4018,
"step": 206800
},
{
"epoch": 66.87136393018746,
"grad_norm": 2.354722738265991,
"learning_rate": 0.001,
"loss": 1.4124,
"step": 206900
},
{
"epoch": 66.90368455074338,
"grad_norm": 2.201211929321289,
"learning_rate": 0.001,
"loss": 1.3906,
"step": 207000
},
{
"epoch": 66.93600517129929,
"grad_norm": 1.7025607824325562,
"learning_rate": 0.001,
"loss": 1.422,
"step": 207100
},
{
"epoch": 66.96832579185521,
"grad_norm": 1.9557732343673706,
"learning_rate": 0.001,
"loss": 1.4151,
"step": 207200
},
{
"epoch": 67.00064641241111,
"grad_norm": 1.8797639608383179,
"learning_rate": 0.001,
"loss": 1.3962,
"step": 207300
},
{
"epoch": 67.03296703296704,
"grad_norm": 1.7835898399353027,
"learning_rate": 0.001,
"loss": 1.3011,
"step": 207400
},
{
"epoch": 67.06528765352294,
"grad_norm": 1.837484359741211,
"learning_rate": 0.001,
"loss": 1.3101,
"step": 207500
},
{
"epoch": 67.09760827407887,
"grad_norm": 1.8803691864013672,
"learning_rate": 0.001,
"loss": 1.308,
"step": 207600
},
{
"epoch": 67.12992889463477,
"grad_norm": 1.8284261226654053,
"learning_rate": 0.001,
"loss": 1.3088,
"step": 207700
},
{
"epoch": 67.1622495151907,
"grad_norm": 2.1951427459716797,
"learning_rate": 0.001,
"loss": 1.3291,
"step": 207800
},
{
"epoch": 67.1945701357466,
"grad_norm": 1.9866607189178467,
"learning_rate": 0.001,
"loss": 1.3332,
"step": 207900
},
{
"epoch": 67.22689075630252,
"grad_norm": 2.5242843627929688,
"learning_rate": 0.001,
"loss": 1.353,
"step": 208000
},
{
"epoch": 67.25921137685843,
"grad_norm": 2.158780336380005,
"learning_rate": 0.001,
"loss": 1.3228,
"step": 208100
},
{
"epoch": 67.29153199741435,
"grad_norm": 2.7524983882904053,
"learning_rate": 0.001,
"loss": 1.331,
"step": 208200
},
{
"epoch": 67.32385261797026,
"grad_norm": 2.2914531230926514,
"learning_rate": 0.001,
"loss": 1.3532,
"step": 208300
},
{
"epoch": 67.35617323852618,
"grad_norm": 2.0430986881256104,
"learning_rate": 0.001,
"loss": 1.3657,
"step": 208400
},
{
"epoch": 67.38849385908209,
"grad_norm": 2.7747621536254883,
"learning_rate": 0.001,
"loss": 1.3503,
"step": 208500
},
{
"epoch": 67.42081447963801,
"grad_norm": 2.5037589073181152,
"learning_rate": 0.001,
"loss": 1.3583,
"step": 208600
},
{
"epoch": 67.45313510019392,
"grad_norm": 1.931546926498413,
"learning_rate": 0.001,
"loss": 1.3666,
"step": 208700
},
{
"epoch": 67.48545572074984,
"grad_norm": 1.8246216773986816,
"learning_rate": 0.001,
"loss": 1.362,
"step": 208800
},
{
"epoch": 67.51777634130575,
"grad_norm": 2.764183282852173,
"learning_rate": 0.001,
"loss": 1.3682,
"step": 208900
},
{
"epoch": 67.55009696186167,
"grad_norm": 2.9155993461608887,
"learning_rate": 0.001,
"loss": 1.3689,
"step": 209000
},
{
"epoch": 67.58241758241758,
"grad_norm": 1.8072984218597412,
"learning_rate": 0.001,
"loss": 1.3907,
"step": 209100
},
{
"epoch": 67.6147382029735,
"grad_norm": 2.9374847412109375,
"learning_rate": 0.001,
"loss": 1.3778,
"step": 209200
},
{
"epoch": 67.6470588235294,
"grad_norm": 2.64780592918396,
"learning_rate": 0.001,
"loss": 1.3905,
"step": 209300
},
{
"epoch": 67.67937944408533,
"grad_norm": 2.061706304550171,
"learning_rate": 0.001,
"loss": 1.393,
"step": 209400
},
{
"epoch": 67.71170006464124,
"grad_norm": 2.0080013275146484,
"learning_rate": 0.001,
"loss": 1.3761,
"step": 209500
},
{
"epoch": 67.74402068519716,
"grad_norm": 1.6209617853164673,
"learning_rate": 0.001,
"loss": 1.3852,
"step": 209600
},
{
"epoch": 67.77634130575306,
"grad_norm": 2.044245481491089,
"learning_rate": 0.001,
"loss": 1.4122,
"step": 209700
},
{
"epoch": 67.80866192630899,
"grad_norm": 1.3390419483184814,
"learning_rate": 0.001,
"loss": 1.4145,
"step": 209800
},
{
"epoch": 67.8409825468649,
"grad_norm": 1.7002332210540771,
"learning_rate": 0.001,
"loss": 1.388,
"step": 209900
},
{
"epoch": 67.87330316742081,
"grad_norm": 3.0040740966796875,
"learning_rate": 0.001,
"loss": 1.393,
"step": 210000
},
{
"epoch": 67.90562378797672,
"grad_norm": 2.0390803813934326,
"learning_rate": 0.001,
"loss": 1.3944,
"step": 210100
},
{
"epoch": 67.93794440853264,
"grad_norm": 2.327254295349121,
"learning_rate": 0.001,
"loss": 1.4002,
"step": 210200
},
{
"epoch": 67.97026502908855,
"grad_norm": 1.9059028625488281,
"learning_rate": 0.001,
"loss": 1.3971,
"step": 210300
},
{
"epoch": 68.00258564964447,
"grad_norm": 1.3357656002044678,
"learning_rate": 0.001,
"loss": 1.4255,
"step": 210400
},
{
"epoch": 68.0349062702004,
"grad_norm": 1.5476328134536743,
"learning_rate": 0.001,
"loss": 1.3014,
"step": 210500
},
{
"epoch": 68.0672268907563,
"grad_norm": 1.7639280557632446,
"learning_rate": 0.001,
"loss": 1.3068,
"step": 210600
},
{
"epoch": 68.09954751131222,
"grad_norm": 1.6734331846237183,
"learning_rate": 0.001,
"loss": 1.3149,
"step": 210700
},
{
"epoch": 68.13186813186813,
"grad_norm": 1.776329755783081,
"learning_rate": 0.001,
"loss": 1.2975,
"step": 210800
},
{
"epoch": 68.16418875242405,
"grad_norm": 1.520376443862915,
"learning_rate": 0.001,
"loss": 1.311,
"step": 210900
},
{
"epoch": 68.19650937297996,
"grad_norm": 2.158876657485962,
"learning_rate": 0.001,
"loss": 1.3252,
"step": 211000
},
{
"epoch": 68.22882999353588,
"grad_norm": 1.4346355199813843,
"learning_rate": 0.001,
"loss": 1.3211,
"step": 211100
},
{
"epoch": 68.26115061409179,
"grad_norm": 1.6036536693572998,
"learning_rate": 0.001,
"loss": 1.3322,
"step": 211200
},
{
"epoch": 68.29347123464771,
"grad_norm": 1.9127318859100342,
"learning_rate": 0.001,
"loss": 1.3335,
"step": 211300
},
{
"epoch": 68.32579185520362,
"grad_norm": 1.553421974182129,
"learning_rate": 0.001,
"loss": 1.3424,
"step": 211400
},
{
"epoch": 68.35811247575954,
"grad_norm": 1.2245683670043945,
"learning_rate": 0.001,
"loss": 1.343,
"step": 211500
},
{
"epoch": 68.39043309631545,
"grad_norm": 1.4779609441757202,
"learning_rate": 0.001,
"loss": 1.3424,
"step": 211600
},
{
"epoch": 68.42275371687137,
"grad_norm": 1.516800045967102,
"learning_rate": 0.001,
"loss": 1.3462,
"step": 211700
},
{
"epoch": 68.45507433742728,
"grad_norm": 1.5422505140304565,
"learning_rate": 0.001,
"loss": 1.3566,
"step": 211800
},
{
"epoch": 68.4873949579832,
"grad_norm": 1.556553602218628,
"learning_rate": 0.001,
"loss": 1.3554,
"step": 211900
},
{
"epoch": 68.5197155785391,
"grad_norm": 2.1386520862579346,
"learning_rate": 0.001,
"loss": 1.3652,
"step": 212000
},
{
"epoch": 68.55203619909503,
"grad_norm": 1.5340510606765747,
"learning_rate": 0.001,
"loss": 1.3761,
"step": 212100
},
{
"epoch": 68.58435681965094,
"grad_norm": 2.1872169971466064,
"learning_rate": 0.001,
"loss": 1.3684,
"step": 212200
},
{
"epoch": 68.61667744020686,
"grad_norm": 1.7372183799743652,
"learning_rate": 0.001,
"loss": 1.3716,
"step": 212300
},
{
"epoch": 68.64899806076276,
"grad_norm": 1.9826773405075073,
"learning_rate": 0.001,
"loss": 1.3804,
"step": 212400
},
{
"epoch": 68.68131868131869,
"grad_norm": 1.1110419034957886,
"learning_rate": 0.001,
"loss": 1.3517,
"step": 212500
},
{
"epoch": 68.7136393018746,
"grad_norm": 1.7787781953811646,
"learning_rate": 0.001,
"loss": 1.3773,
"step": 212600
},
{
"epoch": 68.74595992243052,
"grad_norm": 2.0240843296051025,
"learning_rate": 0.001,
"loss": 1.385,
"step": 212700
},
{
"epoch": 68.77828054298642,
"grad_norm": 1.1772947311401367,
"learning_rate": 0.001,
"loss": 1.3794,
"step": 212800
},
{
"epoch": 68.81060116354234,
"grad_norm": 1.6726468801498413,
"learning_rate": 0.001,
"loss": 1.3895,
"step": 212900
},
{
"epoch": 68.84292178409825,
"grad_norm": 2.144679069519043,
"learning_rate": 0.001,
"loss": 1.3796,
"step": 213000
},
{
"epoch": 68.87524240465417,
"grad_norm": 1.5072028636932373,
"learning_rate": 0.001,
"loss": 1.3964,
"step": 213100
},
{
"epoch": 68.90756302521008,
"grad_norm": 1.3276915550231934,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 213200
},
{
"epoch": 68.939883645766,
"grad_norm": 1.5651215314865112,
"learning_rate": 0.001,
"loss": 1.4048,
"step": 213300
},
{
"epoch": 68.97220426632191,
"grad_norm": 1.139343023300171,
"learning_rate": 0.001,
"loss": 1.3835,
"step": 213400
},
{
"epoch": 69.00452488687783,
"grad_norm": 1.4380346536636353,
"learning_rate": 0.001,
"loss": 1.4131,
"step": 213500
},
{
"epoch": 69.03684550743374,
"grad_norm": 1.3611998558044434,
"learning_rate": 0.001,
"loss": 1.2713,
"step": 213600
},
{
"epoch": 69.06916612798966,
"grad_norm": 2.2016654014587402,
"learning_rate": 0.001,
"loss": 1.2976,
"step": 213700
},
{
"epoch": 69.10148674854557,
"grad_norm": 1.313778042793274,
"learning_rate": 0.001,
"loss": 1.303,
"step": 213800
},
{
"epoch": 69.13380736910149,
"grad_norm": 1.6348943710327148,
"learning_rate": 0.001,
"loss": 1.3014,
"step": 213900
},
{
"epoch": 69.1661279896574,
"grad_norm": 1.3724998235702515,
"learning_rate": 0.001,
"loss": 1.3346,
"step": 214000
},
{
"epoch": 69.19844861021332,
"grad_norm": 1.3382467031478882,
"learning_rate": 0.001,
"loss": 1.3138,
"step": 214100
},
{
"epoch": 69.23076923076923,
"grad_norm": 1.7528189420700073,
"learning_rate": 0.001,
"loss": 1.307,
"step": 214200
},
{
"epoch": 69.26308985132515,
"grad_norm": 1.160068154335022,
"learning_rate": 0.001,
"loss": 1.3214,
"step": 214300
},
{
"epoch": 69.29541047188106,
"grad_norm": 1.3790607452392578,
"learning_rate": 0.001,
"loss": 1.3213,
"step": 214400
},
{
"epoch": 69.32773109243698,
"grad_norm": 1.7774428129196167,
"learning_rate": 0.001,
"loss": 1.3172,
"step": 214500
},
{
"epoch": 69.36005171299288,
"grad_norm": 1.3554517030715942,
"learning_rate": 0.001,
"loss": 1.3276,
"step": 214600
},
{
"epoch": 69.3923723335488,
"grad_norm": 1.5843979120254517,
"learning_rate": 0.001,
"loss": 1.3305,
"step": 214700
},
{
"epoch": 69.42469295410471,
"grad_norm": 1.7524763345718384,
"learning_rate": 0.001,
"loss": 1.3386,
"step": 214800
},
{
"epoch": 69.45701357466064,
"grad_norm": 2.0899758338928223,
"learning_rate": 0.001,
"loss": 1.3593,
"step": 214900
},
{
"epoch": 69.48933419521654,
"grad_norm": 2.218890905380249,
"learning_rate": 0.001,
"loss": 1.375,
"step": 215000
},
{
"epoch": 69.52165481577246,
"grad_norm": 1.7657575607299805,
"learning_rate": 0.001,
"loss": 1.3616,
"step": 215100
},
{
"epoch": 69.55397543632837,
"grad_norm": 1.7468805313110352,
"learning_rate": 0.001,
"loss": 1.3635,
"step": 215200
},
{
"epoch": 69.5862960568843,
"grad_norm": 2.189852714538574,
"learning_rate": 0.001,
"loss": 1.3652,
"step": 215300
},
{
"epoch": 69.6186166774402,
"grad_norm": 1.7859878540039062,
"learning_rate": 0.001,
"loss": 1.3527,
"step": 215400
},
{
"epoch": 69.65093729799612,
"grad_norm": 1.292675495147705,
"learning_rate": 0.001,
"loss": 1.3743,
"step": 215500
},
{
"epoch": 69.68325791855203,
"grad_norm": 1.5818181037902832,
"learning_rate": 0.001,
"loss": 1.3749,
"step": 215600
},
{
"epoch": 69.71557853910795,
"grad_norm": 1.57987380027771,
"learning_rate": 0.001,
"loss": 1.3835,
"step": 215700
},
{
"epoch": 69.74789915966386,
"grad_norm": 1.5561144351959229,
"learning_rate": 0.001,
"loss": 1.3734,
"step": 215800
},
{
"epoch": 69.78021978021978,
"grad_norm": 1.6584949493408203,
"learning_rate": 0.001,
"loss": 1.3633,
"step": 215900
},
{
"epoch": 69.81254040077569,
"grad_norm": 1.736039161682129,
"learning_rate": 0.001,
"loss": 1.3672,
"step": 216000
},
{
"epoch": 69.84486102133161,
"grad_norm": 1.2760789394378662,
"learning_rate": 0.001,
"loss": 1.3706,
"step": 216100
},
{
"epoch": 69.87718164188752,
"grad_norm": 1.7503100633621216,
"learning_rate": 0.001,
"loss": 1.3709,
"step": 216200
},
{
"epoch": 69.90950226244344,
"grad_norm": 1.3824061155319214,
"learning_rate": 0.001,
"loss": 1.3972,
"step": 216300
},
{
"epoch": 69.94182288299935,
"grad_norm": 1.5360379219055176,
"learning_rate": 0.001,
"loss": 1.3795,
"step": 216400
},
{
"epoch": 69.97414350355527,
"grad_norm": 1.843446135520935,
"learning_rate": 0.001,
"loss": 1.4005,
"step": 216500
},
{
"epoch": 70.00646412411119,
"grad_norm": 1.5242737531661987,
"learning_rate": 0.001,
"loss": 1.3872,
"step": 216600
},
{
"epoch": 70.0387847446671,
"grad_norm": 1.4649934768676758,
"learning_rate": 0.001,
"loss": 1.294,
"step": 216700
},
{
"epoch": 70.07110536522302,
"grad_norm": 1.3466111421585083,
"learning_rate": 0.001,
"loss": 1.2915,
"step": 216800
},
{
"epoch": 70.10342598577893,
"grad_norm": 1.0711795091629028,
"learning_rate": 0.001,
"loss": 1.2955,
"step": 216900
},
{
"epoch": 70.13574660633485,
"grad_norm": 1.2805978059768677,
"learning_rate": 0.001,
"loss": 1.2906,
"step": 217000
},
{
"epoch": 70.16806722689076,
"grad_norm": 1.4977800846099854,
"learning_rate": 0.001,
"loss": 1.3109,
"step": 217100
},
{
"epoch": 70.20038784744668,
"grad_norm": 1.2956316471099854,
"learning_rate": 0.001,
"loss": 1.2928,
"step": 217200
},
{
"epoch": 70.23270846800258,
"grad_norm": 1.3268839120864868,
"learning_rate": 0.001,
"loss": 1.3001,
"step": 217300
},
{
"epoch": 70.2650290885585,
"grad_norm": 1.3315646648406982,
"learning_rate": 0.001,
"loss": 1.3077,
"step": 217400
},
{
"epoch": 70.29734970911441,
"grad_norm": 1.6829092502593994,
"learning_rate": 0.001,
"loss": 1.3241,
"step": 217500
},
{
"epoch": 70.32967032967034,
"grad_norm": 1.6773277521133423,
"learning_rate": 0.001,
"loss": 1.3193,
"step": 217600
},
{
"epoch": 70.36199095022624,
"grad_norm": 1.407239317893982,
"learning_rate": 0.001,
"loss": 1.3409,
"step": 217700
},
{
"epoch": 70.39431157078216,
"grad_norm": 1.7320756912231445,
"learning_rate": 0.001,
"loss": 1.3263,
"step": 217800
},
{
"epoch": 70.42663219133807,
"grad_norm": 1.6086102724075317,
"learning_rate": 0.001,
"loss": 1.3269,
"step": 217900
},
{
"epoch": 70.458952811894,
"grad_norm": 1.3016129732131958,
"learning_rate": 0.001,
"loss": 1.335,
"step": 218000
},
{
"epoch": 70.4912734324499,
"grad_norm": 1.8595035076141357,
"learning_rate": 0.001,
"loss": 1.345,
"step": 218100
},
{
"epoch": 70.52359405300582,
"grad_norm": 1.5059258937835693,
"learning_rate": 0.001,
"loss": 1.3563,
"step": 218200
},
{
"epoch": 70.55591467356173,
"grad_norm": 1.7828933000564575,
"learning_rate": 0.001,
"loss": 1.3489,
"step": 218300
},
{
"epoch": 70.58823529411765,
"grad_norm": 1.5474156141281128,
"learning_rate": 0.001,
"loss": 1.3549,
"step": 218400
},
{
"epoch": 70.62055591467356,
"grad_norm": 1.276524305343628,
"learning_rate": 0.001,
"loss": 1.3553,
"step": 218500
},
{
"epoch": 70.65287653522948,
"grad_norm": 1.3445316553115845,
"learning_rate": 0.001,
"loss": 1.3473,
"step": 218600
},
{
"epoch": 70.68519715578539,
"grad_norm": 1.4307173490524292,
"learning_rate": 0.001,
"loss": 1.358,
"step": 218700
},
{
"epoch": 70.71751777634131,
"grad_norm": 1.488200068473816,
"learning_rate": 0.001,
"loss": 1.354,
"step": 218800
},
{
"epoch": 70.74983839689722,
"grad_norm": 1.5671435594558716,
"learning_rate": 0.001,
"loss": 1.3533,
"step": 218900
},
{
"epoch": 70.78215901745314,
"grad_norm": 1.243046760559082,
"learning_rate": 0.001,
"loss": 1.3584,
"step": 219000
},
{
"epoch": 70.81447963800905,
"grad_norm": 1.329303503036499,
"learning_rate": 0.001,
"loss": 1.3644,
"step": 219100
},
{
"epoch": 70.84680025856497,
"grad_norm": 1.2586477994918823,
"learning_rate": 0.001,
"loss": 1.3829,
"step": 219200
},
{
"epoch": 70.87912087912088,
"grad_norm": 1.5427042245864868,
"learning_rate": 0.001,
"loss": 1.3736,
"step": 219300
},
{
"epoch": 70.9114414996768,
"grad_norm": 1.883329153060913,
"learning_rate": 0.001,
"loss": 1.3865,
"step": 219400
},
{
"epoch": 70.9437621202327,
"grad_norm": 1.3526417016983032,
"learning_rate": 0.001,
"loss": 1.3863,
"step": 219500
},
{
"epoch": 70.97608274078863,
"grad_norm": 1.1791620254516602,
"learning_rate": 0.001,
"loss": 1.3817,
"step": 219600
},
{
"epoch": 71.00840336134453,
"grad_norm": 1.4550600051879883,
"learning_rate": 0.001,
"loss": 1.3615,
"step": 219700
},
{
"epoch": 71.04072398190046,
"grad_norm": 1.37946355342865,
"learning_rate": 0.001,
"loss": 1.2663,
"step": 219800
},
{
"epoch": 71.07304460245636,
"grad_norm": 1.4288150072097778,
"learning_rate": 0.001,
"loss": 1.2903,
"step": 219900
},
{
"epoch": 71.10536522301229,
"grad_norm": 1.9936612844467163,
"learning_rate": 0.001,
"loss": 1.2918,
"step": 220000
},
{
"epoch": 71.13768584356819,
"grad_norm": 1.492072343826294,
"learning_rate": 0.001,
"loss": 1.2894,
"step": 220100
},
{
"epoch": 71.17000646412411,
"grad_norm": 1.6720198392868042,
"learning_rate": 0.001,
"loss": 1.2904,
"step": 220200
},
{
"epoch": 71.20232708468002,
"grad_norm": 1.508172631263733,
"learning_rate": 0.001,
"loss": 1.3064,
"step": 220300
},
{
"epoch": 71.23464770523594,
"grad_norm": 1.0777677297592163,
"learning_rate": 0.001,
"loss": 1.2844,
"step": 220400
},
{
"epoch": 71.26696832579185,
"grad_norm": 1.7317595481872559,
"learning_rate": 0.001,
"loss": 1.3043,
"step": 220500
},
{
"epoch": 71.29928894634777,
"grad_norm": 1.95646333694458,
"learning_rate": 0.001,
"loss": 1.2934,
"step": 220600
},
{
"epoch": 71.33160956690368,
"grad_norm": 1.57399320602417,
"learning_rate": 0.001,
"loss": 1.3161,
"step": 220700
},
{
"epoch": 71.3639301874596,
"grad_norm": 1.6432753801345825,
"learning_rate": 0.001,
"loss": 1.318,
"step": 220800
},
{
"epoch": 71.39625080801551,
"grad_norm": 1.5056768655776978,
"learning_rate": 0.001,
"loss": 1.3235,
"step": 220900
},
{
"epoch": 71.42857142857143,
"grad_norm": 1.7427054643630981,
"learning_rate": 0.001,
"loss": 1.3308,
"step": 221000
},
{
"epoch": 71.46089204912734,
"grad_norm": 1.5216237306594849,
"learning_rate": 0.001,
"loss": 1.3267,
"step": 221100
},
{
"epoch": 71.49321266968326,
"grad_norm": 1.37506902217865,
"learning_rate": 0.001,
"loss": 1.3201,
"step": 221200
},
{
"epoch": 71.52553329023917,
"grad_norm": 4.7747111320495605,
"learning_rate": 0.001,
"loss": 1.3388,
"step": 221300
},
{
"epoch": 71.55785391079509,
"grad_norm": 1.690388798713684,
"learning_rate": 0.001,
"loss": 1.3483,
"step": 221400
},
{
"epoch": 71.590174531351,
"grad_norm": 1.5094951391220093,
"learning_rate": 0.001,
"loss": 1.332,
"step": 221500
},
{
"epoch": 71.62249515190692,
"grad_norm": 1.6856282949447632,
"learning_rate": 0.001,
"loss": 1.3219,
"step": 221600
},
{
"epoch": 71.65481577246283,
"grad_norm": 1.7640798091888428,
"learning_rate": 0.001,
"loss": 1.3439,
"step": 221700
},
{
"epoch": 71.68713639301875,
"grad_norm": 1.9872925281524658,
"learning_rate": 0.001,
"loss": 1.3587,
"step": 221800
},
{
"epoch": 71.71945701357465,
"grad_norm": 2.0306122303009033,
"learning_rate": 0.001,
"loss": 1.3454,
"step": 221900
},
{
"epoch": 71.75177763413058,
"grad_norm": 1.7729268074035645,
"learning_rate": 0.001,
"loss": 1.3727,
"step": 222000
},
{
"epoch": 71.78409825468648,
"grad_norm": 1.2778888940811157,
"learning_rate": 0.001,
"loss": 1.3573,
"step": 222100
},
{
"epoch": 71.8164188752424,
"grad_norm": 1.673628807067871,
"learning_rate": 0.001,
"loss": 1.3682,
"step": 222200
},
{
"epoch": 71.84873949579831,
"grad_norm": 1.5933386087417603,
"learning_rate": 0.001,
"loss": 1.3731,
"step": 222300
},
{
"epoch": 71.88106011635423,
"grad_norm": 1.3162115812301636,
"learning_rate": 0.001,
"loss": 1.373,
"step": 222400
},
{
"epoch": 71.91338073691014,
"grad_norm": 1.569196105003357,
"learning_rate": 0.001,
"loss": 1.3633,
"step": 222500
},
{
"epoch": 71.94570135746606,
"grad_norm": 1.152018666267395,
"learning_rate": 0.001,
"loss": 1.3793,
"step": 222600
},
{
"epoch": 71.97802197802197,
"grad_norm": 1.4814441204071045,
"learning_rate": 0.001,
"loss": 1.3901,
"step": 222700
},
{
"epoch": 72.01034259857789,
"grad_norm": 1.4163578748703003,
"learning_rate": 0.001,
"loss": 1.3384,
"step": 222800
},
{
"epoch": 72.04266321913381,
"grad_norm": 1.1956467628479004,
"learning_rate": 0.001,
"loss": 1.2723,
"step": 222900
},
{
"epoch": 72.07498383968972,
"grad_norm": 1.5386581420898438,
"learning_rate": 0.001,
"loss": 1.2844,
"step": 223000
},
{
"epoch": 72.10730446024564,
"grad_norm": 1.6779135465621948,
"learning_rate": 0.001,
"loss": 1.2884,
"step": 223100
},
{
"epoch": 72.13962508080155,
"grad_norm": 1.844859004020691,
"learning_rate": 0.001,
"loss": 1.2919,
"step": 223200
},
{
"epoch": 72.17194570135747,
"grad_norm": 1.4852981567382812,
"learning_rate": 0.001,
"loss": 1.2842,
"step": 223300
},
{
"epoch": 72.20426632191338,
"grad_norm": 1.4666519165039062,
"learning_rate": 0.001,
"loss": 1.2886,
"step": 223400
},
{
"epoch": 72.2365869424693,
"grad_norm": 1.517770528793335,
"learning_rate": 0.001,
"loss": 1.2893,
"step": 223500
},
{
"epoch": 72.26890756302521,
"grad_norm": 1.9516693353652954,
"learning_rate": 0.001,
"loss": 1.2859,
"step": 223600
},
{
"epoch": 72.30122818358113,
"grad_norm": 1.8538799285888672,
"learning_rate": 0.001,
"loss": 1.3131,
"step": 223700
},
{
"epoch": 72.33354880413704,
"grad_norm": 1.3030380010604858,
"learning_rate": 0.001,
"loss": 1.3164,
"step": 223800
},
{
"epoch": 72.36586942469296,
"grad_norm": 1.5888701677322388,
"learning_rate": 0.001,
"loss": 1.3132,
"step": 223900
},
{
"epoch": 72.39819004524887,
"grad_norm": 1.2761156558990479,
"learning_rate": 0.001,
"loss": 1.3071,
"step": 224000
},
{
"epoch": 72.43051066580479,
"grad_norm": 1.4118460416793823,
"learning_rate": 0.001,
"loss": 1.3191,
"step": 224100
},
{
"epoch": 72.4628312863607,
"grad_norm": 1.9316996335983276,
"learning_rate": 0.001,
"loss": 1.3,
"step": 224200
},
{
"epoch": 72.49515190691662,
"grad_norm": 1.3990607261657715,
"learning_rate": 0.001,
"loss": 1.3264,
"step": 224300
},
{
"epoch": 72.52747252747253,
"grad_norm": 2.499032497406006,
"learning_rate": 0.001,
"loss": 1.3209,
"step": 224400
},
{
"epoch": 72.55979314802845,
"grad_norm": 1.4380666017532349,
"learning_rate": 0.001,
"loss": 1.3185,
"step": 224500
},
{
"epoch": 72.59211376858435,
"grad_norm": 2.1750357151031494,
"learning_rate": 0.001,
"loss": 1.3474,
"step": 224600
},
{
"epoch": 72.62443438914028,
"grad_norm": 1.3416081666946411,
"learning_rate": 0.001,
"loss": 1.3386,
"step": 224700
},
{
"epoch": 72.65675500969618,
"grad_norm": 1.6465892791748047,
"learning_rate": 0.001,
"loss": 1.3424,
"step": 224800
},
{
"epoch": 72.6890756302521,
"grad_norm": 2.0284931659698486,
"learning_rate": 0.001,
"loss": 1.3505,
"step": 224900
},
{
"epoch": 72.72139625080801,
"grad_norm": 1.6026378870010376,
"learning_rate": 0.001,
"loss": 1.343,
"step": 225000
},
{
"epoch": 72.75371687136393,
"grad_norm": 1.9068299531936646,
"learning_rate": 0.001,
"loss": 1.3451,
"step": 225100
},
{
"epoch": 72.78603749191984,
"grad_norm": 1.76893949508667,
"learning_rate": 0.001,
"loss": 1.354,
"step": 225200
},
{
"epoch": 72.81835811247576,
"grad_norm": 1.2659997940063477,
"learning_rate": 0.001,
"loss": 1.3381,
"step": 225300
},
{
"epoch": 72.85067873303167,
"grad_norm": 1.6921870708465576,
"learning_rate": 0.001,
"loss": 1.3553,
"step": 225400
},
{
"epoch": 72.88299935358759,
"grad_norm": 1.335923433303833,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 225500
},
{
"epoch": 72.9153199741435,
"grad_norm": 1.394056797027588,
"learning_rate": 0.001,
"loss": 1.3426,
"step": 225600
},
{
"epoch": 72.94764059469942,
"grad_norm": 1.6350864171981812,
"learning_rate": 0.001,
"loss": 1.3675,
"step": 225700
},
{
"epoch": 72.97996121525533,
"grad_norm": 1.704758882522583,
"learning_rate": 0.001,
"loss": 1.3789,
"step": 225800
},
{
"epoch": 73.01228183581125,
"grad_norm": 1.4936609268188477,
"learning_rate": 0.001,
"loss": 1.316,
"step": 225900
},
{
"epoch": 73.04460245636716,
"grad_norm": 1.3105096817016602,
"learning_rate": 0.001,
"loss": 1.249,
"step": 226000
},
{
"epoch": 73.07692307692308,
"grad_norm": 1.5214767456054688,
"learning_rate": 0.001,
"loss": 1.2599,
"step": 226100
},
{
"epoch": 73.10924369747899,
"grad_norm": 1.9506388902664185,
"learning_rate": 0.001,
"loss": 1.2607,
"step": 226200
},
{
"epoch": 73.14156431803491,
"grad_norm": 1.7868854999542236,
"learning_rate": 0.001,
"loss": 1.2781,
"step": 226300
},
{
"epoch": 73.17388493859082,
"grad_norm": 1.758193016052246,
"learning_rate": 0.001,
"loss": 1.2745,
"step": 226400
},
{
"epoch": 73.20620555914674,
"grad_norm": 1.6787816286087036,
"learning_rate": 0.001,
"loss": 1.2754,
"step": 226500
},
{
"epoch": 73.23852617970265,
"grad_norm": 1.638461709022522,
"learning_rate": 0.001,
"loss": 1.2815,
"step": 226600
},
{
"epoch": 73.27084680025857,
"grad_norm": 1.415128469467163,
"learning_rate": 0.001,
"loss": 1.3014,
"step": 226700
},
{
"epoch": 73.30316742081448,
"grad_norm": 1.5924972295761108,
"learning_rate": 0.001,
"loss": 1.2916,
"step": 226800
},
{
"epoch": 73.3354880413704,
"grad_norm": 1.4342584609985352,
"learning_rate": 0.001,
"loss": 1.3054,
"step": 226900
},
{
"epoch": 73.3678086619263,
"grad_norm": 1.7077903747558594,
"learning_rate": 0.001,
"loss": 1.3224,
"step": 227000
},
{
"epoch": 73.40012928248223,
"grad_norm": 1.7451516389846802,
"learning_rate": 0.001,
"loss": 1.3114,
"step": 227100
},
{
"epoch": 73.43244990303813,
"grad_norm": 1.8831264972686768,
"learning_rate": 0.001,
"loss": 1.3046,
"step": 227200
},
{
"epoch": 73.46477052359405,
"grad_norm": 1.3531708717346191,
"learning_rate": 0.001,
"loss": 1.3241,
"step": 227300
},
{
"epoch": 73.49709114414996,
"grad_norm": 1.7281534671783447,
"learning_rate": 0.001,
"loss": 1.3101,
"step": 227400
},
{
"epoch": 73.52941176470588,
"grad_norm": 2.001070737838745,
"learning_rate": 0.001,
"loss": 1.3167,
"step": 227500
},
{
"epoch": 73.56173238526179,
"grad_norm": 2.234375,
"learning_rate": 0.001,
"loss": 1.3032,
"step": 227600
},
{
"epoch": 73.59405300581771,
"grad_norm": 1.3920390605926514,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 227700
},
{
"epoch": 73.62637362637362,
"grad_norm": 1.7844200134277344,
"learning_rate": 0.001,
"loss": 1.332,
"step": 227800
},
{
"epoch": 73.65869424692954,
"grad_norm": 1.3945051431655884,
"learning_rate": 0.001,
"loss": 1.3281,
"step": 227900
},
{
"epoch": 73.69101486748545,
"grad_norm": 2.268733024597168,
"learning_rate": 0.001,
"loss": 1.3415,
"step": 228000
},
{
"epoch": 73.72333548804137,
"grad_norm": 1.682381510734558,
"learning_rate": 0.001,
"loss": 1.3424,
"step": 228100
},
{
"epoch": 73.75565610859728,
"grad_norm": 1.4384807348251343,
"learning_rate": 0.001,
"loss": 1.344,
"step": 228200
},
{
"epoch": 73.7879767291532,
"grad_norm": 1.9348750114440918,
"learning_rate": 0.001,
"loss": 1.3405,
"step": 228300
},
{
"epoch": 73.82029734970911,
"grad_norm": 1.895323634147644,
"learning_rate": 0.001,
"loss": 1.3485,
"step": 228400
},
{
"epoch": 73.85261797026503,
"grad_norm": 1.6878876686096191,
"learning_rate": 0.001,
"loss": 1.342,
"step": 228500
},
{
"epoch": 73.88493859082094,
"grad_norm": 1.4254131317138672,
"learning_rate": 0.001,
"loss": 1.3479,
"step": 228600
},
{
"epoch": 73.91725921137686,
"grad_norm": 2.2063348293304443,
"learning_rate": 0.001,
"loss": 1.3644,
"step": 228700
},
{
"epoch": 73.94957983193277,
"grad_norm": 1.7034088373184204,
"learning_rate": 0.001,
"loss": 1.3651,
"step": 228800
},
{
"epoch": 73.98190045248869,
"grad_norm": 1.8378410339355469,
"learning_rate": 0.001,
"loss": 1.3628,
"step": 228900
},
{
"epoch": 74.01422107304461,
"grad_norm": 1.528680443763733,
"learning_rate": 0.001,
"loss": 1.2987,
"step": 229000
},
{
"epoch": 74.04654169360052,
"grad_norm": 2.586549997329712,
"learning_rate": 0.001,
"loss": 1.2445,
"step": 229100
},
{
"epoch": 74.07886231415644,
"grad_norm": 1.8505322933197021,
"learning_rate": 0.001,
"loss": 1.259,
"step": 229200
},
{
"epoch": 74.11118293471235,
"grad_norm": 2.220876932144165,
"learning_rate": 0.001,
"loss": 1.2778,
"step": 229300
},
{
"epoch": 74.14350355526827,
"grad_norm": 2.3290762901306152,
"learning_rate": 0.001,
"loss": 1.2634,
"step": 229400
},
{
"epoch": 74.17582417582418,
"grad_norm": 1.517318844795227,
"learning_rate": 0.001,
"loss": 1.26,
"step": 229500
},
{
"epoch": 74.2081447963801,
"grad_norm": 1.7899667024612427,
"learning_rate": 0.001,
"loss": 1.2895,
"step": 229600
},
{
"epoch": 74.240465416936,
"grad_norm": 1.6911143064498901,
"learning_rate": 0.001,
"loss": 1.2937,
"step": 229700
},
{
"epoch": 74.27278603749193,
"grad_norm": 2.1663761138916016,
"learning_rate": 0.001,
"loss": 1.2691,
"step": 229800
},
{
"epoch": 74.30510665804783,
"grad_norm": 1.930943250656128,
"learning_rate": 0.001,
"loss": 1.2927,
"step": 229900
},
{
"epoch": 74.33742727860376,
"grad_norm": 1.7484216690063477,
"learning_rate": 0.001,
"loss": 1.2918,
"step": 230000
},
{
"epoch": 74.36974789915966,
"grad_norm": 2.6189985275268555,
"learning_rate": 0.001,
"loss": 1.2971,
"step": 230100
},
{
"epoch": 74.40206851971558,
"grad_norm": 1.7392945289611816,
"learning_rate": 0.001,
"loss": 1.3088,
"step": 230200
},
{
"epoch": 74.43438914027149,
"grad_norm": 1.9195913076400757,
"learning_rate": 0.001,
"loss": 1.3174,
"step": 230300
},
{
"epoch": 74.46670976082741,
"grad_norm": 2.06931209564209,
"learning_rate": 0.001,
"loss": 1.3164,
"step": 230400
},
{
"epoch": 74.49903038138332,
"grad_norm": 2.035064697265625,
"learning_rate": 0.001,
"loss": 1.3102,
"step": 230500
},
{
"epoch": 74.53135100193924,
"grad_norm": 1.3531410694122314,
"learning_rate": 0.001,
"loss": 1.3013,
"step": 230600
},
{
"epoch": 74.56367162249515,
"grad_norm": 2.2259304523468018,
"learning_rate": 0.001,
"loss": 1.296,
"step": 230700
},
{
"epoch": 74.59599224305107,
"grad_norm": 1.3927483558654785,
"learning_rate": 0.001,
"loss": 1.3141,
"step": 230800
},
{
"epoch": 74.62831286360698,
"grad_norm": 1.8264931440353394,
"learning_rate": 0.001,
"loss": 1.3226,
"step": 230900
},
{
"epoch": 74.6606334841629,
"grad_norm": 1.607061743736267,
"learning_rate": 0.001,
"loss": 1.33,
"step": 231000
},
{
"epoch": 74.69295410471881,
"grad_norm": 1.7336506843566895,
"learning_rate": 0.001,
"loss": 1.3326,
"step": 231100
},
{
"epoch": 74.72527472527473,
"grad_norm": 2.0042037963867188,
"learning_rate": 0.001,
"loss": 1.3291,
"step": 231200
},
{
"epoch": 74.75759534583064,
"grad_norm": 1.6352319717407227,
"learning_rate": 0.001,
"loss": 1.3266,
"step": 231300
},
{
"epoch": 74.78991596638656,
"grad_norm": 1.8481652736663818,
"learning_rate": 0.001,
"loss": 1.346,
"step": 231400
},
{
"epoch": 74.82223658694247,
"grad_norm": 2.5028653144836426,
"learning_rate": 0.001,
"loss": 1.3339,
"step": 231500
},
{
"epoch": 74.85455720749839,
"grad_norm": 1.7493265867233276,
"learning_rate": 0.001,
"loss": 1.3218,
"step": 231600
},
{
"epoch": 74.8868778280543,
"grad_norm": 2.155993700027466,
"learning_rate": 0.001,
"loss": 1.3362,
"step": 231700
},
{
"epoch": 74.91919844861022,
"grad_norm": 1.4427071809768677,
"learning_rate": 0.001,
"loss": 1.3435,
"step": 231800
},
{
"epoch": 74.95151906916612,
"grad_norm": 2.04618763923645,
"learning_rate": 0.001,
"loss": 1.3363,
"step": 231900
},
{
"epoch": 74.98383968972205,
"grad_norm": 2.1985716819763184,
"learning_rate": 0.001,
"loss": 1.3433,
"step": 232000
},
{
"epoch": 75.01616031027795,
"grad_norm": 1.8995875120162964,
"learning_rate": 0.001,
"loss": 1.2547,
"step": 232100
},
{
"epoch": 75.04848093083388,
"grad_norm": 1.994150996208191,
"learning_rate": 0.001,
"loss": 1.2323,
"step": 232200
},
{
"epoch": 75.08080155138978,
"grad_norm": 2.4287238121032715,
"learning_rate": 0.001,
"loss": 1.2461,
"step": 232300
},
{
"epoch": 75.1131221719457,
"grad_norm": 2.103858232498169,
"learning_rate": 0.001,
"loss": 1.2589,
"step": 232400
},
{
"epoch": 75.14544279250161,
"grad_norm": 2.1579859256744385,
"learning_rate": 0.001,
"loss": 1.2667,
"step": 232500
},
{
"epoch": 75.17776341305753,
"grad_norm": 2.698798894882202,
"learning_rate": 0.001,
"loss": 1.2723,
"step": 232600
},
{
"epoch": 75.21008403361344,
"grad_norm": 1.7385883331298828,
"learning_rate": 0.001,
"loss": 1.2695,
"step": 232700
},
{
"epoch": 75.24240465416936,
"grad_norm": 2.474719524383545,
"learning_rate": 0.001,
"loss": 1.2692,
"step": 232800
},
{
"epoch": 75.27472527472527,
"grad_norm": 2.216207265853882,
"learning_rate": 0.001,
"loss": 1.2677,
"step": 232900
},
{
"epoch": 75.30704589528119,
"grad_norm": 2.280482053756714,
"learning_rate": 0.001,
"loss": 1.272,
"step": 233000
},
{
"epoch": 75.3393665158371,
"grad_norm": 1.7663888931274414,
"learning_rate": 0.001,
"loss": 1.2754,
"step": 233100
},
{
"epoch": 75.37168713639302,
"grad_norm": 2.109699249267578,
"learning_rate": 0.001,
"loss": 1.2942,
"step": 233200
},
{
"epoch": 75.40400775694893,
"grad_norm": 1.941871166229248,
"learning_rate": 0.001,
"loss": 1.2891,
"step": 233300
},
{
"epoch": 75.43632837750485,
"grad_norm": 2.5241806507110596,
"learning_rate": 0.001,
"loss": 1.2873,
"step": 233400
},
{
"epoch": 75.46864899806076,
"grad_norm": 2.4268338680267334,
"learning_rate": 0.001,
"loss": 1.303,
"step": 233500
},
{
"epoch": 75.50096961861668,
"grad_norm": 2.0684986114501953,
"learning_rate": 0.001,
"loss": 1.2946,
"step": 233600
},
{
"epoch": 75.53329023917259,
"grad_norm": 2.0574164390563965,
"learning_rate": 0.001,
"loss": 1.3119,
"step": 233700
},
{
"epoch": 75.56561085972851,
"grad_norm": 2.067401647567749,
"learning_rate": 0.001,
"loss": 1.3042,
"step": 233800
},
{
"epoch": 75.59793148028442,
"grad_norm": 2.7004194259643555,
"learning_rate": 0.001,
"loss": 1.3117,
"step": 233900
},
{
"epoch": 75.63025210084034,
"grad_norm": 2.0863163471221924,
"learning_rate": 0.001,
"loss": 1.3333,
"step": 234000
},
{
"epoch": 75.66257272139624,
"grad_norm": 1.922520399093628,
"learning_rate": 0.001,
"loss": 1.3118,
"step": 234100
},
{
"epoch": 75.69489334195217,
"grad_norm": 2.3876094818115234,
"learning_rate": 0.001,
"loss": 1.3296,
"step": 234200
},
{
"epoch": 75.72721396250807,
"grad_norm": 2.248257875442505,
"learning_rate": 0.001,
"loss": 1.3328,
"step": 234300
},
{
"epoch": 75.759534583064,
"grad_norm": 2.286165237426758,
"learning_rate": 0.001,
"loss": 1.3248,
"step": 234400
},
{
"epoch": 75.7918552036199,
"grad_norm": 2.5652499198913574,
"learning_rate": 0.001,
"loss": 1.3373,
"step": 234500
},
{
"epoch": 75.82417582417582,
"grad_norm": 2.686047077178955,
"learning_rate": 0.001,
"loss": 1.3372,
"step": 234600
},
{
"epoch": 75.85649644473173,
"grad_norm": 1.6270856857299805,
"learning_rate": 0.001,
"loss": 1.3453,
"step": 234700
},
{
"epoch": 75.88881706528765,
"grad_norm": 2.0927953720092773,
"learning_rate": 0.001,
"loss": 1.3365,
"step": 234800
},
{
"epoch": 75.92113768584356,
"grad_norm": 2.6588659286499023,
"learning_rate": 0.001,
"loss": 1.3346,
"step": 234900
},
{
"epoch": 75.95345830639948,
"grad_norm": 1.880811095237732,
"learning_rate": 0.001,
"loss": 1.32,
"step": 235000
},
{
"epoch": 75.98577892695539,
"grad_norm": 2.332777261734009,
"learning_rate": 0.001,
"loss": 1.3514,
"step": 235100
},
{
"epoch": 76.01809954751131,
"grad_norm": 1.4552125930786133,
"learning_rate": 0.001,
"loss": 1.2992,
"step": 235200
},
{
"epoch": 76.05042016806723,
"grad_norm": 1.6631439924240112,
"learning_rate": 0.001,
"loss": 1.229,
"step": 235300
},
{
"epoch": 76.08274078862314,
"grad_norm": 1.6396737098693848,
"learning_rate": 0.001,
"loss": 1.2473,
"step": 235400
},
{
"epoch": 76.11506140917906,
"grad_norm": 1.4893487691879272,
"learning_rate": 0.001,
"loss": 1.251,
"step": 235500
},
{
"epoch": 76.14738202973497,
"grad_norm": 1.5711896419525146,
"learning_rate": 0.001,
"loss": 1.2428,
"step": 235600
},
{
"epoch": 76.17970265029089,
"grad_norm": 1.5022379159927368,
"learning_rate": 0.001,
"loss": 1.2523,
"step": 235700
},
{
"epoch": 76.2120232708468,
"grad_norm": 1.4917467832565308,
"learning_rate": 0.001,
"loss": 1.2816,
"step": 235800
},
{
"epoch": 76.24434389140272,
"grad_norm": 2.1330008506774902,
"learning_rate": 0.001,
"loss": 1.2446,
"step": 235900
},
{
"epoch": 76.27666451195863,
"grad_norm": 1.9323478937149048,
"learning_rate": 0.001,
"loss": 1.2773,
"step": 236000
},
{
"epoch": 76.30898513251455,
"grad_norm": 1.798542857170105,
"learning_rate": 0.001,
"loss": 1.2836,
"step": 236100
},
{
"epoch": 76.34130575307046,
"grad_norm": 1.9975454807281494,
"learning_rate": 0.001,
"loss": 1.2751,
"step": 236200
},
{
"epoch": 76.37362637362638,
"grad_norm": 1.5190203189849854,
"learning_rate": 0.001,
"loss": 1.2687,
"step": 236300
},
{
"epoch": 76.40594699418229,
"grad_norm": 2.050550699234009,
"learning_rate": 0.001,
"loss": 1.2856,
"step": 236400
},
{
"epoch": 76.43826761473821,
"grad_norm": 1.8917027711868286,
"learning_rate": 0.001,
"loss": 1.2778,
"step": 236500
},
{
"epoch": 76.47058823529412,
"grad_norm": 1.359777808189392,
"learning_rate": 0.001,
"loss": 1.2847,
"step": 236600
},
{
"epoch": 76.50290885585004,
"grad_norm": 1.5652052164077759,
"learning_rate": 0.001,
"loss": 1.301,
"step": 236700
},
{
"epoch": 76.53522947640595,
"grad_norm": 2.3575685024261475,
"learning_rate": 0.001,
"loss": 1.3116,
"step": 236800
},
{
"epoch": 76.56755009696187,
"grad_norm": 1.5174378156661987,
"learning_rate": 0.001,
"loss": 1.2913,
"step": 236900
},
{
"epoch": 76.59987071751777,
"grad_norm": 1.8151758909225464,
"learning_rate": 0.001,
"loss": 1.2957,
"step": 237000
},
{
"epoch": 76.6321913380737,
"grad_norm": 2.0352110862731934,
"learning_rate": 0.001,
"loss": 1.3008,
"step": 237100
},
{
"epoch": 76.6645119586296,
"grad_norm": 1.5290433168411255,
"learning_rate": 0.001,
"loss": 1.3001,
"step": 237200
},
{
"epoch": 76.69683257918552,
"grad_norm": 1.6776798963546753,
"learning_rate": 0.001,
"loss": 1.2941,
"step": 237300
},
{
"epoch": 76.72915319974143,
"grad_norm": 1.7184401750564575,
"learning_rate": 0.001,
"loss": 1.3165,
"step": 237400
},
{
"epoch": 76.76147382029735,
"grad_norm": 1.907509446144104,
"learning_rate": 0.001,
"loss": 1.3158,
"step": 237500
},
{
"epoch": 76.79379444085326,
"grad_norm": 2.0751779079437256,
"learning_rate": 0.001,
"loss": 1.3261,
"step": 237600
},
{
"epoch": 76.82611506140918,
"grad_norm": 2.162783145904541,
"learning_rate": 0.001,
"loss": 1.3363,
"step": 237700
},
{
"epoch": 76.85843568196509,
"grad_norm": 1.6640735864639282,
"learning_rate": 0.001,
"loss": 1.3185,
"step": 237800
},
{
"epoch": 76.89075630252101,
"grad_norm": 1.5273408889770508,
"learning_rate": 0.001,
"loss": 1.3487,
"step": 237900
},
{
"epoch": 76.92307692307692,
"grad_norm": 1.5014934539794922,
"learning_rate": 0.001,
"loss": 1.34,
"step": 238000
},
{
"epoch": 76.95539754363284,
"grad_norm": 1.6896687746047974,
"learning_rate": 0.001,
"loss": 1.3323,
"step": 238100
},
{
"epoch": 76.98771816418875,
"grad_norm": 1.7303470373153687,
"learning_rate": 0.001,
"loss": 1.3495,
"step": 238200
},
{
"epoch": 77.02003878474467,
"grad_norm": 1.8442621231079102,
"learning_rate": 0.001,
"loss": 1.2774,
"step": 238300
},
{
"epoch": 77.05235940530058,
"grad_norm": 1.4819287061691284,
"learning_rate": 0.001,
"loss": 1.2315,
"step": 238400
},
{
"epoch": 77.0846800258565,
"grad_norm": 1.1895514726638794,
"learning_rate": 0.001,
"loss": 1.2365,
"step": 238500
},
{
"epoch": 77.11700064641241,
"grad_norm": 1.6040140390396118,
"learning_rate": 0.001,
"loss": 1.2295,
"step": 238600
},
{
"epoch": 77.14932126696833,
"grad_norm": 1.6429595947265625,
"learning_rate": 0.001,
"loss": 1.2461,
"step": 238700
},
{
"epoch": 77.18164188752424,
"grad_norm": 2.7958550453186035,
"learning_rate": 0.001,
"loss": 1.2478,
"step": 238800
},
{
"epoch": 77.21396250808016,
"grad_norm": 1.813547968864441,
"learning_rate": 0.001,
"loss": 1.2632,
"step": 238900
},
{
"epoch": 77.24628312863607,
"grad_norm": 1.5957227945327759,
"learning_rate": 0.001,
"loss": 1.2692,
"step": 239000
},
{
"epoch": 77.27860374919199,
"grad_norm": 2.054250955581665,
"learning_rate": 0.001,
"loss": 1.2543,
"step": 239100
},
{
"epoch": 77.3109243697479,
"grad_norm": 1.2453227043151855,
"learning_rate": 0.001,
"loss": 1.273,
"step": 239200
},
{
"epoch": 77.34324499030382,
"grad_norm": 1.8191041946411133,
"learning_rate": 0.001,
"loss": 1.2668,
"step": 239300
},
{
"epoch": 77.37556561085972,
"grad_norm": 1.5118216276168823,
"learning_rate": 0.001,
"loss": 1.2659,
"step": 239400
},
{
"epoch": 77.40788623141565,
"grad_norm": 1.0511139631271362,
"learning_rate": 0.001,
"loss": 1.2665,
"step": 239500
},
{
"epoch": 77.44020685197155,
"grad_norm": 1.2704646587371826,
"learning_rate": 0.001,
"loss": 1.2857,
"step": 239600
},
{
"epoch": 77.47252747252747,
"grad_norm": 2.2298457622528076,
"learning_rate": 0.001,
"loss": 1.2797,
"step": 239700
},
{
"epoch": 77.50484809308338,
"grad_norm": 1.5113145112991333,
"learning_rate": 0.001,
"loss": 1.2943,
"step": 239800
},
{
"epoch": 77.5371687136393,
"grad_norm": 1.9049075841903687,
"learning_rate": 0.001,
"loss": 1.287,
"step": 239900
},
{
"epoch": 77.56948933419521,
"grad_norm": 1.4290504455566406,
"learning_rate": 0.001,
"loss": 1.2926,
"step": 240000
},
{
"epoch": 77.60180995475113,
"grad_norm": 2.210345983505249,
"learning_rate": 0.001,
"loss": 1.2926,
"step": 240100
},
{
"epoch": 77.63413057530704,
"grad_norm": 1.477892518043518,
"learning_rate": 0.001,
"loss": 1.294,
"step": 240200
},
{
"epoch": 77.66645119586296,
"grad_norm": 1.285657525062561,
"learning_rate": 0.001,
"loss": 1.3229,
"step": 240300
},
{
"epoch": 77.69877181641887,
"grad_norm": 1.651516079902649,
"learning_rate": 0.001,
"loss": 1.2973,
"step": 240400
},
{
"epoch": 77.73109243697479,
"grad_norm": 1.464377760887146,
"learning_rate": 0.001,
"loss": 1.2994,
"step": 240500
},
{
"epoch": 77.7634130575307,
"grad_norm": 1.3478065729141235,
"learning_rate": 0.001,
"loss": 1.319,
"step": 240600
},
{
"epoch": 77.79573367808662,
"grad_norm": 1.2298150062561035,
"learning_rate": 0.001,
"loss": 1.3226,
"step": 240700
},
{
"epoch": 77.82805429864253,
"grad_norm": 2.020278215408325,
"learning_rate": 0.001,
"loss": 1.3284,
"step": 240800
},
{
"epoch": 77.86037491919845,
"grad_norm": 1.2401459217071533,
"learning_rate": 0.001,
"loss": 1.3085,
"step": 240900
},
{
"epoch": 77.89269553975436,
"grad_norm": 1.4142014980316162,
"learning_rate": 0.001,
"loss": 1.3168,
"step": 241000
},
{
"epoch": 77.92501616031028,
"grad_norm": 1.6408050060272217,
"learning_rate": 0.001,
"loss": 1.3243,
"step": 241100
},
{
"epoch": 77.95733678086619,
"grad_norm": 1.4897146224975586,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 241200
},
{
"epoch": 77.98965740142211,
"grad_norm": 2.0061614513397217,
"learning_rate": 0.001,
"loss": 1.3167,
"step": 241300
},
{
"epoch": 78.02197802197803,
"grad_norm": 1.3673760890960693,
"learning_rate": 0.001,
"loss": 1.2673,
"step": 241400
},
{
"epoch": 78.05429864253394,
"grad_norm": 1.8199971914291382,
"learning_rate": 0.001,
"loss": 1.2247,
"step": 241500
},
{
"epoch": 78.08661926308986,
"grad_norm": 1.2587751150131226,
"learning_rate": 0.001,
"loss": 1.2225,
"step": 241600
},
{
"epoch": 78.11893988364577,
"grad_norm": 1.8833216428756714,
"learning_rate": 0.001,
"loss": 1.2309,
"step": 241700
},
{
"epoch": 78.15126050420169,
"grad_norm": 1.4985368251800537,
"learning_rate": 0.001,
"loss": 1.2471,
"step": 241800
},
{
"epoch": 78.1835811247576,
"grad_norm": 1.5621936321258545,
"learning_rate": 0.001,
"loss": 1.2422,
"step": 241900
},
{
"epoch": 78.21590174531352,
"grad_norm": 1.7414886951446533,
"learning_rate": 0.001,
"loss": 1.2474,
"step": 242000
},
{
"epoch": 78.24822236586942,
"grad_norm": 1.4623531103134155,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 242100
},
{
"epoch": 78.28054298642535,
"grad_norm": 1.414461374282837,
"learning_rate": 0.001,
"loss": 1.265,
"step": 242200
},
{
"epoch": 78.31286360698125,
"grad_norm": 1.269479513168335,
"learning_rate": 0.001,
"loss": 1.2619,
"step": 242300
},
{
"epoch": 78.34518422753717,
"grad_norm": 1.8102675676345825,
"learning_rate": 0.001,
"loss": 1.2651,
"step": 242400
},
{
"epoch": 78.37750484809308,
"grad_norm": 1.960545539855957,
"learning_rate": 0.001,
"loss": 1.2483,
"step": 242500
},
{
"epoch": 78.409825468649,
"grad_norm": 1.1952399015426636,
"learning_rate": 0.001,
"loss": 1.2729,
"step": 242600
},
{
"epoch": 78.44214608920491,
"grad_norm": 1.357187032699585,
"learning_rate": 0.001,
"loss": 1.2678,
"step": 242700
},
{
"epoch": 78.47446670976083,
"grad_norm": 1.668985366821289,
"learning_rate": 0.001,
"loss": 1.2721,
"step": 242800
},
{
"epoch": 78.50678733031674,
"grad_norm": 1.272112250328064,
"learning_rate": 0.001,
"loss": 1.2756,
"step": 242900
},
{
"epoch": 78.53910795087266,
"grad_norm": 1.2903515100479126,
"learning_rate": 0.001,
"loss": 1.2804,
"step": 243000
},
{
"epoch": 78.57142857142857,
"grad_norm": 1.7097840309143066,
"learning_rate": 0.001,
"loss": 1.2845,
"step": 243100
},
{
"epoch": 78.60374919198449,
"grad_norm": 1.4061191082000732,
"learning_rate": 0.001,
"loss": 1.2745,
"step": 243200
},
{
"epoch": 78.6360698125404,
"grad_norm": 1.44053053855896,
"learning_rate": 0.001,
"loss": 1.2889,
"step": 243300
},
{
"epoch": 78.66839043309632,
"grad_norm": 2.1334850788116455,
"learning_rate": 0.001,
"loss": 1.282,
"step": 243400
},
{
"epoch": 78.70071105365223,
"grad_norm": 1.7297238111495972,
"learning_rate": 0.001,
"loss": 1.295,
"step": 243500
},
{
"epoch": 78.73303167420815,
"grad_norm": 1.1532487869262695,
"learning_rate": 0.001,
"loss": 1.299,
"step": 243600
},
{
"epoch": 78.76535229476406,
"grad_norm": 2.166877031326294,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 243700
},
{
"epoch": 78.79767291531998,
"grad_norm": 1.2987929582595825,
"learning_rate": 0.001,
"loss": 1.3171,
"step": 243800
},
{
"epoch": 78.82999353587589,
"grad_norm": 1.4848920106887817,
"learning_rate": 0.001,
"loss": 1.3014,
"step": 243900
},
{
"epoch": 78.86231415643181,
"grad_norm": 1.5554513931274414,
"learning_rate": 0.001,
"loss": 1.3108,
"step": 244000
},
{
"epoch": 78.89463477698771,
"grad_norm": 1.750722885131836,
"learning_rate": 0.001,
"loss": 1.3258,
"step": 244100
},
{
"epoch": 78.92695539754364,
"grad_norm": 1.9449609518051147,
"learning_rate": 0.001,
"loss": 1.3133,
"step": 244200
},
{
"epoch": 78.95927601809954,
"grad_norm": 1.9386721849441528,
"learning_rate": 0.001,
"loss": 1.3233,
"step": 244300
},
{
"epoch": 78.99159663865547,
"grad_norm": 1.9014012813568115,
"learning_rate": 0.001,
"loss": 1.3221,
"step": 244400
},
{
"epoch": 79.02391725921137,
"grad_norm": 1.2058799266815186,
"learning_rate": 0.001,
"loss": 1.236,
"step": 244500
},
{
"epoch": 79.0562378797673,
"grad_norm": 1.2706985473632812,
"learning_rate": 0.001,
"loss": 1.2129,
"step": 244600
},
{
"epoch": 79.0885585003232,
"grad_norm": 1.4728243350982666,
"learning_rate": 0.001,
"loss": 1.2243,
"step": 244700
},
{
"epoch": 79.12087912087912,
"grad_norm": 1.2134689092636108,
"learning_rate": 0.001,
"loss": 1.2325,
"step": 244800
},
{
"epoch": 79.15319974143503,
"grad_norm": 1.6454464197158813,
"learning_rate": 0.001,
"loss": 1.2395,
"step": 244900
},
{
"epoch": 79.18552036199095,
"grad_norm": 1.9662758111953735,
"learning_rate": 0.001,
"loss": 1.2388,
"step": 245000
},
{
"epoch": 79.21784098254686,
"grad_norm": 1.4569859504699707,
"learning_rate": 0.001,
"loss": 1.2291,
"step": 245100
},
{
"epoch": 79.25016160310278,
"grad_norm": 1.729219675064087,
"learning_rate": 0.001,
"loss": 1.249,
"step": 245200
},
{
"epoch": 79.28248222365869,
"grad_norm": 1.807348608970642,
"learning_rate": 0.001,
"loss": 1.2396,
"step": 245300
},
{
"epoch": 79.31480284421461,
"grad_norm": 1.5521453619003296,
"learning_rate": 0.001,
"loss": 1.2442,
"step": 245400
},
{
"epoch": 79.34712346477052,
"grad_norm": 2.2156636714935303,
"learning_rate": 0.001,
"loss": 1.2616,
"step": 245500
},
{
"epoch": 79.37944408532644,
"grad_norm": 1.3331085443496704,
"learning_rate": 0.001,
"loss": 1.2644,
"step": 245600
},
{
"epoch": 79.41176470588235,
"grad_norm": 1.475218415260315,
"learning_rate": 0.001,
"loss": 1.2595,
"step": 245700
},
{
"epoch": 79.44408532643827,
"grad_norm": 1.450708031654358,
"learning_rate": 0.001,
"loss": 1.2726,
"step": 245800
},
{
"epoch": 79.47640594699418,
"grad_norm": 1.3240978717803955,
"learning_rate": 0.001,
"loss": 1.2772,
"step": 245900
},
{
"epoch": 79.5087265675501,
"grad_norm": 2.481502056121826,
"learning_rate": 0.001,
"loss": 1.264,
"step": 246000
},
{
"epoch": 79.541047188106,
"grad_norm": 2.0001509189605713,
"learning_rate": 0.001,
"loss": 1.2684,
"step": 246100
},
{
"epoch": 79.57336780866193,
"grad_norm": 1.6823323965072632,
"learning_rate": 0.001,
"loss": 1.2738,
"step": 246200
},
{
"epoch": 79.60568842921784,
"grad_norm": 1.9127426147460938,
"learning_rate": 0.001,
"loss": 1.2745,
"step": 246300
},
{
"epoch": 79.63800904977376,
"grad_norm": 2.290731430053711,
"learning_rate": 0.001,
"loss": 1.2999,
"step": 246400
},
{
"epoch": 79.67032967032966,
"grad_norm": 1.918789029121399,
"learning_rate": 0.001,
"loss": 1.2915,
"step": 246500
},
{
"epoch": 79.70265029088559,
"grad_norm": 1.8735263347625732,
"learning_rate": 0.001,
"loss": 1.2894,
"step": 246600
},
{
"epoch": 79.7349709114415,
"grad_norm": 1.90142023563385,
"learning_rate": 0.001,
"loss": 1.2924,
"step": 246700
},
{
"epoch": 79.76729153199742,
"grad_norm": 1.792066216468811,
"learning_rate": 0.001,
"loss": 1.2807,
"step": 246800
},
{
"epoch": 79.79961215255332,
"grad_norm": 1.2293639183044434,
"learning_rate": 0.001,
"loss": 1.2857,
"step": 246900
},
{
"epoch": 79.83193277310924,
"grad_norm": 1.5063929557800293,
"learning_rate": 0.001,
"loss": 1.3013,
"step": 247000
},
{
"epoch": 79.86425339366515,
"grad_norm": 1.3179506063461304,
"learning_rate": 0.001,
"loss": 1.2913,
"step": 247100
},
{
"epoch": 79.89657401422107,
"grad_norm": 1.338300108909607,
"learning_rate": 0.001,
"loss": 1.3201,
"step": 247200
},
{
"epoch": 79.92889463477698,
"grad_norm": 1.3150955438613892,
"learning_rate": 0.001,
"loss": 1.3067,
"step": 247300
},
{
"epoch": 79.9612152553329,
"grad_norm": 1.5713889598846436,
"learning_rate": 0.001,
"loss": 1.3337,
"step": 247400
},
{
"epoch": 79.99353587588882,
"grad_norm": 1.7930541038513184,
"learning_rate": 0.001,
"loss": 1.3121,
"step": 247500
},
{
"epoch": 80.02585649644473,
"grad_norm": 1.610605001449585,
"learning_rate": 0.001,
"loss": 1.2352,
"step": 247600
},
{
"epoch": 80.05817711700065,
"grad_norm": 2.0419564247131348,
"learning_rate": 0.001,
"loss": 1.2096,
"step": 247700
},
{
"epoch": 80.09049773755656,
"grad_norm": 2.3417627811431885,
"learning_rate": 0.001,
"loss": 1.2112,
"step": 247800
},
{
"epoch": 80.12281835811248,
"grad_norm": 1.5265817642211914,
"learning_rate": 0.001,
"loss": 1.2217,
"step": 247900
},
{
"epoch": 80.15513897866839,
"grad_norm": 1.7095226049423218,
"learning_rate": 0.001,
"loss": 1.2166,
"step": 248000
},
{
"epoch": 80.18745959922431,
"grad_norm": 1.6318248510360718,
"learning_rate": 0.001,
"loss": 1.2286,
"step": 248100
},
{
"epoch": 80.21978021978022,
"grad_norm": 1.8089474439620972,
"learning_rate": 0.001,
"loss": 1.2317,
"step": 248200
},
{
"epoch": 80.25210084033614,
"grad_norm": 1.3798012733459473,
"learning_rate": 0.001,
"loss": 1.2327,
"step": 248300
},
{
"epoch": 80.28442146089205,
"grad_norm": 1.7488353252410889,
"learning_rate": 0.001,
"loss": 1.2438,
"step": 248400
},
{
"epoch": 80.31674208144797,
"grad_norm": 1.488607406616211,
"learning_rate": 0.001,
"loss": 1.25,
"step": 248500
},
{
"epoch": 80.34906270200388,
"grad_norm": 2.1224374771118164,
"learning_rate": 0.001,
"loss": 1.2424,
"step": 248600
},
{
"epoch": 80.3813833225598,
"grad_norm": 1.66317880153656,
"learning_rate": 0.001,
"loss": 1.2541,
"step": 248700
},
{
"epoch": 80.4137039431157,
"grad_norm": 1.8994399309158325,
"learning_rate": 0.001,
"loss": 1.2567,
"step": 248800
},
{
"epoch": 80.44602456367163,
"grad_norm": 1.2667827606201172,
"learning_rate": 0.001,
"loss": 1.2605,
"step": 248900
},
{
"epoch": 80.47834518422754,
"grad_norm": 1.685808539390564,
"learning_rate": 0.001,
"loss": 1.2557,
"step": 249000
},
{
"epoch": 80.51066580478346,
"grad_norm": 1.3293232917785645,
"learning_rate": 0.001,
"loss": 1.2535,
"step": 249100
},
{
"epoch": 80.54298642533936,
"grad_norm": 2.5228679180145264,
"learning_rate": 0.001,
"loss": 1.273,
"step": 249200
},
{
"epoch": 80.57530704589529,
"grad_norm": 1.449995756149292,
"learning_rate": 0.001,
"loss": 1.2904,
"step": 249300
},
{
"epoch": 80.6076276664512,
"grad_norm": 1.6519827842712402,
"learning_rate": 0.001,
"loss": 1.2546,
"step": 249400
},
{
"epoch": 80.63994828700712,
"grad_norm": 1.2943912744522095,
"learning_rate": 0.001,
"loss": 1.278,
"step": 249500
},
{
"epoch": 80.67226890756302,
"grad_norm": 1.3893470764160156,
"learning_rate": 0.001,
"loss": 1.2863,
"step": 249600
},
{
"epoch": 80.70458952811894,
"grad_norm": 1.8279205560684204,
"learning_rate": 0.001,
"loss": 1.287,
"step": 249700
},
{
"epoch": 80.73691014867485,
"grad_norm": 1.4569382667541504,
"learning_rate": 0.001,
"loss": 1.2794,
"step": 249800
},
{
"epoch": 80.76923076923077,
"grad_norm": 1.452245831489563,
"learning_rate": 0.001,
"loss": 1.2973,
"step": 249900
},
{
"epoch": 80.80155138978668,
"grad_norm": 1.566039800643921,
"learning_rate": 0.001,
"loss": 1.2916,
"step": 250000
},
{
"epoch": 80.8338720103426,
"grad_norm": 1.5842382907867432,
"learning_rate": 0.001,
"loss": 1.2887,
"step": 250100
},
{
"epoch": 80.86619263089851,
"grad_norm": 1.286839246749878,
"learning_rate": 0.001,
"loss": 1.2955,
"step": 250200
},
{
"epoch": 80.89851325145443,
"grad_norm": 1.8135418891906738,
"learning_rate": 0.001,
"loss": 1.2965,
"step": 250300
},
{
"epoch": 80.93083387201034,
"grad_norm": 1.5368353128433228,
"learning_rate": 0.001,
"loss": 1.2948,
"step": 250400
},
{
"epoch": 80.96315449256626,
"grad_norm": 1.3691718578338623,
"learning_rate": 0.001,
"loss": 1.3264,
"step": 250500
},
{
"epoch": 80.99547511312217,
"grad_norm": 1.9810742139816284,
"learning_rate": 0.001,
"loss": 1.2892,
"step": 250600
},
{
"epoch": 81.02779573367809,
"grad_norm": 1.7202798128128052,
"learning_rate": 0.001,
"loss": 1.2013,
"step": 250700
},
{
"epoch": 81.060116354234,
"grad_norm": 1.5454028844833374,
"learning_rate": 0.001,
"loss": 1.205,
"step": 250800
},
{
"epoch": 81.09243697478992,
"grad_norm": 1.8991540670394897,
"learning_rate": 0.001,
"loss": 1.2132,
"step": 250900
},
{
"epoch": 81.12475759534583,
"grad_norm": 1.3878988027572632,
"learning_rate": 0.001,
"loss": 1.2278,
"step": 251000
},
{
"epoch": 81.15707821590175,
"grad_norm": 1.6878595352172852,
"learning_rate": 0.001,
"loss": 1.2232,
"step": 251100
},
{
"epoch": 81.18939883645766,
"grad_norm": 1.8622406721115112,
"learning_rate": 0.001,
"loss": 1.2336,
"step": 251200
},
{
"epoch": 81.22171945701358,
"grad_norm": 1.7215933799743652,
"learning_rate": 0.001,
"loss": 1.2248,
"step": 251300
},
{
"epoch": 81.25404007756948,
"grad_norm": 1.5063960552215576,
"learning_rate": 0.001,
"loss": 1.2253,
"step": 251400
},
{
"epoch": 81.2863606981254,
"grad_norm": 1.4453834295272827,
"learning_rate": 0.001,
"loss": 1.23,
"step": 251500
},
{
"epoch": 81.31868131868131,
"grad_norm": 2.0164635181427,
"learning_rate": 0.001,
"loss": 1.2498,
"step": 251600
},
{
"epoch": 81.35100193923724,
"grad_norm": 1.4843008518218994,
"learning_rate": 0.001,
"loss": 1.213,
"step": 251700
},
{
"epoch": 81.38332255979314,
"grad_norm": 1.3511826992034912,
"learning_rate": 0.001,
"loss": 1.2365,
"step": 251800
},
{
"epoch": 81.41564318034906,
"grad_norm": 1.81965970993042,
"learning_rate": 0.001,
"loss": 1.2525,
"step": 251900
},
{
"epoch": 81.44796380090497,
"grad_norm": 1.1546568870544434,
"learning_rate": 0.001,
"loss": 1.2577,
"step": 252000
},
{
"epoch": 81.4802844214609,
"grad_norm": 1.674634575843811,
"learning_rate": 0.001,
"loss": 1.2677,
"step": 252100
},
{
"epoch": 81.5126050420168,
"grad_norm": 1.6290024518966675,
"learning_rate": 0.001,
"loss": 1.248,
"step": 252200
},
{
"epoch": 81.54492566257272,
"grad_norm": 1.4020588397979736,
"learning_rate": 0.001,
"loss": 1.258,
"step": 252300
},
{
"epoch": 81.57724628312863,
"grad_norm": 1.9796793460845947,
"learning_rate": 0.001,
"loss": 1.266,
"step": 252400
},
{
"epoch": 81.60956690368455,
"grad_norm": 2.3626046180725098,
"learning_rate": 0.001,
"loss": 1.287,
"step": 252500
},
{
"epoch": 81.64188752424046,
"grad_norm": 1.972294807434082,
"learning_rate": 0.001,
"loss": 1.2659,
"step": 252600
},
{
"epoch": 81.67420814479638,
"grad_norm": 1.8454577922821045,
"learning_rate": 0.001,
"loss": 1.2686,
"step": 252700
},
{
"epoch": 81.70652876535229,
"grad_norm": 1.5707216262817383,
"learning_rate": 0.001,
"loss": 1.2645,
"step": 252800
},
{
"epoch": 81.73884938590821,
"grad_norm": 2.1897661685943604,
"learning_rate": 0.001,
"loss": 1.2955,
"step": 252900
},
{
"epoch": 81.77117000646412,
"grad_norm": 2.0576090812683105,
"learning_rate": 0.001,
"loss": 1.2879,
"step": 253000
},
{
"epoch": 81.80349062702004,
"grad_norm": 1.7895479202270508,
"learning_rate": 0.001,
"loss": 1.2828,
"step": 253100
},
{
"epoch": 81.83581124757595,
"grad_norm": 2.119438886642456,
"learning_rate": 0.001,
"loss": 1.2859,
"step": 253200
},
{
"epoch": 81.86813186813187,
"grad_norm": 1.8811088800430298,
"learning_rate": 0.001,
"loss": 1.3123,
"step": 253300
},
{
"epoch": 81.90045248868778,
"grad_norm": 1.4409716129302979,
"learning_rate": 0.001,
"loss": 1.2987,
"step": 253400
},
{
"epoch": 81.9327731092437,
"grad_norm": 1.7243212461471558,
"learning_rate": 0.001,
"loss": 1.2764,
"step": 253500
},
{
"epoch": 81.9650937297996,
"grad_norm": 1.9806982278823853,
"learning_rate": 0.001,
"loss": 1.2895,
"step": 253600
},
{
"epoch": 81.99741435035553,
"grad_norm": 1.9793446063995361,
"learning_rate": 0.001,
"loss": 1.2932,
"step": 253700
},
{
"epoch": 82.02973497091145,
"grad_norm": 1.556868553161621,
"learning_rate": 0.001,
"loss": 1.2066,
"step": 253800
},
{
"epoch": 82.06205559146736,
"grad_norm": 2.289224147796631,
"learning_rate": 0.001,
"loss": 1.1951,
"step": 253900
},
{
"epoch": 82.09437621202328,
"grad_norm": 2.0421366691589355,
"learning_rate": 0.001,
"loss": 1.1905,
"step": 254000
},
{
"epoch": 82.12669683257919,
"grad_norm": 1.501111388206482,
"learning_rate": 0.001,
"loss": 1.2058,
"step": 254100
},
{
"epoch": 82.1590174531351,
"grad_norm": 1.870614767074585,
"learning_rate": 0.001,
"loss": 1.2213,
"step": 254200
},
{
"epoch": 82.19133807369101,
"grad_norm": 1.9706679582595825,
"learning_rate": 0.001,
"loss": 1.2093,
"step": 254300
},
{
"epoch": 82.22365869424694,
"grad_norm": 1.7341151237487793,
"learning_rate": 0.001,
"loss": 1.2193,
"step": 254400
},
{
"epoch": 82.25597931480284,
"grad_norm": 1.816436529159546,
"learning_rate": 0.001,
"loss": 1.2146,
"step": 254500
},
{
"epoch": 82.28829993535876,
"grad_norm": 1.837285041809082,
"learning_rate": 0.001,
"loss": 1.2314,
"step": 254600
},
{
"epoch": 82.32062055591467,
"grad_norm": 1.6776964664459229,
"learning_rate": 0.001,
"loss": 1.222,
"step": 254700
},
{
"epoch": 82.3529411764706,
"grad_norm": 1.6069055795669556,
"learning_rate": 0.001,
"loss": 1.2429,
"step": 254800
},
{
"epoch": 82.3852617970265,
"grad_norm": 1.5101720094680786,
"learning_rate": 0.001,
"loss": 1.2366,
"step": 254900
},
{
"epoch": 82.41758241758242,
"grad_norm": 2.5875556468963623,
"learning_rate": 0.001,
"loss": 1.2479,
"step": 255000
},
{
"epoch": 82.44990303813833,
"grad_norm": 2.130882740020752,
"learning_rate": 0.001,
"loss": 1.2515,
"step": 255100
},
{
"epoch": 82.48222365869425,
"grad_norm": 1.422727346420288,
"learning_rate": 0.001,
"loss": 1.2506,
"step": 255200
},
{
"epoch": 82.51454427925016,
"grad_norm": 1.9222614765167236,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 255300
},
{
"epoch": 82.54686489980608,
"grad_norm": 3.0412681102752686,
"learning_rate": 0.001,
"loss": 1.2663,
"step": 255400
},
{
"epoch": 82.57918552036199,
"grad_norm": 1.837545394897461,
"learning_rate": 0.001,
"loss": 1.2609,
"step": 255500
},
{
"epoch": 82.61150614091791,
"grad_norm": 2.1308412551879883,
"learning_rate": 0.001,
"loss": 1.2568,
"step": 255600
},
{
"epoch": 82.64382676147382,
"grad_norm": 2.1382863521575928,
"learning_rate": 0.001,
"loss": 1.2525,
"step": 255700
},
{
"epoch": 82.67614738202974,
"grad_norm": 1.825967788696289,
"learning_rate": 0.001,
"loss": 1.2605,
"step": 255800
},
{
"epoch": 82.70846800258565,
"grad_norm": 1.8287477493286133,
"learning_rate": 0.001,
"loss": 1.2894,
"step": 255900
},
{
"epoch": 82.74078862314157,
"grad_norm": 1.9624981880187988,
"learning_rate": 0.001,
"loss": 1.2688,
"step": 256000
},
{
"epoch": 82.77310924369748,
"grad_norm": 2.4718258380889893,
"learning_rate": 0.001,
"loss": 1.2777,
"step": 256100
},
{
"epoch": 82.8054298642534,
"grad_norm": 1.8880257606506348,
"learning_rate": 0.001,
"loss": 1.2586,
"step": 256200
},
{
"epoch": 82.8377504848093,
"grad_norm": 1.5955743789672852,
"learning_rate": 0.001,
"loss": 1.2769,
"step": 256300
},
{
"epoch": 82.87007110536523,
"grad_norm": 1.7947994470596313,
"learning_rate": 0.001,
"loss": 1.2757,
"step": 256400
},
{
"epoch": 82.90239172592113,
"grad_norm": 2.2135133743286133,
"learning_rate": 0.001,
"loss": 1.2686,
"step": 256500
},
{
"epoch": 82.93471234647706,
"grad_norm": 1.7485551834106445,
"learning_rate": 0.001,
"loss": 1.2962,
"step": 256600
},
{
"epoch": 82.96703296703296,
"grad_norm": 1.9167695045471191,
"learning_rate": 0.001,
"loss": 1.2905,
"step": 256700
},
{
"epoch": 82.99935358758889,
"grad_norm": 1.720075249671936,
"learning_rate": 0.001,
"loss": 1.2713,
"step": 256800
},
{
"epoch": 83.03167420814479,
"grad_norm": 2.292396306991577,
"learning_rate": 0.001,
"loss": 1.1707,
"step": 256900
},
{
"epoch": 83.06399482870071,
"grad_norm": 1.627952218055725,
"learning_rate": 0.001,
"loss": 1.1873,
"step": 257000
},
{
"epoch": 83.09631544925662,
"grad_norm": 1.960242509841919,
"learning_rate": 0.001,
"loss": 1.1937,
"step": 257100
},
{
"epoch": 83.12863606981254,
"grad_norm": 1.4049196243286133,
"learning_rate": 0.001,
"loss": 1.2042,
"step": 257200
},
{
"epoch": 83.16095669036845,
"grad_norm": 2.174247980117798,
"learning_rate": 0.001,
"loss": 1.2041,
"step": 257300
},
{
"epoch": 83.19327731092437,
"grad_norm": 1.7446444034576416,
"learning_rate": 0.001,
"loss": 1.2162,
"step": 257400
},
{
"epoch": 83.22559793148028,
"grad_norm": 2.2466647624969482,
"learning_rate": 0.001,
"loss": 1.1843,
"step": 257500
},
{
"epoch": 83.2579185520362,
"grad_norm": 2.1880221366882324,
"learning_rate": 0.001,
"loss": 1.2089,
"step": 257600
},
{
"epoch": 83.29023917259211,
"grad_norm": 1.840790867805481,
"learning_rate": 0.001,
"loss": 1.2136,
"step": 257700
},
{
"epoch": 83.32255979314803,
"grad_norm": 1.8245068788528442,
"learning_rate": 0.001,
"loss": 1.2295,
"step": 257800
},
{
"epoch": 83.35488041370394,
"grad_norm": 2.5724518299102783,
"learning_rate": 0.001,
"loss": 1.2236,
"step": 257900
},
{
"epoch": 83.38720103425986,
"grad_norm": 1.9020832777023315,
"learning_rate": 0.001,
"loss": 1.2394,
"step": 258000
},
{
"epoch": 83.41952165481577,
"grad_norm": 1.76996648311615,
"learning_rate": 0.001,
"loss": 1.2406,
"step": 258100
},
{
"epoch": 83.45184227537169,
"grad_norm": 2.1938960552215576,
"learning_rate": 0.001,
"loss": 1.2445,
"step": 258200
},
{
"epoch": 83.4841628959276,
"grad_norm": 2.2984790802001953,
"learning_rate": 0.001,
"loss": 1.2505,
"step": 258300
},
{
"epoch": 83.51648351648352,
"grad_norm": 1.9285473823547363,
"learning_rate": 0.001,
"loss": 1.266,
"step": 258400
},
{
"epoch": 83.54880413703943,
"grad_norm": 1.770233392715454,
"learning_rate": 0.001,
"loss": 1.2508,
"step": 258500
},
{
"epoch": 83.58112475759535,
"grad_norm": 1.5901305675506592,
"learning_rate": 0.001,
"loss": 1.2521,
"step": 258600
},
{
"epoch": 83.61344537815125,
"grad_norm": 2.1101291179656982,
"learning_rate": 0.001,
"loss": 1.2566,
"step": 258700
},
{
"epoch": 83.64576599870718,
"grad_norm": 2.0092530250549316,
"learning_rate": 0.001,
"loss": 1.2601,
"step": 258800
},
{
"epoch": 83.67808661926308,
"grad_norm": 2.428415298461914,
"learning_rate": 0.001,
"loss": 1.272,
"step": 258900
},
{
"epoch": 83.710407239819,
"grad_norm": 2.0060224533081055,
"learning_rate": 0.001,
"loss": 1.2866,
"step": 259000
},
{
"epoch": 83.74272786037491,
"grad_norm": 3.634275436401367,
"learning_rate": 0.001,
"loss": 1.2685,
"step": 259100
},
{
"epoch": 83.77504848093083,
"grad_norm": 1.6629226207733154,
"learning_rate": 0.001,
"loss": 1.2629,
"step": 259200
},
{
"epoch": 83.80736910148674,
"grad_norm": 2.2785284519195557,
"learning_rate": 0.001,
"loss": 1.2617,
"step": 259300
},
{
"epoch": 83.83968972204266,
"grad_norm": 2.1297974586486816,
"learning_rate": 0.001,
"loss": 1.2808,
"step": 259400
},
{
"epoch": 83.87201034259857,
"grad_norm": 1.719080924987793,
"learning_rate": 0.001,
"loss": 1.2701,
"step": 259500
},
{
"epoch": 83.9043309631545,
"grad_norm": 2.1362926959991455,
"learning_rate": 0.001,
"loss": 1.2789,
"step": 259600
},
{
"epoch": 83.9366515837104,
"grad_norm": 1.7237249612808228,
"learning_rate": 0.001,
"loss": 1.2776,
"step": 259700
},
{
"epoch": 83.96897220426632,
"grad_norm": 2.223992347717285,
"learning_rate": 0.001,
"loss": 1.2811,
"step": 259800
},
{
"epoch": 84.00129282482224,
"grad_norm": 1.4170386791229248,
"learning_rate": 0.001,
"loss": 1.2828,
"step": 259900
},
{
"epoch": 84.03361344537815,
"grad_norm": 1.6659284830093384,
"learning_rate": 0.001,
"loss": 1.1881,
"step": 260000
},
{
"epoch": 84.06593406593407,
"grad_norm": 2.675107002258301,
"learning_rate": 0.001,
"loss": 1.1863,
"step": 260100
},
{
"epoch": 84.09825468648998,
"grad_norm": 1.7558014392852783,
"learning_rate": 0.001,
"loss": 1.1711,
"step": 260200
},
{
"epoch": 84.1305753070459,
"grad_norm": 1.7978248596191406,
"learning_rate": 0.001,
"loss": 1.1957,
"step": 260300
},
{
"epoch": 84.16289592760181,
"grad_norm": 1.5442531108856201,
"learning_rate": 0.001,
"loss": 1.2006,
"step": 260400
},
{
"epoch": 84.19521654815773,
"grad_norm": 1.9399360418319702,
"learning_rate": 0.001,
"loss": 1.2176,
"step": 260500
},
{
"epoch": 84.22753716871364,
"grad_norm": 1.6368488073349,
"learning_rate": 0.001,
"loss": 1.2071,
"step": 260600
},
{
"epoch": 84.25985778926956,
"grad_norm": 1.65631103515625,
"learning_rate": 0.001,
"loss": 1.1956,
"step": 260700
},
{
"epoch": 84.29217840982547,
"grad_norm": 1.7855615615844727,
"learning_rate": 0.001,
"loss": 1.2152,
"step": 260800
},
{
"epoch": 84.32449903038139,
"grad_norm": 1.879173755645752,
"learning_rate": 0.001,
"loss": 1.2267,
"step": 260900
},
{
"epoch": 84.3568196509373,
"grad_norm": 2.325134038925171,
"learning_rate": 0.001,
"loss": 1.2399,
"step": 261000
},
{
"epoch": 84.38914027149322,
"grad_norm": 1.6414111852645874,
"learning_rate": 0.001,
"loss": 1.2183,
"step": 261100
},
{
"epoch": 84.42146089204913,
"grad_norm": 1.7910301685333252,
"learning_rate": 0.001,
"loss": 1.2269,
"step": 261200
},
{
"epoch": 84.45378151260505,
"grad_norm": 1.5484957695007324,
"learning_rate": 0.001,
"loss": 1.2264,
"step": 261300
},
{
"epoch": 84.48610213316095,
"grad_norm": 2.113661527633667,
"learning_rate": 0.001,
"loss": 1.2232,
"step": 261400
},
{
"epoch": 84.51842275371688,
"grad_norm": 1.4829312562942505,
"learning_rate": 0.001,
"loss": 1.2486,
"step": 261500
},
{
"epoch": 84.55074337427278,
"grad_norm": 1.8122377395629883,
"learning_rate": 0.001,
"loss": 1.232,
"step": 261600
},
{
"epoch": 84.5830639948287,
"grad_norm": 1.4297584295272827,
"learning_rate": 0.001,
"loss": 1.2386,
"step": 261700
},
{
"epoch": 84.61538461538461,
"grad_norm": 2.0205373764038086,
"learning_rate": 0.001,
"loss": 1.2506,
"step": 261800
},
{
"epoch": 84.64770523594053,
"grad_norm": 1.655685544013977,
"learning_rate": 0.001,
"loss": 1.2424,
"step": 261900
},
{
"epoch": 84.68002585649644,
"grad_norm": 2.893658399581909,
"learning_rate": 0.001,
"loss": 1.2462,
"step": 262000
},
{
"epoch": 84.71234647705236,
"grad_norm": 2.2161877155303955,
"learning_rate": 0.001,
"loss": 1.2411,
"step": 262100
},
{
"epoch": 84.74466709760827,
"grad_norm": 1.4535630941390991,
"learning_rate": 0.001,
"loss": 1.2521,
"step": 262200
},
{
"epoch": 84.7769877181642,
"grad_norm": 2.2844653129577637,
"learning_rate": 0.001,
"loss": 1.2678,
"step": 262300
},
{
"epoch": 84.8093083387201,
"grad_norm": 2.162257432937622,
"learning_rate": 0.001,
"loss": 1.2596,
"step": 262400
},
{
"epoch": 84.84162895927602,
"grad_norm": 1.8454498052597046,
"learning_rate": 0.001,
"loss": 1.2757,
"step": 262500
},
{
"epoch": 84.87394957983193,
"grad_norm": 1.4504950046539307,
"learning_rate": 0.001,
"loss": 1.2718,
"step": 262600
},
{
"epoch": 84.90627020038785,
"grad_norm": 1.8487071990966797,
"learning_rate": 0.001,
"loss": 1.2746,
"step": 262700
},
{
"epoch": 84.93859082094376,
"grad_norm": 2.1302402019500732,
"learning_rate": 0.001,
"loss": 1.2723,
"step": 262800
},
{
"epoch": 84.97091144149968,
"grad_norm": 2.300407648086548,
"learning_rate": 0.001,
"loss": 1.2888,
"step": 262900
},
{
"epoch": 85.00323206205559,
"grad_norm": 1.430111050605774,
"learning_rate": 0.001,
"loss": 1.2833,
"step": 263000
},
{
"epoch": 85.03555268261151,
"grad_norm": 1.5040006637573242,
"learning_rate": 0.001,
"loss": 1.177,
"step": 263100
},
{
"epoch": 85.06787330316742,
"grad_norm": 1.4833869934082031,
"learning_rate": 0.001,
"loss": 1.1778,
"step": 263200
},
{
"epoch": 85.10019392372334,
"grad_norm": 1.8925707340240479,
"learning_rate": 0.001,
"loss": 1.1796,
"step": 263300
},
{
"epoch": 85.13251454427925,
"grad_norm": 1.7188044786453247,
"learning_rate": 0.001,
"loss": 1.1676,
"step": 263400
},
{
"epoch": 85.16483516483517,
"grad_norm": 1.8766733407974243,
"learning_rate": 0.001,
"loss": 1.1869,
"step": 263500
},
{
"epoch": 85.19715578539108,
"grad_norm": 1.5815516710281372,
"learning_rate": 0.001,
"loss": 1.1921,
"step": 263600
},
{
"epoch": 85.229476405947,
"grad_norm": 2.0210530757904053,
"learning_rate": 0.001,
"loss": 1.2048,
"step": 263700
},
{
"epoch": 85.2617970265029,
"grad_norm": 1.6629260778427124,
"learning_rate": 0.001,
"loss": 1.2358,
"step": 263800
},
{
"epoch": 85.29411764705883,
"grad_norm": 1.2752668857574463,
"learning_rate": 0.001,
"loss": 1.2006,
"step": 263900
},
{
"epoch": 85.32643826761473,
"grad_norm": 1.6782549619674683,
"learning_rate": 0.001,
"loss": 1.2229,
"step": 264000
},
{
"epoch": 85.35875888817066,
"grad_norm": 1.7548514604568481,
"learning_rate": 0.001,
"loss": 1.2279,
"step": 264100
},
{
"epoch": 85.39107950872656,
"grad_norm": 1.534002423286438,
"learning_rate": 0.001,
"loss": 1.2161,
"step": 264200
},
{
"epoch": 85.42340012928248,
"grad_norm": 1.6671688556671143,
"learning_rate": 0.001,
"loss": 1.23,
"step": 264300
},
{
"epoch": 85.45572074983839,
"grad_norm": 1.7113094329833984,
"learning_rate": 0.001,
"loss": 1.2288,
"step": 264400
},
{
"epoch": 85.48804137039431,
"grad_norm": 1.709190011024475,
"learning_rate": 0.001,
"loss": 1.2163,
"step": 264500
},
{
"epoch": 85.52036199095022,
"grad_norm": 1.7736314535140991,
"learning_rate": 0.001,
"loss": 1.2368,
"step": 264600
},
{
"epoch": 85.55268261150614,
"grad_norm": 1.317912220954895,
"learning_rate": 0.001,
"loss": 1.2442,
"step": 264700
},
{
"epoch": 85.58500323206205,
"grad_norm": 1.6481040716171265,
"learning_rate": 0.001,
"loss": 1.2399,
"step": 264800
},
{
"epoch": 85.61732385261797,
"grad_norm": 1.806247591972351,
"learning_rate": 0.001,
"loss": 1.2449,
"step": 264900
},
{
"epoch": 85.64964447317388,
"grad_norm": 1.9776191711425781,
"learning_rate": 0.001,
"loss": 1.2268,
"step": 265000
},
{
"epoch": 85.6819650937298,
"grad_norm": 1.5968152284622192,
"learning_rate": 0.001,
"loss": 1.2466,
"step": 265100
},
{
"epoch": 85.71428571428571,
"grad_norm": 1.5977829694747925,
"learning_rate": 0.001,
"loss": 1.2251,
"step": 265200
},
{
"epoch": 85.74660633484163,
"grad_norm": 1.4744341373443604,
"learning_rate": 0.001,
"loss": 1.2618,
"step": 265300
},
{
"epoch": 85.77892695539754,
"grad_norm": 1.6004712581634521,
"learning_rate": 0.001,
"loss": 1.2541,
"step": 265400
},
{
"epoch": 85.81124757595346,
"grad_norm": 1.8422083854675293,
"learning_rate": 0.001,
"loss": 1.2681,
"step": 265500
},
{
"epoch": 85.84356819650937,
"grad_norm": 1.2561287879943848,
"learning_rate": 0.001,
"loss": 1.2552,
"step": 265600
},
{
"epoch": 85.87588881706529,
"grad_norm": 1.9980125427246094,
"learning_rate": 0.001,
"loss": 1.265,
"step": 265700
},
{
"epoch": 85.9082094376212,
"grad_norm": 1.6248669624328613,
"learning_rate": 0.001,
"loss": 1.2585,
"step": 265800
},
{
"epoch": 85.94053005817712,
"grad_norm": 1.5957036018371582,
"learning_rate": 0.001,
"loss": 1.2756,
"step": 265900
},
{
"epoch": 85.97285067873302,
"grad_norm": 1.9643137454986572,
"learning_rate": 0.001,
"loss": 1.2729,
"step": 266000
},
{
"epoch": 86.00517129928895,
"grad_norm": 1.8060412406921387,
"learning_rate": 0.001,
"loss": 1.2612,
"step": 266100
},
{
"epoch": 86.03749191984487,
"grad_norm": 1.5058979988098145,
"learning_rate": 0.001,
"loss": 1.1647,
"step": 266200
},
{
"epoch": 86.06981254040078,
"grad_norm": 1.47739839553833,
"learning_rate": 0.001,
"loss": 1.1664,
"step": 266300
},
{
"epoch": 86.1021331609567,
"grad_norm": 1.7648849487304688,
"learning_rate": 0.001,
"loss": 1.1676,
"step": 266400
},
{
"epoch": 86.1344537815126,
"grad_norm": 2.075975179672241,
"learning_rate": 0.001,
"loss": 1.1901,
"step": 266500
},
{
"epoch": 86.16677440206853,
"grad_norm": 1.394399642944336,
"learning_rate": 0.001,
"loss": 1.1891,
"step": 266600
},
{
"epoch": 86.19909502262443,
"grad_norm": 1.4004300832748413,
"learning_rate": 0.001,
"loss": 1.1836,
"step": 266700
},
{
"epoch": 86.23141564318036,
"grad_norm": 2.235933303833008,
"learning_rate": 0.001,
"loss": 1.1876,
"step": 266800
},
{
"epoch": 86.26373626373626,
"grad_norm": 1.395911693572998,
"learning_rate": 0.001,
"loss": 1.2057,
"step": 266900
},
{
"epoch": 86.29605688429218,
"grad_norm": 1.5776115655899048,
"learning_rate": 0.001,
"loss": 1.2151,
"step": 267000
},
{
"epoch": 86.32837750484809,
"grad_norm": 2.331512451171875,
"learning_rate": 0.001,
"loss": 1.204,
"step": 267100
},
{
"epoch": 86.36069812540401,
"grad_norm": 1.41651451587677,
"learning_rate": 0.001,
"loss": 1.2105,
"step": 267200
},
{
"epoch": 86.39301874595992,
"grad_norm": 1.1959208250045776,
"learning_rate": 0.001,
"loss": 1.2182,
"step": 267300
},
{
"epoch": 86.42533936651584,
"grad_norm": 1.5867549180984497,
"learning_rate": 0.001,
"loss": 1.226,
"step": 267400
},
{
"epoch": 86.45765998707175,
"grad_norm": 1.6116405725479126,
"learning_rate": 0.001,
"loss": 1.2006,
"step": 267500
},
{
"epoch": 86.48998060762767,
"grad_norm": 1.4707088470458984,
"learning_rate": 0.001,
"loss": 1.2255,
"step": 267600
},
{
"epoch": 86.52230122818358,
"grad_norm": 1.641430139541626,
"learning_rate": 0.001,
"loss": 1.2294,
"step": 267700
},
{
"epoch": 86.5546218487395,
"grad_norm": 1.7087689638137817,
"learning_rate": 0.001,
"loss": 1.2211,
"step": 267800
},
{
"epoch": 86.58694246929541,
"grad_norm": 1.8745614290237427,
"learning_rate": 0.001,
"loss": 1.2417,
"step": 267900
},
{
"epoch": 86.61926308985133,
"grad_norm": 1.6811854839324951,
"learning_rate": 0.001,
"loss": 1.2286,
"step": 268000
},
{
"epoch": 86.65158371040724,
"grad_norm": 1.4782730340957642,
"learning_rate": 0.001,
"loss": 1.2382,
"step": 268100
},
{
"epoch": 86.68390433096316,
"grad_norm": 1.927930235862732,
"learning_rate": 0.001,
"loss": 1.2499,
"step": 268200
},
{
"epoch": 86.71622495151907,
"grad_norm": 1.6725027561187744,
"learning_rate": 0.001,
"loss": 1.2356,
"step": 268300
},
{
"epoch": 86.74854557207499,
"grad_norm": 1.7192740440368652,
"learning_rate": 0.001,
"loss": 1.2512,
"step": 268400
},
{
"epoch": 86.7808661926309,
"grad_norm": 1.5314967632293701,
"learning_rate": 0.001,
"loss": 1.2408,
"step": 268500
},
{
"epoch": 86.81318681318682,
"grad_norm": 1.8034896850585938,
"learning_rate": 0.001,
"loss": 1.2595,
"step": 268600
},
{
"epoch": 86.84550743374272,
"grad_norm": 1.245100975036621,
"learning_rate": 0.001,
"loss": 1.2543,
"step": 268700
},
{
"epoch": 86.87782805429865,
"grad_norm": 2.7486352920532227,
"learning_rate": 0.001,
"loss": 1.2456,
"step": 268800
},
{
"epoch": 86.91014867485455,
"grad_norm": 1.4387123584747314,
"learning_rate": 0.001,
"loss": 1.2536,
"step": 268900
},
{
"epoch": 86.94246929541048,
"grad_norm": 1.3550703525543213,
"learning_rate": 0.001,
"loss": 1.2516,
"step": 269000
},
{
"epoch": 86.97478991596638,
"grad_norm": 1.9709012508392334,
"learning_rate": 0.001,
"loss": 1.2724,
"step": 269100
},
{
"epoch": 87.0071105365223,
"grad_norm": 1.8654589653015137,
"learning_rate": 0.001,
"loss": 1.2617,
"step": 269200
},
{
"epoch": 87.03943115707821,
"grad_norm": 1.6749591827392578,
"learning_rate": 0.001,
"loss": 1.1557,
"step": 269300
},
{
"epoch": 87.07175177763413,
"grad_norm": 1.3441767692565918,
"learning_rate": 0.001,
"loss": 1.1659,
"step": 269400
},
{
"epoch": 87.10407239819004,
"grad_norm": 1.4837168455123901,
"learning_rate": 0.001,
"loss": 1.1694,
"step": 269500
},
{
"epoch": 87.13639301874596,
"grad_norm": 1.3991222381591797,
"learning_rate": 0.001,
"loss": 1.179,
"step": 269600
},
{
"epoch": 87.16871363930187,
"grad_norm": 1.1834875345230103,
"learning_rate": 0.001,
"loss": 1.1607,
"step": 269700
},
{
"epoch": 87.20103425985779,
"grad_norm": 1.8037092685699463,
"learning_rate": 0.001,
"loss": 1.1802,
"step": 269800
},
{
"epoch": 87.2333548804137,
"grad_norm": 1.2468910217285156,
"learning_rate": 0.001,
"loss": 1.1921,
"step": 269900
},
{
"epoch": 87.26567550096962,
"grad_norm": 1.94520902633667,
"learning_rate": 0.001,
"loss": 1.1904,
"step": 270000
},
{
"epoch": 87.29799612152553,
"grad_norm": 1.4701403379440308,
"learning_rate": 0.001,
"loss": 1.202,
"step": 270100
},
{
"epoch": 87.33031674208145,
"grad_norm": 1.37777841091156,
"learning_rate": 0.001,
"loss": 1.2069,
"step": 270200
},
{
"epoch": 87.36263736263736,
"grad_norm": 1.7511200904846191,
"learning_rate": 0.001,
"loss": 1.2119,
"step": 270300
},
{
"epoch": 87.39495798319328,
"grad_norm": 1.3382188081741333,
"learning_rate": 0.001,
"loss": 1.2133,
"step": 270400
},
{
"epoch": 87.42727860374919,
"grad_norm": 1.5276641845703125,
"learning_rate": 0.001,
"loss": 1.1879,
"step": 270500
},
{
"epoch": 87.45959922430511,
"grad_norm": 1.6079522371292114,
"learning_rate": 0.001,
"loss": 1.219,
"step": 270600
},
{
"epoch": 87.49191984486102,
"grad_norm": 1.564125657081604,
"learning_rate": 0.001,
"loss": 1.2191,
"step": 270700
},
{
"epoch": 87.52424046541694,
"grad_norm": 1.3489891290664673,
"learning_rate": 0.001,
"loss": 1.2099,
"step": 270800
},
{
"epoch": 87.55656108597285,
"grad_norm": 1.456994891166687,
"learning_rate": 0.001,
"loss": 1.2107,
"step": 270900
},
{
"epoch": 87.58888170652877,
"grad_norm": 1.3025860786437988,
"learning_rate": 0.001,
"loss": 1.2268,
"step": 271000
},
{
"epoch": 87.62120232708467,
"grad_norm": 3.015007972717285,
"learning_rate": 0.001,
"loss": 1.2385,
"step": 271100
},
{
"epoch": 87.6535229476406,
"grad_norm": 1.3912714719772339,
"learning_rate": 0.001,
"loss": 1.2338,
"step": 271200
},
{
"epoch": 87.6858435681965,
"grad_norm": 1.419293761253357,
"learning_rate": 0.001,
"loss": 1.2603,
"step": 271300
},
{
"epoch": 87.71816418875243,
"grad_norm": 1.6812208890914917,
"learning_rate": 0.001,
"loss": 1.2414,
"step": 271400
},
{
"epoch": 87.75048480930833,
"grad_norm": 1.2488850355148315,
"learning_rate": 0.001,
"loss": 1.2462,
"step": 271500
},
{
"epoch": 87.78280542986425,
"grad_norm": 1.5979465246200562,
"learning_rate": 0.001,
"loss": 1.2491,
"step": 271600
},
{
"epoch": 87.81512605042016,
"grad_norm": 1.653700351715088,
"learning_rate": 0.001,
"loss": 1.2597,
"step": 271700
},
{
"epoch": 87.84744667097608,
"grad_norm": 1.933099389076233,
"learning_rate": 0.001,
"loss": 1.2401,
"step": 271800
},
{
"epoch": 87.87976729153199,
"grad_norm": 1.5535463094711304,
"learning_rate": 0.001,
"loss": 1.23,
"step": 271900
},
{
"epoch": 87.91208791208791,
"grad_norm": 1.6965819597244263,
"learning_rate": 0.001,
"loss": 1.2601,
"step": 272000
},
{
"epoch": 87.94440853264382,
"grad_norm": 1.4667446613311768,
"learning_rate": 0.001,
"loss": 1.2626,
"step": 272100
},
{
"epoch": 87.97672915319974,
"grad_norm": 2.0227420330047607,
"learning_rate": 0.001,
"loss": 1.2639,
"step": 272200
},
{
"epoch": 88.00904977375566,
"grad_norm": 1.4765963554382324,
"learning_rate": 0.001,
"loss": 1.2238,
"step": 272300
},
{
"epoch": 88.04137039431157,
"grad_norm": 1.793649435043335,
"learning_rate": 0.001,
"loss": 1.1591,
"step": 272400
},
{
"epoch": 88.07369101486749,
"grad_norm": 2.400162696838379,
"learning_rate": 0.001,
"loss": 1.1521,
"step": 272500
},
{
"epoch": 88.1060116354234,
"grad_norm": 1.7438671588897705,
"learning_rate": 0.001,
"loss": 1.1577,
"step": 272600
},
{
"epoch": 88.13833225597932,
"grad_norm": 1.7920339107513428,
"learning_rate": 0.001,
"loss": 1.184,
"step": 272700
},
{
"epoch": 88.17065287653523,
"grad_norm": 1.8441482782363892,
"learning_rate": 0.001,
"loss": 1.1824,
"step": 272800
},
{
"epoch": 88.20297349709115,
"grad_norm": 1.4053528308868408,
"learning_rate": 0.001,
"loss": 1.1754,
"step": 272900
},
{
"epoch": 88.23529411764706,
"grad_norm": 1.9038269519805908,
"learning_rate": 0.001,
"loss": 1.1905,
"step": 273000
},
{
"epoch": 88.26761473820298,
"grad_norm": 1.4530909061431885,
"learning_rate": 0.001,
"loss": 1.1835,
"step": 273100
},
{
"epoch": 88.29993535875889,
"grad_norm": 1.4575316905975342,
"learning_rate": 0.001,
"loss": 1.1903,
"step": 273200
},
{
"epoch": 88.33225597931481,
"grad_norm": 1.709861159324646,
"learning_rate": 0.001,
"loss": 1.198,
"step": 273300
},
{
"epoch": 88.36457659987072,
"grad_norm": 1.7516298294067383,
"learning_rate": 0.001,
"loss": 1.1878,
"step": 273400
},
{
"epoch": 88.39689722042664,
"grad_norm": 1.1928800344467163,
"learning_rate": 0.001,
"loss": 1.1845,
"step": 273500
},
{
"epoch": 88.42921784098255,
"grad_norm": 1.709099292755127,
"learning_rate": 0.001,
"loss": 1.2039,
"step": 273600
},
{
"epoch": 88.46153846153847,
"grad_norm": 1.385904312133789,
"learning_rate": 0.001,
"loss": 1.2043,
"step": 273700
},
{
"epoch": 88.49385908209437,
"grad_norm": 1.6170157194137573,
"learning_rate": 0.001,
"loss": 1.204,
"step": 273800
},
{
"epoch": 88.5261797026503,
"grad_norm": 2.177598476409912,
"learning_rate": 0.001,
"loss": 1.2059,
"step": 273900
},
{
"epoch": 88.5585003232062,
"grad_norm": 1.483108639717102,
"learning_rate": 0.001,
"loss": 1.2129,
"step": 274000
},
{
"epoch": 88.59082094376213,
"grad_norm": 1.579835295677185,
"learning_rate": 0.001,
"loss": 1.2201,
"step": 274100
},
{
"epoch": 88.62314156431803,
"grad_norm": 1.5038883686065674,
"learning_rate": 0.001,
"loss": 1.2275,
"step": 274200
},
{
"epoch": 88.65546218487395,
"grad_norm": 1.4175972938537598,
"learning_rate": 0.001,
"loss": 1.2286,
"step": 274300
},
{
"epoch": 88.68778280542986,
"grad_norm": 1.6123201847076416,
"learning_rate": 0.001,
"loss": 1.2265,
"step": 274400
},
{
"epoch": 88.72010342598578,
"grad_norm": 1.1493910551071167,
"learning_rate": 0.001,
"loss": 1.2208,
"step": 274500
},
{
"epoch": 88.75242404654169,
"grad_norm": 1.391905426979065,
"learning_rate": 0.001,
"loss": 1.2416,
"step": 274600
},
{
"epoch": 88.78474466709761,
"grad_norm": 1.8348945379257202,
"learning_rate": 0.001,
"loss": 1.235,
"step": 274700
},
{
"epoch": 88.81706528765352,
"grad_norm": 1.6627116203308105,
"learning_rate": 0.001,
"loss": 1.2438,
"step": 274800
},
{
"epoch": 88.84938590820944,
"grad_norm": 1.9870775938034058,
"learning_rate": 0.001,
"loss": 1.2364,
"step": 274900
},
{
"epoch": 88.88170652876535,
"grad_norm": 2.135324478149414,
"learning_rate": 0.001,
"loss": 1.2473,
"step": 275000
},
{
"epoch": 88.91402714932127,
"grad_norm": 2.056169033050537,
"learning_rate": 0.001,
"loss": 1.2485,
"step": 275100
},
{
"epoch": 88.94634776987718,
"grad_norm": 1.3009580373764038,
"learning_rate": 0.001,
"loss": 1.2538,
"step": 275200
},
{
"epoch": 88.9786683904331,
"grad_norm": 1.1325513124465942,
"learning_rate": 0.001,
"loss": 1.2549,
"step": 275300
},
{
"epoch": 89.01098901098901,
"grad_norm": 1.875545620918274,
"learning_rate": 0.001,
"loss": 1.2179,
"step": 275400
},
{
"epoch": 89.04330963154493,
"grad_norm": 1.6356204748153687,
"learning_rate": 0.001,
"loss": 1.1624,
"step": 275500
},
{
"epoch": 89.07563025210084,
"grad_norm": 1.6611065864562988,
"learning_rate": 0.001,
"loss": 1.1568,
"step": 275600
},
{
"epoch": 89.10795087265676,
"grad_norm": 1.744854211807251,
"learning_rate": 0.001,
"loss": 1.1693,
"step": 275700
},
{
"epoch": 89.14027149321267,
"grad_norm": 1.3900748491287231,
"learning_rate": 0.001,
"loss": 1.1565,
"step": 275800
},
{
"epoch": 89.17259211376859,
"grad_norm": 1.3018250465393066,
"learning_rate": 0.001,
"loss": 1.1738,
"step": 275900
},
{
"epoch": 89.2049127343245,
"grad_norm": 1.2930840253829956,
"learning_rate": 0.001,
"loss": 1.1621,
"step": 276000
},
{
"epoch": 89.23723335488042,
"grad_norm": 1.5405704975128174,
"learning_rate": 0.001,
"loss": 1.1807,
"step": 276100
},
{
"epoch": 89.26955397543632,
"grad_norm": 2.3899145126342773,
"learning_rate": 0.001,
"loss": 1.1647,
"step": 276200
},
{
"epoch": 89.30187459599225,
"grad_norm": 1.6378835439682007,
"learning_rate": 0.001,
"loss": 1.2026,
"step": 276300
},
{
"epoch": 89.33419521654815,
"grad_norm": 1.6792488098144531,
"learning_rate": 0.001,
"loss": 1.18,
"step": 276400
},
{
"epoch": 89.36651583710407,
"grad_norm": 2.291990280151367,
"learning_rate": 0.001,
"loss": 1.1842,
"step": 276500
},
{
"epoch": 89.39883645765998,
"grad_norm": 1.4593669176101685,
"learning_rate": 0.001,
"loss": 1.2003,
"step": 276600
},
{
"epoch": 89.4311570782159,
"grad_norm": 1.4892886877059937,
"learning_rate": 0.001,
"loss": 1.1874,
"step": 276700
},
{
"epoch": 89.46347769877181,
"grad_norm": 1.5590368509292603,
"learning_rate": 0.001,
"loss": 1.1991,
"step": 276800
},
{
"epoch": 89.49579831932773,
"grad_norm": 1.8275721073150635,
"learning_rate": 0.001,
"loss": 1.1896,
"step": 276900
},
{
"epoch": 89.52811893988364,
"grad_norm": 1.4684890508651733,
"learning_rate": 0.001,
"loss": 1.2036,
"step": 277000
},
{
"epoch": 89.56043956043956,
"grad_norm": 1.6872565746307373,
"learning_rate": 0.001,
"loss": 1.2127,
"step": 277100
},
{
"epoch": 89.59276018099547,
"grad_norm": 1.5722708702087402,
"learning_rate": 0.001,
"loss": 1.2219,
"step": 277200
},
{
"epoch": 89.62508080155139,
"grad_norm": 1.6149400472640991,
"learning_rate": 0.001,
"loss": 1.2185,
"step": 277300
},
{
"epoch": 89.6574014221073,
"grad_norm": 2.0069503784179688,
"learning_rate": 0.001,
"loss": 1.2172,
"step": 277400
},
{
"epoch": 89.68972204266322,
"grad_norm": 1.6077826023101807,
"learning_rate": 0.001,
"loss": 1.2148,
"step": 277500
},
{
"epoch": 89.72204266321913,
"grad_norm": 2.0763955116271973,
"learning_rate": 0.001,
"loss": 1.2339,
"step": 277600
},
{
"epoch": 89.75436328377505,
"grad_norm": 2.2948250770568848,
"learning_rate": 0.001,
"loss": 1.2174,
"step": 277700
},
{
"epoch": 89.78668390433096,
"grad_norm": 1.7497351169586182,
"learning_rate": 0.001,
"loss": 1.2169,
"step": 277800
},
{
"epoch": 89.81900452488688,
"grad_norm": 1.3225711584091187,
"learning_rate": 0.001,
"loss": 1.2359,
"step": 277900
},
{
"epoch": 89.85132514544279,
"grad_norm": 2.303297281265259,
"learning_rate": 0.001,
"loss": 1.2395,
"step": 278000
},
{
"epoch": 89.88364576599871,
"grad_norm": 1.5867317914962769,
"learning_rate": 0.001,
"loss": 1.2494,
"step": 278100
},
{
"epoch": 89.91596638655462,
"grad_norm": 2.0131945610046387,
"learning_rate": 0.001,
"loss": 1.2354,
"step": 278200
},
{
"epoch": 89.94828700711054,
"grad_norm": 2.566061019897461,
"learning_rate": 0.001,
"loss": 1.244,
"step": 278300
},
{
"epoch": 89.98060762766644,
"grad_norm": 1.5201761722564697,
"learning_rate": 0.001,
"loss": 1.2486,
"step": 278400
},
{
"epoch": 90.01292824822237,
"grad_norm": 1.5201029777526855,
"learning_rate": 0.001,
"loss": 1.1924,
"step": 278500
},
{
"epoch": 90.04524886877829,
"grad_norm": 2.2412655353546143,
"learning_rate": 0.001,
"loss": 1.1479,
"step": 278600
},
{
"epoch": 90.0775694893342,
"grad_norm": 1.3718254566192627,
"learning_rate": 0.001,
"loss": 1.1611,
"step": 278700
},
{
"epoch": 90.10989010989012,
"grad_norm": 1.3264670372009277,
"learning_rate": 0.001,
"loss": 1.1573,
"step": 278800
},
{
"epoch": 90.14221073044602,
"grad_norm": 2.397085189819336,
"learning_rate": 0.001,
"loss": 1.1617,
"step": 278900
},
{
"epoch": 90.17453135100195,
"grad_norm": 1.645229458808899,
"learning_rate": 0.001,
"loss": 1.1594,
"step": 279000
},
{
"epoch": 90.20685197155785,
"grad_norm": 1.3829667568206787,
"learning_rate": 0.001,
"loss": 1.1772,
"step": 279100
},
{
"epoch": 90.23917259211377,
"grad_norm": 1.2982192039489746,
"learning_rate": 0.001,
"loss": 1.166,
"step": 279200
},
{
"epoch": 90.27149321266968,
"grad_norm": 1.889725685119629,
"learning_rate": 0.001,
"loss": 1.1722,
"step": 279300
},
{
"epoch": 90.3038138332256,
"grad_norm": 1.9517900943756104,
"learning_rate": 0.001,
"loss": 1.1658,
"step": 279400
},
{
"epoch": 90.33613445378151,
"grad_norm": 1.5264050960540771,
"learning_rate": 0.001,
"loss": 1.1812,
"step": 279500
},
{
"epoch": 90.36845507433743,
"grad_norm": 1.7745048999786377,
"learning_rate": 0.001,
"loss": 1.1828,
"step": 279600
},
{
"epoch": 90.40077569489334,
"grad_norm": 1.7343498468399048,
"learning_rate": 0.001,
"loss": 1.1999,
"step": 279700
},
{
"epoch": 90.43309631544926,
"grad_norm": 1.9775893688201904,
"learning_rate": 0.001,
"loss": 1.1927,
"step": 279800
},
{
"epoch": 90.46541693600517,
"grad_norm": 1.539505124092102,
"learning_rate": 0.001,
"loss": 1.1905,
"step": 279900
},
{
"epoch": 90.49773755656109,
"grad_norm": 1.4885227680206299,
"learning_rate": 0.001,
"loss": 1.2113,
"step": 280000
},
{
"epoch": 90.530058177117,
"grad_norm": 2.354274034500122,
"learning_rate": 0.001,
"loss": 1.1966,
"step": 280100
},
{
"epoch": 90.56237879767292,
"grad_norm": 1.6391592025756836,
"learning_rate": 0.001,
"loss": 1.2162,
"step": 280200
},
{
"epoch": 90.59469941822883,
"grad_norm": 1.4249579906463623,
"learning_rate": 0.001,
"loss": 1.1971,
"step": 280300
},
{
"epoch": 90.62702003878475,
"grad_norm": 1.3456435203552246,
"learning_rate": 0.001,
"loss": 1.2086,
"step": 280400
},
{
"epoch": 90.65934065934066,
"grad_norm": 1.8398938179016113,
"learning_rate": 0.001,
"loss": 1.2132,
"step": 280500
},
{
"epoch": 90.69166127989658,
"grad_norm": 1.3585916757583618,
"learning_rate": 0.001,
"loss": 1.2132,
"step": 280600
},
{
"epoch": 90.72398190045249,
"grad_norm": 1.7988615036010742,
"learning_rate": 0.001,
"loss": 1.2183,
"step": 280700
},
{
"epoch": 90.75630252100841,
"grad_norm": 1.2848037481307983,
"learning_rate": 0.001,
"loss": 1.2222,
"step": 280800
},
{
"epoch": 90.78862314156432,
"grad_norm": 1.5872573852539062,
"learning_rate": 0.001,
"loss": 1.2307,
"step": 280900
},
{
"epoch": 90.82094376212024,
"grad_norm": 2.1241719722747803,
"learning_rate": 0.001,
"loss": 1.2165,
"step": 281000
},
{
"epoch": 90.85326438267614,
"grad_norm": 1.4638140201568604,
"learning_rate": 0.001,
"loss": 1.2404,
"step": 281100
},
{
"epoch": 90.88558500323207,
"grad_norm": 1.294054627418518,
"learning_rate": 0.001,
"loss": 1.2369,
"step": 281200
},
{
"epoch": 90.91790562378797,
"grad_norm": 1.7885204553604126,
"learning_rate": 0.001,
"loss": 1.2221,
"step": 281300
},
{
"epoch": 90.9502262443439,
"grad_norm": 1.8226639032363892,
"learning_rate": 0.001,
"loss": 1.2324,
"step": 281400
},
{
"epoch": 90.9825468648998,
"grad_norm": 2.0065526962280273,
"learning_rate": 0.001,
"loss": 1.2467,
"step": 281500
},
{
"epoch": 91.01486748545572,
"grad_norm": 1.553343653678894,
"learning_rate": 0.001,
"loss": 1.1705,
"step": 281600
},
{
"epoch": 91.04718810601163,
"grad_norm": 1.574761152267456,
"learning_rate": 0.001,
"loss": 1.1325,
"step": 281700
},
{
"epoch": 91.07950872656755,
"grad_norm": 2.1614627838134766,
"learning_rate": 0.001,
"loss": 1.139,
"step": 281800
},
{
"epoch": 91.11182934712346,
"grad_norm": 1.8416796922683716,
"learning_rate": 0.001,
"loss": 1.1524,
"step": 281900
},
{
"epoch": 91.14414996767938,
"grad_norm": 2.1431214809417725,
"learning_rate": 0.001,
"loss": 1.1487,
"step": 282000
},
{
"epoch": 91.17647058823529,
"grad_norm": 1.7777167558670044,
"learning_rate": 0.001,
"loss": 1.1679,
"step": 282100
},
{
"epoch": 91.20879120879121,
"grad_norm": 1.6724756956100464,
"learning_rate": 0.001,
"loss": 1.1428,
"step": 282200
},
{
"epoch": 91.24111182934712,
"grad_norm": 1.779128074645996,
"learning_rate": 0.001,
"loss": 1.1642,
"step": 282300
},
{
"epoch": 91.27343244990304,
"grad_norm": 1.6053786277770996,
"learning_rate": 0.001,
"loss": 1.169,
"step": 282400
},
{
"epoch": 91.30575307045895,
"grad_norm": 1.982901930809021,
"learning_rate": 0.001,
"loss": 1.172,
"step": 282500
},
{
"epoch": 91.33807369101487,
"grad_norm": 2.143629550933838,
"learning_rate": 0.001,
"loss": 1.1728,
"step": 282600
},
{
"epoch": 91.37039431157078,
"grad_norm": 1.7679647207260132,
"learning_rate": 0.001,
"loss": 1.1881,
"step": 282700
},
{
"epoch": 91.4027149321267,
"grad_norm": 2.009580373764038,
"learning_rate": 0.001,
"loss": 1.1857,
"step": 282800
},
{
"epoch": 91.4350355526826,
"grad_norm": 1.971984624862671,
"learning_rate": 0.001,
"loss": 1.1823,
"step": 282900
},
{
"epoch": 91.46735617323853,
"grad_norm": 1.5214993953704834,
"learning_rate": 0.001,
"loss": 1.1773,
"step": 283000
},
{
"epoch": 91.49967679379444,
"grad_norm": 1.8038649559020996,
"learning_rate": 0.001,
"loss": 1.1941,
"step": 283100
},
{
"epoch": 91.53199741435036,
"grad_norm": 1.5943965911865234,
"learning_rate": 0.001,
"loss": 1.1726,
"step": 283200
},
{
"epoch": 91.56431803490626,
"grad_norm": 2.3694846630096436,
"learning_rate": 0.001,
"loss": 1.2078,
"step": 283300
},
{
"epoch": 91.59663865546219,
"grad_norm": 1.6778090000152588,
"learning_rate": 0.001,
"loss": 1.2191,
"step": 283400
},
{
"epoch": 91.6289592760181,
"grad_norm": 1.5933371782302856,
"learning_rate": 0.001,
"loss": 1.1945,
"step": 283500
},
{
"epoch": 91.66127989657402,
"grad_norm": 1.8515619039535522,
"learning_rate": 0.001,
"loss": 1.2358,
"step": 283600
},
{
"epoch": 91.69360051712992,
"grad_norm": 1.7395762205123901,
"learning_rate": 0.001,
"loss": 1.2278,
"step": 283700
},
{
"epoch": 91.72592113768584,
"grad_norm": 2.1333162784576416,
"learning_rate": 0.001,
"loss": 1.2081,
"step": 283800
},
{
"epoch": 91.75824175824175,
"grad_norm": 2.038357734680176,
"learning_rate": 0.001,
"loss": 1.2195,
"step": 283900
},
{
"epoch": 91.79056237879767,
"grad_norm": 1.819784164428711,
"learning_rate": 0.001,
"loss": 1.2284,
"step": 284000
},
{
"epoch": 91.82288299935358,
"grad_norm": 2.1269073486328125,
"learning_rate": 0.001,
"loss": 1.2231,
"step": 284100
},
{
"epoch": 91.8552036199095,
"grad_norm": 2.2647945880889893,
"learning_rate": 0.001,
"loss": 1.2115,
"step": 284200
},
{
"epoch": 91.88752424046541,
"grad_norm": 1.5195850133895874,
"learning_rate": 0.001,
"loss": 1.2324,
"step": 284300
},
{
"epoch": 91.91984486102133,
"grad_norm": 2.303753137588501,
"learning_rate": 0.001,
"loss": 1.2391,
"step": 284400
},
{
"epoch": 91.95216548157724,
"grad_norm": 1.922239899635315,
"learning_rate": 0.001,
"loss": 1.2234,
"step": 284500
},
{
"epoch": 91.98448610213316,
"grad_norm": 1.695598840713501,
"learning_rate": 0.001,
"loss": 1.2452,
"step": 284600
},
{
"epoch": 92.01680672268908,
"grad_norm": 1.7118265628814697,
"learning_rate": 0.001,
"loss": 1.1588,
"step": 284700
},
{
"epoch": 92.04912734324499,
"grad_norm": 1.8453078269958496,
"learning_rate": 0.001,
"loss": 1.1266,
"step": 284800
},
{
"epoch": 92.08144796380091,
"grad_norm": 2.1100013256073,
"learning_rate": 0.001,
"loss": 1.137,
"step": 284900
},
{
"epoch": 92.11376858435682,
"grad_norm": 1.8456408977508545,
"learning_rate": 0.001,
"loss": 1.1368,
"step": 285000
},
{
"epoch": 92.14608920491274,
"grad_norm": 1.7554715871810913,
"learning_rate": 0.001,
"loss": 1.1544,
"step": 285100
},
{
"epoch": 92.17840982546865,
"grad_norm": 1.618524432182312,
"learning_rate": 0.001,
"loss": 1.1593,
"step": 285200
},
{
"epoch": 92.21073044602457,
"grad_norm": 1.7603201866149902,
"learning_rate": 0.001,
"loss": 1.1633,
"step": 285300
},
{
"epoch": 92.24305106658048,
"grad_norm": 1.407418966293335,
"learning_rate": 0.001,
"loss": 1.153,
"step": 285400
},
{
"epoch": 92.2753716871364,
"grad_norm": 2.6295015811920166,
"learning_rate": 0.001,
"loss": 1.1683,
"step": 285500
},
{
"epoch": 92.3076923076923,
"grad_norm": 1.855170726776123,
"learning_rate": 0.001,
"loss": 1.1654,
"step": 285600
},
{
"epoch": 92.34001292824823,
"grad_norm": 3.7528069019317627,
"learning_rate": 0.001,
"loss": 1.1533,
"step": 285700
},
{
"epoch": 92.37233354880414,
"grad_norm": 2.2886791229248047,
"learning_rate": 0.001,
"loss": 1.1604,
"step": 285800
},
{
"epoch": 92.40465416936006,
"grad_norm": 2.4651310443878174,
"learning_rate": 0.001,
"loss": 1.1765,
"step": 285900
},
{
"epoch": 92.43697478991596,
"grad_norm": 2.6569302082061768,
"learning_rate": 0.001,
"loss": 1.184,
"step": 286000
},
{
"epoch": 92.46929541047189,
"grad_norm": 2.0062756538391113,
"learning_rate": 0.001,
"loss": 1.2011,
"step": 286100
},
{
"epoch": 92.5016160310278,
"grad_norm": 1.9018144607543945,
"learning_rate": 0.001,
"loss": 1.1914,
"step": 286200
},
{
"epoch": 92.53393665158372,
"grad_norm": 1.7802410125732422,
"learning_rate": 0.001,
"loss": 1.1921,
"step": 286300
},
{
"epoch": 92.56625727213962,
"grad_norm": 1.9713374376296997,
"learning_rate": 0.001,
"loss": 1.1795,
"step": 286400
},
{
"epoch": 92.59857789269554,
"grad_norm": 2.6365981101989746,
"learning_rate": 0.001,
"loss": 1.1892,
"step": 286500
},
{
"epoch": 92.63089851325145,
"grad_norm": 2.187910318374634,
"learning_rate": 0.001,
"loss": 1.2051,
"step": 286600
},
{
"epoch": 92.66321913380737,
"grad_norm": 2.039344072341919,
"learning_rate": 0.001,
"loss": 1.1984,
"step": 286700
},
{
"epoch": 92.69553975436328,
"grad_norm": 2.032949209213257,
"learning_rate": 0.001,
"loss": 1.2181,
"step": 286800
},
{
"epoch": 92.7278603749192,
"grad_norm": 2.804405689239502,
"learning_rate": 0.001,
"loss": 1.1987,
"step": 286900
},
{
"epoch": 92.76018099547511,
"grad_norm": 1.8193217515945435,
"learning_rate": 0.001,
"loss": 1.1995,
"step": 287000
},
{
"epoch": 92.79250161603103,
"grad_norm": 2.1374568939208984,
"learning_rate": 0.001,
"loss": 1.2147,
"step": 287100
},
{
"epoch": 92.82482223658694,
"grad_norm": 2.0475242137908936,
"learning_rate": 0.001,
"loss": 1.2137,
"step": 287200
},
{
"epoch": 92.85714285714286,
"grad_norm": 1.8530298471450806,
"learning_rate": 0.001,
"loss": 1.2203,
"step": 287300
},
{
"epoch": 92.88946347769877,
"grad_norm": 2.1449034214019775,
"learning_rate": 0.001,
"loss": 1.2357,
"step": 287400
},
{
"epoch": 92.92178409825469,
"grad_norm": 1.8257231712341309,
"learning_rate": 0.001,
"loss": 1.2428,
"step": 287500
},
{
"epoch": 92.9541047188106,
"grad_norm": 1.6015028953552246,
"learning_rate": 0.001,
"loss": 1.2322,
"step": 287600
},
{
"epoch": 92.98642533936652,
"grad_norm": 2.003861665725708,
"learning_rate": 0.001,
"loss": 1.2222,
"step": 287700
},
{
"epoch": 93.01874595992243,
"grad_norm": 1.8797523975372314,
"learning_rate": 0.001,
"loss": 1.1834,
"step": 287800
},
{
"epoch": 93.05106658047835,
"grad_norm": 1.5017789602279663,
"learning_rate": 0.001,
"loss": 1.1172,
"step": 287900
},
{
"epoch": 93.08338720103426,
"grad_norm": 1.3459205627441406,
"learning_rate": 0.001,
"loss": 1.1426,
"step": 288000
},
{
"epoch": 93.11570782159018,
"grad_norm": 1.5411944389343262,
"learning_rate": 0.001,
"loss": 1.1388,
"step": 288100
},
{
"epoch": 93.14802844214609,
"grad_norm": 2.0234344005584717,
"learning_rate": 0.001,
"loss": 1.1572,
"step": 288200
},
{
"epoch": 93.180349062702,
"grad_norm": 1.6879246234893799,
"learning_rate": 0.001,
"loss": 1.1414,
"step": 288300
},
{
"epoch": 93.21266968325791,
"grad_norm": 1.4526965618133545,
"learning_rate": 0.001,
"loss": 1.1476,
"step": 288400
},
{
"epoch": 93.24499030381384,
"grad_norm": 2.114469528198242,
"learning_rate": 0.001,
"loss": 1.1461,
"step": 288500
},
{
"epoch": 93.27731092436974,
"grad_norm": 2.7266154289245605,
"learning_rate": 0.001,
"loss": 1.1739,
"step": 288600
},
{
"epoch": 93.30963154492567,
"grad_norm": 2.0473294258117676,
"learning_rate": 0.001,
"loss": 1.1502,
"step": 288700
},
{
"epoch": 93.34195216548157,
"grad_norm": 1.4588580131530762,
"learning_rate": 0.001,
"loss": 1.1757,
"step": 288800
},
{
"epoch": 93.3742727860375,
"grad_norm": 1.7957432270050049,
"learning_rate": 0.001,
"loss": 1.1789,
"step": 288900
},
{
"epoch": 93.4065934065934,
"grad_norm": 1.9788367748260498,
"learning_rate": 0.001,
"loss": 1.1766,
"step": 289000
},
{
"epoch": 93.43891402714932,
"grad_norm": 1.7873634099960327,
"learning_rate": 0.001,
"loss": 1.1648,
"step": 289100
},
{
"epoch": 93.47123464770523,
"grad_norm": 1.6952708959579468,
"learning_rate": 0.001,
"loss": 1.1728,
"step": 289200
},
{
"epoch": 93.50355526826115,
"grad_norm": 1.3903928995132446,
"learning_rate": 0.001,
"loss": 1.1731,
"step": 289300
},
{
"epoch": 93.53587588881706,
"grad_norm": 1.5076650381088257,
"learning_rate": 0.001,
"loss": 1.185,
"step": 289400
},
{
"epoch": 93.56819650937298,
"grad_norm": 2.4555656909942627,
"learning_rate": 0.001,
"loss": 1.1829,
"step": 289500
},
{
"epoch": 93.60051712992889,
"grad_norm": 2.349330425262451,
"learning_rate": 0.001,
"loss": 1.1894,
"step": 289600
},
{
"epoch": 93.63283775048481,
"grad_norm": 1.7061920166015625,
"learning_rate": 0.001,
"loss": 1.1822,
"step": 289700
},
{
"epoch": 93.66515837104072,
"grad_norm": 1.7948163747787476,
"learning_rate": 0.001,
"loss": 1.1854,
"step": 289800
},
{
"epoch": 93.69747899159664,
"grad_norm": 1.5287566184997559,
"learning_rate": 0.001,
"loss": 1.1888,
"step": 289900
},
{
"epoch": 93.72979961215255,
"grad_norm": 1.590301513671875,
"learning_rate": 0.001,
"loss": 1.2022,
"step": 290000
},
{
"epoch": 93.76212023270847,
"grad_norm": 1.5148197412490845,
"learning_rate": 0.001,
"loss": 1.2101,
"step": 290100
},
{
"epoch": 93.79444085326438,
"grad_norm": 1.9384305477142334,
"learning_rate": 0.001,
"loss": 1.2059,
"step": 290200
},
{
"epoch": 93.8267614738203,
"grad_norm": 1.7551034688949585,
"learning_rate": 0.001,
"loss": 1.2084,
"step": 290300
},
{
"epoch": 93.8590820943762,
"grad_norm": 1.8605059385299683,
"learning_rate": 0.001,
"loss": 1.2075,
"step": 290400
},
{
"epoch": 93.89140271493213,
"grad_norm": 1.7120189666748047,
"learning_rate": 0.001,
"loss": 1.2183,
"step": 290500
},
{
"epoch": 93.92372333548803,
"grad_norm": 1.8584399223327637,
"learning_rate": 0.001,
"loss": 1.2287,
"step": 290600
},
{
"epoch": 93.95604395604396,
"grad_norm": 1.8144450187683105,
"learning_rate": 0.001,
"loss": 1.2198,
"step": 290700
},
{
"epoch": 93.98836457659988,
"grad_norm": 2.061105966567993,
"learning_rate": 0.001,
"loss": 1.2181,
"step": 290800
},
{
"epoch": 94.02068519715579,
"grad_norm": 2.702474594116211,
"learning_rate": 0.001,
"loss": 1.1685,
"step": 290900
},
{
"epoch": 94.0530058177117,
"grad_norm": 1.64632248878479,
"learning_rate": 0.001,
"loss": 1.1171,
"step": 291000
},
{
"epoch": 94.08532643826761,
"grad_norm": 2.041278839111328,
"learning_rate": 0.001,
"loss": 1.1365,
"step": 291100
},
{
"epoch": 94.11764705882354,
"grad_norm": 1.4763818979263306,
"learning_rate": 0.001,
"loss": 1.1433,
"step": 291200
},
{
"epoch": 94.14996767937944,
"grad_norm": 1.36812424659729,
"learning_rate": 0.001,
"loss": 1.118,
"step": 291300
},
{
"epoch": 94.18228829993537,
"grad_norm": 1.6154983043670654,
"learning_rate": 0.001,
"loss": 1.1485,
"step": 291400
},
{
"epoch": 94.21460892049127,
"grad_norm": 1.8375099897384644,
"learning_rate": 0.001,
"loss": 1.1424,
"step": 291500
},
{
"epoch": 94.2469295410472,
"grad_norm": 1.8397036790847778,
"learning_rate": 0.001,
"loss": 1.1528,
"step": 291600
},
{
"epoch": 94.2792501616031,
"grad_norm": 1.466280460357666,
"learning_rate": 0.001,
"loss": 1.1516,
"step": 291700
},
{
"epoch": 94.31157078215902,
"grad_norm": 2.9650964736938477,
"learning_rate": 0.001,
"loss": 1.1675,
"step": 291800
},
{
"epoch": 94.34389140271493,
"grad_norm": 2.0993025302886963,
"learning_rate": 0.001,
"loss": 1.1409,
"step": 291900
},
{
"epoch": 94.37621202327085,
"grad_norm": 1.6603684425354004,
"learning_rate": 0.001,
"loss": 1.168,
"step": 292000
},
{
"epoch": 94.40853264382676,
"grad_norm": 1.7989704608917236,
"learning_rate": 0.001,
"loss": 1.1621,
"step": 292100
},
{
"epoch": 94.44085326438268,
"grad_norm": 1.5147311687469482,
"learning_rate": 0.001,
"loss": 1.1662,
"step": 292200
},
{
"epoch": 94.47317388493859,
"grad_norm": 1.8816263675689697,
"learning_rate": 0.001,
"loss": 1.1662,
"step": 292300
},
{
"epoch": 94.50549450549451,
"grad_norm": 1.3142783641815186,
"learning_rate": 0.001,
"loss": 1.1803,
"step": 292400
},
{
"epoch": 94.53781512605042,
"grad_norm": 2.246133327484131,
"learning_rate": 0.001,
"loss": 1.1873,
"step": 292500
},
{
"epoch": 94.57013574660634,
"grad_norm": 1.5358392000198364,
"learning_rate": 0.001,
"loss": 1.1777,
"step": 292600
},
{
"epoch": 94.60245636716225,
"grad_norm": 1.275778889656067,
"learning_rate": 0.001,
"loss": 1.172,
"step": 292700
},
{
"epoch": 94.63477698771817,
"grad_norm": 1.631397008895874,
"learning_rate": 0.001,
"loss": 1.1943,
"step": 292800
},
{
"epoch": 94.66709760827408,
"grad_norm": 1.462730884552002,
"learning_rate": 0.001,
"loss": 1.1937,
"step": 292900
},
{
"epoch": 94.69941822883,
"grad_norm": 2.0201778411865234,
"learning_rate": 0.001,
"loss": 1.2112,
"step": 293000
},
{
"epoch": 94.7317388493859,
"grad_norm": 1.8298077583312988,
"learning_rate": 0.001,
"loss": 1.2012,
"step": 293100
},
{
"epoch": 94.76405946994183,
"grad_norm": 1.727700114250183,
"learning_rate": 0.001,
"loss": 1.1993,
"step": 293200
},
{
"epoch": 94.79638009049773,
"grad_norm": 1.6637187004089355,
"learning_rate": 0.001,
"loss": 1.1984,
"step": 293300
},
{
"epoch": 94.82870071105366,
"grad_norm": 1.911285400390625,
"learning_rate": 0.001,
"loss": 1.2174,
"step": 293400
},
{
"epoch": 94.86102133160956,
"grad_norm": 1.6867454051971436,
"learning_rate": 0.001,
"loss": 1.2025,
"step": 293500
},
{
"epoch": 94.89334195216549,
"grad_norm": 1.5231786966323853,
"learning_rate": 0.001,
"loss": 1.2163,
"step": 293600
},
{
"epoch": 94.9256625727214,
"grad_norm": 2.033961296081543,
"learning_rate": 0.001,
"loss": 1.2061,
"step": 293700
},
{
"epoch": 94.95798319327731,
"grad_norm": 2.617823362350464,
"learning_rate": 0.001,
"loss": 1.2122,
"step": 293800
},
{
"epoch": 94.99030381383322,
"grad_norm": 1.7739336490631104,
"learning_rate": 0.001,
"loss": 1.2211,
"step": 293900
},
{
"epoch": 95.02262443438914,
"grad_norm": 2.470210552215576,
"learning_rate": 0.001,
"loss": 1.1605,
"step": 294000
},
{
"epoch": 95.05494505494505,
"grad_norm": 1.1751980781555176,
"learning_rate": 0.001,
"loss": 1.1296,
"step": 294100
},
{
"epoch": 95.08726567550097,
"grad_norm": 1.6907886266708374,
"learning_rate": 0.001,
"loss": 1.1378,
"step": 294200
},
{
"epoch": 95.11958629605688,
"grad_norm": 2.201791524887085,
"learning_rate": 0.001,
"loss": 1.1211,
"step": 294300
},
{
"epoch": 95.1519069166128,
"grad_norm": 1.7307604551315308,
"learning_rate": 0.001,
"loss": 1.1174,
"step": 294400
},
{
"epoch": 95.18422753716871,
"grad_norm": 1.2891641855239868,
"learning_rate": 0.001,
"loss": 1.1371,
"step": 294500
},
{
"epoch": 95.21654815772463,
"grad_norm": 1.5916978120803833,
"learning_rate": 0.001,
"loss": 1.1486,
"step": 294600
},
{
"epoch": 95.24886877828054,
"grad_norm": 1.6448651552200317,
"learning_rate": 0.001,
"loss": 1.143,
"step": 294700
},
{
"epoch": 95.28118939883646,
"grad_norm": 2.0232772827148438,
"learning_rate": 0.001,
"loss": 1.1474,
"step": 294800
},
{
"epoch": 95.31351001939237,
"grad_norm": 1.5678558349609375,
"learning_rate": 0.001,
"loss": 1.1452,
"step": 294900
},
{
"epoch": 95.34583063994829,
"grad_norm": 2.3138344287872314,
"learning_rate": 0.001,
"loss": 1.1451,
"step": 295000
},
{
"epoch": 95.3781512605042,
"grad_norm": 1.411215901374817,
"learning_rate": 0.001,
"loss": 1.1632,
"step": 295100
},
{
"epoch": 95.41047188106012,
"grad_norm": 1.6272342205047607,
"learning_rate": 0.001,
"loss": 1.1544,
"step": 295200
},
{
"epoch": 95.44279250161603,
"grad_norm": 2.5069704055786133,
"learning_rate": 0.001,
"loss": 1.1573,
"step": 295300
},
{
"epoch": 95.47511312217195,
"grad_norm": 1.6448183059692383,
"learning_rate": 0.001,
"loss": 1.167,
"step": 295400
},
{
"epoch": 95.50743374272786,
"grad_norm": 1.4927773475646973,
"learning_rate": 0.001,
"loss": 1.1704,
"step": 295500
},
{
"epoch": 95.53975436328378,
"grad_norm": 1.57712984085083,
"learning_rate": 0.001,
"loss": 1.1789,
"step": 295600
},
{
"epoch": 95.57207498383968,
"grad_norm": 1.4393914937973022,
"learning_rate": 0.001,
"loss": 1.1751,
"step": 295700
},
{
"epoch": 95.6043956043956,
"grad_norm": 1.8276870250701904,
"learning_rate": 0.001,
"loss": 1.1739,
"step": 295800
},
{
"epoch": 95.63671622495151,
"grad_norm": 1.1675989627838135,
"learning_rate": 0.001,
"loss": 1.1823,
"step": 295900
},
{
"epoch": 95.66903684550743,
"grad_norm": 1.2009097337722778,
"learning_rate": 0.001,
"loss": 1.1933,
"step": 296000
},
{
"epoch": 95.70135746606334,
"grad_norm": 1.5777475833892822,
"learning_rate": 0.001,
"loss": 1.1854,
"step": 296100
},
{
"epoch": 95.73367808661926,
"grad_norm": 1.2625652551651,
"learning_rate": 0.001,
"loss": 1.1898,
"step": 296200
},
{
"epoch": 95.76599870717517,
"grad_norm": 1.355389952659607,
"learning_rate": 0.001,
"loss": 1.1894,
"step": 296300
},
{
"epoch": 95.7983193277311,
"grad_norm": 1.6914863586425781,
"learning_rate": 0.001,
"loss": 1.2022,
"step": 296400
},
{
"epoch": 95.830639948287,
"grad_norm": 1.6860562562942505,
"learning_rate": 0.001,
"loss": 1.2021,
"step": 296500
},
{
"epoch": 95.86296056884292,
"grad_norm": 1.7306995391845703,
"learning_rate": 0.001,
"loss": 1.2158,
"step": 296600
},
{
"epoch": 95.89528118939883,
"grad_norm": 1.8452569246292114,
"learning_rate": 0.001,
"loss": 1.1822,
"step": 296700
},
{
"epoch": 95.92760180995475,
"grad_norm": 1.5447825193405151,
"learning_rate": 0.001,
"loss": 1.1996,
"step": 296800
},
{
"epoch": 95.95992243051066,
"grad_norm": 2.2806003093719482,
"learning_rate": 0.001,
"loss": 1.2088,
"step": 296900
},
{
"epoch": 95.99224305106658,
"grad_norm": 1.7977348566055298,
"learning_rate": 0.001,
"loss": 1.221,
"step": 297000
},
{
"epoch": 96.0245636716225,
"grad_norm": 1.860256314277649,
"learning_rate": 0.001,
"loss": 1.1416,
"step": 297100
},
{
"epoch": 96.05688429217841,
"grad_norm": 1.2851029634475708,
"learning_rate": 0.001,
"loss": 1.1051,
"step": 297200
},
{
"epoch": 96.08920491273433,
"grad_norm": 2.5045924186706543,
"learning_rate": 0.001,
"loss": 1.1224,
"step": 297300
},
{
"epoch": 96.12152553329024,
"grad_norm": 1.9603523015975952,
"learning_rate": 0.001,
"loss": 1.1228,
"step": 297400
},
{
"epoch": 96.15384615384616,
"grad_norm": 1.8318489789962769,
"learning_rate": 0.001,
"loss": 1.1256,
"step": 297500
},
{
"epoch": 96.18616677440207,
"grad_norm": 1.1082842350006104,
"learning_rate": 0.001,
"loss": 1.142,
"step": 297600
},
{
"epoch": 96.21848739495799,
"grad_norm": 2.170994758605957,
"learning_rate": 0.001,
"loss": 1.1279,
"step": 297700
},
{
"epoch": 96.2508080155139,
"grad_norm": 1.8286175727844238,
"learning_rate": 0.001,
"loss": 1.1416,
"step": 297800
},
{
"epoch": 96.28312863606982,
"grad_norm": 1.4053666591644287,
"learning_rate": 0.001,
"loss": 1.1379,
"step": 297900
},
{
"epoch": 96.31544925662573,
"grad_norm": 1.911604404449463,
"learning_rate": 0.001,
"loss": 1.1593,
"step": 298000
},
{
"epoch": 96.34776987718165,
"grad_norm": 1.3394492864608765,
"learning_rate": 0.001,
"loss": 1.1416,
"step": 298100
},
{
"epoch": 96.38009049773756,
"grad_norm": 1.5122565031051636,
"learning_rate": 0.001,
"loss": 1.1487,
"step": 298200
},
{
"epoch": 96.41241111829348,
"grad_norm": 2.1100003719329834,
"learning_rate": 0.001,
"loss": 1.1512,
"step": 298300
},
{
"epoch": 96.44473173884938,
"grad_norm": 1.941153883934021,
"learning_rate": 0.001,
"loss": 1.1693,
"step": 298400
},
{
"epoch": 96.4770523594053,
"grad_norm": 1.6588879823684692,
"learning_rate": 0.001,
"loss": 1.1611,
"step": 298500
},
{
"epoch": 96.50937297996121,
"grad_norm": 1.2245738506317139,
"learning_rate": 0.001,
"loss": 1.1614,
"step": 298600
},
{
"epoch": 96.54169360051714,
"grad_norm": 1.5878503322601318,
"learning_rate": 0.001,
"loss": 1.1777,
"step": 298700
},
{
"epoch": 96.57401422107304,
"grad_norm": 1.4396772384643555,
"learning_rate": 0.001,
"loss": 1.157,
"step": 298800
},
{
"epoch": 96.60633484162896,
"grad_norm": 1.590366244316101,
"learning_rate": 0.001,
"loss": 1.1681,
"step": 298900
},
{
"epoch": 96.63865546218487,
"grad_norm": 1.252959966659546,
"learning_rate": 0.001,
"loss": 1.1765,
"step": 299000
},
{
"epoch": 96.6709760827408,
"grad_norm": 2.057255983352661,
"learning_rate": 0.001,
"loss": 1.1884,
"step": 299100
},
{
"epoch": 96.7032967032967,
"grad_norm": 1.8908469676971436,
"learning_rate": 0.001,
"loss": 1.1843,
"step": 299200
},
{
"epoch": 96.73561732385262,
"grad_norm": 1.616100549697876,
"learning_rate": 0.001,
"loss": 1.1835,
"step": 299300
},
{
"epoch": 96.76793794440853,
"grad_norm": 1.4547150135040283,
"learning_rate": 0.001,
"loss": 1.2027,
"step": 299400
},
{
"epoch": 96.80025856496445,
"grad_norm": 1.4257676601409912,
"learning_rate": 0.001,
"loss": 1.1795,
"step": 299500
},
{
"epoch": 96.83257918552036,
"grad_norm": 1.3480945825576782,
"learning_rate": 0.001,
"loss": 1.2027,
"step": 299600
},
{
"epoch": 96.86489980607628,
"grad_norm": 1.460683822631836,
"learning_rate": 0.001,
"loss": 1.1955,
"step": 299700
},
{
"epoch": 96.89722042663219,
"grad_norm": 2.127368450164795,
"learning_rate": 0.001,
"loss": 1.1796,
"step": 299800
},
{
"epoch": 96.92954104718811,
"grad_norm": 2.2763729095458984,
"learning_rate": 0.001,
"loss": 1.2024,
"step": 299900
},
{
"epoch": 96.96186166774402,
"grad_norm": 1.617999792098999,
"learning_rate": 0.001,
"loss": 1.2059,
"step": 300000
},
{
"epoch": 96.99418228829994,
"grad_norm": 1.829553246498108,
"learning_rate": 0.001,
"loss": 1.1947,
"step": 300100
},
{
"epoch": 97.02650290885585,
"grad_norm": 1.306296944618225,
"learning_rate": 0.001,
"loss": 1.1321,
"step": 300200
},
{
"epoch": 97.05882352941177,
"grad_norm": 1.4006305932998657,
"learning_rate": 0.001,
"loss": 1.1103,
"step": 300300
},
{
"epoch": 97.09114414996768,
"grad_norm": 1.3456957340240479,
"learning_rate": 0.001,
"loss": 1.1028,
"step": 300400
},
{
"epoch": 97.1234647705236,
"grad_norm": 1.495202898979187,
"learning_rate": 0.001,
"loss": 1.099,
"step": 300500
},
{
"epoch": 97.1557853910795,
"grad_norm": 1.624022126197815,
"learning_rate": 0.001,
"loss": 1.1155,
"step": 300600
},
{
"epoch": 97.18810601163543,
"grad_norm": 1.707629680633545,
"learning_rate": 0.001,
"loss": 1.1294,
"step": 300700
},
{
"epoch": 97.22042663219133,
"grad_norm": 1.6954739093780518,
"learning_rate": 0.001,
"loss": 1.1397,
"step": 300800
},
{
"epoch": 97.25274725274726,
"grad_norm": 1.5452526807785034,
"learning_rate": 0.001,
"loss": 1.1212,
"step": 300900
},
{
"epoch": 97.28506787330316,
"grad_norm": 2.1428868770599365,
"learning_rate": 0.001,
"loss": 1.1411,
"step": 301000
},
{
"epoch": 97.31738849385908,
"grad_norm": 1.6986525058746338,
"learning_rate": 0.001,
"loss": 1.1356,
"step": 301100
},
{
"epoch": 97.34970911441499,
"grad_norm": 1.5793352127075195,
"learning_rate": 0.001,
"loss": 1.1409,
"step": 301200
},
{
"epoch": 97.38202973497091,
"grad_norm": 1.5635677576065063,
"learning_rate": 0.001,
"loss": 1.1426,
"step": 301300
},
{
"epoch": 97.41435035552682,
"grad_norm": 1.561034917831421,
"learning_rate": 0.001,
"loss": 1.1528,
"step": 301400
},
{
"epoch": 97.44667097608274,
"grad_norm": 2.120241403579712,
"learning_rate": 0.001,
"loss": 1.162,
"step": 301500
},
{
"epoch": 97.47899159663865,
"grad_norm": 2.110903024673462,
"learning_rate": 0.001,
"loss": 1.1612,
"step": 301600
},
{
"epoch": 97.51131221719457,
"grad_norm": 1.827684998512268,
"learning_rate": 0.001,
"loss": 1.1689,
"step": 301700
},
{
"epoch": 97.54363283775048,
"grad_norm": 1.3731664419174194,
"learning_rate": 0.001,
"loss": 1.154,
"step": 301800
},
{
"epoch": 97.5759534583064,
"grad_norm": 1.240972876548767,
"learning_rate": 0.001,
"loss": 1.1662,
"step": 301900
},
{
"epoch": 97.60827407886231,
"grad_norm": 1.9408973455429077,
"learning_rate": 0.001,
"loss": 1.1606,
"step": 302000
},
{
"epoch": 97.64059469941823,
"grad_norm": 1.7482811212539673,
"learning_rate": 0.001,
"loss": 1.1665,
"step": 302100
},
{
"epoch": 97.67291531997414,
"grad_norm": 2.163902997970581,
"learning_rate": 0.001,
"loss": 1.1741,
"step": 302200
},
{
"epoch": 97.70523594053006,
"grad_norm": 1.2171028852462769,
"learning_rate": 0.001,
"loss": 1.1902,
"step": 302300
},
{
"epoch": 97.73755656108597,
"grad_norm": 1.5814447402954102,
"learning_rate": 0.001,
"loss": 1.1836,
"step": 302400
},
{
"epoch": 97.76987718164189,
"grad_norm": 1.408679723739624,
"learning_rate": 0.001,
"loss": 1.184,
"step": 302500
},
{
"epoch": 97.8021978021978,
"grad_norm": 1.645927906036377,
"learning_rate": 0.001,
"loss": 1.1835,
"step": 302600
},
{
"epoch": 97.83451842275372,
"grad_norm": 1.3949730396270752,
"learning_rate": 0.001,
"loss": 1.1798,
"step": 302700
},
{
"epoch": 97.86683904330962,
"grad_norm": 1.6577595472335815,
"learning_rate": 0.001,
"loss": 1.1858,
"step": 302800
},
{
"epoch": 97.89915966386555,
"grad_norm": 1.934249997138977,
"learning_rate": 0.001,
"loss": 1.2028,
"step": 302900
},
{
"epoch": 97.93148028442145,
"grad_norm": 1.6780946254730225,
"learning_rate": 0.001,
"loss": 1.2043,
"step": 303000
},
{
"epoch": 97.96380090497738,
"grad_norm": 1.8218052387237549,
"learning_rate": 0.001,
"loss": 1.1921,
"step": 303100
},
{
"epoch": 97.99612152553328,
"grad_norm": 2.1109933853149414,
"learning_rate": 0.001,
"loss": 1.1976,
"step": 303200
},
{
"epoch": 98.0284421460892,
"grad_norm": 1.517188310623169,
"learning_rate": 0.001,
"loss": 1.119,
"step": 303300
},
{
"epoch": 98.06076276664513,
"grad_norm": 2.3147544860839844,
"learning_rate": 0.001,
"loss": 1.0996,
"step": 303400
},
{
"epoch": 98.09308338720103,
"grad_norm": 1.553000807762146,
"learning_rate": 0.001,
"loss": 1.0983,
"step": 303500
},
{
"epoch": 98.12540400775696,
"grad_norm": 1.4100111722946167,
"learning_rate": 0.001,
"loss": 1.1109,
"step": 303600
},
{
"epoch": 98.15772462831286,
"grad_norm": 2.305436849594116,
"learning_rate": 0.001,
"loss": 1.1085,
"step": 303700
},
{
"epoch": 98.19004524886878,
"grad_norm": 1.940406322479248,
"learning_rate": 0.001,
"loss": 1.122,
"step": 303800
},
{
"epoch": 98.22236586942469,
"grad_norm": 1.3003724813461304,
"learning_rate": 0.001,
"loss": 1.117,
"step": 303900
},
{
"epoch": 98.25468648998061,
"grad_norm": 1.3235238790512085,
"learning_rate": 0.001,
"loss": 1.1067,
"step": 304000
},
{
"epoch": 98.28700711053652,
"grad_norm": 1.5034582614898682,
"learning_rate": 0.001,
"loss": 1.1329,
"step": 304100
},
{
"epoch": 98.31932773109244,
"grad_norm": 1.7820682525634766,
"learning_rate": 0.001,
"loss": 1.1382,
"step": 304200
},
{
"epoch": 98.35164835164835,
"grad_norm": 1.7481606006622314,
"learning_rate": 0.001,
"loss": 1.1483,
"step": 304300
},
{
"epoch": 98.38396897220427,
"grad_norm": 1.8273428678512573,
"learning_rate": 0.001,
"loss": 1.1556,
"step": 304400
},
{
"epoch": 98.41628959276018,
"grad_norm": 1.975685715675354,
"learning_rate": 0.001,
"loss": 1.1473,
"step": 304500
},
{
"epoch": 98.4486102133161,
"grad_norm": 2.329911231994629,
"learning_rate": 0.001,
"loss": 1.1651,
"step": 304600
},
{
"epoch": 98.48093083387201,
"grad_norm": 2.8156585693359375,
"learning_rate": 0.001,
"loss": 1.1491,
"step": 304700
},
{
"epoch": 98.51325145442793,
"grad_norm": 2.167443037033081,
"learning_rate": 0.001,
"loss": 1.1701,
"step": 304800
},
{
"epoch": 98.54557207498384,
"grad_norm": 2.023167848587036,
"learning_rate": 0.001,
"loss": 1.1603,
"step": 304900
},
{
"epoch": 98.57789269553976,
"grad_norm": 2.202585458755493,
"learning_rate": 0.001,
"loss": 1.1602,
"step": 305000
},
{
"epoch": 98.61021331609567,
"grad_norm": 2.8777387142181396,
"learning_rate": 0.001,
"loss": 1.1638,
"step": 305100
},
{
"epoch": 98.64253393665159,
"grad_norm": 1.5604068040847778,
"learning_rate": 0.001,
"loss": 1.1625,
"step": 305200
},
{
"epoch": 98.6748545572075,
"grad_norm": 1.617970585823059,
"learning_rate": 0.001,
"loss": 1.1686,
"step": 305300
},
{
"epoch": 98.70717517776342,
"grad_norm": 1.7877370119094849,
"learning_rate": 0.001,
"loss": 1.1626,
"step": 305400
},
{
"epoch": 98.73949579831933,
"grad_norm": 2.0654261112213135,
"learning_rate": 0.001,
"loss": 1.1779,
"step": 305500
},
{
"epoch": 98.77181641887525,
"grad_norm": 2.0315310955047607,
"learning_rate": 0.001,
"loss": 1.1987,
"step": 305600
},
{
"epoch": 98.80413703943115,
"grad_norm": 1.6771689653396606,
"learning_rate": 0.001,
"loss": 1.1664,
"step": 305700
},
{
"epoch": 98.83645765998708,
"grad_norm": 1.3909519910812378,
"learning_rate": 0.001,
"loss": 1.1754,
"step": 305800
},
{
"epoch": 98.86877828054298,
"grad_norm": 2.052640914916992,
"learning_rate": 0.001,
"loss": 1.1754,
"step": 305900
},
{
"epoch": 98.9010989010989,
"grad_norm": 1.5628565549850464,
"learning_rate": 0.001,
"loss": 1.1769,
"step": 306000
},
{
"epoch": 98.93341952165481,
"grad_norm": 1.619795322418213,
"learning_rate": 0.001,
"loss": 1.1961,
"step": 306100
},
{
"epoch": 98.96574014221073,
"grad_norm": 2.042397975921631,
"learning_rate": 0.001,
"loss": 1.1972,
"step": 306200
},
{
"epoch": 98.99806076276664,
"grad_norm": 1.8413684368133545,
"learning_rate": 0.001,
"loss": 1.1885,
"step": 306300
},
{
"epoch": 99.03038138332256,
"grad_norm": 2.001903772354126,
"learning_rate": 0.001,
"loss": 1.1038,
"step": 306400
},
{
"epoch": 99.06270200387847,
"grad_norm": 1.5408207178115845,
"learning_rate": 0.001,
"loss": 1.0963,
"step": 306500
},
{
"epoch": 99.09502262443439,
"grad_norm": 1.8227003812789917,
"learning_rate": 0.001,
"loss": 1.0963,
"step": 306600
},
{
"epoch": 99.1273432449903,
"grad_norm": 1.7955845594406128,
"learning_rate": 0.001,
"loss": 1.0903,
"step": 306700
},
{
"epoch": 99.15966386554622,
"grad_norm": 1.5789191722869873,
"learning_rate": 0.001,
"loss": 1.1043,
"step": 306800
},
{
"epoch": 99.19198448610213,
"grad_norm": 2.1326937675476074,
"learning_rate": 0.001,
"loss": 1.1306,
"step": 306900
},
{
"epoch": 99.22430510665805,
"grad_norm": 2.003293514251709,
"learning_rate": 0.001,
"loss": 1.1107,
"step": 307000
},
{
"epoch": 99.25662572721396,
"grad_norm": 1.4671404361724854,
"learning_rate": 0.001,
"loss": 1.1183,
"step": 307100
},
{
"epoch": 99.28894634776988,
"grad_norm": 1.8515692949295044,
"learning_rate": 0.001,
"loss": 1.1388,
"step": 307200
},
{
"epoch": 99.32126696832579,
"grad_norm": 1.925635576248169,
"learning_rate": 0.001,
"loss": 1.1495,
"step": 307300
},
{
"epoch": 99.35358758888171,
"grad_norm": 1.4091397523880005,
"learning_rate": 0.001,
"loss": 1.1281,
"step": 307400
},
{
"epoch": 99.38590820943762,
"grad_norm": 2.091721534729004,
"learning_rate": 0.001,
"loss": 1.1499,
"step": 307500
},
{
"epoch": 99.41822882999354,
"grad_norm": 2.5117223262786865,
"learning_rate": 0.001,
"loss": 1.1406,
"step": 307600
},
{
"epoch": 99.45054945054945,
"grad_norm": 1.422900676727295,
"learning_rate": 0.001,
"loss": 1.1439,
"step": 307700
},
{
"epoch": 99.48287007110537,
"grad_norm": 1.3010830879211426,
"learning_rate": 0.001,
"loss": 1.1704,
"step": 307800
},
{
"epoch": 99.51519069166127,
"grad_norm": 1.8194390535354614,
"learning_rate": 0.001,
"loss": 1.1524,
"step": 307900
},
{
"epoch": 99.5475113122172,
"grad_norm": 2.241621255874634,
"learning_rate": 0.001,
"loss": 1.1472,
"step": 308000
}
],
"logging_steps": 100,
"max_steps": 309400,
"num_input_tokens_seen": 0,
"num_train_epochs": 100,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.6949805082257654e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}