SD21_100K_A_Inv / trainer_state.json
AusmitM's picture
Upload 31 files
a3ff611 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 100.0,
"eval_steps": 20000,
"global_step": 309400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03232062055591468,
"grad_norm": 94.4136962890625,
"learning_rate": 9.900000000000002e-06,
"loss": 5.124,
"step": 100
},
{
"epoch": 0.06464124111182935,
"grad_norm": 30.38750457763672,
"learning_rate": 1.9900000000000003e-05,
"loss": 4.3741,
"step": 200
},
{
"epoch": 0.09696186166774402,
"grad_norm": 13.193641662597656,
"learning_rate": 2.9900000000000002e-05,
"loss": 4.1239,
"step": 300
},
{
"epoch": 0.1292824822236587,
"grad_norm": 6.856531620025635,
"learning_rate": 3.99e-05,
"loss": 4.0501,
"step": 400
},
{
"epoch": 0.16160310277957338,
"grad_norm": 6.676848411560059,
"learning_rate": 4.99e-05,
"loss": 3.996,
"step": 500
},
{
"epoch": 0.19392372333548805,
"grad_norm": 3.3930461406707764,
"learning_rate": 5.9900000000000006e-05,
"loss": 3.9712,
"step": 600
},
{
"epoch": 0.22624434389140272,
"grad_norm": 3.4925343990325928,
"learning_rate": 6.99e-05,
"loss": 3.945,
"step": 700
},
{
"epoch": 0.2585649644473174,
"grad_norm": 2.4270923137664795,
"learning_rate": 7.99e-05,
"loss": 3.9171,
"step": 800
},
{
"epoch": 0.2908855850032321,
"grad_norm": 8.925300598144531,
"learning_rate": 8.989999999999999e-05,
"loss": 3.86,
"step": 900
},
{
"epoch": 0.32320620555914675,
"grad_norm": 3.1117539405822754,
"learning_rate": 9.99e-05,
"loss": 3.8611,
"step": 1000
},
{
"epoch": 0.3555268261150614,
"grad_norm": 3.8932859897613525,
"learning_rate": 0.0001099,
"loss": 3.8336,
"step": 1100
},
{
"epoch": 0.3878474466709761,
"grad_norm": 9.448994636535645,
"learning_rate": 0.00011990000000000001,
"loss": 3.8204,
"step": 1200
},
{
"epoch": 0.42016806722689076,
"grad_norm": 2.309150218963623,
"learning_rate": 0.00012989999999999999,
"loss": 3.7924,
"step": 1300
},
{
"epoch": 0.45248868778280543,
"grad_norm": 2.237368106842041,
"learning_rate": 0.0001399,
"loss": 3.7787,
"step": 1400
},
{
"epoch": 0.4848093083387201,
"grad_norm": 2.2044219970703125,
"learning_rate": 0.0001499,
"loss": 3.7389,
"step": 1500
},
{
"epoch": 0.5171299288946348,
"grad_norm": 10.110157012939453,
"learning_rate": 0.00015989999999999998,
"loss": 3.7319,
"step": 1600
},
{
"epoch": 0.5494505494505495,
"grad_norm": 2.0701940059661865,
"learning_rate": 0.0001699,
"loss": 3.7474,
"step": 1700
},
{
"epoch": 0.5817711700064642,
"grad_norm": 1.965275764465332,
"learning_rate": 0.0001799,
"loss": 3.7112,
"step": 1800
},
{
"epoch": 0.6140917905623788,
"grad_norm": 2.06203293800354,
"learning_rate": 0.0001899,
"loss": 3.7154,
"step": 1900
},
{
"epoch": 0.6464124111182935,
"grad_norm": 2.364362955093384,
"learning_rate": 0.0001999,
"loss": 3.6791,
"step": 2000
},
{
"epoch": 0.6787330316742082,
"grad_norm": 2.3111486434936523,
"learning_rate": 0.0002099,
"loss": 3.6795,
"step": 2100
},
{
"epoch": 0.7110536522301228,
"grad_norm": 2.442371129989624,
"learning_rate": 0.0002199,
"loss": 3.6645,
"step": 2200
},
{
"epoch": 0.7433742727860375,
"grad_norm": 2.105405569076538,
"learning_rate": 0.0002299,
"loss": 3.6598,
"step": 2300
},
{
"epoch": 0.7756948933419522,
"grad_norm": 2.0721235275268555,
"learning_rate": 0.0002399,
"loss": 3.6406,
"step": 2400
},
{
"epoch": 0.8080155138978669,
"grad_norm": 2.167262077331543,
"learning_rate": 0.0002499,
"loss": 3.6207,
"step": 2500
},
{
"epoch": 0.8403361344537815,
"grad_norm": 2.2569775581359863,
"learning_rate": 0.00025990000000000003,
"loss": 3.6442,
"step": 2600
},
{
"epoch": 0.8726567550096962,
"grad_norm": 2.527228832244873,
"learning_rate": 0.0002699,
"loss": 3.5954,
"step": 2700
},
{
"epoch": 0.9049773755656109,
"grad_norm": 2.4030511379241943,
"learning_rate": 0.0002799,
"loss": 3.5998,
"step": 2800
},
{
"epoch": 0.9372979961215255,
"grad_norm": 2.482780694961548,
"learning_rate": 0.0002899,
"loss": 3.5901,
"step": 2900
},
{
"epoch": 0.9696186166774402,
"grad_norm": 2.4679722785949707,
"learning_rate": 0.0002999,
"loss": 3.57,
"step": 3000
},
{
"epoch": 1.0019392372333549,
"grad_norm": 1.5613876581192017,
"learning_rate": 0.0003099,
"loss": 3.5814,
"step": 3100
},
{
"epoch": 1.0342598577892697,
"grad_norm": 2.0592920780181885,
"learning_rate": 0.0003199,
"loss": 3.5201,
"step": 3200
},
{
"epoch": 1.0665804783451842,
"grad_norm": 1.528223991394043,
"learning_rate": 0.00032990000000000005,
"loss": 3.5305,
"step": 3300
},
{
"epoch": 1.098901098901099,
"grad_norm": 1.250991940498352,
"learning_rate": 0.00033989999999999997,
"loss": 3.5284,
"step": 3400
},
{
"epoch": 1.1312217194570136,
"grad_norm": 1.6665195226669312,
"learning_rate": 0.0003499,
"loss": 3.5013,
"step": 3500
},
{
"epoch": 1.1635423400129283,
"grad_norm": 2.1032094955444336,
"learning_rate": 0.0003599,
"loss": 3.545,
"step": 3600
},
{
"epoch": 1.195862960568843,
"grad_norm": 4.082169532775879,
"learning_rate": 0.0003699,
"loss": 3.5364,
"step": 3700
},
{
"epoch": 1.2281835811247577,
"grad_norm": 1.0562688112258911,
"learning_rate": 0.0003799,
"loss": 3.5221,
"step": 3800
},
{
"epoch": 1.2605042016806722,
"grad_norm": 1.779932975769043,
"learning_rate": 0.00038990000000000004,
"loss": 3.5325,
"step": 3900
},
{
"epoch": 1.292824822236587,
"grad_norm": 3.3610312938690186,
"learning_rate": 0.00039989999999999996,
"loss": 3.5088,
"step": 4000
},
{
"epoch": 1.3251454427925016,
"grad_norm": 1.0462590456008911,
"learning_rate": 0.0004099,
"loss": 3.5037,
"step": 4100
},
{
"epoch": 1.3574660633484164,
"grad_norm": 0.9665063619613647,
"learning_rate": 0.0004199,
"loss": 3.5052,
"step": 4200
},
{
"epoch": 1.389786683904331,
"grad_norm": 1.0996367931365967,
"learning_rate": 0.0004299,
"loss": 3.4985,
"step": 4300
},
{
"epoch": 1.4221073044602457,
"grad_norm": 1.4393835067749023,
"learning_rate": 0.0004399,
"loss": 3.5036,
"step": 4400
},
{
"epoch": 1.4544279250161603,
"grad_norm": 1.8453340530395508,
"learning_rate": 0.00044990000000000004,
"loss": 3.5024,
"step": 4500
},
{
"epoch": 1.486748545572075,
"grad_norm": 1.7378803491592407,
"learning_rate": 0.0004599,
"loss": 3.4821,
"step": 4600
},
{
"epoch": 1.5190691661279896,
"grad_norm": 1.6973575353622437,
"learning_rate": 0.0004699,
"loss": 3.4933,
"step": 4700
},
{
"epoch": 1.5513897866839044,
"grad_norm": 1.0343236923217773,
"learning_rate": 0.0004799,
"loss": 3.4748,
"step": 4800
},
{
"epoch": 1.5837104072398192,
"grad_norm": 1.4053367376327515,
"learning_rate": 0.0004899,
"loss": 3.4884,
"step": 4900
},
{
"epoch": 1.6160310277957337,
"grad_norm": 1.3215560913085938,
"learning_rate": 0.0004999000000000001,
"loss": 3.4859,
"step": 5000
},
{
"epoch": 1.6483516483516483,
"grad_norm": 2.2094578742980957,
"learning_rate": 0.0005099,
"loss": 3.4712,
"step": 5100
},
{
"epoch": 1.680672268907563,
"grad_norm": 1.2302595376968384,
"learning_rate": 0.0005199,
"loss": 3.4598,
"step": 5200
},
{
"epoch": 1.7129928894634778,
"grad_norm": 1.5961023569107056,
"learning_rate": 0.0005299,
"loss": 3.4598,
"step": 5300
},
{
"epoch": 1.7453135100193924,
"grad_norm": 1.0842056274414062,
"learning_rate": 0.0005399000000000001,
"loss": 3.4622,
"step": 5400
},
{
"epoch": 1.777634130575307,
"grad_norm": 0.8788201212882996,
"learning_rate": 0.0005499000000000001,
"loss": 3.4446,
"step": 5500
},
{
"epoch": 1.8099547511312217,
"grad_norm": 1.0178688764572144,
"learning_rate": 0.0005599,
"loss": 3.4349,
"step": 5600
},
{
"epoch": 1.8422753716871365,
"grad_norm": 1.478924036026001,
"learning_rate": 0.0005698999999999999,
"loss": 3.4291,
"step": 5700
},
{
"epoch": 1.874595992243051,
"grad_norm": 0.9952222108840942,
"learning_rate": 0.0005799,
"loss": 3.463,
"step": 5800
},
{
"epoch": 1.9069166127989656,
"grad_norm": 1.0364357233047485,
"learning_rate": 0.0005899,
"loss": 3.4442,
"step": 5900
},
{
"epoch": 1.9392372333548804,
"grad_norm": 1.159144639968872,
"learning_rate": 0.0005999,
"loss": 3.4361,
"step": 6000
},
{
"epoch": 1.9715578539107952,
"grad_norm": 1.1590672731399536,
"learning_rate": 0.0006099,
"loss": 3.4418,
"step": 6100
},
{
"epoch": 2.0038784744667097,
"grad_norm": 0.8586789965629578,
"learning_rate": 0.0006199,
"loss": 3.43,
"step": 6200
},
{
"epoch": 2.0361990950226243,
"grad_norm": 0.771930456161499,
"learning_rate": 0.0006299000000000001,
"loss": 3.3556,
"step": 6300
},
{
"epoch": 2.0685197155785393,
"grad_norm": 0.9476509690284729,
"learning_rate": 0.0006399,
"loss": 3.3543,
"step": 6400
},
{
"epoch": 2.100840336134454,
"grad_norm": 0.9481135010719299,
"learning_rate": 0.0006499,
"loss": 3.3565,
"step": 6500
},
{
"epoch": 2.1331609566903684,
"grad_norm": 1.0429929494857788,
"learning_rate": 0.0006599,
"loss": 3.358,
"step": 6600
},
{
"epoch": 2.165481577246283,
"grad_norm": 1.1219472885131836,
"learning_rate": 0.0006699000000000001,
"loss": 3.3608,
"step": 6700
},
{
"epoch": 2.197802197802198,
"grad_norm": 0.7779324650764465,
"learning_rate": 0.0006799,
"loss": 3.3463,
"step": 6800
},
{
"epoch": 2.2301228183581125,
"grad_norm": 0.9975550770759583,
"learning_rate": 0.0006899,
"loss": 3.3443,
"step": 6900
},
{
"epoch": 2.262443438914027,
"grad_norm": 0.814545214176178,
"learning_rate": 0.0006998999999999999,
"loss": 3.3508,
"step": 7000
},
{
"epoch": 2.2947640594699417,
"grad_norm": 1.0870559215545654,
"learning_rate": 0.0007099,
"loss": 3.3362,
"step": 7100
},
{
"epoch": 2.3270846800258567,
"grad_norm": 1.142208456993103,
"learning_rate": 0.0007199,
"loss": 3.3535,
"step": 7200
},
{
"epoch": 2.3594053005817712,
"grad_norm": 1.0361860990524292,
"learning_rate": 0.0007299,
"loss": 3.3394,
"step": 7300
},
{
"epoch": 2.391725921137686,
"grad_norm": 1.1007474660873413,
"learning_rate": 0.0007399,
"loss": 3.3409,
"step": 7400
},
{
"epoch": 2.4240465416936003,
"grad_norm": 1.2482529878616333,
"learning_rate": 0.0007499000000000001,
"loss": 3.3539,
"step": 7500
},
{
"epoch": 2.4563671622495153,
"grad_norm": 1.0279099941253662,
"learning_rate": 0.0007599,
"loss": 3.3537,
"step": 7600
},
{
"epoch": 2.48868778280543,
"grad_norm": 1.0041053295135498,
"learning_rate": 0.0007699,
"loss": 3.3418,
"step": 7700
},
{
"epoch": 2.5210084033613445,
"grad_norm": 1.5385572910308838,
"learning_rate": 0.0007799,
"loss": 3.3214,
"step": 7800
},
{
"epoch": 2.553329023917259,
"grad_norm": 0.8616864681243896,
"learning_rate": 0.0007899000000000001,
"loss": 3.3153,
"step": 7900
},
{
"epoch": 2.585649644473174,
"grad_norm": 0.8564915657043457,
"learning_rate": 0.0007999000000000001,
"loss": 3.3465,
"step": 8000
},
{
"epoch": 2.6179702650290886,
"grad_norm": 1.056174874305725,
"learning_rate": 0.0008099,
"loss": 3.3479,
"step": 8100
},
{
"epoch": 2.650290885585003,
"grad_norm": 1.1474746465682983,
"learning_rate": 0.0008198999999999999,
"loss": 3.3129,
"step": 8200
},
{
"epoch": 2.682611506140918,
"grad_norm": 0.9924173355102539,
"learning_rate": 0.0008299,
"loss": 3.3333,
"step": 8300
},
{
"epoch": 2.7149321266968327,
"grad_norm": 1.0320525169372559,
"learning_rate": 0.0008399,
"loss": 3.3447,
"step": 8400
},
{
"epoch": 2.7472527472527473,
"grad_norm": 0.8991305828094482,
"learning_rate": 0.0008499,
"loss": 3.3354,
"step": 8500
},
{
"epoch": 2.779573367808662,
"grad_norm": 1.1122479438781738,
"learning_rate": 0.0008599,
"loss": 3.3327,
"step": 8600
},
{
"epoch": 2.8118939883645764,
"grad_norm": 1.3306217193603516,
"learning_rate": 0.0008699000000000001,
"loss": 3.3232,
"step": 8700
},
{
"epoch": 2.8442146089204914,
"grad_norm": 1.1624479293823242,
"learning_rate": 0.0008799000000000001,
"loss": 3.3337,
"step": 8800
},
{
"epoch": 2.876535229476406,
"grad_norm": 0.9736101031303406,
"learning_rate": 0.0008899,
"loss": 3.3344,
"step": 8900
},
{
"epoch": 2.9088558500323205,
"grad_norm": 0.8354546427726746,
"learning_rate": 0.0008999,
"loss": 3.309,
"step": 9000
},
{
"epoch": 2.9411764705882355,
"grad_norm": 1.0423007011413574,
"learning_rate": 0.0009099,
"loss": 3.3371,
"step": 9100
},
{
"epoch": 2.97349709114415,
"grad_norm": 1.0594122409820557,
"learning_rate": 0.0009199000000000001,
"loss": 3.3351,
"step": 9200
},
{
"epoch": 3.0058177117000646,
"grad_norm": 1.0311226844787598,
"learning_rate": 0.0009299,
"loss": 3.3134,
"step": 9300
},
{
"epoch": 3.038138332255979,
"grad_norm": 0.8921378254890442,
"learning_rate": 0.0009399,
"loss": 3.2091,
"step": 9400
},
{
"epoch": 3.070458952811894,
"grad_norm": 1.0563340187072754,
"learning_rate": 0.0009498999999999999,
"loss": 3.2276,
"step": 9500
},
{
"epoch": 3.1027795733678087,
"grad_norm": 0.9416050314903259,
"learning_rate": 0.0009599,
"loss": 3.2211,
"step": 9600
},
{
"epoch": 3.1351001939237233,
"grad_norm": 1.0302422046661377,
"learning_rate": 0.0009699,
"loss": 3.2241,
"step": 9700
},
{
"epoch": 3.167420814479638,
"grad_norm": 0.9644986987113953,
"learning_rate": 0.0009799,
"loss": 3.2215,
"step": 9800
},
{
"epoch": 3.199741435035553,
"grad_norm": 0.8126795887947083,
"learning_rate": 0.0009899,
"loss": 3.203,
"step": 9900
},
{
"epoch": 3.2320620555914674,
"grad_norm": 0.8086977005004883,
"learning_rate": 0.0009999,
"loss": 3.2374,
"step": 10000
},
{
"epoch": 3.264382676147382,
"grad_norm": 0.9846087098121643,
"learning_rate": 0.001,
"loss": 3.2431,
"step": 10100
},
{
"epoch": 3.2967032967032965,
"grad_norm": 1.3535737991333008,
"learning_rate": 0.001,
"loss": 3.227,
"step": 10200
},
{
"epoch": 3.3290239172592115,
"grad_norm": 0.9775270223617554,
"learning_rate": 0.001,
"loss": 3.2288,
"step": 10300
},
{
"epoch": 3.361344537815126,
"grad_norm": 1.1156953573226929,
"learning_rate": 0.001,
"loss": 3.2141,
"step": 10400
},
{
"epoch": 3.3936651583710407,
"grad_norm": 1.0679025650024414,
"learning_rate": 0.001,
"loss": 3.2245,
"step": 10500
},
{
"epoch": 3.425985778926955,
"grad_norm": 0.6540291905403137,
"learning_rate": 0.001,
"loss": 3.2456,
"step": 10600
},
{
"epoch": 3.45830639948287,
"grad_norm": 1.0269280672073364,
"learning_rate": 0.001,
"loss": 3.2473,
"step": 10700
},
{
"epoch": 3.490627020038785,
"grad_norm": 1.029340386390686,
"learning_rate": 0.001,
"loss": 3.2423,
"step": 10800
},
{
"epoch": 3.5229476405946993,
"grad_norm": 0.9062139391899109,
"learning_rate": 0.001,
"loss": 3.2239,
"step": 10900
},
{
"epoch": 3.555268261150614,
"grad_norm": 0.9112501740455627,
"learning_rate": 0.001,
"loss": 3.2181,
"step": 11000
},
{
"epoch": 3.587588881706529,
"grad_norm": 0.9416837692260742,
"learning_rate": 0.001,
"loss": 3.2337,
"step": 11100
},
{
"epoch": 3.6199095022624435,
"grad_norm": 1.097151517868042,
"learning_rate": 0.001,
"loss": 3.2356,
"step": 11200
},
{
"epoch": 3.652230122818358,
"grad_norm": 0.7672462463378906,
"learning_rate": 0.001,
"loss": 3.2566,
"step": 11300
},
{
"epoch": 3.684550743374273,
"grad_norm": 0.9438802599906921,
"learning_rate": 0.001,
"loss": 3.2223,
"step": 11400
},
{
"epoch": 3.7168713639301876,
"grad_norm": 0.958585798740387,
"learning_rate": 0.001,
"loss": 3.1993,
"step": 11500
},
{
"epoch": 3.749191984486102,
"grad_norm": 0.962122917175293,
"learning_rate": 0.001,
"loss": 3.2214,
"step": 11600
},
{
"epoch": 3.7815126050420167,
"grad_norm": 0.9139586091041565,
"learning_rate": 0.001,
"loss": 3.2304,
"step": 11700
},
{
"epoch": 3.8138332255979313,
"grad_norm": 0.9144065976142883,
"learning_rate": 0.001,
"loss": 3.2293,
"step": 11800
},
{
"epoch": 3.8461538461538463,
"grad_norm": 1.07357656955719,
"learning_rate": 0.001,
"loss": 3.2099,
"step": 11900
},
{
"epoch": 3.878474466709761,
"grad_norm": 1.0424174070358276,
"learning_rate": 0.001,
"loss": 3.2267,
"step": 12000
},
{
"epoch": 3.9107950872656754,
"grad_norm": 0.8743547797203064,
"learning_rate": 0.001,
"loss": 3.2174,
"step": 12100
},
{
"epoch": 3.9431157078215904,
"grad_norm": 1.0850470066070557,
"learning_rate": 0.001,
"loss": 3.2094,
"step": 12200
},
{
"epoch": 3.975436328377505,
"grad_norm": 1.1467704772949219,
"learning_rate": 0.001,
"loss": 3.2313,
"step": 12300
},
{
"epoch": 4.0077569489334195,
"grad_norm": 1.1973806619644165,
"learning_rate": 0.001,
"loss": 3.2122,
"step": 12400
},
{
"epoch": 4.040077569489334,
"grad_norm": 0.83821040391922,
"learning_rate": 0.001,
"loss": 3.0902,
"step": 12500
},
{
"epoch": 4.072398190045249,
"grad_norm": 1.1615233421325684,
"learning_rate": 0.001,
"loss": 3.0969,
"step": 12600
},
{
"epoch": 4.104718810601163,
"grad_norm": 0.8355074524879456,
"learning_rate": 0.001,
"loss": 3.0995,
"step": 12700
},
{
"epoch": 4.137039431157079,
"grad_norm": 0.9700531959533691,
"learning_rate": 0.001,
"loss": 3.0991,
"step": 12800
},
{
"epoch": 4.169360051712993,
"grad_norm": 0.9404054284095764,
"learning_rate": 0.001,
"loss": 3.1024,
"step": 12900
},
{
"epoch": 4.201680672268908,
"grad_norm": 1.3212854862213135,
"learning_rate": 0.001,
"loss": 3.0951,
"step": 13000
},
{
"epoch": 4.234001292824822,
"grad_norm": 0.8313888907432556,
"learning_rate": 0.001,
"loss": 3.1069,
"step": 13100
},
{
"epoch": 4.266321913380737,
"grad_norm": 0.840483546257019,
"learning_rate": 0.001,
"loss": 3.0888,
"step": 13200
},
{
"epoch": 4.298642533936651,
"grad_norm": 1.115980625152588,
"learning_rate": 0.001,
"loss": 3.1097,
"step": 13300
},
{
"epoch": 4.330963154492566,
"grad_norm": 0.7574254274368286,
"learning_rate": 0.001,
"loss": 3.1054,
"step": 13400
},
{
"epoch": 4.3632837750484805,
"grad_norm": 0.7509799003601074,
"learning_rate": 0.001,
"loss": 3.1317,
"step": 13500
},
{
"epoch": 4.395604395604396,
"grad_norm": 0.8572108149528503,
"learning_rate": 0.001,
"loss": 3.1047,
"step": 13600
},
{
"epoch": 4.4279250161603105,
"grad_norm": 1.0784690380096436,
"learning_rate": 0.001,
"loss": 3.1183,
"step": 13700
},
{
"epoch": 4.460245636716225,
"grad_norm": 0.8089799284934998,
"learning_rate": 0.001,
"loss": 3.1231,
"step": 13800
},
{
"epoch": 4.49256625727214,
"grad_norm": 0.8911246061325073,
"learning_rate": 0.001,
"loss": 3.1244,
"step": 13900
},
{
"epoch": 4.524886877828054,
"grad_norm": 1.1788434982299805,
"learning_rate": 0.001,
"loss": 3.1191,
"step": 14000
},
{
"epoch": 4.557207498383969,
"grad_norm": 0.9430946707725525,
"learning_rate": 0.001,
"loss": 3.1014,
"step": 14100
},
{
"epoch": 4.589528118939883,
"grad_norm": 1.1967456340789795,
"learning_rate": 0.001,
"loss": 3.114,
"step": 14200
},
{
"epoch": 4.621848739495798,
"grad_norm": 1.2368634939193726,
"learning_rate": 0.001,
"loss": 3.117,
"step": 14300
},
{
"epoch": 4.654169360051713,
"grad_norm": 1.0995897054672241,
"learning_rate": 0.001,
"loss": 3.1105,
"step": 14400
},
{
"epoch": 4.686489980607628,
"grad_norm": 1.1519954204559326,
"learning_rate": 0.001,
"loss": 3.1174,
"step": 14500
},
{
"epoch": 4.7188106011635425,
"grad_norm": 0.6269290447235107,
"learning_rate": 0.001,
"loss": 3.1243,
"step": 14600
},
{
"epoch": 4.751131221719457,
"grad_norm": 0.797653317451477,
"learning_rate": 0.001,
"loss": 3.1171,
"step": 14700
},
{
"epoch": 4.783451842275372,
"grad_norm": 0.844874382019043,
"learning_rate": 0.001,
"loss": 3.1265,
"step": 14800
},
{
"epoch": 4.815772462831286,
"grad_norm": 1.2588833570480347,
"learning_rate": 0.001,
"loss": 3.1136,
"step": 14900
},
{
"epoch": 4.848093083387201,
"grad_norm": 0.7878243327140808,
"learning_rate": 0.001,
"loss": 3.0769,
"step": 15000
},
{
"epoch": 4.880413703943116,
"grad_norm": 0.7327384352684021,
"learning_rate": 0.001,
"loss": 3.1036,
"step": 15100
},
{
"epoch": 4.912734324499031,
"grad_norm": 0.8009315133094788,
"learning_rate": 0.001,
"loss": 3.112,
"step": 15200
},
{
"epoch": 4.945054945054945,
"grad_norm": 0.8452103734016418,
"learning_rate": 0.001,
"loss": 3.1162,
"step": 15300
},
{
"epoch": 4.97737556561086,
"grad_norm": 0.8712530732154846,
"learning_rate": 0.001,
"loss": 3.1105,
"step": 15400
},
{
"epoch": 5.009696186166774,
"grad_norm": 0.9652482271194458,
"learning_rate": 0.001,
"loss": 3.0834,
"step": 15500
},
{
"epoch": 5.042016806722689,
"grad_norm": 0.859568178653717,
"learning_rate": 0.001,
"loss": 2.9717,
"step": 15600
},
{
"epoch": 5.0743374272786035,
"grad_norm": 0.7683727145195007,
"learning_rate": 0.001,
"loss": 2.9795,
"step": 15700
},
{
"epoch": 5.106658047834518,
"grad_norm": 1.091881275177002,
"learning_rate": 0.001,
"loss": 2.9931,
"step": 15800
},
{
"epoch": 5.1389786683904335,
"grad_norm": 0.7646498680114746,
"learning_rate": 0.001,
"loss": 2.9805,
"step": 15900
},
{
"epoch": 5.171299288946348,
"grad_norm": 0.8571824431419373,
"learning_rate": 0.001,
"loss": 3.0124,
"step": 16000
},
{
"epoch": 5.203619909502263,
"grad_norm": 0.8407406210899353,
"learning_rate": 0.001,
"loss": 2.9923,
"step": 16100
},
{
"epoch": 5.235940530058177,
"grad_norm": 0.8803465962409973,
"learning_rate": 0.001,
"loss": 2.9907,
"step": 16200
},
{
"epoch": 5.268261150614092,
"grad_norm": 1.067539930343628,
"learning_rate": 0.001,
"loss": 3.0015,
"step": 16300
},
{
"epoch": 5.300581771170006,
"grad_norm": 1.130493402481079,
"learning_rate": 0.001,
"loss": 2.9913,
"step": 16400
},
{
"epoch": 5.332902391725921,
"grad_norm": 1.1129696369171143,
"learning_rate": 0.001,
"loss": 3.0305,
"step": 16500
},
{
"epoch": 5.365223012281835,
"grad_norm": 1.0208700895309448,
"learning_rate": 0.001,
"loss": 3.0207,
"step": 16600
},
{
"epoch": 5.397543632837751,
"grad_norm": 0.9395743608474731,
"learning_rate": 0.001,
"loss": 3.0008,
"step": 16700
},
{
"epoch": 5.429864253393665,
"grad_norm": 0.9820137619972229,
"learning_rate": 0.001,
"loss": 3.0232,
"step": 16800
},
{
"epoch": 5.46218487394958,
"grad_norm": 0.9509523510932922,
"learning_rate": 0.001,
"loss": 3.0245,
"step": 16900
},
{
"epoch": 5.4945054945054945,
"grad_norm": 0.9762880206108093,
"learning_rate": 0.001,
"loss": 3.0071,
"step": 17000
},
{
"epoch": 5.526826115061409,
"grad_norm": 0.9137035608291626,
"learning_rate": 0.001,
"loss": 3.028,
"step": 17100
},
{
"epoch": 5.559146735617324,
"grad_norm": 0.9402992129325867,
"learning_rate": 0.001,
"loss": 3.0189,
"step": 17200
},
{
"epoch": 5.591467356173238,
"grad_norm": 0.7806065082550049,
"learning_rate": 0.001,
"loss": 3.0051,
"step": 17300
},
{
"epoch": 5.623787976729153,
"grad_norm": 0.7693142294883728,
"learning_rate": 0.001,
"loss": 3.0288,
"step": 17400
},
{
"epoch": 5.656108597285068,
"grad_norm": 0.8223123550415039,
"learning_rate": 0.001,
"loss": 3.0019,
"step": 17500
},
{
"epoch": 5.688429217840983,
"grad_norm": 0.9738755226135254,
"learning_rate": 0.001,
"loss": 3.0289,
"step": 17600
},
{
"epoch": 5.720749838396897,
"grad_norm": 0.794930636882782,
"learning_rate": 0.001,
"loss": 3.02,
"step": 17700
},
{
"epoch": 5.753070458952812,
"grad_norm": 0.8179877996444702,
"learning_rate": 0.001,
"loss": 3.0317,
"step": 17800
},
{
"epoch": 5.785391079508726,
"grad_norm": 0.883860170841217,
"learning_rate": 0.001,
"loss": 3.0336,
"step": 17900
},
{
"epoch": 5.817711700064641,
"grad_norm": 1.0176060199737549,
"learning_rate": 0.001,
"loss": 3.0055,
"step": 18000
},
{
"epoch": 5.850032320620556,
"grad_norm": 0.7133760452270508,
"learning_rate": 0.001,
"loss": 3.0179,
"step": 18100
},
{
"epoch": 5.882352941176471,
"grad_norm": 0.8435713052749634,
"learning_rate": 0.001,
"loss": 3.0179,
"step": 18200
},
{
"epoch": 5.914673561732386,
"grad_norm": 1.1324796676635742,
"learning_rate": 0.001,
"loss": 3.034,
"step": 18300
},
{
"epoch": 5.9469941822883,
"grad_norm": 0.8907880187034607,
"learning_rate": 0.001,
"loss": 3.0362,
"step": 18400
},
{
"epoch": 5.979314802844215,
"grad_norm": 0.7963925004005432,
"learning_rate": 0.001,
"loss": 3.0192,
"step": 18500
},
{
"epoch": 6.011635423400129,
"grad_norm": 0.9743801355361938,
"learning_rate": 0.001,
"loss": 2.9673,
"step": 18600
},
{
"epoch": 6.043956043956044,
"grad_norm": 0.934952437877655,
"learning_rate": 0.001,
"loss": 2.9047,
"step": 18700
},
{
"epoch": 6.076276664511958,
"grad_norm": 0.9375582933425903,
"learning_rate": 0.001,
"loss": 2.8872,
"step": 18800
},
{
"epoch": 6.108597285067873,
"grad_norm": 1.1716747283935547,
"learning_rate": 0.001,
"loss": 2.9023,
"step": 18900
},
{
"epoch": 6.140917905623788,
"grad_norm": 0.8983238935470581,
"learning_rate": 0.001,
"loss": 2.8882,
"step": 19000
},
{
"epoch": 6.173238526179703,
"grad_norm": 0.8296594619750977,
"learning_rate": 0.001,
"loss": 2.9151,
"step": 19100
},
{
"epoch": 6.2055591467356175,
"grad_norm": 0.7835599184036255,
"learning_rate": 0.001,
"loss": 2.8974,
"step": 19200
},
{
"epoch": 6.237879767291532,
"grad_norm": 1.4942642450332642,
"learning_rate": 0.001,
"loss": 2.9239,
"step": 19300
},
{
"epoch": 6.270200387847447,
"grad_norm": 1.0689419507980347,
"learning_rate": 0.001,
"loss": 2.9084,
"step": 19400
},
{
"epoch": 6.302521008403361,
"grad_norm": 0.8567982316017151,
"learning_rate": 0.001,
"loss": 2.9226,
"step": 19500
},
{
"epoch": 6.334841628959276,
"grad_norm": 0.758380115032196,
"learning_rate": 0.001,
"loss": 2.9312,
"step": 19600
},
{
"epoch": 6.36716224951519,
"grad_norm": 0.796357274055481,
"learning_rate": 0.001,
"loss": 2.9264,
"step": 19700
},
{
"epoch": 6.399482870071106,
"grad_norm": 1.1560801267623901,
"learning_rate": 0.001,
"loss": 2.9138,
"step": 19800
},
{
"epoch": 6.43180349062702,
"grad_norm": 0.9434789419174194,
"learning_rate": 0.001,
"loss": 2.9422,
"step": 19900
},
{
"epoch": 6.464124111182935,
"grad_norm": 0.6994492411613464,
"learning_rate": 0.001,
"loss": 2.9191,
"step": 20000
},
{
"epoch": 6.496444731738849,
"grad_norm": 0.7224216461181641,
"learning_rate": 0.001,
"loss": 2.8486,
"step": 20100
},
{
"epoch": 6.528765352294764,
"grad_norm": 1.068552851676941,
"learning_rate": 0.001,
"loss": 2.8749,
"step": 20200
},
{
"epoch": 6.5610859728506785,
"grad_norm": 0.9315723180770874,
"learning_rate": 0.001,
"loss": 2.8656,
"step": 20300
},
{
"epoch": 6.593406593406593,
"grad_norm": 0.8688428997993469,
"learning_rate": 0.001,
"loss": 2.8861,
"step": 20400
},
{
"epoch": 6.625727213962508,
"grad_norm": 0.9467467069625854,
"learning_rate": 0.001,
"loss": 2.8766,
"step": 20500
},
{
"epoch": 6.658047834518423,
"grad_norm": 0.8501046895980835,
"learning_rate": 0.001,
"loss": 2.8756,
"step": 20600
},
{
"epoch": 6.690368455074338,
"grad_norm": 0.8660249710083008,
"learning_rate": 0.001,
"loss": 2.884,
"step": 20700
},
{
"epoch": 6.722689075630252,
"grad_norm": 1.0461561679840088,
"learning_rate": 0.001,
"loss": 2.8805,
"step": 20800
},
{
"epoch": 6.755009696186167,
"grad_norm": 1.0024206638336182,
"learning_rate": 0.001,
"loss": 2.89,
"step": 20900
},
{
"epoch": 6.787330316742081,
"grad_norm": 0.9066793918609619,
"learning_rate": 0.001,
"loss": 2.8821,
"step": 21000
},
{
"epoch": 6.819650937297996,
"grad_norm": 0.9832000136375427,
"learning_rate": 0.001,
"loss": 2.8885,
"step": 21100
},
{
"epoch": 6.85197155785391,
"grad_norm": 0.9620049595832825,
"learning_rate": 0.001,
"loss": 2.8827,
"step": 21200
},
{
"epoch": 6.884292178409826,
"grad_norm": 0.8055165410041809,
"learning_rate": 0.001,
"loss": 2.8836,
"step": 21300
},
{
"epoch": 6.91661279896574,
"grad_norm": 0.9188902378082275,
"learning_rate": 0.001,
"loss": 2.8839,
"step": 21400
},
{
"epoch": 6.948933419521655,
"grad_norm": 0.9314457178115845,
"learning_rate": 0.001,
"loss": 2.8865,
"step": 21500
},
{
"epoch": 6.98125404007757,
"grad_norm": 0.9447315335273743,
"learning_rate": 0.001,
"loss": 2.883,
"step": 21600
},
{
"epoch": 7.013574660633484,
"grad_norm": 1.0738145112991333,
"learning_rate": 0.001,
"loss": 2.8576,
"step": 21700
},
{
"epoch": 7.045895281189399,
"grad_norm": 1.0047656297683716,
"learning_rate": 0.001,
"loss": 2.8245,
"step": 21800
},
{
"epoch": 7.078215901745313,
"grad_norm": 1.1342542171478271,
"learning_rate": 0.001,
"loss": 2.8145,
"step": 21900
},
{
"epoch": 7.110536522301228,
"grad_norm": 1.0721144676208496,
"learning_rate": 0.001,
"loss": 2.8396,
"step": 22000
},
{
"epoch": 7.142857142857143,
"grad_norm": 1.0003174543380737,
"learning_rate": 0.001,
"loss": 2.8265,
"step": 22100
},
{
"epoch": 7.175177763413058,
"grad_norm": 0.967972993850708,
"learning_rate": 0.001,
"loss": 2.8274,
"step": 22200
},
{
"epoch": 7.207498383968972,
"grad_norm": 0.9438908696174622,
"learning_rate": 0.001,
"loss": 2.8265,
"step": 22300
},
{
"epoch": 7.239819004524887,
"grad_norm": 0.9553608298301697,
"learning_rate": 0.001,
"loss": 2.8524,
"step": 22400
},
{
"epoch": 7.2721396250808015,
"grad_norm": 1.0461164712905884,
"learning_rate": 0.001,
"loss": 2.8562,
"step": 22500
},
{
"epoch": 7.304460245636716,
"grad_norm": 1.0306490659713745,
"learning_rate": 0.001,
"loss": 2.8297,
"step": 22600
},
{
"epoch": 7.336780866192631,
"grad_norm": 0.8757615089416504,
"learning_rate": 0.001,
"loss": 2.8544,
"step": 22700
},
{
"epoch": 7.369101486748546,
"grad_norm": 0.8251903057098389,
"learning_rate": 0.001,
"loss": 2.8401,
"step": 22800
},
{
"epoch": 7.401422107304461,
"grad_norm": 1.0933467149734497,
"learning_rate": 0.001,
"loss": 2.8338,
"step": 22900
},
{
"epoch": 7.433742727860375,
"grad_norm": 0.90103679895401,
"learning_rate": 0.001,
"loss": 2.8784,
"step": 23000
},
{
"epoch": 7.46606334841629,
"grad_norm": 1.121545433998108,
"learning_rate": 0.001,
"loss": 2.8682,
"step": 23100
},
{
"epoch": 7.498383968972204,
"grad_norm": 1.1099205017089844,
"learning_rate": 0.001,
"loss": 2.8866,
"step": 23200
},
{
"epoch": 7.530704589528119,
"grad_norm": 1.0353825092315674,
"learning_rate": 0.001,
"loss": 2.8642,
"step": 23300
},
{
"epoch": 7.563025210084033,
"grad_norm": 1.17049241065979,
"learning_rate": 0.001,
"loss": 2.883,
"step": 23400
},
{
"epoch": 7.595345830639948,
"grad_norm": 0.9848366975784302,
"learning_rate": 0.001,
"loss": 2.8821,
"step": 23500
},
{
"epoch": 7.6276664511958625,
"grad_norm": 1.0439475774765015,
"learning_rate": 0.001,
"loss": 2.8574,
"step": 23600
},
{
"epoch": 7.659987071751778,
"grad_norm": 1.0100176334381104,
"learning_rate": 0.001,
"loss": 2.8823,
"step": 23700
},
{
"epoch": 7.6923076923076925,
"grad_norm": 0.8114297986030579,
"learning_rate": 0.001,
"loss": 2.8861,
"step": 23800
},
{
"epoch": 7.724628312863607,
"grad_norm": 0.9616294503211975,
"learning_rate": 0.001,
"loss": 2.8623,
"step": 23900
},
{
"epoch": 7.756948933419522,
"grad_norm": 1.0478956699371338,
"learning_rate": 0.001,
"loss": 2.8811,
"step": 24000
},
{
"epoch": 7.789269553975436,
"grad_norm": 0.7646477222442627,
"learning_rate": 0.001,
"loss": 2.8786,
"step": 24100
},
{
"epoch": 7.821590174531351,
"grad_norm": 1.2913520336151123,
"learning_rate": 0.001,
"loss": 2.8812,
"step": 24200
},
{
"epoch": 7.853910795087265,
"grad_norm": 1.0038334131240845,
"learning_rate": 0.001,
"loss": 2.8881,
"step": 24300
},
{
"epoch": 7.886231415643181,
"grad_norm": 0.9279253482818604,
"learning_rate": 0.001,
"loss": 2.8725,
"step": 24400
},
{
"epoch": 7.918552036199095,
"grad_norm": 1.2101916074752808,
"learning_rate": 0.001,
"loss": 2.8985,
"step": 24500
},
{
"epoch": 7.95087265675501,
"grad_norm": 1.0921961069107056,
"learning_rate": 0.001,
"loss": 2.8677,
"step": 24600
},
{
"epoch": 7.983193277310924,
"grad_norm": 0.8465441465377808,
"learning_rate": 0.001,
"loss": 2.8784,
"step": 24700
},
{
"epoch": 8.015513897866839,
"grad_norm": 1.211400032043457,
"learning_rate": 0.001,
"loss": 2.8045,
"step": 24800
},
{
"epoch": 8.047834518422754,
"grad_norm": 0.9247515201568604,
"learning_rate": 0.001,
"loss": 2.7496,
"step": 24900
},
{
"epoch": 8.080155138978668,
"grad_norm": 1.4464887380599976,
"learning_rate": 0.001,
"loss": 2.7513,
"step": 25000
},
{
"epoch": 8.112475759534583,
"grad_norm": 1.3413829803466797,
"learning_rate": 0.001,
"loss": 2.7659,
"step": 25100
},
{
"epoch": 8.144796380090497,
"grad_norm": 1.0862212181091309,
"learning_rate": 0.001,
"loss": 2.7621,
"step": 25200
},
{
"epoch": 8.177117000646412,
"grad_norm": 1.6204279661178589,
"learning_rate": 0.001,
"loss": 2.763,
"step": 25300
},
{
"epoch": 8.209437621202326,
"grad_norm": 1.328371286392212,
"learning_rate": 0.001,
"loss": 2.7572,
"step": 25400
},
{
"epoch": 8.241758241758241,
"grad_norm": 1.3516168594360352,
"learning_rate": 0.001,
"loss": 2.766,
"step": 25500
},
{
"epoch": 8.274078862314157,
"grad_norm": 1.499612808227539,
"learning_rate": 0.001,
"loss": 2.7966,
"step": 25600
},
{
"epoch": 8.306399482870072,
"grad_norm": 1.4105859994888306,
"learning_rate": 0.001,
"loss": 2.7939,
"step": 25700
},
{
"epoch": 8.338720103425986,
"grad_norm": 1.1324214935302734,
"learning_rate": 0.001,
"loss": 2.7883,
"step": 25800
},
{
"epoch": 8.371040723981901,
"grad_norm": 1.451233148574829,
"learning_rate": 0.001,
"loss": 2.7555,
"step": 25900
},
{
"epoch": 8.403361344537815,
"grad_norm": 1.1795775890350342,
"learning_rate": 0.001,
"loss": 2.7797,
"step": 26000
},
{
"epoch": 8.43568196509373,
"grad_norm": 1.3906598091125488,
"learning_rate": 0.001,
"loss": 2.7877,
"step": 26100
},
{
"epoch": 8.468002585649645,
"grad_norm": 1.5041311979293823,
"learning_rate": 0.001,
"loss": 2.7946,
"step": 26200
},
{
"epoch": 8.50032320620556,
"grad_norm": 1.1948392391204834,
"learning_rate": 0.001,
"loss": 2.8193,
"step": 26300
},
{
"epoch": 8.532643826761474,
"grad_norm": 1.5433799028396606,
"learning_rate": 0.001,
"loss": 2.8171,
"step": 26400
},
{
"epoch": 8.564964447317388,
"grad_norm": 1.172419786453247,
"learning_rate": 0.001,
"loss": 2.7877,
"step": 26500
},
{
"epoch": 8.597285067873303,
"grad_norm": 1.047184944152832,
"learning_rate": 0.001,
"loss": 2.8103,
"step": 26600
},
{
"epoch": 8.629605688429217,
"grad_norm": 1.2895269393920898,
"learning_rate": 0.001,
"loss": 2.8001,
"step": 26700
},
{
"epoch": 8.661926308985132,
"grad_norm": 1.5003352165222168,
"learning_rate": 0.001,
"loss": 2.8126,
"step": 26800
},
{
"epoch": 8.694246929541046,
"grad_norm": 1.3400719165802002,
"learning_rate": 0.001,
"loss": 2.8173,
"step": 26900
},
{
"epoch": 8.726567550096961,
"grad_norm": 1.139378547668457,
"learning_rate": 0.001,
"loss": 2.8185,
"step": 27000
},
{
"epoch": 8.758888170652877,
"grad_norm": 1.3214950561523438,
"learning_rate": 0.001,
"loss": 2.7963,
"step": 27100
},
{
"epoch": 8.791208791208792,
"grad_norm": 1.3089869022369385,
"learning_rate": 0.001,
"loss": 2.8069,
"step": 27200
},
{
"epoch": 8.823529411764707,
"grad_norm": 1.0301275253295898,
"learning_rate": 0.001,
"loss": 2.8172,
"step": 27300
},
{
"epoch": 8.855850032320621,
"grad_norm": 2.0230872631073,
"learning_rate": 0.001,
"loss": 2.8163,
"step": 27400
},
{
"epoch": 8.888170652876536,
"grad_norm": 1.0308681726455688,
"learning_rate": 0.001,
"loss": 2.8241,
"step": 27500
},
{
"epoch": 8.92049127343245,
"grad_norm": 1.0888649225234985,
"learning_rate": 0.001,
"loss": 2.8251,
"step": 27600
},
{
"epoch": 8.952811893988365,
"grad_norm": 1.191940426826477,
"learning_rate": 0.001,
"loss": 2.8301,
"step": 27700
},
{
"epoch": 8.98513251454428,
"grad_norm": 0.9693983197212219,
"learning_rate": 0.001,
"loss": 2.8207,
"step": 27800
},
{
"epoch": 9.017453135100194,
"grad_norm": 1.3811169862747192,
"learning_rate": 0.001,
"loss": 2.7349,
"step": 27900
},
{
"epoch": 9.049773755656108,
"grad_norm": 1.1540076732635498,
"learning_rate": 0.001,
"loss": 2.6591,
"step": 28000
},
{
"epoch": 9.082094376212023,
"grad_norm": 1.6411869525909424,
"learning_rate": 0.001,
"loss": 2.6966,
"step": 28100
},
{
"epoch": 9.114414996767938,
"grad_norm": 1.3888397216796875,
"learning_rate": 0.001,
"loss": 2.6626,
"step": 28200
},
{
"epoch": 9.146735617323852,
"grad_norm": 1.1449769735336304,
"learning_rate": 0.001,
"loss": 2.715,
"step": 28300
},
{
"epoch": 9.179056237879767,
"grad_norm": 1.1566001176834106,
"learning_rate": 0.001,
"loss": 2.7052,
"step": 28400
},
{
"epoch": 9.211376858435681,
"grad_norm": 1.7325108051300049,
"learning_rate": 0.001,
"loss": 2.7084,
"step": 28500
},
{
"epoch": 9.243697478991596,
"grad_norm": 1.7383753061294556,
"learning_rate": 0.001,
"loss": 2.7083,
"step": 28600
},
{
"epoch": 9.276018099547512,
"grad_norm": 1.0822782516479492,
"learning_rate": 0.001,
"loss": 2.7432,
"step": 28700
},
{
"epoch": 9.308338720103427,
"grad_norm": 1.3996139764785767,
"learning_rate": 0.001,
"loss": 2.7223,
"step": 28800
},
{
"epoch": 9.340659340659341,
"grad_norm": 0.9429243803024292,
"learning_rate": 0.001,
"loss": 2.7267,
"step": 28900
},
{
"epoch": 9.372979961215256,
"grad_norm": 1.4174455404281616,
"learning_rate": 0.001,
"loss": 2.7325,
"step": 29000
},
{
"epoch": 9.40530058177117,
"grad_norm": 1.6818008422851562,
"learning_rate": 0.001,
"loss": 2.7286,
"step": 29100
},
{
"epoch": 9.437621202327085,
"grad_norm": 1.2587084770202637,
"learning_rate": 0.001,
"loss": 2.7395,
"step": 29200
},
{
"epoch": 9.469941822883,
"grad_norm": 1.0367249250411987,
"learning_rate": 0.001,
"loss": 2.737,
"step": 29300
},
{
"epoch": 9.502262443438914,
"grad_norm": 1.4259616136550903,
"learning_rate": 0.001,
"loss": 2.7208,
"step": 29400
},
{
"epoch": 9.534583063994829,
"grad_norm": 0.9334765076637268,
"learning_rate": 0.001,
"loss": 2.7471,
"step": 29500
},
{
"epoch": 9.566903684550743,
"grad_norm": 1.2142447233200073,
"learning_rate": 0.001,
"loss": 2.7435,
"step": 29600
},
{
"epoch": 9.599224305106658,
"grad_norm": 1.038910150527954,
"learning_rate": 0.001,
"loss": 2.7578,
"step": 29700
},
{
"epoch": 9.631544925662572,
"grad_norm": 1.0953419208526611,
"learning_rate": 0.001,
"loss": 2.7527,
"step": 29800
},
{
"epoch": 9.663865546218487,
"grad_norm": 1.251102328300476,
"learning_rate": 0.001,
"loss": 2.7422,
"step": 29900
},
{
"epoch": 9.696186166774401,
"grad_norm": 0.9764232635498047,
"learning_rate": 0.001,
"loss": 2.7373,
"step": 30000
},
{
"epoch": 9.728506787330316,
"grad_norm": 1.2931694984436035,
"learning_rate": 0.001,
"loss": 2.7496,
"step": 30100
},
{
"epoch": 9.760827407886232,
"grad_norm": 1.287348985671997,
"learning_rate": 0.001,
"loss": 2.7474,
"step": 30200
},
{
"epoch": 9.793148028442147,
"grad_norm": 1.186011552810669,
"learning_rate": 0.001,
"loss": 2.7527,
"step": 30300
},
{
"epoch": 9.825468648998061,
"grad_norm": 1.0229896306991577,
"learning_rate": 0.001,
"loss": 2.7593,
"step": 30400
},
{
"epoch": 9.857789269553976,
"grad_norm": 1.1699128150939941,
"learning_rate": 0.001,
"loss": 2.7521,
"step": 30500
},
{
"epoch": 9.89010989010989,
"grad_norm": 1.150494933128357,
"learning_rate": 0.001,
"loss": 2.7564,
"step": 30600
},
{
"epoch": 9.922430510665805,
"grad_norm": 1.6959357261657715,
"learning_rate": 0.001,
"loss": 2.7705,
"step": 30700
},
{
"epoch": 9.95475113122172,
"grad_norm": 1.3214408159255981,
"learning_rate": 0.001,
"loss": 2.7624,
"step": 30800
},
{
"epoch": 9.987071751777634,
"grad_norm": 1.0716474056243896,
"learning_rate": 0.001,
"loss": 2.7677,
"step": 30900
},
{
"epoch": 10.019392372333549,
"grad_norm": 1.2890034914016724,
"learning_rate": 0.001,
"loss": 2.68,
"step": 31000
},
{
"epoch": 10.051712992889463,
"grad_norm": 1.0409849882125854,
"learning_rate": 0.001,
"loss": 2.6201,
"step": 31100
},
{
"epoch": 10.084033613445378,
"grad_norm": 1.0122183561325073,
"learning_rate": 0.001,
"loss": 2.642,
"step": 31200
},
{
"epoch": 10.116354234001292,
"grad_norm": 0.7843957543373108,
"learning_rate": 0.001,
"loss": 2.6478,
"step": 31300
},
{
"epoch": 10.148674854557207,
"grad_norm": 0.8205838799476624,
"learning_rate": 0.001,
"loss": 2.6287,
"step": 31400
},
{
"epoch": 10.180995475113122,
"grad_norm": 0.863365888595581,
"learning_rate": 0.001,
"loss": 2.6559,
"step": 31500
},
{
"epoch": 10.213316095669036,
"grad_norm": 0.9866477847099304,
"learning_rate": 0.001,
"loss": 2.6472,
"step": 31600
},
{
"epoch": 10.24563671622495,
"grad_norm": 0.9632337093353271,
"learning_rate": 0.001,
"loss": 2.6516,
"step": 31700
},
{
"epoch": 10.277957336780867,
"grad_norm": 1.1927064657211304,
"learning_rate": 0.001,
"loss": 2.6616,
"step": 31800
},
{
"epoch": 10.310277957336782,
"grad_norm": 0.9914249777793884,
"learning_rate": 0.001,
"loss": 2.6617,
"step": 31900
},
{
"epoch": 10.342598577892696,
"grad_norm": 1.2201826572418213,
"learning_rate": 0.001,
"loss": 2.6579,
"step": 32000
},
{
"epoch": 10.37491919844861,
"grad_norm": 0.955346941947937,
"learning_rate": 0.001,
"loss": 2.6688,
"step": 32100
},
{
"epoch": 10.407239819004525,
"grad_norm": 1.2246835231781006,
"learning_rate": 0.001,
"loss": 2.6984,
"step": 32200
},
{
"epoch": 10.43956043956044,
"grad_norm": 1.0954246520996094,
"learning_rate": 0.001,
"loss": 2.6782,
"step": 32300
},
{
"epoch": 10.471881060116354,
"grad_norm": 0.9353541731834412,
"learning_rate": 0.001,
"loss": 2.6917,
"step": 32400
},
{
"epoch": 10.504201680672269,
"grad_norm": 1.1944663524627686,
"learning_rate": 0.001,
"loss": 2.6831,
"step": 32500
},
{
"epoch": 10.536522301228183,
"grad_norm": 1.0694438219070435,
"learning_rate": 0.001,
"loss": 2.6841,
"step": 32600
},
{
"epoch": 10.568842921784098,
"grad_norm": 1.4416074752807617,
"learning_rate": 0.001,
"loss": 2.6845,
"step": 32700
},
{
"epoch": 10.601163542340013,
"grad_norm": 1.1764634847640991,
"learning_rate": 0.001,
"loss": 2.6719,
"step": 32800
},
{
"epoch": 10.633484162895927,
"grad_norm": 0.9815175533294678,
"learning_rate": 0.001,
"loss": 2.6963,
"step": 32900
},
{
"epoch": 10.665804783451842,
"grad_norm": 0.7032698392868042,
"learning_rate": 0.001,
"loss": 2.6736,
"step": 33000
},
{
"epoch": 10.698125404007756,
"grad_norm": 1.076450228691101,
"learning_rate": 0.001,
"loss": 2.6897,
"step": 33100
},
{
"epoch": 10.73044602456367,
"grad_norm": 1.0582174062728882,
"learning_rate": 0.001,
"loss": 2.7031,
"step": 33200
},
{
"epoch": 10.762766645119587,
"grad_norm": 1.106713891029358,
"learning_rate": 0.001,
"loss": 2.7188,
"step": 33300
},
{
"epoch": 10.795087265675502,
"grad_norm": 0.9909460544586182,
"learning_rate": 0.001,
"loss": 2.6978,
"step": 33400
},
{
"epoch": 10.827407886231416,
"grad_norm": 1.056771993637085,
"learning_rate": 0.001,
"loss": 2.7152,
"step": 33500
},
{
"epoch": 10.85972850678733,
"grad_norm": 1.1948317289352417,
"learning_rate": 0.001,
"loss": 2.7073,
"step": 33600
},
{
"epoch": 10.892049127343245,
"grad_norm": 0.899512767791748,
"learning_rate": 0.001,
"loss": 2.7133,
"step": 33700
},
{
"epoch": 10.92436974789916,
"grad_norm": 0.9636661410331726,
"learning_rate": 0.001,
"loss": 2.6947,
"step": 33800
},
{
"epoch": 10.956690368455074,
"grad_norm": 0.9970238208770752,
"learning_rate": 0.001,
"loss": 2.7089,
"step": 33900
},
{
"epoch": 10.989010989010989,
"grad_norm": 1.0537792444229126,
"learning_rate": 0.001,
"loss": 2.7068,
"step": 34000
},
{
"epoch": 11.021331609566904,
"grad_norm": 1.3103049993515015,
"learning_rate": 0.001,
"loss": 2.6557,
"step": 34100
},
{
"epoch": 11.053652230122818,
"grad_norm": 0.9231497049331665,
"learning_rate": 0.001,
"loss": 2.5867,
"step": 34200
},
{
"epoch": 11.085972850678733,
"grad_norm": 1.1425038576126099,
"learning_rate": 0.001,
"loss": 2.5834,
"step": 34300
},
{
"epoch": 11.118293471234647,
"grad_norm": 1.3606386184692383,
"learning_rate": 0.001,
"loss": 2.6062,
"step": 34400
},
{
"epoch": 11.150614091790562,
"grad_norm": 1.0751980543136597,
"learning_rate": 0.001,
"loss": 2.5946,
"step": 34500
},
{
"epoch": 11.182934712346476,
"grad_norm": 1.1164932250976562,
"learning_rate": 0.001,
"loss": 2.6077,
"step": 34600
},
{
"epoch": 11.215255332902391,
"grad_norm": 1.1553841829299927,
"learning_rate": 0.001,
"loss": 2.6126,
"step": 34700
},
{
"epoch": 11.247575953458306,
"grad_norm": 1.4459047317504883,
"learning_rate": 0.001,
"loss": 2.5955,
"step": 34800
},
{
"epoch": 11.279896574014222,
"grad_norm": 1.1452357769012451,
"learning_rate": 0.001,
"loss": 2.609,
"step": 34900
},
{
"epoch": 11.312217194570136,
"grad_norm": 1.210909366607666,
"learning_rate": 0.001,
"loss": 2.606,
"step": 35000
},
{
"epoch": 11.344537815126051,
"grad_norm": 1.2074875831604004,
"learning_rate": 0.001,
"loss": 2.6085,
"step": 35100
},
{
"epoch": 11.376858435681966,
"grad_norm": 1.050931453704834,
"learning_rate": 0.001,
"loss": 2.6086,
"step": 35200
},
{
"epoch": 11.40917905623788,
"grad_norm": 1.1592390537261963,
"learning_rate": 0.001,
"loss": 2.62,
"step": 35300
},
{
"epoch": 11.441499676793795,
"grad_norm": 0.913329541683197,
"learning_rate": 0.001,
"loss": 2.6126,
"step": 35400
},
{
"epoch": 11.47382029734971,
"grad_norm": 0.9915534853935242,
"learning_rate": 0.001,
"loss": 2.6374,
"step": 35500
},
{
"epoch": 11.506140917905624,
"grad_norm": 0.9962285757064819,
"learning_rate": 0.001,
"loss": 2.6664,
"step": 35600
},
{
"epoch": 11.538461538461538,
"grad_norm": 1.0693681240081787,
"learning_rate": 0.001,
"loss": 2.6353,
"step": 35700
},
{
"epoch": 11.570782159017453,
"grad_norm": 1.021084189414978,
"learning_rate": 0.001,
"loss": 2.6456,
"step": 35800
},
{
"epoch": 11.603102779573367,
"grad_norm": 1.0230939388275146,
"learning_rate": 0.001,
"loss": 2.6261,
"step": 35900
},
{
"epoch": 11.635423400129282,
"grad_norm": 0.9915379285812378,
"learning_rate": 0.001,
"loss": 2.6338,
"step": 36000
},
{
"epoch": 11.667744020685197,
"grad_norm": 1.1837557554244995,
"learning_rate": 0.001,
"loss": 2.6526,
"step": 36100
},
{
"epoch": 11.700064641241111,
"grad_norm": 1.0220301151275635,
"learning_rate": 0.001,
"loss": 2.6398,
"step": 36200
},
{
"epoch": 11.732385261797026,
"grad_norm": 1.0417808294296265,
"learning_rate": 0.001,
"loss": 2.6442,
"step": 36300
},
{
"epoch": 11.764705882352942,
"grad_norm": 1.0188359022140503,
"learning_rate": 0.001,
"loss": 2.6421,
"step": 36400
},
{
"epoch": 11.797026502908857,
"grad_norm": 1.1374149322509766,
"learning_rate": 0.001,
"loss": 2.635,
"step": 36500
},
{
"epoch": 11.829347123464771,
"grad_norm": 0.9831808805465698,
"learning_rate": 0.001,
"loss": 2.6698,
"step": 36600
},
{
"epoch": 11.861667744020686,
"grad_norm": 1.0103954076766968,
"learning_rate": 0.001,
"loss": 2.6554,
"step": 36700
},
{
"epoch": 11.8939883645766,
"grad_norm": 1.2161500453948975,
"learning_rate": 0.001,
"loss": 2.6435,
"step": 36800
},
{
"epoch": 11.926308985132515,
"grad_norm": 1.1546547412872314,
"learning_rate": 0.001,
"loss": 2.6743,
"step": 36900
},
{
"epoch": 11.95862960568843,
"grad_norm": 0.9428865313529968,
"learning_rate": 0.001,
"loss": 2.6472,
"step": 37000
},
{
"epoch": 11.990950226244344,
"grad_norm": 1.1618021726608276,
"learning_rate": 0.001,
"loss": 2.6731,
"step": 37100
},
{
"epoch": 12.023270846800258,
"grad_norm": 1.301104187965393,
"learning_rate": 0.001,
"loss": 2.5516,
"step": 37200
},
{
"epoch": 12.055591467356173,
"grad_norm": 0.9387957453727722,
"learning_rate": 0.001,
"loss": 2.5173,
"step": 37300
},
{
"epoch": 12.087912087912088,
"grad_norm": 1.0211634635925293,
"learning_rate": 0.001,
"loss": 2.5258,
"step": 37400
},
{
"epoch": 12.120232708468002,
"grad_norm": 1.263898253440857,
"learning_rate": 0.001,
"loss": 2.522,
"step": 37500
},
{
"epoch": 12.152553329023917,
"grad_norm": 1.0215619802474976,
"learning_rate": 0.001,
"loss": 2.5351,
"step": 37600
},
{
"epoch": 12.184873949579831,
"grad_norm": 1.094390630722046,
"learning_rate": 0.001,
"loss": 2.5529,
"step": 37700
},
{
"epoch": 12.217194570135746,
"grad_norm": 1.04618501663208,
"learning_rate": 0.001,
"loss": 2.5752,
"step": 37800
},
{
"epoch": 12.24951519069166,
"grad_norm": 0.859111487865448,
"learning_rate": 0.001,
"loss": 2.5695,
"step": 37900
},
{
"epoch": 12.281835811247577,
"grad_norm": 1.2946406602859497,
"learning_rate": 0.001,
"loss": 2.5731,
"step": 38000
},
{
"epoch": 12.314156431803491,
"grad_norm": 1.108901858329773,
"learning_rate": 0.001,
"loss": 2.5704,
"step": 38100
},
{
"epoch": 12.346477052359406,
"grad_norm": 1.4274181127548218,
"learning_rate": 0.001,
"loss": 2.56,
"step": 38200
},
{
"epoch": 12.37879767291532,
"grad_norm": 1.0157301425933838,
"learning_rate": 0.001,
"loss": 2.5483,
"step": 38300
},
{
"epoch": 12.411118293471235,
"grad_norm": 0.9874992966651917,
"learning_rate": 0.001,
"loss": 2.5837,
"step": 38400
},
{
"epoch": 12.44343891402715,
"grad_norm": 1.0054415464401245,
"learning_rate": 0.001,
"loss": 2.5891,
"step": 38500
},
{
"epoch": 12.475759534583064,
"grad_norm": 0.8881959915161133,
"learning_rate": 0.001,
"loss": 2.5765,
"step": 38600
},
{
"epoch": 12.508080155138979,
"grad_norm": 0.997611403465271,
"learning_rate": 0.001,
"loss": 2.5813,
"step": 38700
},
{
"epoch": 12.540400775694893,
"grad_norm": 1.2662944793701172,
"learning_rate": 0.001,
"loss": 2.5845,
"step": 38800
},
{
"epoch": 12.572721396250808,
"grad_norm": 1.206846833229065,
"learning_rate": 0.001,
"loss": 2.5873,
"step": 38900
},
{
"epoch": 12.605042016806722,
"grad_norm": 1.4133809804916382,
"learning_rate": 0.001,
"loss": 2.6056,
"step": 39000
},
{
"epoch": 12.637362637362637,
"grad_norm": 1.0198854207992554,
"learning_rate": 0.001,
"loss": 2.6078,
"step": 39100
},
{
"epoch": 12.669683257918551,
"grad_norm": 1.1534576416015625,
"learning_rate": 0.001,
"loss": 2.6034,
"step": 39200
},
{
"epoch": 12.702003878474466,
"grad_norm": 0.9695433974266052,
"learning_rate": 0.001,
"loss": 2.5871,
"step": 39300
},
{
"epoch": 12.73432449903038,
"grad_norm": 1.0158796310424805,
"learning_rate": 0.001,
"loss": 2.6296,
"step": 39400
},
{
"epoch": 12.766645119586297,
"grad_norm": 1.290499210357666,
"learning_rate": 0.001,
"loss": 2.6158,
"step": 39500
},
{
"epoch": 12.798965740142211,
"grad_norm": 1.1555784940719604,
"learning_rate": 0.001,
"loss": 2.6042,
"step": 39600
},
{
"epoch": 12.831286360698126,
"grad_norm": 0.9990980625152588,
"learning_rate": 0.001,
"loss": 2.6166,
"step": 39700
},
{
"epoch": 12.86360698125404,
"grad_norm": 1.0118318796157837,
"learning_rate": 0.001,
"loss": 2.6057,
"step": 39800
},
{
"epoch": 12.895927601809955,
"grad_norm": 1.1140888929367065,
"learning_rate": 0.001,
"loss": 2.6055,
"step": 39900
},
{
"epoch": 12.92824822236587,
"grad_norm": 0.9500235319137573,
"learning_rate": 0.001,
"loss": 2.6168,
"step": 40000
},
{
"epoch": 12.960568842921784,
"grad_norm": 1.2661901712417603,
"learning_rate": 0.001,
"loss": 2.6192,
"step": 40100
},
{
"epoch": 12.992889463477699,
"grad_norm": 1.0866193771362305,
"learning_rate": 0.001,
"loss": 2.6135,
"step": 40200
},
{
"epoch": 13.025210084033613,
"grad_norm": 1.2633984088897705,
"learning_rate": 0.001,
"loss": 2.4986,
"step": 40300
},
{
"epoch": 13.057530704589528,
"grad_norm": 1.2532984018325806,
"learning_rate": 0.001,
"loss": 2.4877,
"step": 40400
},
{
"epoch": 13.089851325145442,
"grad_norm": 0.9414843916893005,
"learning_rate": 0.001,
"loss": 2.5059,
"step": 40500
},
{
"epoch": 13.122171945701357,
"grad_norm": 0.8735656142234802,
"learning_rate": 0.001,
"loss": 2.4852,
"step": 40600
},
{
"epoch": 13.154492566257272,
"grad_norm": 0.8817693591117859,
"learning_rate": 0.001,
"loss": 2.4978,
"step": 40700
},
{
"epoch": 13.186813186813186,
"grad_norm": 0.9837480783462524,
"learning_rate": 0.001,
"loss": 2.5209,
"step": 40800
},
{
"epoch": 13.2191338073691,
"grad_norm": 0.951636016368866,
"learning_rate": 0.001,
"loss": 2.514,
"step": 40900
},
{
"epoch": 13.251454427925015,
"grad_norm": 0.9579486846923828,
"learning_rate": 0.001,
"loss": 2.4985,
"step": 41000
},
{
"epoch": 13.283775048480932,
"grad_norm": 0.9534587860107422,
"learning_rate": 0.001,
"loss": 2.4987,
"step": 41100
},
{
"epoch": 13.316095669036846,
"grad_norm": 1.0087788105010986,
"learning_rate": 0.001,
"loss": 2.5169,
"step": 41200
},
{
"epoch": 13.34841628959276,
"grad_norm": 1.015776515007019,
"learning_rate": 0.001,
"loss": 2.5123,
"step": 41300
},
{
"epoch": 13.380736910148675,
"grad_norm": 1.0848371982574463,
"learning_rate": 0.001,
"loss": 2.5282,
"step": 41400
},
{
"epoch": 13.41305753070459,
"grad_norm": 1.1065797805786133,
"learning_rate": 0.001,
"loss": 2.5395,
"step": 41500
},
{
"epoch": 13.445378151260504,
"grad_norm": 1.040236234664917,
"learning_rate": 0.001,
"loss": 2.5397,
"step": 41600
},
{
"epoch": 13.477698771816419,
"grad_norm": 1.1812667846679688,
"learning_rate": 0.001,
"loss": 2.5326,
"step": 41700
},
{
"epoch": 13.510019392372334,
"grad_norm": 1.0985037088394165,
"learning_rate": 0.001,
"loss": 2.5523,
"step": 41800
},
{
"epoch": 13.542340012928248,
"grad_norm": 1.080064296722412,
"learning_rate": 0.001,
"loss": 2.5387,
"step": 41900
},
{
"epoch": 13.574660633484163,
"grad_norm": 1.0306446552276611,
"learning_rate": 0.001,
"loss": 2.5369,
"step": 42000
},
{
"epoch": 13.606981254040077,
"grad_norm": 0.9848862886428833,
"learning_rate": 0.001,
"loss": 2.5596,
"step": 42100
},
{
"epoch": 13.639301874595992,
"grad_norm": 0.9932380318641663,
"learning_rate": 0.001,
"loss": 2.5515,
"step": 42200
},
{
"epoch": 13.671622495151906,
"grad_norm": 1.0285083055496216,
"learning_rate": 0.001,
"loss": 2.5555,
"step": 42300
},
{
"epoch": 13.70394311570782,
"grad_norm": 1.123618721961975,
"learning_rate": 0.001,
"loss": 2.5401,
"step": 42400
},
{
"epoch": 13.736263736263737,
"grad_norm": 0.9957885146141052,
"learning_rate": 0.001,
"loss": 2.5543,
"step": 42500
},
{
"epoch": 13.768584356819652,
"grad_norm": 1.0012941360473633,
"learning_rate": 0.001,
"loss": 2.5686,
"step": 42600
},
{
"epoch": 13.800904977375566,
"grad_norm": 1.053511381149292,
"learning_rate": 0.001,
"loss": 2.5572,
"step": 42700
},
{
"epoch": 13.83322559793148,
"grad_norm": 1.0062965154647827,
"learning_rate": 0.001,
"loss": 2.5783,
"step": 42800
},
{
"epoch": 13.865546218487395,
"grad_norm": 0.8813051581382751,
"learning_rate": 0.001,
"loss": 2.5697,
"step": 42900
},
{
"epoch": 13.89786683904331,
"grad_norm": 1.2018564939498901,
"learning_rate": 0.001,
"loss": 2.5927,
"step": 43000
},
{
"epoch": 13.930187459599225,
"grad_norm": 1.1996262073516846,
"learning_rate": 0.001,
"loss": 2.5796,
"step": 43100
},
{
"epoch": 13.96250808015514,
"grad_norm": 1.088511347770691,
"learning_rate": 0.001,
"loss": 2.5746,
"step": 43200
},
{
"epoch": 13.994828700711054,
"grad_norm": 1.128455400466919,
"learning_rate": 0.001,
"loss": 2.5757,
"step": 43300
},
{
"epoch": 14.027149321266968,
"grad_norm": 1.0083279609680176,
"learning_rate": 0.001,
"loss": 2.4543,
"step": 43400
},
{
"epoch": 14.059469941822883,
"grad_norm": 0.801069974899292,
"learning_rate": 0.001,
"loss": 2.439,
"step": 43500
},
{
"epoch": 14.091790562378797,
"grad_norm": 1.0841903686523438,
"learning_rate": 0.001,
"loss": 2.45,
"step": 43600
},
{
"epoch": 14.124111182934712,
"grad_norm": 1.1224544048309326,
"learning_rate": 0.001,
"loss": 2.4582,
"step": 43700
},
{
"epoch": 14.156431803490626,
"grad_norm": 1.295549988746643,
"learning_rate": 0.001,
"loss": 2.4543,
"step": 43800
},
{
"epoch": 14.188752424046541,
"grad_norm": 1.1939016580581665,
"learning_rate": 0.001,
"loss": 2.4752,
"step": 43900
},
{
"epoch": 14.221073044602456,
"grad_norm": 1.0265029668807983,
"learning_rate": 0.001,
"loss": 2.4714,
"step": 44000
},
{
"epoch": 14.25339366515837,
"grad_norm": 0.8389987349510193,
"learning_rate": 0.001,
"loss": 2.4543,
"step": 44100
},
{
"epoch": 14.285714285714286,
"grad_norm": 0.8589958548545837,
"learning_rate": 0.001,
"loss": 2.4577,
"step": 44200
},
{
"epoch": 14.318034906270201,
"grad_norm": 1.4587507247924805,
"learning_rate": 0.001,
"loss": 2.4898,
"step": 44300
},
{
"epoch": 14.350355526826116,
"grad_norm": 1.1461809873580933,
"learning_rate": 0.001,
"loss": 2.482,
"step": 44400
},
{
"epoch": 14.38267614738203,
"grad_norm": 0.9660089015960693,
"learning_rate": 0.001,
"loss": 2.4885,
"step": 44500
},
{
"epoch": 14.414996767937945,
"grad_norm": 1.0140084028244019,
"learning_rate": 0.001,
"loss": 2.4967,
"step": 44600
},
{
"epoch": 14.44731738849386,
"grad_norm": 1.0041033029556274,
"learning_rate": 0.001,
"loss": 2.482,
"step": 44700
},
{
"epoch": 14.479638009049774,
"grad_norm": 1.1929504871368408,
"learning_rate": 0.001,
"loss": 2.4936,
"step": 44800
},
{
"epoch": 14.511958629605688,
"grad_norm": 0.9118449091911316,
"learning_rate": 0.001,
"loss": 2.5111,
"step": 44900
},
{
"epoch": 14.544279250161603,
"grad_norm": 1.134766936302185,
"learning_rate": 0.001,
"loss": 2.4912,
"step": 45000
},
{
"epoch": 14.576599870717518,
"grad_norm": 1.055952548980713,
"learning_rate": 0.001,
"loss": 2.4914,
"step": 45100
},
{
"epoch": 14.608920491273432,
"grad_norm": 0.9826187491416931,
"learning_rate": 0.001,
"loss": 2.5003,
"step": 45200
},
{
"epoch": 14.641241111829347,
"grad_norm": 1.1690611839294434,
"learning_rate": 0.001,
"loss": 2.5108,
"step": 45300
},
{
"epoch": 14.673561732385261,
"grad_norm": 1.432464361190796,
"learning_rate": 0.001,
"loss": 2.4994,
"step": 45400
},
{
"epoch": 14.705882352941176,
"grad_norm": 0.9632334113121033,
"learning_rate": 0.001,
"loss": 2.5067,
"step": 45500
},
{
"epoch": 14.738202973497092,
"grad_norm": 1.0171823501586914,
"learning_rate": 0.001,
"loss": 2.5184,
"step": 45600
},
{
"epoch": 14.770523594053007,
"grad_norm": 1.2903285026550293,
"learning_rate": 0.001,
"loss": 2.5358,
"step": 45700
},
{
"epoch": 14.802844214608921,
"grad_norm": 1.0135929584503174,
"learning_rate": 0.001,
"loss": 2.5115,
"step": 45800
},
{
"epoch": 14.835164835164836,
"grad_norm": 1.3158890008926392,
"learning_rate": 0.001,
"loss": 2.542,
"step": 45900
},
{
"epoch": 14.86748545572075,
"grad_norm": 1.061798095703125,
"learning_rate": 0.001,
"loss": 2.5355,
"step": 46000
},
{
"epoch": 14.899806076276665,
"grad_norm": 0.8953672647476196,
"learning_rate": 0.001,
"loss": 2.5211,
"step": 46100
},
{
"epoch": 14.93212669683258,
"grad_norm": 1.2173285484313965,
"learning_rate": 0.001,
"loss": 2.5493,
"step": 46200
},
{
"epoch": 14.964447317388494,
"grad_norm": 0.8660424947738647,
"learning_rate": 0.001,
"loss": 2.5273,
"step": 46300
},
{
"epoch": 14.996767937944409,
"grad_norm": 1.432119369506836,
"learning_rate": 0.001,
"loss": 2.5365,
"step": 46400
},
{
"epoch": 15.029088558500323,
"grad_norm": 1.1390154361724854,
"learning_rate": 0.001,
"loss": 2.4147,
"step": 46500
},
{
"epoch": 15.061409179056238,
"grad_norm": 0.9568191766738892,
"learning_rate": 0.001,
"loss": 2.3977,
"step": 46600
},
{
"epoch": 15.093729799612152,
"grad_norm": 1.104967474937439,
"learning_rate": 0.001,
"loss": 2.4257,
"step": 46700
},
{
"epoch": 15.126050420168067,
"grad_norm": 1.2843493223190308,
"learning_rate": 0.001,
"loss": 2.3913,
"step": 46800
},
{
"epoch": 15.158371040723981,
"grad_norm": 1.081358790397644,
"learning_rate": 0.001,
"loss": 2.4021,
"step": 46900
},
{
"epoch": 15.190691661279896,
"grad_norm": 1.2713805437088013,
"learning_rate": 0.001,
"loss": 2.4117,
"step": 47000
},
{
"epoch": 15.22301228183581,
"grad_norm": 0.9084070324897766,
"learning_rate": 0.001,
"loss": 2.4243,
"step": 47100
},
{
"epoch": 15.255332902391725,
"grad_norm": 1.2753498554229736,
"learning_rate": 0.001,
"loss": 2.4543,
"step": 47200
},
{
"epoch": 15.287653522947641,
"grad_norm": 1.1518727540969849,
"learning_rate": 0.001,
"loss": 2.4297,
"step": 47300
},
{
"epoch": 15.319974143503556,
"grad_norm": 1.4574276208877563,
"learning_rate": 0.001,
"loss": 2.4394,
"step": 47400
},
{
"epoch": 15.35229476405947,
"grad_norm": 1.1068369150161743,
"learning_rate": 0.001,
"loss": 2.4553,
"step": 47500
},
{
"epoch": 15.384615384615385,
"grad_norm": 1.5481140613555908,
"learning_rate": 0.001,
"loss": 2.435,
"step": 47600
},
{
"epoch": 15.4169360051713,
"grad_norm": 1.0796083211898804,
"learning_rate": 0.001,
"loss": 2.4464,
"step": 47700
},
{
"epoch": 15.449256625727214,
"grad_norm": 1.4825586080551147,
"learning_rate": 0.001,
"loss": 2.4576,
"step": 47800
},
{
"epoch": 15.481577246283129,
"grad_norm": 1.120011568069458,
"learning_rate": 0.001,
"loss": 2.4536,
"step": 47900
},
{
"epoch": 15.513897866839043,
"grad_norm": 0.9324677586555481,
"learning_rate": 0.001,
"loss": 2.4663,
"step": 48000
},
{
"epoch": 15.546218487394958,
"grad_norm": 1.296181321144104,
"learning_rate": 0.001,
"loss": 2.4766,
"step": 48100
},
{
"epoch": 15.578539107950872,
"grad_norm": 1.2365769147872925,
"learning_rate": 0.001,
"loss": 2.4607,
"step": 48200
},
{
"epoch": 15.610859728506787,
"grad_norm": 1.240800380706787,
"learning_rate": 0.001,
"loss": 2.4644,
"step": 48300
},
{
"epoch": 15.643180349062701,
"grad_norm": 1.4814155101776123,
"learning_rate": 0.001,
"loss": 2.4784,
"step": 48400
},
{
"epoch": 15.675500969618616,
"grad_norm": 0.9514549970626831,
"learning_rate": 0.001,
"loss": 2.4892,
"step": 48500
},
{
"epoch": 15.70782159017453,
"grad_norm": 1.3035391569137573,
"learning_rate": 0.001,
"loss": 2.4723,
"step": 48600
},
{
"epoch": 15.740142210730447,
"grad_norm": 1.1913089752197266,
"learning_rate": 0.001,
"loss": 2.4802,
"step": 48700
},
{
"epoch": 15.772462831286362,
"grad_norm": 1.4761337041854858,
"learning_rate": 0.001,
"loss": 2.4862,
"step": 48800
},
{
"epoch": 15.804783451842276,
"grad_norm": 0.8643561601638794,
"learning_rate": 0.001,
"loss": 2.4809,
"step": 48900
},
{
"epoch": 15.83710407239819,
"grad_norm": 1.069809913635254,
"learning_rate": 0.001,
"loss": 2.4954,
"step": 49000
},
{
"epoch": 15.869424692954105,
"grad_norm": 1.0733697414398193,
"learning_rate": 0.001,
"loss": 2.4837,
"step": 49100
},
{
"epoch": 15.90174531351002,
"grad_norm": 1.1564855575561523,
"learning_rate": 0.001,
"loss": 2.4911,
"step": 49200
},
{
"epoch": 15.934065934065934,
"grad_norm": 1.0230194330215454,
"learning_rate": 0.001,
"loss": 2.4951,
"step": 49300
},
{
"epoch": 15.966386554621849,
"grad_norm": 0.9934403896331787,
"learning_rate": 0.001,
"loss": 2.498,
"step": 49400
},
{
"epoch": 15.998707175177763,
"grad_norm": 0.9659956097602844,
"learning_rate": 0.001,
"loss": 2.4897,
"step": 49500
},
{
"epoch": 16.031027795733678,
"grad_norm": 1.4045450687408447,
"learning_rate": 0.001,
"loss": 2.3623,
"step": 49600
},
{
"epoch": 16.063348416289593,
"grad_norm": 1.362570881843567,
"learning_rate": 0.001,
"loss": 2.3642,
"step": 49700
},
{
"epoch": 16.095669036845507,
"grad_norm": 1.4277714490890503,
"learning_rate": 0.001,
"loss": 2.3641,
"step": 49800
},
{
"epoch": 16.12798965740142,
"grad_norm": 1.408735990524292,
"learning_rate": 0.001,
"loss": 2.3843,
"step": 49900
},
{
"epoch": 16.160310277957336,
"grad_norm": 1.2235407829284668,
"learning_rate": 0.001,
"loss": 2.3742,
"step": 50000
},
{
"epoch": 16.19263089851325,
"grad_norm": 1.425565242767334,
"learning_rate": 0.001,
"loss": 2.3827,
"step": 50100
},
{
"epoch": 16.224951519069165,
"grad_norm": 1.2481727600097656,
"learning_rate": 0.001,
"loss": 2.3899,
"step": 50200
},
{
"epoch": 16.25727213962508,
"grad_norm": 1.2243844270706177,
"learning_rate": 0.001,
"loss": 2.3893,
"step": 50300
},
{
"epoch": 16.289592760180994,
"grad_norm": 1.8857603073120117,
"learning_rate": 0.001,
"loss": 2.3921,
"step": 50400
},
{
"epoch": 16.32191338073691,
"grad_norm": 1.1694952249526978,
"learning_rate": 0.001,
"loss": 2.3869,
"step": 50500
},
{
"epoch": 16.354234001292824,
"grad_norm": 1.39084792137146,
"learning_rate": 0.001,
"loss": 2.4093,
"step": 50600
},
{
"epoch": 16.386554621848738,
"grad_norm": 1.175167202949524,
"learning_rate": 0.001,
"loss": 2.4108,
"step": 50700
},
{
"epoch": 16.418875242404653,
"grad_norm": 1.2448058128356934,
"learning_rate": 0.001,
"loss": 2.4234,
"step": 50800
},
{
"epoch": 16.451195862960567,
"grad_norm": 1.566321849822998,
"learning_rate": 0.001,
"loss": 2.4209,
"step": 50900
},
{
"epoch": 16.483516483516482,
"grad_norm": 1.3612757921218872,
"learning_rate": 0.001,
"loss": 2.4261,
"step": 51000
},
{
"epoch": 16.5158371040724,
"grad_norm": 1.2318320274353027,
"learning_rate": 0.001,
"loss": 2.4114,
"step": 51100
},
{
"epoch": 16.548157724628314,
"grad_norm": 1.5574370622634888,
"learning_rate": 0.001,
"loss": 2.4152,
"step": 51200
},
{
"epoch": 16.58047834518423,
"grad_norm": 1.347001075744629,
"learning_rate": 0.001,
"loss": 2.4476,
"step": 51300
},
{
"epoch": 16.612798965740144,
"grad_norm": 1.343788504600525,
"learning_rate": 0.001,
"loss": 2.4289,
"step": 51400
},
{
"epoch": 16.645119586296058,
"grad_norm": 1.179029107093811,
"learning_rate": 0.001,
"loss": 2.427,
"step": 51500
},
{
"epoch": 16.677440206851973,
"grad_norm": 1.2791211605072021,
"learning_rate": 0.001,
"loss": 2.433,
"step": 51600
},
{
"epoch": 16.709760827407887,
"grad_norm": 1.183396816253662,
"learning_rate": 0.001,
"loss": 2.4398,
"step": 51700
},
{
"epoch": 16.742081447963802,
"grad_norm": 1.3307725191116333,
"learning_rate": 0.001,
"loss": 2.4396,
"step": 51800
},
{
"epoch": 16.774402068519716,
"grad_norm": 1.4831727743148804,
"learning_rate": 0.001,
"loss": 2.4427,
"step": 51900
},
{
"epoch": 16.80672268907563,
"grad_norm": 1.2639983892440796,
"learning_rate": 0.001,
"loss": 2.4597,
"step": 52000
},
{
"epoch": 16.839043309631545,
"grad_norm": 1.4711253643035889,
"learning_rate": 0.001,
"loss": 2.4426,
"step": 52100
},
{
"epoch": 16.87136393018746,
"grad_norm": 1.704418420791626,
"learning_rate": 0.001,
"loss": 2.4416,
"step": 52200
},
{
"epoch": 16.903684550743375,
"grad_norm": 1.6160430908203125,
"learning_rate": 0.001,
"loss": 2.4515,
"step": 52300
},
{
"epoch": 16.93600517129929,
"grad_norm": 1.2862269878387451,
"learning_rate": 0.001,
"loss": 2.4437,
"step": 52400
},
{
"epoch": 16.968325791855204,
"grad_norm": 1.3109511137008667,
"learning_rate": 0.001,
"loss": 2.4758,
"step": 52500
},
{
"epoch": 17.00064641241112,
"grad_norm": 1.0594240427017212,
"learning_rate": 0.001,
"loss": 2.4494,
"step": 52600
},
{
"epoch": 17.032967032967033,
"grad_norm": 1.245521903038025,
"learning_rate": 0.001,
"loss": 2.3124,
"step": 52700
},
{
"epoch": 17.065287653522947,
"grad_norm": 1.4481103420257568,
"learning_rate": 0.001,
"loss": 2.3213,
"step": 52800
},
{
"epoch": 17.097608274078862,
"grad_norm": 1.3900834321975708,
"learning_rate": 0.001,
"loss": 2.306,
"step": 52900
},
{
"epoch": 17.129928894634777,
"grad_norm": 1.735656499862671,
"learning_rate": 0.001,
"loss": 2.3283,
"step": 53000
},
{
"epoch": 17.16224951519069,
"grad_norm": 2.0901541709899902,
"learning_rate": 0.001,
"loss": 2.3559,
"step": 53100
},
{
"epoch": 17.194570135746606,
"grad_norm": 1.581668496131897,
"learning_rate": 0.001,
"loss": 2.3499,
"step": 53200
},
{
"epoch": 17.22689075630252,
"grad_norm": 1.2505109310150146,
"learning_rate": 0.001,
"loss": 2.3687,
"step": 53300
},
{
"epoch": 17.259211376858435,
"grad_norm": 1.596501111984253,
"learning_rate": 0.001,
"loss": 2.356,
"step": 53400
},
{
"epoch": 17.29153199741435,
"grad_norm": 1.9706599712371826,
"learning_rate": 0.001,
"loss": 2.356,
"step": 53500
},
{
"epoch": 17.323852617970264,
"grad_norm": 1.9692374467849731,
"learning_rate": 0.001,
"loss": 2.3571,
"step": 53600
},
{
"epoch": 17.35617323852618,
"grad_norm": 1.86066734790802,
"learning_rate": 0.001,
"loss": 2.3735,
"step": 53700
},
{
"epoch": 17.388493859082093,
"grad_norm": 1.219201922416687,
"learning_rate": 0.001,
"loss": 2.376,
"step": 53800
},
{
"epoch": 17.420814479638008,
"grad_norm": 1.489042043685913,
"learning_rate": 0.001,
"loss": 2.3818,
"step": 53900
},
{
"epoch": 17.453135100193922,
"grad_norm": 1.5569493770599365,
"learning_rate": 0.001,
"loss": 2.3736,
"step": 54000
},
{
"epoch": 17.485455720749837,
"grad_norm": 1.4509748220443726,
"learning_rate": 0.001,
"loss": 2.3853,
"step": 54100
},
{
"epoch": 17.517776341305755,
"grad_norm": 1.6958837509155273,
"learning_rate": 0.001,
"loss": 2.3815,
"step": 54200
},
{
"epoch": 17.55009696186167,
"grad_norm": 1.4118542671203613,
"learning_rate": 0.001,
"loss": 2.3897,
"step": 54300
},
{
"epoch": 17.582417582417584,
"grad_norm": 1.9448574781417847,
"learning_rate": 0.001,
"loss": 2.4188,
"step": 54400
},
{
"epoch": 17.6147382029735,
"grad_norm": 1.5198105573654175,
"learning_rate": 0.001,
"loss": 2.3924,
"step": 54500
},
{
"epoch": 17.647058823529413,
"grad_norm": 1.7049535512924194,
"learning_rate": 0.001,
"loss": 2.4061,
"step": 54600
},
{
"epoch": 17.679379444085328,
"grad_norm": 1.584429144859314,
"learning_rate": 0.001,
"loss": 2.3826,
"step": 54700
},
{
"epoch": 17.711700064641242,
"grad_norm": 1.6980667114257812,
"learning_rate": 0.001,
"loss": 2.4035,
"step": 54800
},
{
"epoch": 17.744020685197157,
"grad_norm": 1.6391141414642334,
"learning_rate": 0.001,
"loss": 2.4301,
"step": 54900
},
{
"epoch": 17.77634130575307,
"grad_norm": 1.6878070831298828,
"learning_rate": 0.001,
"loss": 2.4132,
"step": 55000
},
{
"epoch": 17.808661926308986,
"grad_norm": 2.0005171298980713,
"learning_rate": 0.001,
"loss": 2.4109,
"step": 55100
},
{
"epoch": 17.8409825468649,
"grad_norm": 2.002854108810425,
"learning_rate": 0.001,
"loss": 2.4095,
"step": 55200
},
{
"epoch": 17.873303167420815,
"grad_norm": 1.7028067111968994,
"learning_rate": 0.001,
"loss": 2.422,
"step": 55300
},
{
"epoch": 17.90562378797673,
"grad_norm": 1.8584864139556885,
"learning_rate": 0.001,
"loss": 2.4003,
"step": 55400
},
{
"epoch": 17.937944408532644,
"grad_norm": 1.6013153791427612,
"learning_rate": 0.001,
"loss": 2.4068,
"step": 55500
},
{
"epoch": 17.97026502908856,
"grad_norm": 1.375704288482666,
"learning_rate": 0.001,
"loss": 2.4275,
"step": 55600
},
{
"epoch": 18.002585649644473,
"grad_norm": 1.4080970287322998,
"learning_rate": 0.001,
"loss": 2.4447,
"step": 55700
},
{
"epoch": 18.034906270200388,
"grad_norm": 1.0998477935791016,
"learning_rate": 0.001,
"loss": 2.2773,
"step": 55800
},
{
"epoch": 18.067226890756302,
"grad_norm": 1.1358418464660645,
"learning_rate": 0.001,
"loss": 2.2797,
"step": 55900
},
{
"epoch": 18.099547511312217,
"grad_norm": 1.4835680723190308,
"learning_rate": 0.001,
"loss": 2.2864,
"step": 56000
},
{
"epoch": 18.13186813186813,
"grad_norm": 0.9853982329368591,
"learning_rate": 0.001,
"loss": 2.3065,
"step": 56100
},
{
"epoch": 18.164188752424046,
"grad_norm": 1.0776718854904175,
"learning_rate": 0.001,
"loss": 2.3334,
"step": 56200
},
{
"epoch": 18.19650937297996,
"grad_norm": 1.2667760848999023,
"learning_rate": 0.001,
"loss": 2.3165,
"step": 56300
},
{
"epoch": 18.228829993535875,
"grad_norm": 1.05290687084198,
"learning_rate": 0.001,
"loss": 2.3336,
"step": 56400
},
{
"epoch": 18.26115061409179,
"grad_norm": 0.8548942804336548,
"learning_rate": 0.001,
"loss": 2.3282,
"step": 56500
},
{
"epoch": 18.293471234647704,
"grad_norm": 1.513579249382019,
"learning_rate": 0.001,
"loss": 2.3231,
"step": 56600
},
{
"epoch": 18.32579185520362,
"grad_norm": 1.1067171096801758,
"learning_rate": 0.001,
"loss": 2.3423,
"step": 56700
},
{
"epoch": 18.358112475759533,
"grad_norm": 1.2587110996246338,
"learning_rate": 0.001,
"loss": 2.3068,
"step": 56800
},
{
"epoch": 18.390433096315448,
"grad_norm": 1.1359119415283203,
"learning_rate": 0.001,
"loss": 2.3377,
"step": 56900
},
{
"epoch": 18.422753716871362,
"grad_norm": 1.1252273321151733,
"learning_rate": 0.001,
"loss": 2.3459,
"step": 57000
},
{
"epoch": 18.455074337427277,
"grad_norm": 1.3988776206970215,
"learning_rate": 0.001,
"loss": 2.3555,
"step": 57100
},
{
"epoch": 18.48739495798319,
"grad_norm": 1.3088960647583008,
"learning_rate": 0.001,
"loss": 2.3194,
"step": 57200
},
{
"epoch": 18.51971557853911,
"grad_norm": 1.3026854991912842,
"learning_rate": 0.001,
"loss": 2.3658,
"step": 57300
},
{
"epoch": 18.552036199095024,
"grad_norm": 1.1831040382385254,
"learning_rate": 0.001,
"loss": 2.3763,
"step": 57400
},
{
"epoch": 18.58435681965094,
"grad_norm": 0.9830296635627747,
"learning_rate": 0.001,
"loss": 2.3726,
"step": 57500
},
{
"epoch": 18.616677440206853,
"grad_norm": 1.1031728982925415,
"learning_rate": 0.001,
"loss": 2.3551,
"step": 57600
},
{
"epoch": 18.648998060762768,
"grad_norm": 1.4140679836273193,
"learning_rate": 0.001,
"loss": 2.3739,
"step": 57700
},
{
"epoch": 18.681318681318682,
"grad_norm": 1.1503195762634277,
"learning_rate": 0.001,
"loss": 2.3626,
"step": 57800
},
{
"epoch": 18.713639301874597,
"grad_norm": 1.269458293914795,
"learning_rate": 0.001,
"loss": 2.3516,
"step": 57900
},
{
"epoch": 18.74595992243051,
"grad_norm": 0.9836449027061462,
"learning_rate": 0.001,
"loss": 2.382,
"step": 58000
},
{
"epoch": 18.778280542986426,
"grad_norm": 1.0975106954574585,
"learning_rate": 0.001,
"loss": 2.3987,
"step": 58100
},
{
"epoch": 18.81060116354234,
"grad_norm": 1.3973971605300903,
"learning_rate": 0.001,
"loss": 2.3876,
"step": 58200
},
{
"epoch": 18.842921784098255,
"grad_norm": 1.6158678531646729,
"learning_rate": 0.001,
"loss": 2.3745,
"step": 58300
},
{
"epoch": 18.87524240465417,
"grad_norm": 1.4859204292297363,
"learning_rate": 0.001,
"loss": 2.3941,
"step": 58400
},
{
"epoch": 18.907563025210084,
"grad_norm": 1.0631186962127686,
"learning_rate": 0.001,
"loss": 2.3618,
"step": 58500
},
{
"epoch": 18.939883645766,
"grad_norm": 1.0279467105865479,
"learning_rate": 0.001,
"loss": 2.3963,
"step": 58600
},
{
"epoch": 18.972204266321913,
"grad_norm": 1.0480549335479736,
"learning_rate": 0.001,
"loss": 2.3945,
"step": 58700
},
{
"epoch": 19.004524886877828,
"grad_norm": 1.1009399890899658,
"learning_rate": 0.001,
"loss": 2.3892,
"step": 58800
},
{
"epoch": 19.036845507433743,
"grad_norm": 1.3645431995391846,
"learning_rate": 0.001,
"loss": 2.2606,
"step": 58900
},
{
"epoch": 19.069166127989657,
"grad_norm": 1.3152440786361694,
"learning_rate": 0.001,
"loss": 2.2503,
"step": 59000
},
{
"epoch": 19.10148674854557,
"grad_norm": 1.4546312093734741,
"learning_rate": 0.001,
"loss": 2.2665,
"step": 59100
},
{
"epoch": 19.133807369101486,
"grad_norm": 1.2125812768936157,
"learning_rate": 0.001,
"loss": 2.2907,
"step": 59200
},
{
"epoch": 19.1661279896574,
"grad_norm": 1.1450135707855225,
"learning_rate": 0.001,
"loss": 2.2718,
"step": 59300
},
{
"epoch": 19.198448610213315,
"grad_norm": 0.9613053202629089,
"learning_rate": 0.001,
"loss": 2.2803,
"step": 59400
},
{
"epoch": 19.23076923076923,
"grad_norm": 1.0520044565200806,
"learning_rate": 0.001,
"loss": 2.281,
"step": 59500
},
{
"epoch": 19.263089851325145,
"grad_norm": 1.1664214134216309,
"learning_rate": 0.001,
"loss": 2.301,
"step": 59600
},
{
"epoch": 19.29541047188106,
"grad_norm": 1.1827281713485718,
"learning_rate": 0.001,
"loss": 2.2904,
"step": 59700
},
{
"epoch": 19.327731092436974,
"grad_norm": 1.176406979560852,
"learning_rate": 0.001,
"loss": 2.2924,
"step": 59800
},
{
"epoch": 19.360051712992888,
"grad_norm": 1.2200160026550293,
"learning_rate": 0.001,
"loss": 2.3197,
"step": 59900
},
{
"epoch": 19.392372333548803,
"grad_norm": 1.1691123247146606,
"learning_rate": 0.001,
"loss": 2.3033,
"step": 60000
},
{
"epoch": 19.424692954104717,
"grad_norm": 1.2286561727523804,
"learning_rate": 0.001,
"loss": 2.3164,
"step": 60100
},
{
"epoch": 19.457013574660632,
"grad_norm": 1.1665754318237305,
"learning_rate": 0.001,
"loss": 2.3034,
"step": 60200
},
{
"epoch": 19.489334195216546,
"grad_norm": 0.9893099665641785,
"learning_rate": 0.001,
"loss": 2.3287,
"step": 60300
},
{
"epoch": 19.521654815772465,
"grad_norm": 1.4509778022766113,
"learning_rate": 0.001,
"loss": 2.3249,
"step": 60400
},
{
"epoch": 19.55397543632838,
"grad_norm": 1.0169960260391235,
"learning_rate": 0.001,
"loss": 2.3198,
"step": 60500
},
{
"epoch": 19.586296056884294,
"grad_norm": 0.9599341750144958,
"learning_rate": 0.001,
"loss": 2.3205,
"step": 60600
},
{
"epoch": 19.618616677440208,
"grad_norm": 1.3709796667099,
"learning_rate": 0.001,
"loss": 2.3335,
"step": 60700
},
{
"epoch": 19.650937297996123,
"grad_norm": 1.3136423826217651,
"learning_rate": 0.001,
"loss": 2.3336,
"step": 60800
},
{
"epoch": 19.683257918552037,
"grad_norm": 1.39202880859375,
"learning_rate": 0.001,
"loss": 2.3263,
"step": 60900
},
{
"epoch": 19.715578539107952,
"grad_norm": 1.1869385242462158,
"learning_rate": 0.001,
"loss": 2.3412,
"step": 61000
},
{
"epoch": 19.747899159663866,
"grad_norm": 1.2061609029769897,
"learning_rate": 0.001,
"loss": 2.3265,
"step": 61100
},
{
"epoch": 19.78021978021978,
"grad_norm": 1.0782445669174194,
"learning_rate": 0.001,
"loss": 2.3344,
"step": 61200
},
{
"epoch": 19.812540400775696,
"grad_norm": 1.2751260995864868,
"learning_rate": 0.001,
"loss": 2.3609,
"step": 61300
},
{
"epoch": 19.84486102133161,
"grad_norm": 0.8370693325996399,
"learning_rate": 0.001,
"loss": 2.3305,
"step": 61400
},
{
"epoch": 19.877181641887525,
"grad_norm": 1.0448507070541382,
"learning_rate": 0.001,
"loss": 2.3659,
"step": 61500
},
{
"epoch": 19.90950226244344,
"grad_norm": 1.565977692604065,
"learning_rate": 0.001,
"loss": 2.3694,
"step": 61600
},
{
"epoch": 19.941822882999354,
"grad_norm": 1.0472636222839355,
"learning_rate": 0.001,
"loss": 2.3815,
"step": 61700
},
{
"epoch": 19.97414350355527,
"grad_norm": 1.2321832180023193,
"learning_rate": 0.001,
"loss": 2.3747,
"step": 61800
},
{
"epoch": 20.006464124111183,
"grad_norm": 1.3486043214797974,
"learning_rate": 0.001,
"loss": 2.341,
"step": 61900
},
{
"epoch": 20.038784744667097,
"grad_norm": 1.2104501724243164,
"learning_rate": 0.001,
"loss": 2.2406,
"step": 62000
},
{
"epoch": 20.071105365223012,
"grad_norm": 1.1732680797576904,
"learning_rate": 0.001,
"loss": 2.217,
"step": 62100
},
{
"epoch": 20.103425985778927,
"grad_norm": 1.1484770774841309,
"learning_rate": 0.001,
"loss": 2.2429,
"step": 62200
},
{
"epoch": 20.13574660633484,
"grad_norm": 0.9160681366920471,
"learning_rate": 0.001,
"loss": 2.229,
"step": 62300
},
{
"epoch": 20.168067226890756,
"grad_norm": 1.2878648042678833,
"learning_rate": 0.001,
"loss": 2.251,
"step": 62400
},
{
"epoch": 20.20038784744667,
"grad_norm": 0.9854994416236877,
"learning_rate": 0.001,
"loss": 2.2597,
"step": 62500
},
{
"epoch": 20.232708468002585,
"grad_norm": 0.8945751786231995,
"learning_rate": 0.001,
"loss": 2.2564,
"step": 62600
},
{
"epoch": 20.2650290885585,
"grad_norm": 1.1483877897262573,
"learning_rate": 0.001,
"loss": 2.2496,
"step": 62700
},
{
"epoch": 20.297349709114414,
"grad_norm": 1.2519859075546265,
"learning_rate": 0.001,
"loss": 2.2538,
"step": 62800
},
{
"epoch": 20.32967032967033,
"grad_norm": 1.1303123235702515,
"learning_rate": 0.001,
"loss": 2.254,
"step": 62900
},
{
"epoch": 20.361990950226243,
"grad_norm": 1.1656547784805298,
"learning_rate": 0.001,
"loss": 2.2556,
"step": 63000
},
{
"epoch": 20.394311570782158,
"grad_norm": 1.0333799123764038,
"learning_rate": 0.001,
"loss": 2.2945,
"step": 63100
},
{
"epoch": 20.426632191338072,
"grad_norm": 1.1399365663528442,
"learning_rate": 0.001,
"loss": 2.2859,
"step": 63200
},
{
"epoch": 20.458952811893987,
"grad_norm": 1.2444775104522705,
"learning_rate": 0.001,
"loss": 2.2911,
"step": 63300
},
{
"epoch": 20.4912734324499,
"grad_norm": 0.9173609018325806,
"learning_rate": 0.001,
"loss": 2.2929,
"step": 63400
},
{
"epoch": 20.52359405300582,
"grad_norm": 1.079097032546997,
"learning_rate": 0.001,
"loss": 2.2886,
"step": 63500
},
{
"epoch": 20.555914673561734,
"grad_norm": 1.1917749643325806,
"learning_rate": 0.001,
"loss": 2.2985,
"step": 63600
},
{
"epoch": 20.58823529411765,
"grad_norm": 1.770207166671753,
"learning_rate": 0.001,
"loss": 2.291,
"step": 63700
},
{
"epoch": 20.620555914673563,
"grad_norm": 1.3492732048034668,
"learning_rate": 0.001,
"loss": 2.3069,
"step": 63800
},
{
"epoch": 20.652876535229478,
"grad_norm": 1.407410740852356,
"learning_rate": 0.001,
"loss": 2.2951,
"step": 63900
},
{
"epoch": 20.685197155785392,
"grad_norm": 0.997594952583313,
"learning_rate": 0.001,
"loss": 2.3101,
"step": 64000
},
{
"epoch": 20.717517776341307,
"grad_norm": 0.9661346673965454,
"learning_rate": 0.001,
"loss": 2.3111,
"step": 64100
},
{
"epoch": 20.74983839689722,
"grad_norm": 0.9275293946266174,
"learning_rate": 0.001,
"loss": 2.2979,
"step": 64200
},
{
"epoch": 20.782159017453136,
"grad_norm": 1.0585098266601562,
"learning_rate": 0.001,
"loss": 2.3409,
"step": 64300
},
{
"epoch": 20.81447963800905,
"grad_norm": 0.9181944727897644,
"learning_rate": 0.001,
"loss": 2.3337,
"step": 64400
},
{
"epoch": 20.846800258564965,
"grad_norm": 1.199889063835144,
"learning_rate": 0.001,
"loss": 2.3133,
"step": 64500
},
{
"epoch": 20.87912087912088,
"grad_norm": 0.9734306931495667,
"learning_rate": 0.001,
"loss": 2.3142,
"step": 64600
},
{
"epoch": 20.911441499676794,
"grad_norm": 1.0917556285858154,
"learning_rate": 0.001,
"loss": 2.3156,
"step": 64700
},
{
"epoch": 20.94376212023271,
"grad_norm": 1.1194589138031006,
"learning_rate": 0.001,
"loss": 2.3354,
"step": 64800
},
{
"epoch": 20.976082740788623,
"grad_norm": 1.3248929977416992,
"learning_rate": 0.001,
"loss": 2.3181,
"step": 64900
},
{
"epoch": 21.008403361344538,
"grad_norm": 0.9312332272529602,
"learning_rate": 0.001,
"loss": 2.3122,
"step": 65000
},
{
"epoch": 21.040723981900452,
"grad_norm": 1.1081972122192383,
"learning_rate": 0.001,
"loss": 2.1918,
"step": 65100
},
{
"epoch": 21.073044602456367,
"grad_norm": 0.9267410635948181,
"learning_rate": 0.001,
"loss": 2.1872,
"step": 65200
},
{
"epoch": 21.10536522301228,
"grad_norm": 0.9274157881736755,
"learning_rate": 0.001,
"loss": 2.1976,
"step": 65300
},
{
"epoch": 21.137685843568196,
"grad_norm": 0.9890946745872498,
"learning_rate": 0.001,
"loss": 2.2216,
"step": 65400
},
{
"epoch": 21.17000646412411,
"grad_norm": 1.1272910833358765,
"learning_rate": 0.001,
"loss": 2.2268,
"step": 65500
},
{
"epoch": 21.202327084680025,
"grad_norm": 1.047804594039917,
"learning_rate": 0.001,
"loss": 2.2316,
"step": 65600
},
{
"epoch": 21.23464770523594,
"grad_norm": 1.1816391944885254,
"learning_rate": 0.001,
"loss": 2.2298,
"step": 65700
},
{
"epoch": 21.266968325791854,
"grad_norm": 1.0794835090637207,
"learning_rate": 0.001,
"loss": 2.227,
"step": 65800
},
{
"epoch": 21.29928894634777,
"grad_norm": 1.1223770380020142,
"learning_rate": 0.001,
"loss": 2.2317,
"step": 65900
},
{
"epoch": 21.331609566903683,
"grad_norm": 1.1707874536514282,
"learning_rate": 0.001,
"loss": 2.243,
"step": 66000
},
{
"epoch": 21.363930187459598,
"grad_norm": 0.9382168054580688,
"learning_rate": 0.001,
"loss": 2.2329,
"step": 66100
},
{
"epoch": 21.396250808015512,
"grad_norm": 0.9318244457244873,
"learning_rate": 0.001,
"loss": 2.2347,
"step": 66200
},
{
"epoch": 21.428571428571427,
"grad_norm": 1.050503134727478,
"learning_rate": 0.001,
"loss": 2.2404,
"step": 66300
},
{
"epoch": 21.46089204912734,
"grad_norm": 1.1723307371139526,
"learning_rate": 0.001,
"loss": 2.2697,
"step": 66400
},
{
"epoch": 21.49321266968326,
"grad_norm": 1.2298661470413208,
"learning_rate": 0.001,
"loss": 2.2443,
"step": 66500
},
{
"epoch": 21.525533290239174,
"grad_norm": 1.0454769134521484,
"learning_rate": 0.001,
"loss": 2.2415,
"step": 66600
},
{
"epoch": 21.55785391079509,
"grad_norm": 1.0260510444641113,
"learning_rate": 0.001,
"loss": 2.2695,
"step": 66700
},
{
"epoch": 21.590174531351003,
"grad_norm": 1.1626557111740112,
"learning_rate": 0.001,
"loss": 2.2588,
"step": 66800
},
{
"epoch": 21.622495151906918,
"grad_norm": 0.959842324256897,
"learning_rate": 0.001,
"loss": 2.2597,
"step": 66900
},
{
"epoch": 21.654815772462833,
"grad_norm": 1.1375782489776611,
"learning_rate": 0.001,
"loss": 2.2908,
"step": 67000
},
{
"epoch": 21.687136393018747,
"grad_norm": 0.9817724823951721,
"learning_rate": 0.001,
"loss": 2.2784,
"step": 67100
},
{
"epoch": 21.71945701357466,
"grad_norm": 1.3960691690444946,
"learning_rate": 0.001,
"loss": 2.2897,
"step": 67200
},
{
"epoch": 21.751777634130576,
"grad_norm": 1.14834725856781,
"learning_rate": 0.001,
"loss": 2.2849,
"step": 67300
},
{
"epoch": 21.78409825468649,
"grad_norm": 1.1419909000396729,
"learning_rate": 0.001,
"loss": 2.2785,
"step": 67400
},
{
"epoch": 21.816418875242405,
"grad_norm": 1.1380691528320312,
"learning_rate": 0.001,
"loss": 2.2886,
"step": 67500
},
{
"epoch": 21.84873949579832,
"grad_norm": 1.0933802127838135,
"learning_rate": 0.001,
"loss": 2.2879,
"step": 67600
},
{
"epoch": 21.881060116354234,
"grad_norm": 1.1113094091415405,
"learning_rate": 0.001,
"loss": 2.2924,
"step": 67700
},
{
"epoch": 21.91338073691015,
"grad_norm": 1.2116150856018066,
"learning_rate": 0.001,
"loss": 2.3067,
"step": 67800
},
{
"epoch": 21.945701357466064,
"grad_norm": 0.9935580492019653,
"learning_rate": 0.001,
"loss": 2.3048,
"step": 67900
},
{
"epoch": 21.978021978021978,
"grad_norm": 1.1998525857925415,
"learning_rate": 0.001,
"loss": 2.3239,
"step": 68000
},
{
"epoch": 22.010342598577893,
"grad_norm": 1.20631742477417,
"learning_rate": 0.001,
"loss": 2.2593,
"step": 68100
},
{
"epoch": 22.042663219133807,
"grad_norm": 1.023864507675171,
"learning_rate": 0.001,
"loss": 2.1791,
"step": 68200
},
{
"epoch": 22.07498383968972,
"grad_norm": 1.0714632272720337,
"learning_rate": 0.001,
"loss": 2.1873,
"step": 68300
},
{
"epoch": 22.107304460245636,
"grad_norm": 1.1849026679992676,
"learning_rate": 0.001,
"loss": 2.1645,
"step": 68400
},
{
"epoch": 22.13962508080155,
"grad_norm": 1.0333696603775024,
"learning_rate": 0.001,
"loss": 2.1697,
"step": 68500
},
{
"epoch": 22.171945701357465,
"grad_norm": 1.240440845489502,
"learning_rate": 0.001,
"loss": 2.1938,
"step": 68600
},
{
"epoch": 22.20426632191338,
"grad_norm": 1.0015324354171753,
"learning_rate": 0.001,
"loss": 2.2063,
"step": 68700
},
{
"epoch": 22.236586942469295,
"grad_norm": 1.1313668489456177,
"learning_rate": 0.001,
"loss": 2.209,
"step": 68800
},
{
"epoch": 22.26890756302521,
"grad_norm": 1.3199909925460815,
"learning_rate": 0.001,
"loss": 2.2065,
"step": 68900
},
{
"epoch": 22.301228183581124,
"grad_norm": 1.0711723566055298,
"learning_rate": 0.001,
"loss": 2.2192,
"step": 69000
},
{
"epoch": 22.33354880413704,
"grad_norm": 1.041747808456421,
"learning_rate": 0.001,
"loss": 2.2037,
"step": 69100
},
{
"epoch": 22.365869424692953,
"grad_norm": 1.3392200469970703,
"learning_rate": 0.001,
"loss": 2.2293,
"step": 69200
},
{
"epoch": 22.398190045248867,
"grad_norm": 1.046087384223938,
"learning_rate": 0.001,
"loss": 2.2145,
"step": 69300
},
{
"epoch": 22.430510665804782,
"grad_norm": 1.1171798706054688,
"learning_rate": 0.001,
"loss": 2.207,
"step": 69400
},
{
"epoch": 22.462831286360696,
"grad_norm": 1.1276259422302246,
"learning_rate": 0.001,
"loss": 2.2183,
"step": 69500
},
{
"epoch": 22.49515190691661,
"grad_norm": 1.1094915866851807,
"learning_rate": 0.001,
"loss": 2.2105,
"step": 69600
},
{
"epoch": 22.52747252747253,
"grad_norm": 1.0818486213684082,
"learning_rate": 0.001,
"loss": 2.2343,
"step": 69700
},
{
"epoch": 22.559793148028444,
"grad_norm": 1.2103203535079956,
"learning_rate": 0.001,
"loss": 2.2288,
"step": 69800
},
{
"epoch": 22.59211376858436,
"grad_norm": 1.234008550643921,
"learning_rate": 0.001,
"loss": 2.2322,
"step": 69900
},
{
"epoch": 22.624434389140273,
"grad_norm": 0.994259238243103,
"learning_rate": 0.001,
"loss": 2.2492,
"step": 70000
},
{
"epoch": 22.656755009696187,
"grad_norm": 1.5072773694992065,
"learning_rate": 0.001,
"loss": 2.2561,
"step": 70100
},
{
"epoch": 22.689075630252102,
"grad_norm": 1.2016674280166626,
"learning_rate": 0.001,
"loss": 2.2574,
"step": 70200
},
{
"epoch": 22.721396250808017,
"grad_norm": 1.2648993730545044,
"learning_rate": 0.001,
"loss": 2.2433,
"step": 70300
},
{
"epoch": 22.75371687136393,
"grad_norm": 1.1488635540008545,
"learning_rate": 0.001,
"loss": 2.2484,
"step": 70400
},
{
"epoch": 22.786037491919846,
"grad_norm": 1.3801467418670654,
"learning_rate": 0.001,
"loss": 2.2532,
"step": 70500
},
{
"epoch": 22.81835811247576,
"grad_norm": 1.04673171043396,
"learning_rate": 0.001,
"loss": 2.2578,
"step": 70600
},
{
"epoch": 22.850678733031675,
"grad_norm": 0.8600795865058899,
"learning_rate": 0.001,
"loss": 2.2686,
"step": 70700
},
{
"epoch": 22.88299935358759,
"grad_norm": 1.0992897748947144,
"learning_rate": 0.001,
"loss": 2.2633,
"step": 70800
},
{
"epoch": 22.915319974143504,
"grad_norm": 1.3973101377487183,
"learning_rate": 0.001,
"loss": 2.2632,
"step": 70900
},
{
"epoch": 22.94764059469942,
"grad_norm": 1.2797373533248901,
"learning_rate": 0.001,
"loss": 2.2624,
"step": 71000
},
{
"epoch": 22.979961215255333,
"grad_norm": 1.142982006072998,
"learning_rate": 0.001,
"loss": 2.2924,
"step": 71100
},
{
"epoch": 23.012281835811248,
"grad_norm": 1.1095130443572998,
"learning_rate": 0.001,
"loss": 2.2068,
"step": 71200
},
{
"epoch": 23.044602456367162,
"grad_norm": 1.2436374425888062,
"learning_rate": 0.001,
"loss": 2.1178,
"step": 71300
},
{
"epoch": 23.076923076923077,
"grad_norm": 1.511630654335022,
"learning_rate": 0.001,
"loss": 2.1486,
"step": 71400
},
{
"epoch": 23.10924369747899,
"grad_norm": 0.9946317076683044,
"learning_rate": 0.001,
"loss": 2.1527,
"step": 71500
},
{
"epoch": 23.141564318034906,
"grad_norm": 1.2775388956069946,
"learning_rate": 0.001,
"loss": 2.1587,
"step": 71600
},
{
"epoch": 23.17388493859082,
"grad_norm": 1.0382200479507446,
"learning_rate": 0.001,
"loss": 2.1525,
"step": 71700
},
{
"epoch": 23.206205559146735,
"grad_norm": 1.2592073678970337,
"learning_rate": 0.001,
"loss": 2.1755,
"step": 71800
},
{
"epoch": 23.23852617970265,
"grad_norm": 1.5339597463607788,
"learning_rate": 0.001,
"loss": 2.1797,
"step": 71900
},
{
"epoch": 23.270846800258564,
"grad_norm": 1.19203782081604,
"learning_rate": 0.001,
"loss": 2.1559,
"step": 72000
},
{
"epoch": 23.30316742081448,
"grad_norm": 1.60515296459198,
"learning_rate": 0.001,
"loss": 2.1823,
"step": 72100
},
{
"epoch": 23.335488041370393,
"grad_norm": 1.2635244131088257,
"learning_rate": 0.001,
"loss": 2.1879,
"step": 72200
},
{
"epoch": 23.367808661926308,
"grad_norm": 1.1627428531646729,
"learning_rate": 0.001,
"loss": 2.1869,
"step": 72300
},
{
"epoch": 23.400129282482222,
"grad_norm": 1.6652495861053467,
"learning_rate": 0.001,
"loss": 2.2076,
"step": 72400
},
{
"epoch": 23.432449903038137,
"grad_norm": 1.048194169998169,
"learning_rate": 0.001,
"loss": 2.1801,
"step": 72500
},
{
"epoch": 23.46477052359405,
"grad_norm": 1.3235198259353638,
"learning_rate": 0.001,
"loss": 2.1952,
"step": 72600
},
{
"epoch": 23.49709114414997,
"grad_norm": 1.5227469205856323,
"learning_rate": 0.001,
"loss": 2.2016,
"step": 72700
},
{
"epoch": 23.529411764705884,
"grad_norm": 1.2690283060073853,
"learning_rate": 0.001,
"loss": 2.1891,
"step": 72800
},
{
"epoch": 23.5617323852618,
"grad_norm": 1.524413824081421,
"learning_rate": 0.001,
"loss": 2.2064,
"step": 72900
},
{
"epoch": 23.594053005817713,
"grad_norm": 1.07742178440094,
"learning_rate": 0.001,
"loss": 2.2007,
"step": 73000
},
{
"epoch": 23.626373626373628,
"grad_norm": 1.1482139825820923,
"learning_rate": 0.001,
"loss": 2.2018,
"step": 73100
},
{
"epoch": 23.658694246929542,
"grad_norm": 1.5760014057159424,
"learning_rate": 0.001,
"loss": 2.2167,
"step": 73200
},
{
"epoch": 23.691014867485457,
"grad_norm": 1.315416693687439,
"learning_rate": 0.001,
"loss": 2.2259,
"step": 73300
},
{
"epoch": 23.72333548804137,
"grad_norm": 1.3581657409667969,
"learning_rate": 0.001,
"loss": 2.2281,
"step": 73400
},
{
"epoch": 23.755656108597286,
"grad_norm": 1.3128376007080078,
"learning_rate": 0.001,
"loss": 2.2248,
"step": 73500
},
{
"epoch": 23.7879767291532,
"grad_norm": 1.2848385572433472,
"learning_rate": 0.001,
"loss": 2.25,
"step": 73600
},
{
"epoch": 23.820297349709115,
"grad_norm": 1.3273836374282837,
"learning_rate": 0.001,
"loss": 2.2414,
"step": 73700
},
{
"epoch": 23.85261797026503,
"grad_norm": 1.5152201652526855,
"learning_rate": 0.001,
"loss": 2.2267,
"step": 73800
},
{
"epoch": 23.884938590820944,
"grad_norm": 1.1121904850006104,
"learning_rate": 0.001,
"loss": 2.2398,
"step": 73900
},
{
"epoch": 23.91725921137686,
"grad_norm": 1.3406445980072021,
"learning_rate": 0.001,
"loss": 2.2347,
"step": 74000
},
{
"epoch": 23.949579831932773,
"grad_norm": 1.1728599071502686,
"learning_rate": 0.001,
"loss": 2.252,
"step": 74100
},
{
"epoch": 23.981900452488688,
"grad_norm": 1.477712869644165,
"learning_rate": 0.001,
"loss": 2.2578,
"step": 74200
},
{
"epoch": 24.014221073044602,
"grad_norm": 1.252150297164917,
"learning_rate": 0.001,
"loss": 2.1751,
"step": 74300
},
{
"epoch": 24.046541693600517,
"grad_norm": 1.0892266035079956,
"learning_rate": 0.001,
"loss": 2.0975,
"step": 74400
},
{
"epoch": 24.07886231415643,
"grad_norm": 1.5348308086395264,
"learning_rate": 0.001,
"loss": 2.1171,
"step": 74500
},
{
"epoch": 24.111182934712346,
"grad_norm": 1.2735607624053955,
"learning_rate": 0.001,
"loss": 2.1248,
"step": 74600
},
{
"epoch": 24.14350355526826,
"grad_norm": 1.4522603750228882,
"learning_rate": 0.001,
"loss": 2.1353,
"step": 74700
},
{
"epoch": 24.175824175824175,
"grad_norm": 1.1594727039337158,
"learning_rate": 0.001,
"loss": 2.1203,
"step": 74800
},
{
"epoch": 24.20814479638009,
"grad_norm": 1.1483879089355469,
"learning_rate": 0.001,
"loss": 2.1202,
"step": 74900
},
{
"epoch": 24.240465416936004,
"grad_norm": 1.5544039011001587,
"learning_rate": 0.001,
"loss": 2.1497,
"step": 75000
},
{
"epoch": 24.27278603749192,
"grad_norm": 1.1212824583053589,
"learning_rate": 0.001,
"loss": 2.1311,
"step": 75100
},
{
"epoch": 24.305106658047833,
"grad_norm": 1.3615349531173706,
"learning_rate": 0.001,
"loss": 2.1655,
"step": 75200
},
{
"epoch": 24.337427278603748,
"grad_norm": 1.7518833875656128,
"learning_rate": 0.001,
"loss": 2.1492,
"step": 75300
},
{
"epoch": 24.369747899159663,
"grad_norm": 1.6515196561813354,
"learning_rate": 0.001,
"loss": 2.1704,
"step": 75400
},
{
"epoch": 24.402068519715577,
"grad_norm": 1.1430226564407349,
"learning_rate": 0.001,
"loss": 2.1821,
"step": 75500
},
{
"epoch": 24.43438914027149,
"grad_norm": 1.417546033859253,
"learning_rate": 0.001,
"loss": 2.1521,
"step": 75600
},
{
"epoch": 24.466709760827406,
"grad_norm": 1.274763584136963,
"learning_rate": 0.001,
"loss": 2.178,
"step": 75700
},
{
"epoch": 24.49903038138332,
"grad_norm": 1.2675822973251343,
"learning_rate": 0.001,
"loss": 2.1669,
"step": 75800
},
{
"epoch": 24.53135100193924,
"grad_norm": 1.223568081855774,
"learning_rate": 0.001,
"loss": 2.1686,
"step": 75900
},
{
"epoch": 24.563671622495153,
"grad_norm": 1.2633905410766602,
"learning_rate": 0.001,
"loss": 2.1922,
"step": 76000
},
{
"epoch": 24.595992243051068,
"grad_norm": 1.2331293821334839,
"learning_rate": 0.001,
"loss": 2.1862,
"step": 76100
},
{
"epoch": 24.628312863606983,
"grad_norm": 1.312574028968811,
"learning_rate": 0.001,
"loss": 2.193,
"step": 76200
},
{
"epoch": 24.660633484162897,
"grad_norm": 1.2204712629318237,
"learning_rate": 0.001,
"loss": 2.2033,
"step": 76300
},
{
"epoch": 24.69295410471881,
"grad_norm": 1.5437800884246826,
"learning_rate": 0.001,
"loss": 2.1897,
"step": 76400
},
{
"epoch": 24.725274725274726,
"grad_norm": 1.1601932048797607,
"learning_rate": 0.001,
"loss": 2.1951,
"step": 76500
},
{
"epoch": 24.75759534583064,
"grad_norm": 1.2676047086715698,
"learning_rate": 0.001,
"loss": 2.2037,
"step": 76600
},
{
"epoch": 24.789915966386555,
"grad_norm": 1.3851171731948853,
"learning_rate": 0.001,
"loss": 2.2063,
"step": 76700
},
{
"epoch": 24.82223658694247,
"grad_norm": 1.4091224670410156,
"learning_rate": 0.001,
"loss": 2.2055,
"step": 76800
},
{
"epoch": 24.854557207498384,
"grad_norm": 1.5163503885269165,
"learning_rate": 0.001,
"loss": 2.195,
"step": 76900
},
{
"epoch": 24.8868778280543,
"grad_norm": 1.767259955406189,
"learning_rate": 0.001,
"loss": 2.2134,
"step": 77000
},
{
"epoch": 24.919198448610214,
"grad_norm": 1.4628684520721436,
"learning_rate": 0.001,
"loss": 2.2266,
"step": 77100
},
{
"epoch": 24.951519069166128,
"grad_norm": 1.3239951133728027,
"learning_rate": 0.001,
"loss": 2.2268,
"step": 77200
},
{
"epoch": 24.983839689722043,
"grad_norm": 1.3385649919509888,
"learning_rate": 0.001,
"loss": 2.2221,
"step": 77300
},
{
"epoch": 25.016160310277957,
"grad_norm": 2.364973783493042,
"learning_rate": 0.001,
"loss": 2.138,
"step": 77400
},
{
"epoch": 25.048480930833872,
"grad_norm": 1.813896656036377,
"learning_rate": 0.001,
"loss": 2.0917,
"step": 77500
},
{
"epoch": 25.080801551389786,
"grad_norm": 2.0273733139038086,
"learning_rate": 0.001,
"loss": 2.0815,
"step": 77600
},
{
"epoch": 25.1131221719457,
"grad_norm": 1.9387582540512085,
"learning_rate": 0.001,
"loss": 2.1073,
"step": 77700
},
{
"epoch": 25.145442792501616,
"grad_norm": 1.9063010215759277,
"learning_rate": 0.001,
"loss": 2.1143,
"step": 77800
},
{
"epoch": 25.17776341305753,
"grad_norm": 1.8928618431091309,
"learning_rate": 0.001,
"loss": 2.1038,
"step": 77900
},
{
"epoch": 25.210084033613445,
"grad_norm": 1.899552583694458,
"learning_rate": 0.001,
"loss": 2.1039,
"step": 78000
},
{
"epoch": 25.24240465416936,
"grad_norm": 1.945588231086731,
"learning_rate": 0.001,
"loss": 2.0994,
"step": 78100
},
{
"epoch": 25.274725274725274,
"grad_norm": 1.9876327514648438,
"learning_rate": 0.001,
"loss": 2.1436,
"step": 78200
},
{
"epoch": 25.30704589528119,
"grad_norm": 2.129936456680298,
"learning_rate": 0.001,
"loss": 2.113,
"step": 78300
},
{
"epoch": 25.339366515837103,
"grad_norm": 1.7983858585357666,
"learning_rate": 0.001,
"loss": 2.114,
"step": 78400
},
{
"epoch": 25.371687136393017,
"grad_norm": 2.5916574001312256,
"learning_rate": 0.001,
"loss": 2.1116,
"step": 78500
},
{
"epoch": 25.404007756948932,
"grad_norm": 1.6096268892288208,
"learning_rate": 0.001,
"loss": 2.1336,
"step": 78600
},
{
"epoch": 25.436328377504847,
"grad_norm": 2.4804930686950684,
"learning_rate": 0.001,
"loss": 2.1351,
"step": 78700
},
{
"epoch": 25.46864899806076,
"grad_norm": 1.5392675399780273,
"learning_rate": 0.001,
"loss": 2.1542,
"step": 78800
},
{
"epoch": 25.50096961861668,
"grad_norm": 1.8824764490127563,
"learning_rate": 0.001,
"loss": 2.1527,
"step": 78900
},
{
"epoch": 25.533290239172594,
"grad_norm": 2.1411380767822266,
"learning_rate": 0.001,
"loss": 2.1505,
"step": 79000
},
{
"epoch": 25.56561085972851,
"grad_norm": 2.1214542388916016,
"learning_rate": 0.001,
"loss": 2.1562,
"step": 79100
},
{
"epoch": 25.597931480284423,
"grad_norm": 2.041605234146118,
"learning_rate": 0.001,
"loss": 2.1686,
"step": 79200
},
{
"epoch": 25.630252100840337,
"grad_norm": 1.676820993423462,
"learning_rate": 0.001,
"loss": 2.1571,
"step": 79300
},
{
"epoch": 25.662572721396252,
"grad_norm": 1.79721200466156,
"learning_rate": 0.001,
"loss": 2.1817,
"step": 79400
},
{
"epoch": 25.694893341952167,
"grad_norm": 1.4796991348266602,
"learning_rate": 0.001,
"loss": 2.1801,
"step": 79500
},
{
"epoch": 25.72721396250808,
"grad_norm": 2.2555365562438965,
"learning_rate": 0.001,
"loss": 2.1785,
"step": 79600
},
{
"epoch": 25.759534583063996,
"grad_norm": 2.7121639251708984,
"learning_rate": 0.001,
"loss": 2.1737,
"step": 79700
},
{
"epoch": 25.79185520361991,
"grad_norm": 2.044637680053711,
"learning_rate": 0.001,
"loss": 2.1911,
"step": 79800
},
{
"epoch": 25.824175824175825,
"grad_norm": 1.9338701963424683,
"learning_rate": 0.001,
"loss": 2.1627,
"step": 79900
},
{
"epoch": 25.85649644473174,
"grad_norm": 2.463005781173706,
"learning_rate": 0.001,
"loss": 2.1955,
"step": 80000
},
{
"epoch": 25.888817065287654,
"grad_norm": 2.6166152954101562,
"learning_rate": 0.001,
"loss": 2.1743,
"step": 80100
},
{
"epoch": 25.92113768584357,
"grad_norm": 2.0384294986724854,
"learning_rate": 0.001,
"loss": 2.2037,
"step": 80200
},
{
"epoch": 25.953458306399483,
"grad_norm": 1.685394287109375,
"learning_rate": 0.001,
"loss": 2.1862,
"step": 80300
},
{
"epoch": 25.985778926955398,
"grad_norm": 1.6715224981307983,
"learning_rate": 0.001,
"loss": 2.2114,
"step": 80400
},
{
"epoch": 26.018099547511312,
"grad_norm": 1.4495394229888916,
"learning_rate": 0.001,
"loss": 2.1076,
"step": 80500
},
{
"epoch": 26.050420168067227,
"grad_norm": 1.3092225790023804,
"learning_rate": 0.001,
"loss": 2.0463,
"step": 80600
},
{
"epoch": 26.08274078862314,
"grad_norm": 1.4169902801513672,
"learning_rate": 0.001,
"loss": 2.0543,
"step": 80700
},
{
"epoch": 26.115061409179056,
"grad_norm": 1.2692700624465942,
"learning_rate": 0.001,
"loss": 2.0541,
"step": 80800
},
{
"epoch": 26.14738202973497,
"grad_norm": 1.4265687465667725,
"learning_rate": 0.001,
"loss": 2.086,
"step": 80900
},
{
"epoch": 26.179702650290885,
"grad_norm": 1.3067816495895386,
"learning_rate": 0.001,
"loss": 2.0886,
"step": 81000
},
{
"epoch": 26.2120232708468,
"grad_norm": 1.217885971069336,
"learning_rate": 0.001,
"loss": 2.0945,
"step": 81100
},
{
"epoch": 26.244343891402714,
"grad_norm": 1.2995250225067139,
"learning_rate": 0.001,
"loss": 2.0904,
"step": 81200
},
{
"epoch": 26.27666451195863,
"grad_norm": 1.3405325412750244,
"learning_rate": 0.001,
"loss": 2.0948,
"step": 81300
},
{
"epoch": 26.308985132514543,
"grad_norm": 0.9935649633407593,
"learning_rate": 0.001,
"loss": 2.1113,
"step": 81400
},
{
"epoch": 26.341305753070458,
"grad_norm": 1.5722565650939941,
"learning_rate": 0.001,
"loss": 2.1159,
"step": 81500
},
{
"epoch": 26.373626373626372,
"grad_norm": 1.4917912483215332,
"learning_rate": 0.001,
"loss": 2.0941,
"step": 81600
},
{
"epoch": 26.405946994182287,
"grad_norm": 1.4292240142822266,
"learning_rate": 0.001,
"loss": 2.1237,
"step": 81700
},
{
"epoch": 26.4382676147382,
"grad_norm": 1.2860208749771118,
"learning_rate": 0.001,
"loss": 2.1273,
"step": 81800
},
{
"epoch": 26.470588235294116,
"grad_norm": 1.6937307119369507,
"learning_rate": 0.001,
"loss": 2.1174,
"step": 81900
},
{
"epoch": 26.50290885585003,
"grad_norm": 1.2667704820632935,
"learning_rate": 0.001,
"loss": 2.1197,
"step": 82000
},
{
"epoch": 26.53522947640595,
"grad_norm": 1.7752530574798584,
"learning_rate": 0.001,
"loss": 2.1358,
"step": 82100
},
{
"epoch": 26.567550096961863,
"grad_norm": 1.4119603633880615,
"learning_rate": 0.001,
"loss": 2.1247,
"step": 82200
},
{
"epoch": 26.599870717517778,
"grad_norm": 1.3704473972320557,
"learning_rate": 0.001,
"loss": 2.1381,
"step": 82300
},
{
"epoch": 26.632191338073692,
"grad_norm": 1.3702034950256348,
"learning_rate": 0.001,
"loss": 2.1407,
"step": 82400
},
{
"epoch": 26.664511958629607,
"grad_norm": 1.2412974834442139,
"learning_rate": 0.001,
"loss": 2.1443,
"step": 82500
},
{
"epoch": 26.69683257918552,
"grad_norm": 1.1445008516311646,
"learning_rate": 0.001,
"loss": 2.1509,
"step": 82600
},
{
"epoch": 26.729153199741436,
"grad_norm": 1.217456340789795,
"learning_rate": 0.001,
"loss": 2.14,
"step": 82700
},
{
"epoch": 26.76147382029735,
"grad_norm": 1.5372179746627808,
"learning_rate": 0.001,
"loss": 2.1676,
"step": 82800
},
{
"epoch": 26.793794440853265,
"grad_norm": 1.5856389999389648,
"learning_rate": 0.001,
"loss": 2.1575,
"step": 82900
},
{
"epoch": 26.82611506140918,
"grad_norm": 1.4989391565322876,
"learning_rate": 0.001,
"loss": 2.1596,
"step": 83000
},
{
"epoch": 26.858435681965094,
"grad_norm": 1.3020153045654297,
"learning_rate": 0.001,
"loss": 2.173,
"step": 83100
},
{
"epoch": 26.89075630252101,
"grad_norm": 1.2349504232406616,
"learning_rate": 0.001,
"loss": 2.1653,
"step": 83200
},
{
"epoch": 26.923076923076923,
"grad_norm": 1.3083016872406006,
"learning_rate": 0.001,
"loss": 2.1624,
"step": 83300
},
{
"epoch": 26.955397543632838,
"grad_norm": 1.6010842323303223,
"learning_rate": 0.001,
"loss": 2.1658,
"step": 83400
},
{
"epoch": 26.987718164188752,
"grad_norm": 1.087155818939209,
"learning_rate": 0.001,
"loss": 2.1564,
"step": 83500
},
{
"epoch": 27.020038784744667,
"grad_norm": 1.358615756034851,
"learning_rate": 0.001,
"loss": 2.0963,
"step": 83600
},
{
"epoch": 27.05235940530058,
"grad_norm": 1.1136542558670044,
"learning_rate": 0.001,
"loss": 2.0385,
"step": 83700
},
{
"epoch": 27.084680025856496,
"grad_norm": 1.32275390625,
"learning_rate": 0.001,
"loss": 2.0445,
"step": 83800
},
{
"epoch": 27.11700064641241,
"grad_norm": 1.2127283811569214,
"learning_rate": 0.001,
"loss": 2.0571,
"step": 83900
},
{
"epoch": 27.149321266968325,
"grad_norm": 1.7168000936508179,
"learning_rate": 0.001,
"loss": 2.0538,
"step": 84000
},
{
"epoch": 27.18164188752424,
"grad_norm": 0.990697979927063,
"learning_rate": 0.001,
"loss": 2.08,
"step": 84100
},
{
"epoch": 27.213962508080154,
"grad_norm": 1.3970186710357666,
"learning_rate": 0.001,
"loss": 2.0379,
"step": 84200
},
{
"epoch": 27.24628312863607,
"grad_norm": 1.258516550064087,
"learning_rate": 0.001,
"loss": 2.0596,
"step": 84300
},
{
"epoch": 27.278603749191983,
"grad_norm": 1.6113446950912476,
"learning_rate": 0.001,
"loss": 2.0657,
"step": 84400
},
{
"epoch": 27.310924369747898,
"grad_norm": 1.2201672792434692,
"learning_rate": 0.001,
"loss": 2.0685,
"step": 84500
},
{
"epoch": 27.343244990303813,
"grad_norm": 1.1018732786178589,
"learning_rate": 0.001,
"loss": 2.1003,
"step": 84600
},
{
"epoch": 27.375565610859727,
"grad_norm": 1.2542102336883545,
"learning_rate": 0.001,
"loss": 2.0918,
"step": 84700
},
{
"epoch": 27.40788623141564,
"grad_norm": 1.2258397340774536,
"learning_rate": 0.001,
"loss": 2.0821,
"step": 84800
},
{
"epoch": 27.440206851971556,
"grad_norm": 0.9618849158287048,
"learning_rate": 0.001,
"loss": 2.0826,
"step": 84900
},
{
"epoch": 27.47252747252747,
"grad_norm": 1.1504327058792114,
"learning_rate": 0.001,
"loss": 2.095,
"step": 85000
},
{
"epoch": 27.50484809308339,
"grad_norm": 1.3419139385223389,
"learning_rate": 0.001,
"loss": 2.1173,
"step": 85100
},
{
"epoch": 27.537168713639304,
"grad_norm": 1.3046069145202637,
"learning_rate": 0.001,
"loss": 2.1214,
"step": 85200
},
{
"epoch": 27.569489334195218,
"grad_norm": 1.1985609531402588,
"learning_rate": 0.001,
"loss": 2.1168,
"step": 85300
},
{
"epoch": 27.601809954751133,
"grad_norm": 1.0604177713394165,
"learning_rate": 0.001,
"loss": 2.11,
"step": 85400
},
{
"epoch": 27.634130575307047,
"grad_norm": 1.3282095193862915,
"learning_rate": 0.001,
"loss": 2.1155,
"step": 85500
},
{
"epoch": 27.66645119586296,
"grad_norm": 1.3031256198883057,
"learning_rate": 0.001,
"loss": 2.1115,
"step": 85600
},
{
"epoch": 27.698771816418876,
"grad_norm": 1.1260483264923096,
"learning_rate": 0.001,
"loss": 2.1306,
"step": 85700
},
{
"epoch": 27.73109243697479,
"grad_norm": 1.288153052330017,
"learning_rate": 0.001,
"loss": 2.1186,
"step": 85800
},
{
"epoch": 27.763413057530705,
"grad_norm": 1.1125783920288086,
"learning_rate": 0.001,
"loss": 2.1173,
"step": 85900
},
{
"epoch": 27.79573367808662,
"grad_norm": 1.61385178565979,
"learning_rate": 0.001,
"loss": 2.1265,
"step": 86000
},
{
"epoch": 27.828054298642535,
"grad_norm": 0.9980874061584473,
"learning_rate": 0.001,
"loss": 2.1223,
"step": 86100
},
{
"epoch": 27.86037491919845,
"grad_norm": 0.7647992372512817,
"learning_rate": 0.001,
"loss": 2.1427,
"step": 86200
},
{
"epoch": 27.892695539754364,
"grad_norm": 1.3897727727890015,
"learning_rate": 0.001,
"loss": 2.1298,
"step": 86300
},
{
"epoch": 27.92501616031028,
"grad_norm": 1.217728853225708,
"learning_rate": 0.001,
"loss": 2.1337,
"step": 86400
},
{
"epoch": 27.957336780866193,
"grad_norm": 1.0582938194274902,
"learning_rate": 0.001,
"loss": 2.1467,
"step": 86500
},
{
"epoch": 27.989657401422107,
"grad_norm": 1.0420843362808228,
"learning_rate": 0.001,
"loss": 2.1387,
"step": 86600
},
{
"epoch": 28.021978021978022,
"grad_norm": 1.2445080280303955,
"learning_rate": 0.001,
"loss": 2.0747,
"step": 86700
},
{
"epoch": 28.054298642533936,
"grad_norm": 1.015954613685608,
"learning_rate": 0.001,
"loss": 2.0105,
"step": 86800
},
{
"epoch": 28.08661926308985,
"grad_norm": 1.0916825532913208,
"learning_rate": 0.001,
"loss": 2.0008,
"step": 86900
},
{
"epoch": 28.118939883645766,
"grad_norm": 1.141888976097107,
"learning_rate": 0.001,
"loss": 2.026,
"step": 87000
},
{
"epoch": 28.15126050420168,
"grad_norm": 1.2229506969451904,
"learning_rate": 0.001,
"loss": 2.0125,
"step": 87100
},
{
"epoch": 28.183581124757595,
"grad_norm": 0.9546578526496887,
"learning_rate": 0.001,
"loss": 2.0334,
"step": 87200
},
{
"epoch": 28.21590174531351,
"grad_norm": 1.0811423063278198,
"learning_rate": 0.001,
"loss": 2.0476,
"step": 87300
},
{
"epoch": 28.248222365869424,
"grad_norm": 1.2767107486724854,
"learning_rate": 0.001,
"loss": 2.0474,
"step": 87400
},
{
"epoch": 28.28054298642534,
"grad_norm": 0.9923326969146729,
"learning_rate": 0.001,
"loss": 2.0579,
"step": 87500
},
{
"epoch": 28.312863606981253,
"grad_norm": 1.0915007591247559,
"learning_rate": 0.001,
"loss": 2.0425,
"step": 87600
},
{
"epoch": 28.345184227537167,
"grad_norm": 1.1557867527008057,
"learning_rate": 0.001,
"loss": 2.0387,
"step": 87700
},
{
"epoch": 28.377504848093082,
"grad_norm": 1.1348267793655396,
"learning_rate": 0.001,
"loss": 2.0857,
"step": 87800
},
{
"epoch": 28.409825468648997,
"grad_norm": 1.1359829902648926,
"learning_rate": 0.001,
"loss": 2.0654,
"step": 87900
},
{
"epoch": 28.44214608920491,
"grad_norm": 1.2314882278442383,
"learning_rate": 0.001,
"loss": 2.0608,
"step": 88000
},
{
"epoch": 28.474466709760826,
"grad_norm": 1.2598603963851929,
"learning_rate": 0.001,
"loss": 2.0714,
"step": 88100
},
{
"epoch": 28.50678733031674,
"grad_norm": 1.1337944269180298,
"learning_rate": 0.001,
"loss": 2.0763,
"step": 88200
},
{
"epoch": 28.53910795087266,
"grad_norm": 1.1325387954711914,
"learning_rate": 0.001,
"loss": 2.0914,
"step": 88300
},
{
"epoch": 28.571428571428573,
"grad_norm": 1.2363420724868774,
"learning_rate": 0.001,
"loss": 2.0917,
"step": 88400
},
{
"epoch": 28.603749191984488,
"grad_norm": 1.3564585447311401,
"learning_rate": 0.001,
"loss": 2.1034,
"step": 88500
},
{
"epoch": 28.636069812540402,
"grad_norm": 1.3790624141693115,
"learning_rate": 0.001,
"loss": 2.0954,
"step": 88600
},
{
"epoch": 28.668390433096317,
"grad_norm": 0.9206322431564331,
"learning_rate": 0.001,
"loss": 2.0993,
"step": 88700
},
{
"epoch": 28.70071105365223,
"grad_norm": 1.0365331172943115,
"learning_rate": 0.001,
"loss": 2.1008,
"step": 88800
},
{
"epoch": 28.733031674208146,
"grad_norm": 1.0946674346923828,
"learning_rate": 0.001,
"loss": 2.1036,
"step": 88900
},
{
"epoch": 28.76535229476406,
"grad_norm": 1.2759056091308594,
"learning_rate": 0.001,
"loss": 2.1125,
"step": 89000
},
{
"epoch": 28.797672915319975,
"grad_norm": 1.6298202276229858,
"learning_rate": 0.001,
"loss": 2.1104,
"step": 89100
},
{
"epoch": 28.82999353587589,
"grad_norm": 1.1779752969741821,
"learning_rate": 0.001,
"loss": 2.1118,
"step": 89200
},
{
"epoch": 28.862314156431804,
"grad_norm": 1.2181776762008667,
"learning_rate": 0.001,
"loss": 2.1097,
"step": 89300
},
{
"epoch": 28.89463477698772,
"grad_norm": 1.2061901092529297,
"learning_rate": 0.001,
"loss": 2.1001,
"step": 89400
},
{
"epoch": 28.926955397543633,
"grad_norm": 1.1980839967727661,
"learning_rate": 0.001,
"loss": 2.1211,
"step": 89500
},
{
"epoch": 28.959276018099548,
"grad_norm": 1.0727787017822266,
"learning_rate": 0.001,
"loss": 2.1142,
"step": 89600
},
{
"epoch": 28.991596638655462,
"grad_norm": 1.0199949741363525,
"learning_rate": 0.001,
"loss": 2.1295,
"step": 89700
},
{
"epoch": 29.023917259211377,
"grad_norm": 1.209977626800537,
"learning_rate": 0.001,
"loss": 2.035,
"step": 89800
},
{
"epoch": 29.05623787976729,
"grad_norm": 1.1219505071640015,
"learning_rate": 0.001,
"loss": 1.9767,
"step": 89900
},
{
"epoch": 29.088558500323206,
"grad_norm": 1.0901050567626953,
"learning_rate": 0.001,
"loss": 1.9869,
"step": 90000
},
{
"epoch": 29.12087912087912,
"grad_norm": 1.3818155527114868,
"learning_rate": 0.001,
"loss": 2.0048,
"step": 90100
},
{
"epoch": 29.153199741435035,
"grad_norm": 1.2119892835617065,
"learning_rate": 0.001,
"loss": 2.0016,
"step": 90200
},
{
"epoch": 29.18552036199095,
"grad_norm": 1.0864521265029907,
"learning_rate": 0.001,
"loss": 2.0227,
"step": 90300
},
{
"epoch": 29.217840982546864,
"grad_norm": 1.1717758178710938,
"learning_rate": 0.001,
"loss": 2.0127,
"step": 90400
},
{
"epoch": 29.25016160310278,
"grad_norm": 1.2278838157653809,
"learning_rate": 0.001,
"loss": 2.0122,
"step": 90500
},
{
"epoch": 29.282482223658693,
"grad_norm": 1.002016544342041,
"learning_rate": 0.001,
"loss": 2.02,
"step": 90600
},
{
"epoch": 29.314802844214608,
"grad_norm": 1.3417019844055176,
"learning_rate": 0.001,
"loss": 2.0371,
"step": 90700
},
{
"epoch": 29.347123464770522,
"grad_norm": 0.9367550611495972,
"learning_rate": 0.001,
"loss": 2.0433,
"step": 90800
},
{
"epoch": 29.379444085326437,
"grad_norm": 0.9104499220848083,
"learning_rate": 0.001,
"loss": 2.0271,
"step": 90900
},
{
"epoch": 29.41176470588235,
"grad_norm": 1.0613069534301758,
"learning_rate": 0.001,
"loss": 2.0518,
"step": 91000
},
{
"epoch": 29.444085326438266,
"grad_norm": 0.9659221768379211,
"learning_rate": 0.001,
"loss": 2.0541,
"step": 91100
},
{
"epoch": 29.47640594699418,
"grad_norm": 1.1530755758285522,
"learning_rate": 0.001,
"loss": 2.0486,
"step": 91200
},
{
"epoch": 29.5087265675501,
"grad_norm": 1.2209970951080322,
"learning_rate": 0.001,
"loss": 2.0591,
"step": 91300
},
{
"epoch": 29.541047188106013,
"grad_norm": 1.1463372707366943,
"learning_rate": 0.001,
"loss": 2.0773,
"step": 91400
},
{
"epoch": 29.573367808661928,
"grad_norm": 1.2101801633834839,
"learning_rate": 0.001,
"loss": 2.0641,
"step": 91500
},
{
"epoch": 29.605688429217842,
"grad_norm": 1.11674165725708,
"learning_rate": 0.001,
"loss": 2.0439,
"step": 91600
},
{
"epoch": 29.638009049773757,
"grad_norm": 0.9217492938041687,
"learning_rate": 0.001,
"loss": 2.0683,
"step": 91700
},
{
"epoch": 29.67032967032967,
"grad_norm": 0.9988812804222107,
"learning_rate": 0.001,
"loss": 2.0841,
"step": 91800
},
{
"epoch": 29.702650290885586,
"grad_norm": 1.1292462348937988,
"learning_rate": 0.001,
"loss": 2.086,
"step": 91900
},
{
"epoch": 29.7349709114415,
"grad_norm": 1.1154932975769043,
"learning_rate": 0.001,
"loss": 2.0777,
"step": 92000
},
{
"epoch": 29.767291531997415,
"grad_norm": 1.351989984512329,
"learning_rate": 0.001,
"loss": 2.0771,
"step": 92100
},
{
"epoch": 29.79961215255333,
"grad_norm": 1.0719629526138306,
"learning_rate": 0.001,
"loss": 2.0989,
"step": 92200
},
{
"epoch": 29.831932773109244,
"grad_norm": 1.2590843439102173,
"learning_rate": 0.001,
"loss": 2.0894,
"step": 92300
},
{
"epoch": 29.86425339366516,
"grad_norm": 1.1782222986221313,
"learning_rate": 0.001,
"loss": 2.0924,
"step": 92400
},
{
"epoch": 29.896574014221073,
"grad_norm": 1.0151180028915405,
"learning_rate": 0.001,
"loss": 2.0971,
"step": 92500
},
{
"epoch": 29.928894634776988,
"grad_norm": 1.0597012042999268,
"learning_rate": 0.001,
"loss": 2.0982,
"step": 92600
},
{
"epoch": 29.961215255332903,
"grad_norm": 1.2898638248443604,
"learning_rate": 0.001,
"loss": 2.0902,
"step": 92700
},
{
"epoch": 29.993535875888817,
"grad_norm": 1.1568880081176758,
"learning_rate": 0.001,
"loss": 2.1153,
"step": 92800
},
{
"epoch": 30.02585649644473,
"grad_norm": 1.0892854928970337,
"learning_rate": 0.001,
"loss": 1.9812,
"step": 92900
},
{
"epoch": 30.058177117000646,
"grad_norm": 1.165239691734314,
"learning_rate": 0.001,
"loss": 1.9629,
"step": 93000
},
{
"epoch": 30.09049773755656,
"grad_norm": 1.393893837928772,
"learning_rate": 0.001,
"loss": 1.9599,
"step": 93100
},
{
"epoch": 30.122818358112475,
"grad_norm": 0.9506264328956604,
"learning_rate": 0.001,
"loss": 1.9668,
"step": 93200
},
{
"epoch": 30.15513897866839,
"grad_norm": 1.2412270307540894,
"learning_rate": 0.001,
"loss": 1.9742,
"step": 93300
},
{
"epoch": 30.187459599224304,
"grad_norm": 1.3401048183441162,
"learning_rate": 0.001,
"loss": 2.0028,
"step": 93400
},
{
"epoch": 30.21978021978022,
"grad_norm": 1.3545461893081665,
"learning_rate": 0.001,
"loss": 1.9896,
"step": 93500
},
{
"epoch": 30.252100840336134,
"grad_norm": 1.7124943733215332,
"learning_rate": 0.001,
"loss": 2.0128,
"step": 93600
},
{
"epoch": 30.284421460892048,
"grad_norm": 1.5949537754058838,
"learning_rate": 0.001,
"loss": 2.0048,
"step": 93700
},
{
"epoch": 30.316742081447963,
"grad_norm": 1.0007027387619019,
"learning_rate": 0.001,
"loss": 1.986,
"step": 93800
},
{
"epoch": 30.349062702003877,
"grad_norm": 1.0602052211761475,
"learning_rate": 0.001,
"loss": 2.0132,
"step": 93900
},
{
"epoch": 30.381383322559792,
"grad_norm": 1.114917516708374,
"learning_rate": 0.001,
"loss": 2.0205,
"step": 94000
},
{
"epoch": 30.413703943115706,
"grad_norm": 1.368137240409851,
"learning_rate": 0.001,
"loss": 2.0398,
"step": 94100
},
{
"epoch": 30.44602456367162,
"grad_norm": 1.0994625091552734,
"learning_rate": 0.001,
"loss": 2.0159,
"step": 94200
},
{
"epoch": 30.478345184227535,
"grad_norm": 1.3286597728729248,
"learning_rate": 0.001,
"loss": 2.0524,
"step": 94300
},
{
"epoch": 30.51066580478345,
"grad_norm": 1.1345680952072144,
"learning_rate": 0.001,
"loss": 2.0544,
"step": 94400
},
{
"epoch": 30.542986425339368,
"grad_norm": 1.373855710029602,
"learning_rate": 0.001,
"loss": 2.0427,
"step": 94500
},
{
"epoch": 30.575307045895283,
"grad_norm": 1.3303792476654053,
"learning_rate": 0.001,
"loss": 2.053,
"step": 94600
},
{
"epoch": 30.607627666451197,
"grad_norm": 1.0612677335739136,
"learning_rate": 0.001,
"loss": 2.0572,
"step": 94700
},
{
"epoch": 30.639948287007112,
"grad_norm": 1.22883939743042,
"learning_rate": 0.001,
"loss": 2.0184,
"step": 94800
},
{
"epoch": 30.672268907563026,
"grad_norm": 1.1892231702804565,
"learning_rate": 0.001,
"loss": 2.0367,
"step": 94900
},
{
"epoch": 30.70458952811894,
"grad_norm": 1.2804641723632812,
"learning_rate": 0.001,
"loss": 2.0526,
"step": 95000
},
{
"epoch": 30.736910148674855,
"grad_norm": 1.4119378328323364,
"learning_rate": 0.001,
"loss": 2.0782,
"step": 95100
},
{
"epoch": 30.76923076923077,
"grad_norm": 1.3576971292495728,
"learning_rate": 0.001,
"loss": 2.0558,
"step": 95200
},
{
"epoch": 30.801551389786685,
"grad_norm": 1.6085991859436035,
"learning_rate": 0.001,
"loss": 2.0634,
"step": 95300
},
{
"epoch": 30.8338720103426,
"grad_norm": 1.2733300924301147,
"learning_rate": 0.001,
"loss": 2.0831,
"step": 95400
},
{
"epoch": 30.866192630898514,
"grad_norm": 1.2089987993240356,
"learning_rate": 0.001,
"loss": 2.0758,
"step": 95500
},
{
"epoch": 30.89851325145443,
"grad_norm": 1.0193136930465698,
"learning_rate": 0.001,
"loss": 2.0593,
"step": 95600
},
{
"epoch": 30.930833872010343,
"grad_norm": 1.3752954006195068,
"learning_rate": 0.001,
"loss": 2.0988,
"step": 95700
},
{
"epoch": 30.963154492566257,
"grad_norm": 1.3775840997695923,
"learning_rate": 0.001,
"loss": 2.0713,
"step": 95800
},
{
"epoch": 30.995475113122172,
"grad_norm": 1.3531081676483154,
"learning_rate": 0.001,
"loss": 2.0581,
"step": 95900
},
{
"epoch": 31.027795733678087,
"grad_norm": 1.4643930196762085,
"learning_rate": 0.001,
"loss": 1.9396,
"step": 96000
},
{
"epoch": 31.060116354234,
"grad_norm": 1.622295618057251,
"learning_rate": 0.001,
"loss": 1.9453,
"step": 96100
},
{
"epoch": 31.092436974789916,
"grad_norm": 1.37124764919281,
"learning_rate": 0.001,
"loss": 1.9667,
"step": 96200
},
{
"epoch": 31.12475759534583,
"grad_norm": 1.5041195154190063,
"learning_rate": 0.001,
"loss": 1.9684,
"step": 96300
},
{
"epoch": 31.157078215901745,
"grad_norm": 1.3748421669006348,
"learning_rate": 0.001,
"loss": 1.9641,
"step": 96400
},
{
"epoch": 31.18939883645766,
"grad_norm": 1.1635115146636963,
"learning_rate": 0.001,
"loss": 1.9473,
"step": 96500
},
{
"epoch": 31.221719457013574,
"grad_norm": 1.4115711450576782,
"learning_rate": 0.001,
"loss": 1.978,
"step": 96600
},
{
"epoch": 31.25404007756949,
"grad_norm": 1.1416429281234741,
"learning_rate": 0.001,
"loss": 1.9737,
"step": 96700
},
{
"epoch": 31.286360698125403,
"grad_norm": 1.2036595344543457,
"learning_rate": 0.001,
"loss": 1.9879,
"step": 96800
},
{
"epoch": 31.318681318681318,
"grad_norm": 1.4464876651763916,
"learning_rate": 0.001,
"loss": 1.9954,
"step": 96900
},
{
"epoch": 31.351001939237232,
"grad_norm": 1.212299108505249,
"learning_rate": 0.001,
"loss": 1.9863,
"step": 97000
},
{
"epoch": 31.383322559793147,
"grad_norm": 1.3392387628555298,
"learning_rate": 0.001,
"loss": 1.988,
"step": 97100
},
{
"epoch": 31.41564318034906,
"grad_norm": 1.2326029539108276,
"learning_rate": 0.001,
"loss": 2.0119,
"step": 97200
},
{
"epoch": 31.447963800904976,
"grad_norm": 1.7774823904037476,
"learning_rate": 0.001,
"loss": 2.0086,
"step": 97300
},
{
"epoch": 31.48028442146089,
"grad_norm": 1.0509564876556396,
"learning_rate": 0.001,
"loss": 2.0065,
"step": 97400
},
{
"epoch": 31.51260504201681,
"grad_norm": 1.7675199508666992,
"learning_rate": 0.001,
"loss": 2.0276,
"step": 97500
},
{
"epoch": 31.544925662572723,
"grad_norm": 1.4120814800262451,
"learning_rate": 0.001,
"loss": 2.0211,
"step": 97600
},
{
"epoch": 31.577246283128638,
"grad_norm": 1.2935658693313599,
"learning_rate": 0.001,
"loss": 2.0228,
"step": 97700
},
{
"epoch": 31.609566903684552,
"grad_norm": 1.2025492191314697,
"learning_rate": 0.001,
"loss": 2.0197,
"step": 97800
},
{
"epoch": 31.641887524240467,
"grad_norm": 1.3041045665740967,
"learning_rate": 0.001,
"loss": 2.0296,
"step": 97900
},
{
"epoch": 31.67420814479638,
"grad_norm": 1.280734658241272,
"learning_rate": 0.001,
"loss": 2.0354,
"step": 98000
},
{
"epoch": 31.706528765352296,
"grad_norm": 1.2508794069290161,
"learning_rate": 0.001,
"loss": 2.0405,
"step": 98100
},
{
"epoch": 31.73884938590821,
"grad_norm": 1.4229962825775146,
"learning_rate": 0.001,
"loss": 2.0246,
"step": 98200
},
{
"epoch": 31.771170006464125,
"grad_norm": 1.2830511331558228,
"learning_rate": 0.001,
"loss": 2.0466,
"step": 98300
},
{
"epoch": 31.80349062702004,
"grad_norm": 1.3752238750457764,
"learning_rate": 0.001,
"loss": 2.0429,
"step": 98400
},
{
"epoch": 31.835811247575954,
"grad_norm": 1.176871657371521,
"learning_rate": 0.001,
"loss": 2.0491,
"step": 98500
},
{
"epoch": 31.86813186813187,
"grad_norm": 1.1471308469772339,
"learning_rate": 0.001,
"loss": 2.0492,
"step": 98600
},
{
"epoch": 31.900452488687783,
"grad_norm": 1.2065348625183105,
"learning_rate": 0.001,
"loss": 2.0649,
"step": 98700
},
{
"epoch": 31.932773109243698,
"grad_norm": 1.600611925125122,
"learning_rate": 0.001,
"loss": 2.0569,
"step": 98800
},
{
"epoch": 31.965093729799612,
"grad_norm": 1.4030392169952393,
"learning_rate": 0.001,
"loss": 2.049,
"step": 98900
},
{
"epoch": 31.997414350355527,
"grad_norm": 1.5414460897445679,
"learning_rate": 0.001,
"loss": 2.0468,
"step": 99000
},
{
"epoch": 32.02973497091144,
"grad_norm": 1.6330195665359497,
"learning_rate": 0.001,
"loss": 1.9566,
"step": 99100
},
{
"epoch": 32.062055591467356,
"grad_norm": 1.6512374877929688,
"learning_rate": 0.001,
"loss": 1.9254,
"step": 99200
},
{
"epoch": 32.09437621202327,
"grad_norm": 1.8136323690414429,
"learning_rate": 0.001,
"loss": 1.9135,
"step": 99300
},
{
"epoch": 32.126696832579185,
"grad_norm": 1.2204009294509888,
"learning_rate": 0.001,
"loss": 1.9369,
"step": 99400
},
{
"epoch": 32.1590174531351,
"grad_norm": 1.2248504161834717,
"learning_rate": 0.001,
"loss": 1.9482,
"step": 99500
},
{
"epoch": 32.191338073691014,
"grad_norm": 1.2195435762405396,
"learning_rate": 0.001,
"loss": 1.9549,
"step": 99600
},
{
"epoch": 32.22365869424693,
"grad_norm": 1.4469695091247559,
"learning_rate": 0.001,
"loss": 1.9544,
"step": 99700
},
{
"epoch": 32.25597931480284,
"grad_norm": 1.5263967514038086,
"learning_rate": 0.001,
"loss": 1.9625,
"step": 99800
},
{
"epoch": 32.28829993535876,
"grad_norm": 1.549071192741394,
"learning_rate": 0.001,
"loss": 1.9679,
"step": 99900
},
{
"epoch": 32.32062055591467,
"grad_norm": 1.6799860000610352,
"learning_rate": 0.001,
"loss": 1.9806,
"step": 100000
},
{
"epoch": 32.35294117647059,
"grad_norm": 1.532671332359314,
"learning_rate": 0.001,
"loss": 1.9814,
"step": 100100
},
{
"epoch": 32.3852617970265,
"grad_norm": 1.2696813344955444,
"learning_rate": 0.001,
"loss": 1.9811,
"step": 100200
},
{
"epoch": 32.417582417582416,
"grad_norm": 1.7718359231948853,
"learning_rate": 0.001,
"loss": 1.9756,
"step": 100300
},
{
"epoch": 32.44990303813833,
"grad_norm": 1.7090567350387573,
"learning_rate": 0.001,
"loss": 1.9795,
"step": 100400
},
{
"epoch": 32.482223658694245,
"grad_norm": 1.541060209274292,
"learning_rate": 0.001,
"loss": 1.9743,
"step": 100500
},
{
"epoch": 32.51454427925016,
"grad_norm": 1.5212962627410889,
"learning_rate": 0.001,
"loss": 1.9764,
"step": 100600
},
{
"epoch": 32.546864899806074,
"grad_norm": 1.3458622694015503,
"learning_rate": 0.001,
"loss": 1.9994,
"step": 100700
},
{
"epoch": 32.57918552036199,
"grad_norm": 1.728844404220581,
"learning_rate": 0.001,
"loss": 1.9841,
"step": 100800
},
{
"epoch": 32.6115061409179,
"grad_norm": 1.5122705698013306,
"learning_rate": 0.001,
"loss": 1.9987,
"step": 100900
},
{
"epoch": 32.64382676147382,
"grad_norm": 1.3499188423156738,
"learning_rate": 0.001,
"loss": 2.0089,
"step": 101000
},
{
"epoch": 32.67614738202973,
"grad_norm": 1.3995920419692993,
"learning_rate": 0.001,
"loss": 1.9964,
"step": 101100
},
{
"epoch": 32.70846800258565,
"grad_norm": 1.3177103996276855,
"learning_rate": 0.001,
"loss": 2.016,
"step": 101200
},
{
"epoch": 32.74078862314156,
"grad_norm": 1.4601622819900513,
"learning_rate": 0.001,
"loss": 2.0178,
"step": 101300
},
{
"epoch": 32.773109243697476,
"grad_norm": 1.145120620727539,
"learning_rate": 0.001,
"loss": 2.0273,
"step": 101400
},
{
"epoch": 32.80542986425339,
"grad_norm": 1.5704030990600586,
"learning_rate": 0.001,
"loss": 2.0331,
"step": 101500
},
{
"epoch": 32.837750484809305,
"grad_norm": 1.4068289995193481,
"learning_rate": 0.001,
"loss": 2.0277,
"step": 101600
},
{
"epoch": 32.87007110536522,
"grad_norm": 1.481192946434021,
"learning_rate": 0.001,
"loss": 2.0254,
"step": 101700
},
{
"epoch": 32.902391725921134,
"grad_norm": 1.383341670036316,
"learning_rate": 0.001,
"loss": 2.0301,
"step": 101800
},
{
"epoch": 32.93471234647705,
"grad_norm": 1.6861649751663208,
"learning_rate": 0.001,
"loss": 2.0389,
"step": 101900
},
{
"epoch": 32.967032967032964,
"grad_norm": 1.4227408170700073,
"learning_rate": 0.001,
"loss": 2.0338,
"step": 102000
},
{
"epoch": 32.999353587588885,
"grad_norm": 1.8178858757019043,
"learning_rate": 0.001,
"loss": 2.0282,
"step": 102100
},
{
"epoch": 33.0316742081448,
"grad_norm": 1.7232046127319336,
"learning_rate": 0.001,
"loss": 1.9,
"step": 102200
},
{
"epoch": 33.063994828700714,
"grad_norm": 1.6832845211029053,
"learning_rate": 0.001,
"loss": 1.889,
"step": 102300
},
{
"epoch": 33.09631544925663,
"grad_norm": 1.3407903909683228,
"learning_rate": 0.001,
"loss": 1.9024,
"step": 102400
},
{
"epoch": 33.12863606981254,
"grad_norm": 1.867085337638855,
"learning_rate": 0.001,
"loss": 1.9042,
"step": 102500
},
{
"epoch": 33.16095669036846,
"grad_norm": 1.544847846031189,
"learning_rate": 0.001,
"loss": 1.9177,
"step": 102600
},
{
"epoch": 33.19327731092437,
"grad_norm": 1.7738703489303589,
"learning_rate": 0.001,
"loss": 1.9193,
"step": 102700
},
{
"epoch": 33.22559793148029,
"grad_norm": 1.6412625312805176,
"learning_rate": 0.001,
"loss": 1.9312,
"step": 102800
},
{
"epoch": 33.2579185520362,
"grad_norm": 1.9249825477600098,
"learning_rate": 0.001,
"loss": 1.9378,
"step": 102900
},
{
"epoch": 33.290239172592116,
"grad_norm": 1.7050279378890991,
"learning_rate": 0.001,
"loss": 1.9439,
"step": 103000
},
{
"epoch": 33.32255979314803,
"grad_norm": 1.336020588874817,
"learning_rate": 0.001,
"loss": 1.939,
"step": 103100
},
{
"epoch": 33.354880413703945,
"grad_norm": 1.7792845964431763,
"learning_rate": 0.001,
"loss": 1.956,
"step": 103200
},
{
"epoch": 33.38720103425986,
"grad_norm": 1.70310640335083,
"learning_rate": 0.001,
"loss": 1.9585,
"step": 103300
},
{
"epoch": 33.419521654815775,
"grad_norm": 1.6857690811157227,
"learning_rate": 0.001,
"loss": 1.9793,
"step": 103400
},
{
"epoch": 33.45184227537169,
"grad_norm": 1.8818587064743042,
"learning_rate": 0.001,
"loss": 1.9775,
"step": 103500
},
{
"epoch": 33.484162895927604,
"grad_norm": 1.6532014608383179,
"learning_rate": 0.001,
"loss": 1.9674,
"step": 103600
},
{
"epoch": 33.51648351648352,
"grad_norm": 1.7631884813308716,
"learning_rate": 0.001,
"loss": 1.9702,
"step": 103700
},
{
"epoch": 33.54880413703943,
"grad_norm": 1.3442251682281494,
"learning_rate": 0.001,
"loss": 1.9769,
"step": 103800
},
{
"epoch": 33.58112475759535,
"grad_norm": 1.660846471786499,
"learning_rate": 0.001,
"loss": 1.9718,
"step": 103900
},
{
"epoch": 33.61344537815126,
"grad_norm": 1.6144530773162842,
"learning_rate": 0.001,
"loss": 1.9929,
"step": 104000
},
{
"epoch": 33.645765998707176,
"grad_norm": 1.7088271379470825,
"learning_rate": 0.001,
"loss": 1.9805,
"step": 104100
},
{
"epoch": 33.67808661926309,
"grad_norm": 1.7659376859664917,
"learning_rate": 0.001,
"loss": 1.9927,
"step": 104200
},
{
"epoch": 33.710407239819006,
"grad_norm": 1.9894053936004639,
"learning_rate": 0.001,
"loss": 1.988,
"step": 104300
},
{
"epoch": 33.74272786037492,
"grad_norm": 1.8328500986099243,
"learning_rate": 0.001,
"loss": 2.0087,
"step": 104400
},
{
"epoch": 33.775048480930835,
"grad_norm": 1.4711683988571167,
"learning_rate": 0.001,
"loss": 2.0038,
"step": 104500
},
{
"epoch": 33.80736910148675,
"grad_norm": 1.4590061902999878,
"learning_rate": 0.001,
"loss": 1.9973,
"step": 104600
},
{
"epoch": 33.839689722042664,
"grad_norm": 1.496267318725586,
"learning_rate": 0.001,
"loss": 2.016,
"step": 104700
},
{
"epoch": 33.87201034259858,
"grad_norm": 1.4315950870513916,
"learning_rate": 0.001,
"loss": 2.0161,
"step": 104800
},
{
"epoch": 33.90433096315449,
"grad_norm": 1.8336668014526367,
"learning_rate": 0.001,
"loss": 2.0096,
"step": 104900
},
{
"epoch": 33.93665158371041,
"grad_norm": 1.4270433187484741,
"learning_rate": 0.001,
"loss": 2.0209,
"step": 105000
},
{
"epoch": 33.96897220426632,
"grad_norm": 1.973786473274231,
"learning_rate": 0.001,
"loss": 2.0477,
"step": 105100
},
{
"epoch": 34.00129282482224,
"grad_norm": 1.3552184104919434,
"learning_rate": 0.001,
"loss": 2.0127,
"step": 105200
},
{
"epoch": 34.03361344537815,
"grad_norm": 1.5386784076690674,
"learning_rate": 0.001,
"loss": 1.8884,
"step": 105300
},
{
"epoch": 34.065934065934066,
"grad_norm": 1.4204431772232056,
"learning_rate": 0.001,
"loss": 1.8889,
"step": 105400
},
{
"epoch": 34.09825468648998,
"grad_norm": 1.4803218841552734,
"learning_rate": 0.001,
"loss": 1.8837,
"step": 105500
},
{
"epoch": 34.130575307045895,
"grad_norm": 1.8072541952133179,
"learning_rate": 0.001,
"loss": 1.909,
"step": 105600
},
{
"epoch": 34.16289592760181,
"grad_norm": 1.662071943283081,
"learning_rate": 0.001,
"loss": 1.9122,
"step": 105700
},
{
"epoch": 34.195216548157724,
"grad_norm": 1.8705205917358398,
"learning_rate": 0.001,
"loss": 1.9154,
"step": 105800
},
{
"epoch": 34.22753716871364,
"grad_norm": 1.4515314102172852,
"learning_rate": 0.001,
"loss": 1.9121,
"step": 105900
},
{
"epoch": 34.25985778926955,
"grad_norm": 1.758162260055542,
"learning_rate": 0.001,
"loss": 1.9201,
"step": 106000
},
{
"epoch": 34.29217840982547,
"grad_norm": 1.2175607681274414,
"learning_rate": 0.001,
"loss": 1.9219,
"step": 106100
},
{
"epoch": 34.32449903038138,
"grad_norm": 1.7785415649414062,
"learning_rate": 0.001,
"loss": 1.9343,
"step": 106200
},
{
"epoch": 34.3568196509373,
"grad_norm": 1.673640489578247,
"learning_rate": 0.001,
"loss": 1.9479,
"step": 106300
},
{
"epoch": 34.38914027149321,
"grad_norm": 1.2646243572235107,
"learning_rate": 0.001,
"loss": 1.9262,
"step": 106400
},
{
"epoch": 34.421460892049126,
"grad_norm": 1.467193841934204,
"learning_rate": 0.001,
"loss": 1.9508,
"step": 106500
},
{
"epoch": 34.45378151260504,
"grad_norm": 1.7420579195022583,
"learning_rate": 0.001,
"loss": 1.943,
"step": 106600
},
{
"epoch": 34.486102133160955,
"grad_norm": 1.6206345558166504,
"learning_rate": 0.001,
"loss": 1.9342,
"step": 106700
},
{
"epoch": 34.51842275371687,
"grad_norm": 1.926108479499817,
"learning_rate": 0.001,
"loss": 1.9569,
"step": 106800
},
{
"epoch": 34.550743374272784,
"grad_norm": 1.3895916938781738,
"learning_rate": 0.001,
"loss": 1.9495,
"step": 106900
},
{
"epoch": 34.5830639948287,
"grad_norm": 1.3170398473739624,
"learning_rate": 0.001,
"loss": 1.9565,
"step": 107000
},
{
"epoch": 34.61538461538461,
"grad_norm": 1.467783808708191,
"learning_rate": 0.001,
"loss": 1.976,
"step": 107100
},
{
"epoch": 34.64770523594053,
"grad_norm": 1.54426908493042,
"learning_rate": 0.001,
"loss": 1.9787,
"step": 107200
},
{
"epoch": 34.68002585649644,
"grad_norm": 1.6267613172531128,
"learning_rate": 0.001,
"loss": 1.971,
"step": 107300
},
{
"epoch": 34.71234647705236,
"grad_norm": 1.6680964231491089,
"learning_rate": 0.001,
"loss": 1.9693,
"step": 107400
},
{
"epoch": 34.74466709760827,
"grad_norm": 2.0082974433898926,
"learning_rate": 0.001,
"loss": 1.9713,
"step": 107500
},
{
"epoch": 34.776987718164186,
"grad_norm": 1.5093539953231812,
"learning_rate": 0.001,
"loss": 1.9793,
"step": 107600
},
{
"epoch": 34.8093083387201,
"grad_norm": 1.2009838819503784,
"learning_rate": 0.001,
"loss": 1.9879,
"step": 107700
},
{
"epoch": 34.841628959276015,
"grad_norm": 1.7011646032333374,
"learning_rate": 0.001,
"loss": 1.984,
"step": 107800
},
{
"epoch": 34.87394957983193,
"grad_norm": 2.0037291049957275,
"learning_rate": 0.001,
"loss": 1.9889,
"step": 107900
},
{
"epoch": 34.906270200387844,
"grad_norm": 1.5127506256103516,
"learning_rate": 0.001,
"loss": 1.9867,
"step": 108000
},
{
"epoch": 34.93859082094376,
"grad_norm": 1.4229438304901123,
"learning_rate": 0.001,
"loss": 2.0055,
"step": 108100
},
{
"epoch": 34.97091144149967,
"grad_norm": 1.0902063846588135,
"learning_rate": 0.001,
"loss": 2.0062,
"step": 108200
},
{
"epoch": 35.003232062055595,
"grad_norm": 1.591429352760315,
"learning_rate": 0.001,
"loss": 1.9947,
"step": 108300
},
{
"epoch": 35.03555268261151,
"grad_norm": 1.3210487365722656,
"learning_rate": 0.001,
"loss": 1.8631,
"step": 108400
},
{
"epoch": 35.067873303167424,
"grad_norm": 1.823500394821167,
"learning_rate": 0.001,
"loss": 1.8618,
"step": 108500
},
{
"epoch": 35.10019392372334,
"grad_norm": 1.0470653772354126,
"learning_rate": 0.001,
"loss": 1.8604,
"step": 108600
},
{
"epoch": 35.13251454427925,
"grad_norm": 1.7314393520355225,
"learning_rate": 0.001,
"loss": 1.8956,
"step": 108700
},
{
"epoch": 35.16483516483517,
"grad_norm": 1.1611652374267578,
"learning_rate": 0.001,
"loss": 1.8885,
"step": 108800
},
{
"epoch": 35.19715578539108,
"grad_norm": 1.1454203128814697,
"learning_rate": 0.001,
"loss": 1.906,
"step": 108900
},
{
"epoch": 35.229476405947,
"grad_norm": 1.153507947921753,
"learning_rate": 0.001,
"loss": 1.8947,
"step": 109000
},
{
"epoch": 35.26179702650291,
"grad_norm": 1.3420323133468628,
"learning_rate": 0.001,
"loss": 1.92,
"step": 109100
},
{
"epoch": 35.294117647058826,
"grad_norm": 1.4737643003463745,
"learning_rate": 0.001,
"loss": 1.9021,
"step": 109200
},
{
"epoch": 35.32643826761474,
"grad_norm": 0.9562607407569885,
"learning_rate": 0.001,
"loss": 1.9076,
"step": 109300
},
{
"epoch": 35.358758888170655,
"grad_norm": 1.198275089263916,
"learning_rate": 0.001,
"loss": 1.9077,
"step": 109400
},
{
"epoch": 35.39107950872657,
"grad_norm": 1.1489148139953613,
"learning_rate": 0.001,
"loss": 1.9048,
"step": 109500
},
{
"epoch": 35.423400129282484,
"grad_norm": 2.1524202823638916,
"learning_rate": 0.001,
"loss": 1.9074,
"step": 109600
},
{
"epoch": 35.4557207498384,
"grad_norm": 1.2565486431121826,
"learning_rate": 0.001,
"loss": 1.9379,
"step": 109700
},
{
"epoch": 35.48804137039431,
"grad_norm": 1.3871768712997437,
"learning_rate": 0.001,
"loss": 1.9275,
"step": 109800
},
{
"epoch": 35.52036199095023,
"grad_norm": 1.1737405061721802,
"learning_rate": 0.001,
"loss": 1.9491,
"step": 109900
},
{
"epoch": 35.55268261150614,
"grad_norm": 1.262061357498169,
"learning_rate": 0.001,
"loss": 1.9558,
"step": 110000
},
{
"epoch": 35.58500323206206,
"grad_norm": 1.2257012128829956,
"learning_rate": 0.001,
"loss": 1.9361,
"step": 110100
},
{
"epoch": 35.61732385261797,
"grad_norm": 1.172411561012268,
"learning_rate": 0.001,
"loss": 1.9339,
"step": 110200
},
{
"epoch": 35.649644473173886,
"grad_norm": 1.9752097129821777,
"learning_rate": 0.001,
"loss": 1.9706,
"step": 110300
},
{
"epoch": 35.6819650937298,
"grad_norm": 1.097636103630066,
"learning_rate": 0.001,
"loss": 1.9552,
"step": 110400
},
{
"epoch": 35.714285714285715,
"grad_norm": 1.2427012920379639,
"learning_rate": 0.001,
"loss": 1.9491,
"step": 110500
},
{
"epoch": 35.74660633484163,
"grad_norm": 1.4779609441757202,
"learning_rate": 0.001,
"loss": 1.9561,
"step": 110600
},
{
"epoch": 35.778926955397544,
"grad_norm": 1.4894160032272339,
"learning_rate": 0.001,
"loss": 1.9566,
"step": 110700
},
{
"epoch": 35.81124757595346,
"grad_norm": 1.5670257806777954,
"learning_rate": 0.001,
"loss": 1.9446,
"step": 110800
},
{
"epoch": 35.84356819650937,
"grad_norm": 1.08371102809906,
"learning_rate": 0.001,
"loss": 1.9911,
"step": 110900
},
{
"epoch": 35.87588881706529,
"grad_norm": 1.3103485107421875,
"learning_rate": 0.001,
"loss": 1.9627,
"step": 111000
},
{
"epoch": 35.9082094376212,
"grad_norm": 1.6154662370681763,
"learning_rate": 0.001,
"loss": 1.9735,
"step": 111100
},
{
"epoch": 35.94053005817712,
"grad_norm": 1.6229771375656128,
"learning_rate": 0.001,
"loss": 1.9785,
"step": 111200
},
{
"epoch": 35.97285067873303,
"grad_norm": 1.8419809341430664,
"learning_rate": 0.001,
"loss": 1.9811,
"step": 111300
},
{
"epoch": 36.005171299288946,
"grad_norm": 0.9819174408912659,
"learning_rate": 0.001,
"loss": 1.9883,
"step": 111400
},
{
"epoch": 36.03749191984486,
"grad_norm": 1.2235621213912964,
"learning_rate": 0.001,
"loss": 1.8435,
"step": 111500
},
{
"epoch": 36.069812540400775,
"grad_norm": 1.4074771404266357,
"learning_rate": 0.001,
"loss": 1.8563,
"step": 111600
},
{
"epoch": 36.10213316095669,
"grad_norm": 1.066227674484253,
"learning_rate": 0.001,
"loss": 1.8614,
"step": 111700
},
{
"epoch": 36.134453781512605,
"grad_norm": 1.0207966566085815,
"learning_rate": 0.001,
"loss": 1.8534,
"step": 111800
},
{
"epoch": 36.16677440206852,
"grad_norm": 1.1360985040664673,
"learning_rate": 0.001,
"loss": 1.8739,
"step": 111900
},
{
"epoch": 36.199095022624434,
"grad_norm": 0.963874876499176,
"learning_rate": 0.001,
"loss": 1.8814,
"step": 112000
},
{
"epoch": 36.23141564318035,
"grad_norm": 1.484297513961792,
"learning_rate": 0.001,
"loss": 1.8973,
"step": 112100
},
{
"epoch": 36.26373626373626,
"grad_norm": 1.1659079790115356,
"learning_rate": 0.001,
"loss": 1.8755,
"step": 112200
},
{
"epoch": 36.29605688429218,
"grad_norm": 1.292757511138916,
"learning_rate": 0.001,
"loss": 1.8907,
"step": 112300
},
{
"epoch": 36.32837750484809,
"grad_norm": 1.2694268226623535,
"learning_rate": 0.001,
"loss": 1.9052,
"step": 112400
},
{
"epoch": 36.36069812540401,
"grad_norm": 1.2071782350540161,
"learning_rate": 0.001,
"loss": 1.9001,
"step": 112500
},
{
"epoch": 36.39301874595992,
"grad_norm": 1.1544362306594849,
"learning_rate": 0.001,
"loss": 1.897,
"step": 112600
},
{
"epoch": 36.425339366515836,
"grad_norm": 1.2569639682769775,
"learning_rate": 0.001,
"loss": 1.8928,
"step": 112700
},
{
"epoch": 36.45765998707175,
"grad_norm": 1.5046840906143188,
"learning_rate": 0.001,
"loss": 1.9227,
"step": 112800
},
{
"epoch": 36.489980607627665,
"grad_norm": 1.3405776023864746,
"learning_rate": 0.001,
"loss": 1.931,
"step": 112900
},
{
"epoch": 36.52230122818358,
"grad_norm": 1.0977368354797363,
"learning_rate": 0.001,
"loss": 1.9025,
"step": 113000
},
{
"epoch": 36.554621848739494,
"grad_norm": 1.0207895040512085,
"learning_rate": 0.001,
"loss": 1.9232,
"step": 113100
},
{
"epoch": 36.58694246929541,
"grad_norm": 0.9704816937446594,
"learning_rate": 0.001,
"loss": 1.9225,
"step": 113200
},
{
"epoch": 36.61926308985132,
"grad_norm": 1.2907272577285767,
"learning_rate": 0.001,
"loss": 1.9476,
"step": 113300
},
{
"epoch": 36.65158371040724,
"grad_norm": 1.250844120979309,
"learning_rate": 0.001,
"loss": 1.9309,
"step": 113400
},
{
"epoch": 36.68390433096315,
"grad_norm": 1.179707646369934,
"learning_rate": 0.001,
"loss": 1.931,
"step": 113500
},
{
"epoch": 36.71622495151907,
"grad_norm": 1.287279725074768,
"learning_rate": 0.001,
"loss": 1.93,
"step": 113600
},
{
"epoch": 36.74854557207498,
"grad_norm": 1.0894107818603516,
"learning_rate": 0.001,
"loss": 1.9475,
"step": 113700
},
{
"epoch": 36.780866192630896,
"grad_norm": 1.185888648033142,
"learning_rate": 0.001,
"loss": 1.94,
"step": 113800
},
{
"epoch": 36.81318681318681,
"grad_norm": 1.1891213655471802,
"learning_rate": 0.001,
"loss": 1.9537,
"step": 113900
},
{
"epoch": 36.845507433742725,
"grad_norm": 1.5178064107894897,
"learning_rate": 0.001,
"loss": 1.9468,
"step": 114000
},
{
"epoch": 36.87782805429864,
"grad_norm": 1.2908250093460083,
"learning_rate": 0.001,
"loss": 1.9463,
"step": 114100
},
{
"epoch": 36.910148674854554,
"grad_norm": 1.1198320388793945,
"learning_rate": 0.001,
"loss": 1.9632,
"step": 114200
},
{
"epoch": 36.94246929541047,
"grad_norm": 1.1957751512527466,
"learning_rate": 0.001,
"loss": 1.9557,
"step": 114300
},
{
"epoch": 36.97478991596638,
"grad_norm": 1.448580265045166,
"learning_rate": 0.001,
"loss": 1.9608,
"step": 114400
},
{
"epoch": 37.007110536522305,
"grad_norm": 1.33165442943573,
"learning_rate": 0.001,
"loss": 1.9341,
"step": 114500
},
{
"epoch": 37.03943115707822,
"grad_norm": 1.670316219329834,
"learning_rate": 0.001,
"loss": 1.8265,
"step": 114600
},
{
"epoch": 37.071751777634134,
"grad_norm": 1.0109338760375977,
"learning_rate": 0.001,
"loss": 1.835,
"step": 114700
},
{
"epoch": 37.10407239819005,
"grad_norm": 1.2939589023590088,
"learning_rate": 0.001,
"loss": 1.8395,
"step": 114800
},
{
"epoch": 37.13639301874596,
"grad_norm": 1.208137035369873,
"learning_rate": 0.001,
"loss": 1.8421,
"step": 114900
},
{
"epoch": 37.16871363930188,
"grad_norm": 1.5535569190979004,
"learning_rate": 0.001,
"loss": 1.8426,
"step": 115000
},
{
"epoch": 37.20103425985779,
"grad_norm": 1.3210135698318481,
"learning_rate": 0.001,
"loss": 1.8498,
"step": 115100
},
{
"epoch": 37.23335488041371,
"grad_norm": 1.1312525272369385,
"learning_rate": 0.001,
"loss": 1.8843,
"step": 115200
},
{
"epoch": 37.26567550096962,
"grad_norm": 1.3623147010803223,
"learning_rate": 0.001,
"loss": 1.8632,
"step": 115300
},
{
"epoch": 37.297996121525536,
"grad_norm": 1.2084800004959106,
"learning_rate": 0.001,
"loss": 1.8752,
"step": 115400
},
{
"epoch": 37.33031674208145,
"grad_norm": 1.254078984260559,
"learning_rate": 0.001,
"loss": 1.8789,
"step": 115500
},
{
"epoch": 37.362637362637365,
"grad_norm": 1.6026792526245117,
"learning_rate": 0.001,
"loss": 1.8707,
"step": 115600
},
{
"epoch": 37.39495798319328,
"grad_norm": 1.4884287118911743,
"learning_rate": 0.001,
"loss": 1.8919,
"step": 115700
},
{
"epoch": 37.427278603749194,
"grad_norm": 1.3817373514175415,
"learning_rate": 0.001,
"loss": 1.9042,
"step": 115800
},
{
"epoch": 37.45959922430511,
"grad_norm": 1.039664626121521,
"learning_rate": 0.001,
"loss": 1.889,
"step": 115900
},
{
"epoch": 37.49191984486102,
"grad_norm": 1.3237055540084839,
"learning_rate": 0.001,
"loss": 1.9004,
"step": 116000
},
{
"epoch": 37.52424046541694,
"grad_norm": 1.173531413078308,
"learning_rate": 0.001,
"loss": 1.8903,
"step": 116100
},
{
"epoch": 37.55656108597285,
"grad_norm": 1.696763038635254,
"learning_rate": 0.001,
"loss": 1.8968,
"step": 116200
},
{
"epoch": 37.58888170652877,
"grad_norm": 1.1118332147598267,
"learning_rate": 0.001,
"loss": 1.9124,
"step": 116300
},
{
"epoch": 37.62120232708468,
"grad_norm": 1.162667989730835,
"learning_rate": 0.001,
"loss": 1.916,
"step": 116400
},
{
"epoch": 37.653522947640596,
"grad_norm": 1.2187628746032715,
"learning_rate": 0.001,
"loss": 1.9133,
"step": 116500
},
{
"epoch": 37.68584356819651,
"grad_norm": 1.3280999660491943,
"learning_rate": 0.001,
"loss": 1.9118,
"step": 116600
},
{
"epoch": 37.718164188752425,
"grad_norm": 1.41241455078125,
"learning_rate": 0.001,
"loss": 1.9308,
"step": 116700
},
{
"epoch": 37.75048480930834,
"grad_norm": 1.185173511505127,
"learning_rate": 0.001,
"loss": 1.9146,
"step": 116800
},
{
"epoch": 37.782805429864254,
"grad_norm": 1.3146615028381348,
"learning_rate": 0.001,
"loss": 1.9481,
"step": 116900
},
{
"epoch": 37.81512605042017,
"grad_norm": 1.258465051651001,
"learning_rate": 0.001,
"loss": 1.9228,
"step": 117000
},
{
"epoch": 37.84744667097608,
"grad_norm": 1.0832951068878174,
"learning_rate": 0.001,
"loss": 1.9423,
"step": 117100
},
{
"epoch": 37.879767291532,
"grad_norm": 1.3820182085037231,
"learning_rate": 0.001,
"loss": 1.9135,
"step": 117200
},
{
"epoch": 37.91208791208791,
"grad_norm": 1.4527101516723633,
"learning_rate": 0.001,
"loss": 1.9549,
"step": 117300
},
{
"epoch": 37.94440853264383,
"grad_norm": 1.1982343196868896,
"learning_rate": 0.001,
"loss": 1.9517,
"step": 117400
},
{
"epoch": 37.97672915319974,
"grad_norm": 1.3673944473266602,
"learning_rate": 0.001,
"loss": 1.9575,
"step": 117500
},
{
"epoch": 38.009049773755656,
"grad_norm": 1.5035516023635864,
"learning_rate": 0.001,
"loss": 1.9109,
"step": 117600
},
{
"epoch": 38.04137039431157,
"grad_norm": 1.4764431715011597,
"learning_rate": 0.001,
"loss": 1.821,
"step": 117700
},
{
"epoch": 38.073691014867485,
"grad_norm": 1.0759299993515015,
"learning_rate": 0.001,
"loss": 1.8158,
"step": 117800
},
{
"epoch": 38.1060116354234,
"grad_norm": 1.0882259607315063,
"learning_rate": 0.001,
"loss": 1.8302,
"step": 117900
},
{
"epoch": 38.138332255979314,
"grad_norm": 1.246886968612671,
"learning_rate": 0.001,
"loss": 1.8436,
"step": 118000
},
{
"epoch": 38.17065287653523,
"grad_norm": 1.2654645442962646,
"learning_rate": 0.001,
"loss": 1.8389,
"step": 118100
},
{
"epoch": 38.20297349709114,
"grad_norm": 1.3032474517822266,
"learning_rate": 0.001,
"loss": 1.8522,
"step": 118200
},
{
"epoch": 38.23529411764706,
"grad_norm": 1.343469500541687,
"learning_rate": 0.001,
"loss": 1.8295,
"step": 118300
},
{
"epoch": 38.26761473820297,
"grad_norm": 1.4752659797668457,
"learning_rate": 0.001,
"loss": 1.8532,
"step": 118400
},
{
"epoch": 38.29993535875889,
"grad_norm": 1.1999928951263428,
"learning_rate": 0.001,
"loss": 1.8559,
"step": 118500
},
{
"epoch": 38.3322559793148,
"grad_norm": 1.4642833471298218,
"learning_rate": 0.001,
"loss": 1.863,
"step": 118600
},
{
"epoch": 38.364576599870716,
"grad_norm": 1.3492789268493652,
"learning_rate": 0.001,
"loss": 1.8608,
"step": 118700
},
{
"epoch": 38.39689722042663,
"grad_norm": 1.05475652217865,
"learning_rate": 0.001,
"loss": 1.8726,
"step": 118800
},
{
"epoch": 38.429217840982545,
"grad_norm": 1.0771496295928955,
"learning_rate": 0.001,
"loss": 1.8648,
"step": 118900
},
{
"epoch": 38.46153846153846,
"grad_norm": 1.2053498029708862,
"learning_rate": 0.001,
"loss": 1.8638,
"step": 119000
},
{
"epoch": 38.493859082094374,
"grad_norm": 1.025508165359497,
"learning_rate": 0.001,
"loss": 1.8684,
"step": 119100
},
{
"epoch": 38.52617970265029,
"grad_norm": 1.3343952894210815,
"learning_rate": 0.001,
"loss": 1.8904,
"step": 119200
},
{
"epoch": 38.558500323206204,
"grad_norm": 1.3081802129745483,
"learning_rate": 0.001,
"loss": 1.8965,
"step": 119300
},
{
"epoch": 38.59082094376212,
"grad_norm": 1.4606342315673828,
"learning_rate": 0.001,
"loss": 1.8811,
"step": 119400
},
{
"epoch": 38.62314156431803,
"grad_norm": 1.606359839439392,
"learning_rate": 0.001,
"loss": 1.8923,
"step": 119500
},
{
"epoch": 38.65546218487395,
"grad_norm": 1.120044469833374,
"learning_rate": 0.001,
"loss": 1.8905,
"step": 119600
},
{
"epoch": 38.68778280542986,
"grad_norm": 1.1815135478973389,
"learning_rate": 0.001,
"loss": 1.9041,
"step": 119700
},
{
"epoch": 38.720103425985776,
"grad_norm": 1.11037015914917,
"learning_rate": 0.001,
"loss": 1.9106,
"step": 119800
},
{
"epoch": 38.75242404654169,
"grad_norm": 1.1923435926437378,
"learning_rate": 0.001,
"loss": 1.9001,
"step": 119900
},
{
"epoch": 38.784744667097605,
"grad_norm": 1.1960541009902954,
"learning_rate": 0.001,
"loss": 1.9118,
"step": 120000
},
{
"epoch": 38.81706528765352,
"grad_norm": 1.290960669517517,
"learning_rate": 0.001,
"loss": 1.8994,
"step": 120100
},
{
"epoch": 38.849385908209435,
"grad_norm": 1.4803295135498047,
"learning_rate": 0.001,
"loss": 1.9312,
"step": 120200
},
{
"epoch": 38.88170652876535,
"grad_norm": 1.1313281059265137,
"learning_rate": 0.001,
"loss": 1.9301,
"step": 120300
},
{
"epoch": 38.914027149321264,
"grad_norm": 1.1400024890899658,
"learning_rate": 0.001,
"loss": 1.9282,
"step": 120400
},
{
"epoch": 38.94634776987718,
"grad_norm": 1.714592695236206,
"learning_rate": 0.001,
"loss": 1.9147,
"step": 120500
},
{
"epoch": 38.97866839043309,
"grad_norm": 1.5026122331619263,
"learning_rate": 0.001,
"loss": 1.9386,
"step": 120600
},
{
"epoch": 39.010989010989015,
"grad_norm": 1.3159358501434326,
"learning_rate": 0.001,
"loss": 1.8787,
"step": 120700
},
{
"epoch": 39.04330963154493,
"grad_norm": 1.3071198463439941,
"learning_rate": 0.001,
"loss": 1.7981,
"step": 120800
},
{
"epoch": 39.075630252100844,
"grad_norm": 1.2957608699798584,
"learning_rate": 0.001,
"loss": 1.809,
"step": 120900
},
{
"epoch": 39.10795087265676,
"grad_norm": 1.5660688877105713,
"learning_rate": 0.001,
"loss": 1.8187,
"step": 121000
},
{
"epoch": 39.14027149321267,
"grad_norm": 1.9384124279022217,
"learning_rate": 0.001,
"loss": 1.8083,
"step": 121100
},
{
"epoch": 39.17259211376859,
"grad_norm": 1.2615103721618652,
"learning_rate": 0.001,
"loss": 1.8053,
"step": 121200
},
{
"epoch": 39.2049127343245,
"grad_norm": 1.5705621242523193,
"learning_rate": 0.001,
"loss": 1.8048,
"step": 121300
},
{
"epoch": 39.237233354880416,
"grad_norm": 1.4674973487854004,
"learning_rate": 0.001,
"loss": 1.8474,
"step": 121400
},
{
"epoch": 39.26955397543633,
"grad_norm": 1.3472956418991089,
"learning_rate": 0.001,
"loss": 1.8347,
"step": 121500
},
{
"epoch": 39.301874595992246,
"grad_norm": 1.1671059131622314,
"learning_rate": 0.001,
"loss": 1.8212,
"step": 121600
},
{
"epoch": 39.33419521654816,
"grad_norm": 1.453717589378357,
"learning_rate": 0.001,
"loss": 1.8527,
"step": 121700
},
{
"epoch": 39.366515837104075,
"grad_norm": 1.282282829284668,
"learning_rate": 0.001,
"loss": 1.8442,
"step": 121800
},
{
"epoch": 39.39883645765999,
"grad_norm": 1.4532365798950195,
"learning_rate": 0.001,
"loss": 1.838,
"step": 121900
},
{
"epoch": 39.431157078215904,
"grad_norm": 1.4877893924713135,
"learning_rate": 0.001,
"loss": 1.8498,
"step": 122000
},
{
"epoch": 39.46347769877182,
"grad_norm": 1.3164318799972534,
"learning_rate": 0.001,
"loss": 1.8654,
"step": 122100
},
{
"epoch": 39.49579831932773,
"grad_norm": 1.5019081830978394,
"learning_rate": 0.001,
"loss": 1.8596,
"step": 122200
},
{
"epoch": 39.52811893988365,
"grad_norm": 1.4312676191329956,
"learning_rate": 0.001,
"loss": 1.8669,
"step": 122300
},
{
"epoch": 39.56043956043956,
"grad_norm": 1.2259525060653687,
"learning_rate": 0.001,
"loss": 1.8589,
"step": 122400
},
{
"epoch": 39.59276018099548,
"grad_norm": 1.2984930276870728,
"learning_rate": 0.001,
"loss": 1.8755,
"step": 122500
},
{
"epoch": 39.62508080155139,
"grad_norm": 1.4429779052734375,
"learning_rate": 0.001,
"loss": 1.8738,
"step": 122600
},
{
"epoch": 39.657401422107306,
"grad_norm": 1.8335790634155273,
"learning_rate": 0.001,
"loss": 1.8935,
"step": 122700
},
{
"epoch": 39.68972204266322,
"grad_norm": 1.6668318510055542,
"learning_rate": 0.001,
"loss": 1.8973,
"step": 122800
},
{
"epoch": 39.722042663219135,
"grad_norm": 1.3342713117599487,
"learning_rate": 0.001,
"loss": 1.8826,
"step": 122900
},
{
"epoch": 39.75436328377505,
"grad_norm": 1.4650003910064697,
"learning_rate": 0.001,
"loss": 1.8939,
"step": 123000
},
{
"epoch": 39.786683904330964,
"grad_norm": 1.3267019987106323,
"learning_rate": 0.001,
"loss": 1.909,
"step": 123100
},
{
"epoch": 39.81900452488688,
"grad_norm": 1.9128496646881104,
"learning_rate": 0.001,
"loss": 1.9091,
"step": 123200
},
{
"epoch": 39.85132514544279,
"grad_norm": 1.540787696838379,
"learning_rate": 0.001,
"loss": 1.8883,
"step": 123300
},
{
"epoch": 39.88364576599871,
"grad_norm": 1.1803691387176514,
"learning_rate": 0.001,
"loss": 1.9074,
"step": 123400
},
{
"epoch": 39.91596638655462,
"grad_norm": 1.1504863500595093,
"learning_rate": 0.001,
"loss": 1.9048,
"step": 123500
},
{
"epoch": 39.94828700711054,
"grad_norm": 1.9306889772415161,
"learning_rate": 0.001,
"loss": 1.8973,
"step": 123600
},
{
"epoch": 39.98060762766645,
"grad_norm": 1.3253064155578613,
"learning_rate": 0.001,
"loss": 1.9171,
"step": 123700
},
{
"epoch": 40.012928248222366,
"grad_norm": 1.6476284265518188,
"learning_rate": 0.001,
"loss": 1.8672,
"step": 123800
},
{
"epoch": 40.04524886877828,
"grad_norm": 1.8081892728805542,
"learning_rate": 0.001,
"loss": 1.7978,
"step": 123900
},
{
"epoch": 40.077569489334195,
"grad_norm": 1.5348113775253296,
"learning_rate": 0.001,
"loss": 1.7783,
"step": 124000
},
{
"epoch": 40.10989010989011,
"grad_norm": 1.615153431892395,
"learning_rate": 0.001,
"loss": 1.7808,
"step": 124100
},
{
"epoch": 40.142210730446024,
"grad_norm": 1.2413135766983032,
"learning_rate": 0.001,
"loss": 1.7875,
"step": 124200
},
{
"epoch": 40.17453135100194,
"grad_norm": 1.601155161857605,
"learning_rate": 0.001,
"loss": 1.7981,
"step": 124300
},
{
"epoch": 40.20685197155785,
"grad_norm": 1.4046754837036133,
"learning_rate": 0.001,
"loss": 1.8249,
"step": 124400
},
{
"epoch": 40.23917259211377,
"grad_norm": 2.250770092010498,
"learning_rate": 0.001,
"loss": 1.8113,
"step": 124500
},
{
"epoch": 40.27149321266968,
"grad_norm": 1.772512674331665,
"learning_rate": 0.001,
"loss": 1.8117,
"step": 124600
},
{
"epoch": 40.3038138332256,
"grad_norm": 1.3685424327850342,
"learning_rate": 0.001,
"loss": 1.8224,
"step": 124700
},
{
"epoch": 40.33613445378151,
"grad_norm": 1.552601933479309,
"learning_rate": 0.001,
"loss": 1.8236,
"step": 124800
},
{
"epoch": 40.368455074337426,
"grad_norm": 1.4619715213775635,
"learning_rate": 0.001,
"loss": 1.8441,
"step": 124900
},
{
"epoch": 40.40077569489334,
"grad_norm": 1.4244015216827393,
"learning_rate": 0.001,
"loss": 1.8342,
"step": 125000
},
{
"epoch": 40.433096315449255,
"grad_norm": 1.5007152557373047,
"learning_rate": 0.001,
"loss": 1.827,
"step": 125100
},
{
"epoch": 40.46541693600517,
"grad_norm": 1.730450987815857,
"learning_rate": 0.001,
"loss": 1.8571,
"step": 125200
},
{
"epoch": 40.497737556561084,
"grad_norm": 1.6700514554977417,
"learning_rate": 0.001,
"loss": 1.8583,
"step": 125300
},
{
"epoch": 40.530058177117,
"grad_norm": 1.5519285202026367,
"learning_rate": 0.001,
"loss": 1.8541,
"step": 125400
},
{
"epoch": 40.56237879767291,
"grad_norm": 1.6764475107192993,
"learning_rate": 0.001,
"loss": 1.8559,
"step": 125500
},
{
"epoch": 40.59469941822883,
"grad_norm": 1.6330721378326416,
"learning_rate": 0.001,
"loss": 1.8692,
"step": 125600
},
{
"epoch": 40.62702003878474,
"grad_norm": 1.6028348207473755,
"learning_rate": 0.001,
"loss": 1.8617,
"step": 125700
},
{
"epoch": 40.65934065934066,
"grad_norm": 1.3433600664138794,
"learning_rate": 0.001,
"loss": 1.8518,
"step": 125800
},
{
"epoch": 40.69166127989657,
"grad_norm": 1.4976279735565186,
"learning_rate": 0.001,
"loss": 1.8617,
"step": 125900
},
{
"epoch": 40.723981900452486,
"grad_norm": 1.3272761106491089,
"learning_rate": 0.001,
"loss": 1.8815,
"step": 126000
},
{
"epoch": 40.7563025210084,
"grad_norm": 1.7430063486099243,
"learning_rate": 0.001,
"loss": 1.8709,
"step": 126100
},
{
"epoch": 40.788623141564315,
"grad_norm": 1.2516944408416748,
"learning_rate": 0.001,
"loss": 1.8875,
"step": 126200
},
{
"epoch": 40.82094376212023,
"grad_norm": 1.5079940557479858,
"learning_rate": 0.001,
"loss": 1.885,
"step": 126300
},
{
"epoch": 40.853264382676144,
"grad_norm": 1.6306781768798828,
"learning_rate": 0.001,
"loss": 1.8881,
"step": 126400
},
{
"epoch": 40.88558500323206,
"grad_norm": 1.403608798980713,
"learning_rate": 0.001,
"loss": 1.8809,
"step": 126500
},
{
"epoch": 40.91790562378797,
"grad_norm": 1.1745949983596802,
"learning_rate": 0.001,
"loss": 1.8814,
"step": 126600
},
{
"epoch": 40.95022624434389,
"grad_norm": 1.1933189630508423,
"learning_rate": 0.001,
"loss": 1.8946,
"step": 126700
},
{
"epoch": 40.9825468648998,
"grad_norm": 1.845404028892517,
"learning_rate": 0.001,
"loss": 1.9039,
"step": 126800
},
{
"epoch": 41.014867485455724,
"grad_norm": 1.6035813093185425,
"learning_rate": 0.001,
"loss": 1.81,
"step": 126900
},
{
"epoch": 41.04718810601164,
"grad_norm": 1.5612142086029053,
"learning_rate": 0.001,
"loss": 1.772,
"step": 127000
},
{
"epoch": 41.07950872656755,
"grad_norm": 2.1515660285949707,
"learning_rate": 0.001,
"loss": 1.7821,
"step": 127100
},
{
"epoch": 41.11182934712347,
"grad_norm": 1.425368070602417,
"learning_rate": 0.001,
"loss": 1.7765,
"step": 127200
},
{
"epoch": 41.14414996767938,
"grad_norm": 1.5634095668792725,
"learning_rate": 0.001,
"loss": 1.7818,
"step": 127300
},
{
"epoch": 41.1764705882353,
"grad_norm": 1.3987374305725098,
"learning_rate": 0.001,
"loss": 1.7702,
"step": 127400
},
{
"epoch": 41.20879120879121,
"grad_norm": 2.2647273540496826,
"learning_rate": 0.001,
"loss": 1.8,
"step": 127500
},
{
"epoch": 41.241111829347126,
"grad_norm": 2.1380772590637207,
"learning_rate": 0.001,
"loss": 1.8136,
"step": 127600
},
{
"epoch": 41.27343244990304,
"grad_norm": 1.6095609664916992,
"learning_rate": 0.001,
"loss": 1.8197,
"step": 127700
},
{
"epoch": 41.305753070458955,
"grad_norm": 1.7609376907348633,
"learning_rate": 0.001,
"loss": 1.806,
"step": 127800
},
{
"epoch": 41.33807369101487,
"grad_norm": 1.9911599159240723,
"learning_rate": 0.001,
"loss": 1.8001,
"step": 127900
},
{
"epoch": 41.370394311570784,
"grad_norm": 1.6691104173660278,
"learning_rate": 0.001,
"loss": 1.8443,
"step": 128000
},
{
"epoch": 41.4027149321267,
"grad_norm": 1.5380711555480957,
"learning_rate": 0.001,
"loss": 1.8225,
"step": 128100
},
{
"epoch": 41.43503555268261,
"grad_norm": 1.5629456043243408,
"learning_rate": 0.001,
"loss": 1.8127,
"step": 128200
},
{
"epoch": 41.46735617323853,
"grad_norm": 1.8926146030426025,
"learning_rate": 0.001,
"loss": 1.8353,
"step": 128300
},
{
"epoch": 41.49967679379444,
"grad_norm": 1.4129561185836792,
"learning_rate": 0.001,
"loss": 1.8316,
"step": 128400
},
{
"epoch": 41.53199741435036,
"grad_norm": 1.873666524887085,
"learning_rate": 0.001,
"loss": 1.8349,
"step": 128500
},
{
"epoch": 41.56431803490627,
"grad_norm": 1.5797603130340576,
"learning_rate": 0.001,
"loss": 1.8293,
"step": 128600
},
{
"epoch": 41.596638655462186,
"grad_norm": 1.6729636192321777,
"learning_rate": 0.001,
"loss": 1.8204,
"step": 128700
},
{
"epoch": 41.6289592760181,
"grad_norm": 1.4557878971099854,
"learning_rate": 0.001,
"loss": 1.8333,
"step": 128800
},
{
"epoch": 41.661279896574015,
"grad_norm": 1.687487006187439,
"learning_rate": 0.001,
"loss": 1.8383,
"step": 128900
},
{
"epoch": 41.69360051712993,
"grad_norm": 1.804381012916565,
"learning_rate": 0.001,
"loss": 1.8559,
"step": 129000
},
{
"epoch": 41.725921137685845,
"grad_norm": 1.2200418710708618,
"learning_rate": 0.001,
"loss": 1.8529,
"step": 129100
},
{
"epoch": 41.75824175824176,
"grad_norm": 1.5321669578552246,
"learning_rate": 0.001,
"loss": 1.8695,
"step": 129200
},
{
"epoch": 41.790562378797674,
"grad_norm": 1.4247570037841797,
"learning_rate": 0.001,
"loss": 1.8445,
"step": 129300
},
{
"epoch": 41.82288299935359,
"grad_norm": 1.9431339502334595,
"learning_rate": 0.001,
"loss": 1.8713,
"step": 129400
},
{
"epoch": 41.8552036199095,
"grad_norm": 1.9393638372421265,
"learning_rate": 0.001,
"loss": 1.877,
"step": 129500
},
{
"epoch": 41.88752424046542,
"grad_norm": 1.3449734449386597,
"learning_rate": 0.001,
"loss": 1.8618,
"step": 129600
},
{
"epoch": 41.91984486102133,
"grad_norm": 1.9064652919769287,
"learning_rate": 0.001,
"loss": 1.8845,
"step": 129700
},
{
"epoch": 41.95216548157725,
"grad_norm": 1.790702223777771,
"learning_rate": 0.001,
"loss": 1.8672,
"step": 129800
},
{
"epoch": 41.98448610213316,
"grad_norm": 1.5176881551742554,
"learning_rate": 0.001,
"loss": 1.8939,
"step": 129900
},
{
"epoch": 42.016806722689076,
"grad_norm": 1.6592257022857666,
"learning_rate": 0.001,
"loss": 1.7906,
"step": 130000
},
{
"epoch": 42.04912734324499,
"grad_norm": 1.98469877243042,
"learning_rate": 0.001,
"loss": 1.7483,
"step": 130100
},
{
"epoch": 42.081447963800905,
"grad_norm": 1.4336119890213013,
"learning_rate": 0.001,
"loss": 1.7588,
"step": 130200
},
{
"epoch": 42.11376858435682,
"grad_norm": 2.131176710128784,
"learning_rate": 0.001,
"loss": 1.7636,
"step": 130300
},
{
"epoch": 42.146089204912734,
"grad_norm": 1.57058584690094,
"learning_rate": 0.001,
"loss": 1.7595,
"step": 130400
},
{
"epoch": 42.17840982546865,
"grad_norm": 2.4268321990966797,
"learning_rate": 0.001,
"loss": 1.7788,
"step": 130500
},
{
"epoch": 42.21073044602456,
"grad_norm": 1.9531536102294922,
"learning_rate": 0.001,
"loss": 1.7748,
"step": 130600
},
{
"epoch": 42.24305106658048,
"grad_norm": 1.768334150314331,
"learning_rate": 0.001,
"loss": 1.7777,
"step": 130700
},
{
"epoch": 42.27537168713639,
"grad_norm": 1.5845686197280884,
"learning_rate": 0.001,
"loss": 1.782,
"step": 130800
},
{
"epoch": 42.30769230769231,
"grad_norm": 1.7962462902069092,
"learning_rate": 0.001,
"loss": 1.7963,
"step": 130900
},
{
"epoch": 42.34001292824822,
"grad_norm": 1.5864969491958618,
"learning_rate": 0.001,
"loss": 1.7851,
"step": 131000
},
{
"epoch": 42.372333548804136,
"grad_norm": 1.5633211135864258,
"learning_rate": 0.001,
"loss": 1.8111,
"step": 131100
},
{
"epoch": 42.40465416936005,
"grad_norm": 1.7591291666030884,
"learning_rate": 0.001,
"loss": 1.8192,
"step": 131200
},
{
"epoch": 42.436974789915965,
"grad_norm": 1.687812328338623,
"learning_rate": 0.001,
"loss": 1.7977,
"step": 131300
},
{
"epoch": 42.46929541047188,
"grad_norm": 1.776008129119873,
"learning_rate": 0.001,
"loss": 1.8092,
"step": 131400
},
{
"epoch": 42.501616031027794,
"grad_norm": 1.568227767944336,
"learning_rate": 0.001,
"loss": 1.8161,
"step": 131500
},
{
"epoch": 42.53393665158371,
"grad_norm": 1.6084396839141846,
"learning_rate": 0.001,
"loss": 1.8098,
"step": 131600
},
{
"epoch": 42.56625727213962,
"grad_norm": 2.0739943981170654,
"learning_rate": 0.001,
"loss": 1.8194,
"step": 131700
},
{
"epoch": 42.59857789269554,
"grad_norm": 1.7982072830200195,
"learning_rate": 0.001,
"loss": 1.8461,
"step": 131800
},
{
"epoch": 42.63089851325145,
"grad_norm": 1.731550931930542,
"learning_rate": 0.001,
"loss": 1.8189,
"step": 131900
},
{
"epoch": 42.66321913380737,
"grad_norm": 1.6124906539916992,
"learning_rate": 0.001,
"loss": 1.8326,
"step": 132000
},
{
"epoch": 42.69553975436328,
"grad_norm": 2.0933144092559814,
"learning_rate": 0.001,
"loss": 1.8326,
"step": 132100
},
{
"epoch": 42.727860374919196,
"grad_norm": 1.8282585144042969,
"learning_rate": 0.001,
"loss": 1.8413,
"step": 132200
},
{
"epoch": 42.76018099547511,
"grad_norm": 1.7819545269012451,
"learning_rate": 0.001,
"loss": 1.837,
"step": 132300
},
{
"epoch": 42.792501616031025,
"grad_norm": 1.554919719696045,
"learning_rate": 0.001,
"loss": 1.8472,
"step": 132400
},
{
"epoch": 42.82482223658694,
"grad_norm": 1.817638874053955,
"learning_rate": 0.001,
"loss": 1.8546,
"step": 132500
},
{
"epoch": 42.857142857142854,
"grad_norm": 1.4772257804870605,
"learning_rate": 0.001,
"loss": 1.8549,
"step": 132600
},
{
"epoch": 42.88946347769877,
"grad_norm": 1.5371580123901367,
"learning_rate": 0.001,
"loss": 1.8443,
"step": 132700
},
{
"epoch": 42.92178409825468,
"grad_norm": 2.0546774864196777,
"learning_rate": 0.001,
"loss": 1.8805,
"step": 132800
},
{
"epoch": 42.9541047188106,
"grad_norm": 1.7636181116104126,
"learning_rate": 0.001,
"loss": 1.8686,
"step": 132900
},
{
"epoch": 42.98642533936652,
"grad_norm": 2.0109689235687256,
"learning_rate": 0.001,
"loss": 1.8597,
"step": 133000
},
{
"epoch": 43.018745959922434,
"grad_norm": 1.4981175661087036,
"learning_rate": 0.001,
"loss": 1.7994,
"step": 133100
},
{
"epoch": 43.05106658047835,
"grad_norm": 1.4558171033859253,
"learning_rate": 0.001,
"loss": 1.7275,
"step": 133200
},
{
"epoch": 43.08338720103426,
"grad_norm": 1.6762676239013672,
"learning_rate": 0.001,
"loss": 1.7289,
"step": 133300
},
{
"epoch": 43.11570782159018,
"grad_norm": 1.6511410474777222,
"learning_rate": 0.001,
"loss": 1.7406,
"step": 133400
},
{
"epoch": 43.14802844214609,
"grad_norm": 1.166945457458496,
"learning_rate": 0.001,
"loss": 1.7587,
"step": 133500
},
{
"epoch": 43.18034906270201,
"grad_norm": 1.2843260765075684,
"learning_rate": 0.001,
"loss": 1.7729,
"step": 133600
},
{
"epoch": 43.21266968325792,
"grad_norm": 1.9249118566513062,
"learning_rate": 0.001,
"loss": 1.7864,
"step": 133700
},
{
"epoch": 43.244990303813836,
"grad_norm": 1.5876914262771606,
"learning_rate": 0.001,
"loss": 1.7532,
"step": 133800
},
{
"epoch": 43.27731092436975,
"grad_norm": 0.9783575534820557,
"learning_rate": 0.001,
"loss": 1.7701,
"step": 133900
},
{
"epoch": 43.309631544925665,
"grad_norm": 1.4633513689041138,
"learning_rate": 0.001,
"loss": 1.7906,
"step": 134000
},
{
"epoch": 43.34195216548158,
"grad_norm": 1.3617346286773682,
"learning_rate": 0.001,
"loss": 1.7917,
"step": 134100
},
{
"epoch": 43.374272786037494,
"grad_norm": 1.201565146446228,
"learning_rate": 0.001,
"loss": 1.7905,
"step": 134200
},
{
"epoch": 43.40659340659341,
"grad_norm": 1.3246203660964966,
"learning_rate": 0.001,
"loss": 1.7993,
"step": 134300
},
{
"epoch": 43.43891402714932,
"grad_norm": 1.2301287651062012,
"learning_rate": 0.001,
"loss": 1.7808,
"step": 134400
},
{
"epoch": 43.47123464770524,
"grad_norm": 1.4977960586547852,
"learning_rate": 0.001,
"loss": 1.808,
"step": 134500
},
{
"epoch": 43.50355526826115,
"grad_norm": 1.3553272485733032,
"learning_rate": 0.001,
"loss": 1.7998,
"step": 134600
},
{
"epoch": 43.53587588881707,
"grad_norm": 1.44217848777771,
"learning_rate": 0.001,
"loss": 1.7985,
"step": 134700
},
{
"epoch": 43.56819650937298,
"grad_norm": 1.1187900304794312,
"learning_rate": 0.001,
"loss": 1.8194,
"step": 134800
},
{
"epoch": 43.600517129928896,
"grad_norm": 1.4122588634490967,
"learning_rate": 0.001,
"loss": 1.798,
"step": 134900
},
{
"epoch": 43.63283775048481,
"grad_norm": 1.8867087364196777,
"learning_rate": 0.001,
"loss": 1.8276,
"step": 135000
},
{
"epoch": 43.665158371040725,
"grad_norm": 1.6213741302490234,
"learning_rate": 0.001,
"loss": 1.803,
"step": 135100
},
{
"epoch": 43.69747899159664,
"grad_norm": 1.4863262176513672,
"learning_rate": 0.001,
"loss": 1.8304,
"step": 135200
},
{
"epoch": 43.729799612152554,
"grad_norm": 1.1993991136550903,
"learning_rate": 0.001,
"loss": 1.8263,
"step": 135300
},
{
"epoch": 43.76212023270847,
"grad_norm": 1.2402936220169067,
"learning_rate": 0.001,
"loss": 1.8377,
"step": 135400
},
{
"epoch": 43.79444085326438,
"grad_norm": 1.1922273635864258,
"learning_rate": 0.001,
"loss": 1.8451,
"step": 135500
},
{
"epoch": 43.8267614738203,
"grad_norm": 1.2159183025360107,
"learning_rate": 0.001,
"loss": 1.8202,
"step": 135600
},
{
"epoch": 43.85908209437621,
"grad_norm": 1.1227339506149292,
"learning_rate": 0.001,
"loss": 1.836,
"step": 135700
},
{
"epoch": 43.89140271493213,
"grad_norm": 1.4446237087249756,
"learning_rate": 0.001,
"loss": 1.8453,
"step": 135800
},
{
"epoch": 43.92372333548804,
"grad_norm": 1.2879489660263062,
"learning_rate": 0.001,
"loss": 1.8418,
"step": 135900
},
{
"epoch": 43.956043956043956,
"grad_norm": 1.248571515083313,
"learning_rate": 0.001,
"loss": 1.8395,
"step": 136000
},
{
"epoch": 43.98836457659987,
"grad_norm": 1.248064398765564,
"learning_rate": 0.001,
"loss": 1.8493,
"step": 136100
},
{
"epoch": 44.020685197155785,
"grad_norm": 1.336883783340454,
"learning_rate": 0.001,
"loss": 1.7783,
"step": 136200
},
{
"epoch": 44.0530058177117,
"grad_norm": 1.620754361152649,
"learning_rate": 0.001,
"loss": 1.7312,
"step": 136300
},
{
"epoch": 44.085326438267614,
"grad_norm": 1.142549753189087,
"learning_rate": 0.001,
"loss": 1.7199,
"step": 136400
},
{
"epoch": 44.11764705882353,
"grad_norm": 1.7734692096710205,
"learning_rate": 0.001,
"loss": 1.743,
"step": 136500
},
{
"epoch": 44.14996767937944,
"grad_norm": 1.630759835243225,
"learning_rate": 0.001,
"loss": 1.731,
"step": 136600
},
{
"epoch": 44.18228829993536,
"grad_norm": 1.2140966653823853,
"learning_rate": 0.001,
"loss": 1.7513,
"step": 136700
},
{
"epoch": 44.21460892049127,
"grad_norm": 1.1699700355529785,
"learning_rate": 0.001,
"loss": 1.7448,
"step": 136800
},
{
"epoch": 44.24692954104719,
"grad_norm": 1.9535216093063354,
"learning_rate": 0.001,
"loss": 1.7415,
"step": 136900
},
{
"epoch": 44.2792501616031,
"grad_norm": 1.5197383165359497,
"learning_rate": 0.001,
"loss": 1.768,
"step": 137000
},
{
"epoch": 44.311570782159016,
"grad_norm": 1.1582030057907104,
"learning_rate": 0.001,
"loss": 1.749,
"step": 137100
},
{
"epoch": 44.34389140271493,
"grad_norm": 1.2484724521636963,
"learning_rate": 0.001,
"loss": 1.7691,
"step": 137200
},
{
"epoch": 44.376212023270845,
"grad_norm": 1.4032336473464966,
"learning_rate": 0.001,
"loss": 1.7582,
"step": 137300
},
{
"epoch": 44.40853264382676,
"grad_norm": 1.0077219009399414,
"learning_rate": 0.001,
"loss": 1.7642,
"step": 137400
},
{
"epoch": 44.440853264382675,
"grad_norm": 1.2438334226608276,
"learning_rate": 0.001,
"loss": 1.7674,
"step": 137500
},
{
"epoch": 44.47317388493859,
"grad_norm": 1.2038356065750122,
"learning_rate": 0.001,
"loss": 1.7954,
"step": 137600
},
{
"epoch": 44.505494505494504,
"grad_norm": 1.57408607006073,
"learning_rate": 0.001,
"loss": 1.7872,
"step": 137700
},
{
"epoch": 44.53781512605042,
"grad_norm": 1.263692855834961,
"learning_rate": 0.001,
"loss": 1.7949,
"step": 137800
},
{
"epoch": 44.57013574660633,
"grad_norm": 1.847970962524414,
"learning_rate": 0.001,
"loss": 1.809,
"step": 137900
},
{
"epoch": 44.60245636716225,
"grad_norm": 1.1731334924697876,
"learning_rate": 0.001,
"loss": 1.7889,
"step": 138000
},
{
"epoch": 44.63477698771816,
"grad_norm": 1.5566082000732422,
"learning_rate": 0.001,
"loss": 1.79,
"step": 138100
},
{
"epoch": 44.66709760827408,
"grad_norm": 1.4474111795425415,
"learning_rate": 0.001,
"loss": 1.809,
"step": 138200
},
{
"epoch": 44.69941822882999,
"grad_norm": 1.4985994100570679,
"learning_rate": 0.001,
"loss": 1.8243,
"step": 138300
},
{
"epoch": 44.731738849385906,
"grad_norm": 1.215576410293579,
"learning_rate": 0.001,
"loss": 1.8165,
"step": 138400
},
{
"epoch": 44.76405946994182,
"grad_norm": 1.4949811697006226,
"learning_rate": 0.001,
"loss": 1.8177,
"step": 138500
},
{
"epoch": 44.796380090497735,
"grad_norm": 1.0145524740219116,
"learning_rate": 0.001,
"loss": 1.808,
"step": 138600
},
{
"epoch": 44.82870071105365,
"grad_norm": 1.6265182495117188,
"learning_rate": 0.001,
"loss": 1.8245,
"step": 138700
},
{
"epoch": 44.861021331609564,
"grad_norm": 1.18686842918396,
"learning_rate": 0.001,
"loss": 1.8194,
"step": 138800
},
{
"epoch": 44.89334195216548,
"grad_norm": 1.5669019222259521,
"learning_rate": 0.001,
"loss": 1.823,
"step": 138900
},
{
"epoch": 44.92566257272139,
"grad_norm": 1.338637113571167,
"learning_rate": 0.001,
"loss": 1.8264,
"step": 139000
},
{
"epoch": 44.95798319327731,
"grad_norm": 1.6436604261398315,
"learning_rate": 0.001,
"loss": 1.8453,
"step": 139100
},
{
"epoch": 44.99030381383322,
"grad_norm": 2.3497281074523926,
"learning_rate": 0.001,
"loss": 1.8322,
"step": 139200
},
{
"epoch": 45.022624434389144,
"grad_norm": 1.2802793979644775,
"learning_rate": 0.001,
"loss": 1.7512,
"step": 139300
},
{
"epoch": 45.05494505494506,
"grad_norm": 1.2619880437850952,
"learning_rate": 0.001,
"loss": 1.7113,
"step": 139400
},
{
"epoch": 45.08726567550097,
"grad_norm": 1.8655284643173218,
"learning_rate": 0.001,
"loss": 1.7134,
"step": 139500
},
{
"epoch": 45.11958629605689,
"grad_norm": 1.3278272151947021,
"learning_rate": 0.001,
"loss": 1.7256,
"step": 139600
},
{
"epoch": 45.1519069166128,
"grad_norm": 1.3971140384674072,
"learning_rate": 0.001,
"loss": 1.6979,
"step": 139700
},
{
"epoch": 45.18422753716872,
"grad_norm": 1.2571136951446533,
"learning_rate": 0.001,
"loss": 1.7232,
"step": 139800
},
{
"epoch": 45.21654815772463,
"grad_norm": 1.362691879272461,
"learning_rate": 0.001,
"loss": 1.7353,
"step": 139900
},
{
"epoch": 45.248868778280546,
"grad_norm": 1.193507194519043,
"learning_rate": 0.001,
"loss": 1.7433,
"step": 140000
},
{
"epoch": 45.28118939883646,
"grad_norm": 1.516035795211792,
"learning_rate": 0.001,
"loss": 1.7305,
"step": 140100
},
{
"epoch": 45.313510019392375,
"grad_norm": 1.4402673244476318,
"learning_rate": 0.001,
"loss": 1.7452,
"step": 140200
},
{
"epoch": 45.34583063994829,
"grad_norm": 1.6060928106307983,
"learning_rate": 0.001,
"loss": 1.7449,
"step": 140300
},
{
"epoch": 45.378151260504204,
"grad_norm": 1.4396388530731201,
"learning_rate": 0.001,
"loss": 1.7543,
"step": 140400
},
{
"epoch": 45.41047188106012,
"grad_norm": 1.294250726699829,
"learning_rate": 0.001,
"loss": 1.7701,
"step": 140500
},
{
"epoch": 45.44279250161603,
"grad_norm": 1.2159461975097656,
"learning_rate": 0.001,
"loss": 1.7674,
"step": 140600
},
{
"epoch": 45.47511312217195,
"grad_norm": 1.2815823554992676,
"learning_rate": 0.001,
"loss": 1.7722,
"step": 140700
},
{
"epoch": 45.50743374272786,
"grad_norm": 1.1248795986175537,
"learning_rate": 0.001,
"loss": 1.7878,
"step": 140800
},
{
"epoch": 45.53975436328378,
"grad_norm": 1.2497678995132446,
"learning_rate": 0.001,
"loss": 1.7736,
"step": 140900
},
{
"epoch": 45.57207498383969,
"grad_norm": 1.6308103799819946,
"learning_rate": 0.001,
"loss": 1.7794,
"step": 141000
},
{
"epoch": 45.604395604395606,
"grad_norm": 1.1841062307357788,
"learning_rate": 0.001,
"loss": 1.7868,
"step": 141100
},
{
"epoch": 45.63671622495152,
"grad_norm": 1.0237959623336792,
"learning_rate": 0.001,
"loss": 1.7914,
"step": 141200
},
{
"epoch": 45.669036845507435,
"grad_norm": 1.1166704893112183,
"learning_rate": 0.001,
"loss": 1.8011,
"step": 141300
},
{
"epoch": 45.70135746606335,
"grad_norm": 1.2327576875686646,
"learning_rate": 0.001,
"loss": 1.779,
"step": 141400
},
{
"epoch": 45.733678086619264,
"grad_norm": 1.3635004758834839,
"learning_rate": 0.001,
"loss": 1.8076,
"step": 141500
},
{
"epoch": 45.76599870717518,
"grad_norm": 1.2312064170837402,
"learning_rate": 0.001,
"loss": 1.8138,
"step": 141600
},
{
"epoch": 45.79831932773109,
"grad_norm": 1.3265718221664429,
"learning_rate": 0.001,
"loss": 1.8024,
"step": 141700
},
{
"epoch": 45.83063994828701,
"grad_norm": 1.4107000827789307,
"learning_rate": 0.001,
"loss": 1.8003,
"step": 141800
},
{
"epoch": 45.86296056884292,
"grad_norm": 1.5675338506698608,
"learning_rate": 0.001,
"loss": 1.8059,
"step": 141900
},
{
"epoch": 45.89528118939884,
"grad_norm": 1.2969226837158203,
"learning_rate": 0.001,
"loss": 1.8176,
"step": 142000
},
{
"epoch": 45.92760180995475,
"grad_norm": 1.218584656715393,
"learning_rate": 0.001,
"loss": 1.8363,
"step": 142100
},
{
"epoch": 45.959922430510666,
"grad_norm": 1.1808185577392578,
"learning_rate": 0.001,
"loss": 1.8052,
"step": 142200
},
{
"epoch": 45.99224305106658,
"grad_norm": 1.398820400238037,
"learning_rate": 0.001,
"loss": 1.8341,
"step": 142300
},
{
"epoch": 46.024563671622495,
"grad_norm": 1.3547405004501343,
"learning_rate": 0.001,
"loss": 1.7249,
"step": 142400
},
{
"epoch": 46.05688429217841,
"grad_norm": 1.3534218072891235,
"learning_rate": 0.001,
"loss": 1.6994,
"step": 142500
},
{
"epoch": 46.089204912734324,
"grad_norm": 1.7203636169433594,
"learning_rate": 0.001,
"loss": 1.6882,
"step": 142600
},
{
"epoch": 46.12152553329024,
"grad_norm": 1.5423005819320679,
"learning_rate": 0.001,
"loss": 1.6975,
"step": 142700
},
{
"epoch": 46.15384615384615,
"grad_norm": 1.1216588020324707,
"learning_rate": 0.001,
"loss": 1.7135,
"step": 142800
},
{
"epoch": 46.18616677440207,
"grad_norm": 1.2883764505386353,
"learning_rate": 0.001,
"loss": 1.7049,
"step": 142900
},
{
"epoch": 46.21848739495798,
"grad_norm": 1.4580037593841553,
"learning_rate": 0.001,
"loss": 1.7405,
"step": 143000
},
{
"epoch": 46.2508080155139,
"grad_norm": 1.3226218223571777,
"learning_rate": 0.001,
"loss": 1.7324,
"step": 143100
},
{
"epoch": 46.28312863606981,
"grad_norm": 1.4107831716537476,
"learning_rate": 0.001,
"loss": 1.7336,
"step": 143200
},
{
"epoch": 46.315449256625726,
"grad_norm": 1.8357940912246704,
"learning_rate": 0.001,
"loss": 1.7417,
"step": 143300
},
{
"epoch": 46.34776987718164,
"grad_norm": 1.245288610458374,
"learning_rate": 0.001,
"loss": 1.7527,
"step": 143400
},
{
"epoch": 46.380090497737555,
"grad_norm": 1.3753732442855835,
"learning_rate": 0.001,
"loss": 1.737,
"step": 143500
},
{
"epoch": 46.41241111829347,
"grad_norm": 1.384994387626648,
"learning_rate": 0.001,
"loss": 1.7392,
"step": 143600
},
{
"epoch": 46.444731738849384,
"grad_norm": 1.5150701999664307,
"learning_rate": 0.001,
"loss": 1.7538,
"step": 143700
},
{
"epoch": 46.4770523594053,
"grad_norm": 0.9827286601066589,
"learning_rate": 0.001,
"loss": 1.7639,
"step": 143800
},
{
"epoch": 46.50937297996121,
"grad_norm": 1.7500789165496826,
"learning_rate": 0.001,
"loss": 1.7619,
"step": 143900
},
{
"epoch": 46.54169360051713,
"grad_norm": 1.1618425846099854,
"learning_rate": 0.001,
"loss": 1.7543,
"step": 144000
},
{
"epoch": 46.57401422107304,
"grad_norm": 1.3009589910507202,
"learning_rate": 0.001,
"loss": 1.7552,
"step": 144100
},
{
"epoch": 46.60633484162896,
"grad_norm": 1.1285375356674194,
"learning_rate": 0.001,
"loss": 1.7622,
"step": 144200
},
{
"epoch": 46.63865546218487,
"grad_norm": 1.5087919235229492,
"learning_rate": 0.001,
"loss": 1.7607,
"step": 144300
},
{
"epoch": 46.670976082740786,
"grad_norm": 1.759446620941162,
"learning_rate": 0.001,
"loss": 1.7761,
"step": 144400
},
{
"epoch": 46.7032967032967,
"grad_norm": 1.3985189199447632,
"learning_rate": 0.001,
"loss": 1.791,
"step": 144500
},
{
"epoch": 46.735617323852615,
"grad_norm": 1.342124104499817,
"learning_rate": 0.001,
"loss": 1.7956,
"step": 144600
},
{
"epoch": 46.76793794440853,
"grad_norm": 1.409203290939331,
"learning_rate": 0.001,
"loss": 1.7947,
"step": 144700
},
{
"epoch": 46.800258564964444,
"grad_norm": 1.2463408708572388,
"learning_rate": 0.001,
"loss": 1.7921,
"step": 144800
},
{
"epoch": 46.83257918552036,
"grad_norm": 1.3556941747665405,
"learning_rate": 0.001,
"loss": 1.7858,
"step": 144900
},
{
"epoch": 46.864899806076274,
"grad_norm": 1.2283049821853638,
"learning_rate": 0.001,
"loss": 1.7914,
"step": 145000
},
{
"epoch": 46.89722042663219,
"grad_norm": 1.1280850172042847,
"learning_rate": 0.001,
"loss": 1.7992,
"step": 145100
},
{
"epoch": 46.9295410471881,
"grad_norm": 1.629721999168396,
"learning_rate": 0.001,
"loss": 1.8016,
"step": 145200
},
{
"epoch": 46.96186166774402,
"grad_norm": 1.325217604637146,
"learning_rate": 0.001,
"loss": 1.812,
"step": 145300
},
{
"epoch": 46.99418228829994,
"grad_norm": 1.7119725942611694,
"learning_rate": 0.001,
"loss": 1.7949,
"step": 145400
},
{
"epoch": 47.02650290885585,
"grad_norm": 1.1571006774902344,
"learning_rate": 0.001,
"loss": 1.6989,
"step": 145500
},
{
"epoch": 47.05882352941177,
"grad_norm": 1.6925146579742432,
"learning_rate": 0.001,
"loss": 1.6892,
"step": 145600
},
{
"epoch": 47.09114414996768,
"grad_norm": 1.3089109659194946,
"learning_rate": 0.001,
"loss": 1.6835,
"step": 145700
},
{
"epoch": 47.1234647705236,
"grad_norm": 1.3750829696655273,
"learning_rate": 0.001,
"loss": 1.689,
"step": 145800
},
{
"epoch": 47.15578539107951,
"grad_norm": 1.282976508140564,
"learning_rate": 0.001,
"loss": 1.6912,
"step": 145900
},
{
"epoch": 47.188106011635426,
"grad_norm": 1.440115213394165,
"learning_rate": 0.001,
"loss": 1.7282,
"step": 146000
},
{
"epoch": 47.22042663219134,
"grad_norm": 1.5453740358352661,
"learning_rate": 0.001,
"loss": 1.7076,
"step": 146100
},
{
"epoch": 47.252747252747255,
"grad_norm": 1.4297361373901367,
"learning_rate": 0.001,
"loss": 1.716,
"step": 146200
},
{
"epoch": 47.28506787330317,
"grad_norm": 1.2313055992126465,
"learning_rate": 0.001,
"loss": 1.7095,
"step": 146300
},
{
"epoch": 47.317388493859085,
"grad_norm": 1.3750139474868774,
"learning_rate": 0.001,
"loss": 1.7288,
"step": 146400
},
{
"epoch": 47.349709114415,
"grad_norm": 1.1915059089660645,
"learning_rate": 0.001,
"loss": 1.7169,
"step": 146500
},
{
"epoch": 47.382029734970914,
"grad_norm": 1.2173153162002563,
"learning_rate": 0.001,
"loss": 1.7255,
"step": 146600
},
{
"epoch": 47.41435035552683,
"grad_norm": 1.5666117668151855,
"learning_rate": 0.001,
"loss": 1.7329,
"step": 146700
},
{
"epoch": 47.44667097608274,
"grad_norm": 1.4377115964889526,
"learning_rate": 0.001,
"loss": 1.7385,
"step": 146800
},
{
"epoch": 47.47899159663866,
"grad_norm": 1.184586763381958,
"learning_rate": 0.001,
"loss": 1.7473,
"step": 146900
},
{
"epoch": 47.51131221719457,
"grad_norm": 1.4830105304718018,
"learning_rate": 0.001,
"loss": 1.7454,
"step": 147000
},
{
"epoch": 47.543632837750486,
"grad_norm": 1.2405304908752441,
"learning_rate": 0.001,
"loss": 1.7284,
"step": 147100
},
{
"epoch": 47.5759534583064,
"grad_norm": 1.5127040147781372,
"learning_rate": 0.001,
"loss": 1.7545,
"step": 147200
},
{
"epoch": 47.608274078862316,
"grad_norm": 1.4291422367095947,
"learning_rate": 0.001,
"loss": 1.7473,
"step": 147300
},
{
"epoch": 47.64059469941823,
"grad_norm": 1.5448968410491943,
"learning_rate": 0.001,
"loss": 1.7714,
"step": 147400
},
{
"epoch": 47.672915319974145,
"grad_norm": 1.8437740802764893,
"learning_rate": 0.001,
"loss": 1.7797,
"step": 147500
},
{
"epoch": 47.70523594053006,
"grad_norm": 1.2971209287643433,
"learning_rate": 0.001,
"loss": 1.7818,
"step": 147600
},
{
"epoch": 47.737556561085974,
"grad_norm": 1.9197437763214111,
"learning_rate": 0.001,
"loss": 1.7621,
"step": 147700
},
{
"epoch": 47.76987718164189,
"grad_norm": 1.6180408000946045,
"learning_rate": 0.001,
"loss": 1.7608,
"step": 147800
},
{
"epoch": 47.8021978021978,
"grad_norm": 1.7230387926101685,
"learning_rate": 0.001,
"loss": 1.7959,
"step": 147900
},
{
"epoch": 47.83451842275372,
"grad_norm": 1.290977954864502,
"learning_rate": 0.001,
"loss": 1.7735,
"step": 148000
},
{
"epoch": 47.86683904330963,
"grad_norm": 1.6636526584625244,
"learning_rate": 0.001,
"loss": 1.7808,
"step": 148100
},
{
"epoch": 47.89915966386555,
"grad_norm": 1.3727259635925293,
"learning_rate": 0.001,
"loss": 1.7911,
"step": 148200
},
{
"epoch": 47.93148028442146,
"grad_norm": 1.4870383739471436,
"learning_rate": 0.001,
"loss": 1.7784,
"step": 148300
},
{
"epoch": 47.963800904977376,
"grad_norm": 1.2966899871826172,
"learning_rate": 0.001,
"loss": 1.7837,
"step": 148400
},
{
"epoch": 47.99612152553329,
"grad_norm": 1.8222113847732544,
"learning_rate": 0.001,
"loss": 1.7938,
"step": 148500
},
{
"epoch": 48.028442146089205,
"grad_norm": 1.3580232858657837,
"learning_rate": 0.001,
"loss": 1.6822,
"step": 148600
},
{
"epoch": 48.06076276664512,
"grad_norm": 1.2431883811950684,
"learning_rate": 0.001,
"loss": 1.6794,
"step": 148700
},
{
"epoch": 48.093083387201034,
"grad_norm": 1.3935822248458862,
"learning_rate": 0.001,
"loss": 1.6675,
"step": 148800
},
{
"epoch": 48.12540400775695,
"grad_norm": 1.383834958076477,
"learning_rate": 0.001,
"loss": 1.6794,
"step": 148900
},
{
"epoch": 48.15772462831286,
"grad_norm": 1.4718204736709595,
"learning_rate": 0.001,
"loss": 1.6977,
"step": 149000
},
{
"epoch": 48.19004524886878,
"grad_norm": 1.8937287330627441,
"learning_rate": 0.001,
"loss": 1.6792,
"step": 149100
},
{
"epoch": 48.22236586942469,
"grad_norm": 1.2440084218978882,
"learning_rate": 0.001,
"loss": 1.7048,
"step": 149200
},
{
"epoch": 48.25468648998061,
"grad_norm": 1.3202449083328247,
"learning_rate": 0.001,
"loss": 1.7053,
"step": 149300
},
{
"epoch": 48.28700711053652,
"grad_norm": 1.4978262186050415,
"learning_rate": 0.001,
"loss": 1.6938,
"step": 149400
},
{
"epoch": 48.319327731092436,
"grad_norm": 1.9077339172363281,
"learning_rate": 0.001,
"loss": 1.704,
"step": 149500
},
{
"epoch": 48.35164835164835,
"grad_norm": 1.5166305303573608,
"learning_rate": 0.001,
"loss": 1.7122,
"step": 149600
},
{
"epoch": 48.383968972204265,
"grad_norm": 1.4986920356750488,
"learning_rate": 0.001,
"loss": 1.7068,
"step": 149700
},
{
"epoch": 48.41628959276018,
"grad_norm": 1.649933934211731,
"learning_rate": 0.001,
"loss": 1.724,
"step": 149800
},
{
"epoch": 48.448610213316094,
"grad_norm": 1.2541015148162842,
"learning_rate": 0.001,
"loss": 1.7293,
"step": 149900
},
{
"epoch": 48.48093083387201,
"grad_norm": 1.6121021509170532,
"learning_rate": 0.001,
"loss": 1.7224,
"step": 150000
},
{
"epoch": 48.51325145442792,
"grad_norm": 1.4799165725708008,
"learning_rate": 0.001,
"loss": 1.7185,
"step": 150100
},
{
"epoch": 48.54557207498384,
"grad_norm": 1.7521393299102783,
"learning_rate": 0.001,
"loss": 1.7293,
"step": 150200
},
{
"epoch": 48.57789269553975,
"grad_norm": 1.7641196250915527,
"learning_rate": 0.001,
"loss": 1.727,
"step": 150300
},
{
"epoch": 48.61021331609567,
"grad_norm": 1.3735510110855103,
"learning_rate": 0.001,
"loss": 1.7446,
"step": 150400
},
{
"epoch": 48.64253393665158,
"grad_norm": 1.610403060913086,
"learning_rate": 0.001,
"loss": 1.754,
"step": 150500
},
{
"epoch": 48.674854557207496,
"grad_norm": 1.737728476524353,
"learning_rate": 0.001,
"loss": 1.7521,
"step": 150600
},
{
"epoch": 48.70717517776341,
"grad_norm": 1.4922481775283813,
"learning_rate": 0.001,
"loss": 1.7558,
"step": 150700
},
{
"epoch": 48.739495798319325,
"grad_norm": 1.6390092372894287,
"learning_rate": 0.001,
"loss": 1.7731,
"step": 150800
},
{
"epoch": 48.77181641887524,
"grad_norm": 1.9332650899887085,
"learning_rate": 0.001,
"loss": 1.7787,
"step": 150900
},
{
"epoch": 48.804137039431154,
"grad_norm": 1.7974140644073486,
"learning_rate": 0.001,
"loss": 1.7524,
"step": 151000
},
{
"epoch": 48.83645765998707,
"grad_norm": 1.4626390933990479,
"learning_rate": 0.001,
"loss": 1.7654,
"step": 151100
},
{
"epoch": 48.86877828054298,
"grad_norm": 1.7010713815689087,
"learning_rate": 0.001,
"loss": 1.7686,
"step": 151200
},
{
"epoch": 48.9010989010989,
"grad_norm": 1.407484769821167,
"learning_rate": 0.001,
"loss": 1.7644,
"step": 151300
},
{
"epoch": 48.93341952165481,
"grad_norm": 1.2641417980194092,
"learning_rate": 0.001,
"loss": 1.7783,
"step": 151400
},
{
"epoch": 48.96574014221073,
"grad_norm": 1.4772000312805176,
"learning_rate": 0.001,
"loss": 1.7891,
"step": 151500
},
{
"epoch": 48.99806076276664,
"grad_norm": 1.5785305500030518,
"learning_rate": 0.001,
"loss": 1.7565,
"step": 151600
},
{
"epoch": 49.03038138332256,
"grad_norm": 1.3122369050979614,
"learning_rate": 0.001,
"loss": 1.6511,
"step": 151700
},
{
"epoch": 49.06270200387848,
"grad_norm": 1.3335903882980347,
"learning_rate": 0.001,
"loss": 1.6433,
"step": 151800
},
{
"epoch": 49.09502262443439,
"grad_norm": 2.121500253677368,
"learning_rate": 0.001,
"loss": 1.6588,
"step": 151900
},
{
"epoch": 49.12734324499031,
"grad_norm": 1.490792155265808,
"learning_rate": 0.001,
"loss": 1.6737,
"step": 152000
},
{
"epoch": 49.15966386554622,
"grad_norm": 1.4584839344024658,
"learning_rate": 0.001,
"loss": 1.6884,
"step": 152100
},
{
"epoch": 49.191984486102136,
"grad_norm": 1.5305308103561401,
"learning_rate": 0.001,
"loss": 1.6808,
"step": 152200
},
{
"epoch": 49.22430510665805,
"grad_norm": 1.3502559661865234,
"learning_rate": 0.001,
"loss": 1.6932,
"step": 152300
},
{
"epoch": 49.256625727213965,
"grad_norm": 1.68340265750885,
"learning_rate": 0.001,
"loss": 1.6883,
"step": 152400
},
{
"epoch": 49.28894634776988,
"grad_norm": 1.608882188796997,
"learning_rate": 0.001,
"loss": 1.6979,
"step": 152500
},
{
"epoch": 49.321266968325794,
"grad_norm": 1.6637165546417236,
"learning_rate": 0.001,
"loss": 1.7032,
"step": 152600
},
{
"epoch": 49.35358758888171,
"grad_norm": 1.8644802570343018,
"learning_rate": 0.001,
"loss": 1.6843,
"step": 152700
},
{
"epoch": 49.38590820943762,
"grad_norm": 1.5030044317245483,
"learning_rate": 0.001,
"loss": 1.7119,
"step": 152800
},
{
"epoch": 49.41822882999354,
"grad_norm": 1.8687456846237183,
"learning_rate": 0.001,
"loss": 1.7179,
"step": 152900
},
{
"epoch": 49.45054945054945,
"grad_norm": 1.5273497104644775,
"learning_rate": 0.001,
"loss": 1.7096,
"step": 153000
},
{
"epoch": 49.48287007110537,
"grad_norm": 1.5750623941421509,
"learning_rate": 0.001,
"loss": 1.7056,
"step": 153100
},
{
"epoch": 49.51519069166128,
"grad_norm": 1.4431356191635132,
"learning_rate": 0.001,
"loss": 1.7231,
"step": 153200
},
{
"epoch": 49.547511312217196,
"grad_norm": 1.4949909448623657,
"learning_rate": 0.001,
"loss": 1.7241,
"step": 153300
},
{
"epoch": 49.57983193277311,
"grad_norm": 1.5795378684997559,
"learning_rate": 0.001,
"loss": 1.7146,
"step": 153400
},
{
"epoch": 49.612152553329025,
"grad_norm": 1.542120337486267,
"learning_rate": 0.001,
"loss": 1.7281,
"step": 153500
},
{
"epoch": 49.64447317388494,
"grad_norm": 1.3886210918426514,
"learning_rate": 0.001,
"loss": 1.7423,
"step": 153600
},
{
"epoch": 49.676793794440854,
"grad_norm": 2.047396421432495,
"learning_rate": 0.001,
"loss": 1.7369,
"step": 153700
},
{
"epoch": 49.70911441499677,
"grad_norm": 1.8594999313354492,
"learning_rate": 0.001,
"loss": 1.7426,
"step": 153800
},
{
"epoch": 49.74143503555268,
"grad_norm": 1.7180241346359253,
"learning_rate": 0.001,
"loss": 1.7269,
"step": 153900
},
{
"epoch": 49.7737556561086,
"grad_norm": 1.4761431217193604,
"learning_rate": 0.001,
"loss": 1.7451,
"step": 154000
},
{
"epoch": 49.80607627666451,
"grad_norm": 1.7135429382324219,
"learning_rate": 0.001,
"loss": 1.7571,
"step": 154100
},
{
"epoch": 49.83839689722043,
"grad_norm": 1.4928319454193115,
"learning_rate": 0.001,
"loss": 1.7332,
"step": 154200
},
{
"epoch": 49.87071751777634,
"grad_norm": 1.352138638496399,
"learning_rate": 0.001,
"loss": 1.752,
"step": 154300
},
{
"epoch": 49.903038138332256,
"grad_norm": 1.308883786201477,
"learning_rate": 0.001,
"loss": 1.7674,
"step": 154400
},
{
"epoch": 49.93535875888817,
"grad_norm": 1.5187338590621948,
"learning_rate": 0.001,
"loss": 1.7723,
"step": 154500
},
{
"epoch": 49.967679379444085,
"grad_norm": 2.021956205368042,
"learning_rate": 0.001,
"loss": 1.766,
"step": 154600
},
{
"epoch": 50.0,
"grad_norm": 2.2232563495635986,
"learning_rate": 0.001,
"loss": 1.7376,
"step": 154700
},
{
"epoch": 50.032320620555915,
"grad_norm": 1.9615095853805542,
"learning_rate": 0.001,
"loss": 1.6384,
"step": 154800
},
{
"epoch": 50.06464124111183,
"grad_norm": 3.108535051345825,
"learning_rate": 0.001,
"loss": 1.6508,
"step": 154900
},
{
"epoch": 50.096961861667744,
"grad_norm": 2.052661657333374,
"learning_rate": 0.001,
"loss": 1.6476,
"step": 155000
},
{
"epoch": 50.12928248222366,
"grad_norm": 1.971717119216919,
"learning_rate": 0.001,
"loss": 1.6523,
"step": 155100
},
{
"epoch": 50.16160310277957,
"grad_norm": 2.621964454650879,
"learning_rate": 0.001,
"loss": 1.647,
"step": 155200
},
{
"epoch": 50.19392372333549,
"grad_norm": 2.2581889629364014,
"learning_rate": 0.001,
"loss": 1.6595,
"step": 155300
},
{
"epoch": 50.2262443438914,
"grad_norm": 1.6710734367370605,
"learning_rate": 0.001,
"loss": 1.6778,
"step": 155400
},
{
"epoch": 50.25856496444732,
"grad_norm": 2.0659639835357666,
"learning_rate": 0.001,
"loss": 1.6862,
"step": 155500
},
{
"epoch": 50.29088558500323,
"grad_norm": 2.145625114440918,
"learning_rate": 0.001,
"loss": 1.6731,
"step": 155600
},
{
"epoch": 50.323206205559146,
"grad_norm": 2.232274055480957,
"learning_rate": 0.001,
"loss": 1.685,
"step": 155700
},
{
"epoch": 50.35552682611506,
"grad_norm": 2.4370224475860596,
"learning_rate": 0.001,
"loss": 1.6741,
"step": 155800
},
{
"epoch": 50.387847446670975,
"grad_norm": 2.4534711837768555,
"learning_rate": 0.001,
"loss": 1.698,
"step": 155900
},
{
"epoch": 50.42016806722689,
"grad_norm": 2.4986929893493652,
"learning_rate": 0.001,
"loss": 1.6925,
"step": 156000
},
{
"epoch": 50.452488687782804,
"grad_norm": 2.0614304542541504,
"learning_rate": 0.001,
"loss": 1.6956,
"step": 156100
},
{
"epoch": 50.48480930833872,
"grad_norm": 2.957003593444824,
"learning_rate": 0.001,
"loss": 1.7071,
"step": 156200
},
{
"epoch": 50.51712992889463,
"grad_norm": 1.531197190284729,
"learning_rate": 0.001,
"loss": 1.7037,
"step": 156300
},
{
"epoch": 50.54945054945055,
"grad_norm": 1.93813157081604,
"learning_rate": 0.001,
"loss": 1.7091,
"step": 156400
},
{
"epoch": 50.58177117000646,
"grad_norm": 2.1847565174102783,
"learning_rate": 0.001,
"loss": 1.7158,
"step": 156500
},
{
"epoch": 50.61409179056238,
"grad_norm": 1.876352310180664,
"learning_rate": 0.001,
"loss": 1.7317,
"step": 156600
},
{
"epoch": 50.64641241111829,
"grad_norm": 1.9386067390441895,
"learning_rate": 0.001,
"loss": 1.7091,
"step": 156700
},
{
"epoch": 50.678733031674206,
"grad_norm": 1.9585894346237183,
"learning_rate": 0.001,
"loss": 1.7168,
"step": 156800
},
{
"epoch": 50.71105365223012,
"grad_norm": 2.6816301345825195,
"learning_rate": 0.001,
"loss": 1.7478,
"step": 156900
},
{
"epoch": 50.743374272786035,
"grad_norm": 2.5951972007751465,
"learning_rate": 0.001,
"loss": 1.734,
"step": 157000
},
{
"epoch": 50.77569489334195,
"grad_norm": 2.2144699096679688,
"learning_rate": 0.001,
"loss": 1.747,
"step": 157100
},
{
"epoch": 50.808015513897864,
"grad_norm": 2.519739866256714,
"learning_rate": 0.001,
"loss": 1.7408,
"step": 157200
},
{
"epoch": 50.84033613445378,
"grad_norm": 2.1484193801879883,
"learning_rate": 0.001,
"loss": 1.746,
"step": 157300
},
{
"epoch": 50.87265675500969,
"grad_norm": 1.94605553150177,
"learning_rate": 0.001,
"loss": 1.7549,
"step": 157400
},
{
"epoch": 50.90497737556561,
"grad_norm": 2.0627050399780273,
"learning_rate": 0.001,
"loss": 1.7375,
"step": 157500
},
{
"epoch": 50.93729799612152,
"grad_norm": 1.899092674255371,
"learning_rate": 0.001,
"loss": 1.7479,
"step": 157600
},
{
"epoch": 50.96961861667744,
"grad_norm": 2.141066789627075,
"learning_rate": 0.001,
"loss": 1.7563,
"step": 157700
},
{
"epoch": 51.00193923723336,
"grad_norm": 1.6497727632522583,
"learning_rate": 0.001,
"loss": 1.7582,
"step": 157800
},
{
"epoch": 51.03425985778927,
"grad_norm": 1.8637166023254395,
"learning_rate": 0.001,
"loss": 1.6073,
"step": 157900
},
{
"epoch": 51.06658047834519,
"grad_norm": 1.364910364151001,
"learning_rate": 0.001,
"loss": 1.6282,
"step": 158000
},
{
"epoch": 51.0989010989011,
"grad_norm": 1.5030145645141602,
"learning_rate": 0.001,
"loss": 1.6423,
"step": 158100
},
{
"epoch": 51.13122171945702,
"grad_norm": 1.795283555984497,
"learning_rate": 0.001,
"loss": 1.6299,
"step": 158200
},
{
"epoch": 51.16354234001293,
"grad_norm": 1.4750882387161255,
"learning_rate": 0.001,
"loss": 1.649,
"step": 158300
},
{
"epoch": 51.195862960568846,
"grad_norm": 1.403369665145874,
"learning_rate": 0.001,
"loss": 1.674,
"step": 158400
},
{
"epoch": 51.22818358112476,
"grad_norm": 1.5402183532714844,
"learning_rate": 0.001,
"loss": 1.6609,
"step": 158500
},
{
"epoch": 51.260504201680675,
"grad_norm": 1.3701847791671753,
"learning_rate": 0.001,
"loss": 1.6647,
"step": 158600
},
{
"epoch": 51.29282482223659,
"grad_norm": 1.8413647413253784,
"learning_rate": 0.001,
"loss": 1.6594,
"step": 158700
},
{
"epoch": 51.325145442792504,
"grad_norm": 1.3777278661727905,
"learning_rate": 0.001,
"loss": 1.66,
"step": 158800
},
{
"epoch": 51.35746606334842,
"grad_norm": 1.2137295007705688,
"learning_rate": 0.001,
"loss": 1.6628,
"step": 158900
},
{
"epoch": 51.38978668390433,
"grad_norm": 1.5835875272750854,
"learning_rate": 0.001,
"loss": 1.679,
"step": 159000
},
{
"epoch": 51.42210730446025,
"grad_norm": 1.8386844396591187,
"learning_rate": 0.001,
"loss": 1.6919,
"step": 159100
},
{
"epoch": 51.45442792501616,
"grad_norm": 1.222101092338562,
"learning_rate": 0.001,
"loss": 1.6877,
"step": 159200
},
{
"epoch": 51.48674854557208,
"grad_norm": 1.3710792064666748,
"learning_rate": 0.001,
"loss": 1.6906,
"step": 159300
},
{
"epoch": 51.51906916612799,
"grad_norm": 1.6666277647018433,
"learning_rate": 0.001,
"loss": 1.6921,
"step": 159400
},
{
"epoch": 51.551389786683906,
"grad_norm": 1.469344973564148,
"learning_rate": 0.001,
"loss": 1.7079,
"step": 159500
},
{
"epoch": 51.58371040723982,
"grad_norm": 1.5237691402435303,
"learning_rate": 0.001,
"loss": 1.699,
"step": 159600
},
{
"epoch": 51.616031027795735,
"grad_norm": 1.531955599784851,
"learning_rate": 0.001,
"loss": 1.7127,
"step": 159700
},
{
"epoch": 51.64835164835165,
"grad_norm": 1.3699920177459717,
"learning_rate": 0.001,
"loss": 1.7027,
"step": 159800
},
{
"epoch": 51.680672268907564,
"grad_norm": 1.5116556882858276,
"learning_rate": 0.001,
"loss": 1.7133,
"step": 159900
},
{
"epoch": 51.71299288946348,
"grad_norm": 1.7371695041656494,
"learning_rate": 0.001,
"loss": 1.7117,
"step": 160000
},
{
"epoch": 51.74531351001939,
"grad_norm": 1.4322328567504883,
"learning_rate": 0.001,
"loss": 1.7165,
"step": 160100
},
{
"epoch": 51.77763413057531,
"grad_norm": 1.7192296981811523,
"learning_rate": 0.001,
"loss": 1.7403,
"step": 160200
},
{
"epoch": 51.80995475113122,
"grad_norm": 1.9552454948425293,
"learning_rate": 0.001,
"loss": 1.7259,
"step": 160300
},
{
"epoch": 51.84227537168714,
"grad_norm": 1.7755353450775146,
"learning_rate": 0.001,
"loss": 1.7108,
"step": 160400
},
{
"epoch": 51.87459599224305,
"grad_norm": 1.6848562955856323,
"learning_rate": 0.001,
"loss": 1.7365,
"step": 160500
},
{
"epoch": 51.906916612798966,
"grad_norm": 1.2987849712371826,
"learning_rate": 0.001,
"loss": 1.7297,
"step": 160600
},
{
"epoch": 51.93923723335488,
"grad_norm": 1.4124195575714111,
"learning_rate": 0.001,
"loss": 1.7288,
"step": 160700
},
{
"epoch": 51.971557853910795,
"grad_norm": 1.3868930339813232,
"learning_rate": 0.001,
"loss": 1.755,
"step": 160800
},
{
"epoch": 52.00387847446671,
"grad_norm": 1.175879955291748,
"learning_rate": 0.001,
"loss": 1.7681,
"step": 160900
},
{
"epoch": 52.036199095022624,
"grad_norm": 1.1195766925811768,
"learning_rate": 0.001,
"loss": 1.6024,
"step": 161000
},
{
"epoch": 52.06851971557854,
"grad_norm": 1.2098246812820435,
"learning_rate": 0.001,
"loss": 1.5971,
"step": 161100
},
{
"epoch": 52.10084033613445,
"grad_norm": 1.1678169965744019,
"learning_rate": 0.001,
"loss": 1.6382,
"step": 161200
},
{
"epoch": 52.13316095669037,
"grad_norm": 1.3074830770492554,
"learning_rate": 0.001,
"loss": 1.6399,
"step": 161300
},
{
"epoch": 52.16548157724628,
"grad_norm": 1.5640798807144165,
"learning_rate": 0.001,
"loss": 1.6246,
"step": 161400
},
{
"epoch": 52.1978021978022,
"grad_norm": 1.883164882659912,
"learning_rate": 0.001,
"loss": 1.6338,
"step": 161500
},
{
"epoch": 52.23012281835811,
"grad_norm": 1.5124679803848267,
"learning_rate": 0.001,
"loss": 1.6526,
"step": 161600
},
{
"epoch": 52.262443438914026,
"grad_norm": 1.1516411304473877,
"learning_rate": 0.001,
"loss": 1.6522,
"step": 161700
},
{
"epoch": 52.29476405946994,
"grad_norm": 1.2038459777832031,
"learning_rate": 0.001,
"loss": 1.6629,
"step": 161800
},
{
"epoch": 52.327084680025855,
"grad_norm": 1.2261815071105957,
"learning_rate": 0.001,
"loss": 1.6703,
"step": 161900
},
{
"epoch": 52.35940530058177,
"grad_norm": 1.9614273309707642,
"learning_rate": 0.001,
"loss": 1.6717,
"step": 162000
},
{
"epoch": 52.391725921137684,
"grad_norm": 1.4589879512786865,
"learning_rate": 0.001,
"loss": 1.6689,
"step": 162100
},
{
"epoch": 52.4240465416936,
"grad_norm": 1.6024539470672607,
"learning_rate": 0.001,
"loss": 1.6658,
"step": 162200
},
{
"epoch": 52.456367162249514,
"grad_norm": 1.286816120147705,
"learning_rate": 0.001,
"loss": 1.6725,
"step": 162300
},
{
"epoch": 52.48868778280543,
"grad_norm": 1.3517096042633057,
"learning_rate": 0.001,
"loss": 1.6675,
"step": 162400
},
{
"epoch": 52.52100840336134,
"grad_norm": 1.6630303859710693,
"learning_rate": 0.001,
"loss": 1.6732,
"step": 162500
},
{
"epoch": 52.55332902391726,
"grad_norm": 1.4660154581069946,
"learning_rate": 0.001,
"loss": 1.6845,
"step": 162600
},
{
"epoch": 52.58564964447317,
"grad_norm": 1.3145265579223633,
"learning_rate": 0.001,
"loss": 1.6846,
"step": 162700
},
{
"epoch": 52.617970265029086,
"grad_norm": 1.5129708051681519,
"learning_rate": 0.001,
"loss": 1.6995,
"step": 162800
},
{
"epoch": 52.650290885585,
"grad_norm": 1.0445560216903687,
"learning_rate": 0.001,
"loss": 1.71,
"step": 162900
},
{
"epoch": 52.682611506140915,
"grad_norm": 1.1390701532363892,
"learning_rate": 0.001,
"loss": 1.7061,
"step": 163000
},
{
"epoch": 52.71493212669683,
"grad_norm": 1.341044306755066,
"learning_rate": 0.001,
"loss": 1.707,
"step": 163100
},
{
"epoch": 52.747252747252745,
"grad_norm": 1.4199345111846924,
"learning_rate": 0.001,
"loss": 1.7091,
"step": 163200
},
{
"epoch": 52.77957336780866,
"grad_norm": 1.406308650970459,
"learning_rate": 0.001,
"loss": 1.7079,
"step": 163300
},
{
"epoch": 52.811893988364574,
"grad_norm": 1.3773493766784668,
"learning_rate": 0.001,
"loss": 1.7031,
"step": 163400
},
{
"epoch": 52.84421460892049,
"grad_norm": 1.0450148582458496,
"learning_rate": 0.001,
"loss": 1.7244,
"step": 163500
},
{
"epoch": 52.8765352294764,
"grad_norm": 1.5024305582046509,
"learning_rate": 0.001,
"loss": 1.7362,
"step": 163600
},
{
"epoch": 52.90885585003232,
"grad_norm": 1.310044765472412,
"learning_rate": 0.001,
"loss": 1.7239,
"step": 163700
},
{
"epoch": 52.94117647058823,
"grad_norm": 1.3150571584701538,
"learning_rate": 0.001,
"loss": 1.7309,
"step": 163800
},
{
"epoch": 52.97349709114415,
"grad_norm": 1.1534165143966675,
"learning_rate": 0.001,
"loss": 1.739,
"step": 163900
},
{
"epoch": 53.00581771170007,
"grad_norm": 1.321298360824585,
"learning_rate": 0.001,
"loss": 1.7041,
"step": 164000
},
{
"epoch": 53.03813833225598,
"grad_norm": 1.299804925918579,
"learning_rate": 0.001,
"loss": 1.6091,
"step": 164100
},
{
"epoch": 53.0704589528119,
"grad_norm": 1.5923259258270264,
"learning_rate": 0.001,
"loss": 1.5805,
"step": 164200
},
{
"epoch": 53.10277957336781,
"grad_norm": 1.9299325942993164,
"learning_rate": 0.001,
"loss": 1.618,
"step": 164300
},
{
"epoch": 53.135100193923726,
"grad_norm": 1.2914296388626099,
"learning_rate": 0.001,
"loss": 1.6222,
"step": 164400
},
{
"epoch": 53.16742081447964,
"grad_norm": 1.3100395202636719,
"learning_rate": 0.001,
"loss": 1.6342,
"step": 164500
},
{
"epoch": 53.199741435035556,
"grad_norm": 1.1039835214614868,
"learning_rate": 0.001,
"loss": 1.6307,
"step": 164600
},
{
"epoch": 53.23206205559147,
"grad_norm": 1.2956671714782715,
"learning_rate": 0.001,
"loss": 1.6298,
"step": 164700
},
{
"epoch": 53.264382676147385,
"grad_norm": 1.3472180366516113,
"learning_rate": 0.001,
"loss": 1.6535,
"step": 164800
},
{
"epoch": 53.2967032967033,
"grad_norm": 1.6635137796401978,
"learning_rate": 0.001,
"loss": 1.6531,
"step": 164900
},
{
"epoch": 53.329023917259214,
"grad_norm": 1.208890676498413,
"learning_rate": 0.001,
"loss": 1.6571,
"step": 165000
},
{
"epoch": 53.36134453781513,
"grad_norm": 1.1902297735214233,
"learning_rate": 0.001,
"loss": 1.6447,
"step": 165100
},
{
"epoch": 53.39366515837104,
"grad_norm": 1.6317930221557617,
"learning_rate": 0.001,
"loss": 1.662,
"step": 165200
},
{
"epoch": 53.42598577892696,
"grad_norm": 1.278255820274353,
"learning_rate": 0.001,
"loss": 1.6616,
"step": 165300
},
{
"epoch": 53.45830639948287,
"grad_norm": 1.6691334247589111,
"learning_rate": 0.001,
"loss": 1.6593,
"step": 165400
},
{
"epoch": 53.49062702003879,
"grad_norm": 1.4832581281661987,
"learning_rate": 0.001,
"loss": 1.6651,
"step": 165500
},
{
"epoch": 53.5229476405947,
"grad_norm": 1.3829067945480347,
"learning_rate": 0.001,
"loss": 1.6763,
"step": 165600
},
{
"epoch": 53.555268261150616,
"grad_norm": 1.0094833374023438,
"learning_rate": 0.001,
"loss": 1.6801,
"step": 165700
},
{
"epoch": 53.58758888170653,
"grad_norm": 1.3123492002487183,
"learning_rate": 0.001,
"loss": 1.6583,
"step": 165800
},
{
"epoch": 53.619909502262445,
"grad_norm": 1.190656304359436,
"learning_rate": 0.001,
"loss": 1.6765,
"step": 165900
},
{
"epoch": 53.65223012281836,
"grad_norm": 1.2675483226776123,
"learning_rate": 0.001,
"loss": 1.675,
"step": 166000
},
{
"epoch": 53.684550743374274,
"grad_norm": 1.0801315307617188,
"learning_rate": 0.001,
"loss": 1.7036,
"step": 166100
},
{
"epoch": 53.71687136393019,
"grad_norm": 1.2852061986923218,
"learning_rate": 0.001,
"loss": 1.6956,
"step": 166200
},
{
"epoch": 53.7491919844861,
"grad_norm": 1.6678470373153687,
"learning_rate": 0.001,
"loss": 1.6913,
"step": 166300
},
{
"epoch": 53.78151260504202,
"grad_norm": 1.6170557737350464,
"learning_rate": 0.001,
"loss": 1.7111,
"step": 166400
},
{
"epoch": 53.81383322559793,
"grad_norm": 1.2390954494476318,
"learning_rate": 0.001,
"loss": 1.6811,
"step": 166500
},
{
"epoch": 53.84615384615385,
"grad_norm": 1.4333651065826416,
"learning_rate": 0.001,
"loss": 1.6974,
"step": 166600
},
{
"epoch": 53.87847446670976,
"grad_norm": 1.2976824045181274,
"learning_rate": 0.001,
"loss": 1.7074,
"step": 166700
},
{
"epoch": 53.910795087265676,
"grad_norm": 1.3413689136505127,
"learning_rate": 0.001,
"loss": 1.7172,
"step": 166800
},
{
"epoch": 53.94311570782159,
"grad_norm": 1.4473868608474731,
"learning_rate": 0.001,
"loss": 1.7171,
"step": 166900
},
{
"epoch": 53.975436328377505,
"grad_norm": 1.284826636314392,
"learning_rate": 0.001,
"loss": 1.7245,
"step": 167000
},
{
"epoch": 54.00775694893342,
"grad_norm": 1.5123684406280518,
"learning_rate": 0.001,
"loss": 1.7136,
"step": 167100
},
{
"epoch": 54.040077569489334,
"grad_norm": 1.1671874523162842,
"learning_rate": 0.001,
"loss": 1.6013,
"step": 167200
},
{
"epoch": 54.07239819004525,
"grad_norm": 1.1357572078704834,
"learning_rate": 0.001,
"loss": 1.5812,
"step": 167300
},
{
"epoch": 54.10471881060116,
"grad_norm": 1.3527408838272095,
"learning_rate": 0.001,
"loss": 1.5968,
"step": 167400
},
{
"epoch": 54.13703943115708,
"grad_norm": 1.7920830249786377,
"learning_rate": 0.001,
"loss": 1.6004,
"step": 167500
},
{
"epoch": 54.16936005171299,
"grad_norm": 1.3243252038955688,
"learning_rate": 0.001,
"loss": 1.6035,
"step": 167600
},
{
"epoch": 54.20168067226891,
"grad_norm": 1.3215136528015137,
"learning_rate": 0.001,
"loss": 1.6024,
"step": 167700
},
{
"epoch": 54.23400129282482,
"grad_norm": 1.443765640258789,
"learning_rate": 0.001,
"loss": 1.6284,
"step": 167800
},
{
"epoch": 54.266321913380736,
"grad_norm": 1.4005441665649414,
"learning_rate": 0.001,
"loss": 1.6354,
"step": 167900
},
{
"epoch": 54.29864253393665,
"grad_norm": 1.4282481670379639,
"learning_rate": 0.001,
"loss": 1.6293,
"step": 168000
},
{
"epoch": 54.330963154492565,
"grad_norm": 1.4917771816253662,
"learning_rate": 0.001,
"loss": 1.6409,
"step": 168100
},
{
"epoch": 54.36328377504848,
"grad_norm": 1.2454930543899536,
"learning_rate": 0.001,
"loss": 1.6439,
"step": 168200
},
{
"epoch": 54.395604395604394,
"grad_norm": 1.5724455118179321,
"learning_rate": 0.001,
"loss": 1.6449,
"step": 168300
},
{
"epoch": 54.42792501616031,
"grad_norm": 1.3377549648284912,
"learning_rate": 0.001,
"loss": 1.6506,
"step": 168400
},
{
"epoch": 54.46024563671622,
"grad_norm": 1.1285898685455322,
"learning_rate": 0.001,
"loss": 1.6606,
"step": 168500
},
{
"epoch": 54.49256625727214,
"grad_norm": 1.0706590414047241,
"learning_rate": 0.001,
"loss": 1.6509,
"step": 168600
},
{
"epoch": 54.52488687782805,
"grad_norm": 1.223442554473877,
"learning_rate": 0.001,
"loss": 1.6577,
"step": 168700
},
{
"epoch": 54.55720749838397,
"grad_norm": 1.1773663759231567,
"learning_rate": 0.001,
"loss": 1.6662,
"step": 168800
},
{
"epoch": 54.58952811893988,
"grad_norm": 1.327913761138916,
"learning_rate": 0.001,
"loss": 1.6757,
"step": 168900
},
{
"epoch": 54.621848739495796,
"grad_norm": 1.289463758468628,
"learning_rate": 0.001,
"loss": 1.665,
"step": 169000
},
{
"epoch": 54.65416936005171,
"grad_norm": 1.2144544124603271,
"learning_rate": 0.001,
"loss": 1.6782,
"step": 169100
},
{
"epoch": 54.686489980607625,
"grad_norm": 1.2300634384155273,
"learning_rate": 0.001,
"loss": 1.6885,
"step": 169200
},
{
"epoch": 54.71881060116354,
"grad_norm": 1.3137562274932861,
"learning_rate": 0.001,
"loss": 1.6608,
"step": 169300
},
{
"epoch": 54.751131221719454,
"grad_norm": 1.2621924877166748,
"learning_rate": 0.001,
"loss": 1.6889,
"step": 169400
},
{
"epoch": 54.78345184227537,
"grad_norm": 1.669775366783142,
"learning_rate": 0.001,
"loss": 1.6881,
"step": 169500
},
{
"epoch": 54.81577246283128,
"grad_norm": 1.0568089485168457,
"learning_rate": 0.001,
"loss": 1.6953,
"step": 169600
},
{
"epoch": 54.8480930833872,
"grad_norm": 1.399994134902954,
"learning_rate": 0.001,
"loss": 1.6824,
"step": 169700
},
{
"epoch": 54.88041370394311,
"grad_norm": 1.416273832321167,
"learning_rate": 0.001,
"loss": 1.6874,
"step": 169800
},
{
"epoch": 54.91273432449903,
"grad_norm": 1.0285989046096802,
"learning_rate": 0.001,
"loss": 1.6981,
"step": 169900
},
{
"epoch": 54.94505494505494,
"grad_norm": 1.2854260206222534,
"learning_rate": 0.001,
"loss": 1.7173,
"step": 170000
},
{
"epoch": 54.977375565610856,
"grad_norm": 1.643632173538208,
"learning_rate": 0.001,
"loss": 1.7127,
"step": 170100
},
{
"epoch": 55.00969618616678,
"grad_norm": 1.3077548742294312,
"learning_rate": 0.001,
"loss": 1.6783,
"step": 170200
},
{
"epoch": 55.04201680672269,
"grad_norm": 1.3917027711868286,
"learning_rate": 0.001,
"loss": 1.5838,
"step": 170300
},
{
"epoch": 55.07433742727861,
"grad_norm": 1.7883625030517578,
"learning_rate": 0.001,
"loss": 1.5809,
"step": 170400
},
{
"epoch": 55.10665804783452,
"grad_norm": 1.203114628791809,
"learning_rate": 0.001,
"loss": 1.5842,
"step": 170500
},
{
"epoch": 55.138978668390436,
"grad_norm": 1.8607345819473267,
"learning_rate": 0.001,
"loss": 1.5922,
"step": 170600
},
{
"epoch": 55.17129928894635,
"grad_norm": 1.4554990530014038,
"learning_rate": 0.001,
"loss": 1.5769,
"step": 170700
},
{
"epoch": 55.203619909502265,
"grad_norm": 1.9419199228286743,
"learning_rate": 0.001,
"loss": 1.6168,
"step": 170800
},
{
"epoch": 55.23594053005818,
"grad_norm": 1.2351933717727661,
"learning_rate": 0.001,
"loss": 1.636,
"step": 170900
},
{
"epoch": 55.268261150614094,
"grad_norm": 1.5204228162765503,
"learning_rate": 0.001,
"loss": 1.6147,
"step": 171000
},
{
"epoch": 55.30058177117001,
"grad_norm": 1.3980422019958496,
"learning_rate": 0.001,
"loss": 1.6248,
"step": 171100
},
{
"epoch": 55.33290239172592,
"grad_norm": 1.4285188913345337,
"learning_rate": 0.001,
"loss": 1.6258,
"step": 171200
},
{
"epoch": 55.36522301228184,
"grad_norm": 1.2285929918289185,
"learning_rate": 0.001,
"loss": 1.6184,
"step": 171300
},
{
"epoch": 55.39754363283775,
"grad_norm": 1.5338455438613892,
"learning_rate": 0.001,
"loss": 1.6494,
"step": 171400
},
{
"epoch": 55.42986425339367,
"grad_norm": 1.1547831296920776,
"learning_rate": 0.001,
"loss": 1.6404,
"step": 171500
},
{
"epoch": 55.46218487394958,
"grad_norm": 1.5116361379623413,
"learning_rate": 0.001,
"loss": 1.6444,
"step": 171600
},
{
"epoch": 55.494505494505496,
"grad_norm": 1.2315970659255981,
"learning_rate": 0.001,
"loss": 1.6461,
"step": 171700
},
{
"epoch": 55.52682611506141,
"grad_norm": 1.3728704452514648,
"learning_rate": 0.001,
"loss": 1.6427,
"step": 171800
},
{
"epoch": 55.559146735617325,
"grad_norm": 1.6326501369476318,
"learning_rate": 0.001,
"loss": 1.6605,
"step": 171900
},
{
"epoch": 55.59146735617324,
"grad_norm": 1.48503577709198,
"learning_rate": 0.001,
"loss": 1.6611,
"step": 172000
},
{
"epoch": 55.623787976729155,
"grad_norm": 1.2143938541412354,
"learning_rate": 0.001,
"loss": 1.6625,
"step": 172100
},
{
"epoch": 55.65610859728507,
"grad_norm": 1.3616664409637451,
"learning_rate": 0.001,
"loss": 1.6528,
"step": 172200
},
{
"epoch": 55.688429217840984,
"grad_norm": 1.2764620780944824,
"learning_rate": 0.001,
"loss": 1.6576,
"step": 172300
},
{
"epoch": 55.7207498383969,
"grad_norm": 1.3328635692596436,
"learning_rate": 0.001,
"loss": 1.6738,
"step": 172400
},
{
"epoch": 55.75307045895281,
"grad_norm": 1.2406649589538574,
"learning_rate": 0.001,
"loss": 1.6631,
"step": 172500
},
{
"epoch": 55.78539107950873,
"grad_norm": 1.2168680429458618,
"learning_rate": 0.001,
"loss": 1.6711,
"step": 172600
},
{
"epoch": 55.81771170006464,
"grad_norm": 1.2316731214523315,
"learning_rate": 0.001,
"loss": 1.673,
"step": 172700
},
{
"epoch": 55.85003232062056,
"grad_norm": 1.4383785724639893,
"learning_rate": 0.001,
"loss": 1.6822,
"step": 172800
},
{
"epoch": 55.88235294117647,
"grad_norm": 1.5188051462173462,
"learning_rate": 0.001,
"loss": 1.6848,
"step": 172900
},
{
"epoch": 55.914673561732386,
"grad_norm": 1.5495924949645996,
"learning_rate": 0.001,
"loss": 1.6836,
"step": 173000
},
{
"epoch": 55.9469941822883,
"grad_norm": 1.2748793363571167,
"learning_rate": 0.001,
"loss": 1.7036,
"step": 173100
},
{
"epoch": 55.979314802844215,
"grad_norm": 1.400896430015564,
"learning_rate": 0.001,
"loss": 1.6992,
"step": 173200
},
{
"epoch": 56.01163542340013,
"grad_norm": 1.767435073852539,
"learning_rate": 0.001,
"loss": 1.6353,
"step": 173300
},
{
"epoch": 56.043956043956044,
"grad_norm": 1.8928805589675903,
"learning_rate": 0.001,
"loss": 1.5579,
"step": 173400
},
{
"epoch": 56.07627666451196,
"grad_norm": 1.4789179563522339,
"learning_rate": 0.001,
"loss": 1.5707,
"step": 173500
},
{
"epoch": 56.10859728506787,
"grad_norm": 1.8881316184997559,
"learning_rate": 0.001,
"loss": 1.5901,
"step": 173600
},
{
"epoch": 56.14091790562379,
"grad_norm": 1.4080440998077393,
"learning_rate": 0.001,
"loss": 1.5899,
"step": 173700
},
{
"epoch": 56.1732385261797,
"grad_norm": 1.5885981321334839,
"learning_rate": 0.001,
"loss": 1.591,
"step": 173800
},
{
"epoch": 56.20555914673562,
"grad_norm": 1.156201958656311,
"learning_rate": 0.001,
"loss": 1.5842,
"step": 173900
},
{
"epoch": 56.23787976729153,
"grad_norm": 1.5928478240966797,
"learning_rate": 0.001,
"loss": 1.5819,
"step": 174000
},
{
"epoch": 56.270200387847446,
"grad_norm": 1.2533551454544067,
"learning_rate": 0.001,
"loss": 1.6047,
"step": 174100
},
{
"epoch": 56.30252100840336,
"grad_norm": 1.3666526079177856,
"learning_rate": 0.001,
"loss": 1.6154,
"step": 174200
},
{
"epoch": 56.334841628959275,
"grad_norm": 1.387876033782959,
"learning_rate": 0.001,
"loss": 1.6165,
"step": 174300
},
{
"epoch": 56.36716224951519,
"grad_norm": 1.6409032344818115,
"learning_rate": 0.001,
"loss": 1.6202,
"step": 174400
},
{
"epoch": 56.399482870071104,
"grad_norm": 1.8783397674560547,
"learning_rate": 0.001,
"loss": 1.6184,
"step": 174500
},
{
"epoch": 56.43180349062702,
"grad_norm": 1.5471220016479492,
"learning_rate": 0.001,
"loss": 1.6245,
"step": 174600
},
{
"epoch": 56.46412411118293,
"grad_norm": 1.7386776208877563,
"learning_rate": 0.001,
"loss": 1.6154,
"step": 174700
},
{
"epoch": 56.49644473173885,
"grad_norm": 1.4194401502609253,
"learning_rate": 0.001,
"loss": 1.6261,
"step": 174800
},
{
"epoch": 56.52876535229476,
"grad_norm": 1.692075490951538,
"learning_rate": 0.001,
"loss": 1.6293,
"step": 174900
},
{
"epoch": 56.56108597285068,
"grad_norm": 1.5054965019226074,
"learning_rate": 0.001,
"loss": 1.6407,
"step": 175000
},
{
"epoch": 56.59340659340659,
"grad_norm": 2.14496111869812,
"learning_rate": 0.001,
"loss": 1.6589,
"step": 175100
},
{
"epoch": 56.625727213962506,
"grad_norm": 1.6506197452545166,
"learning_rate": 0.001,
"loss": 1.6471,
"step": 175200
},
{
"epoch": 56.65804783451842,
"grad_norm": 1.4753719568252563,
"learning_rate": 0.001,
"loss": 1.669,
"step": 175300
},
{
"epoch": 56.690368455074335,
"grad_norm": 1.2608497142791748,
"learning_rate": 0.001,
"loss": 1.6409,
"step": 175400
},
{
"epoch": 56.72268907563025,
"grad_norm": 1.4934544563293457,
"learning_rate": 0.001,
"loss": 1.661,
"step": 175500
},
{
"epoch": 56.755009696186164,
"grad_norm": 1.8065778017044067,
"learning_rate": 0.001,
"loss": 1.6587,
"step": 175600
},
{
"epoch": 56.78733031674208,
"grad_norm": 1.6587038040161133,
"learning_rate": 0.001,
"loss": 1.6671,
"step": 175700
},
{
"epoch": 56.81965093729799,
"grad_norm": 1.4946959018707275,
"learning_rate": 0.001,
"loss": 1.6765,
"step": 175800
},
{
"epoch": 56.85197155785391,
"grad_norm": 1.2888591289520264,
"learning_rate": 0.001,
"loss": 1.6743,
"step": 175900
},
{
"epoch": 56.88429217840982,
"grad_norm": 1.582378625869751,
"learning_rate": 0.001,
"loss": 1.6815,
"step": 176000
},
{
"epoch": 56.91661279896574,
"grad_norm": 1.1068235635757446,
"learning_rate": 0.001,
"loss": 1.6829,
"step": 176100
},
{
"epoch": 56.94893341952165,
"grad_norm": 1.9266074895858765,
"learning_rate": 0.001,
"loss": 1.6805,
"step": 176200
},
{
"epoch": 56.981254040077566,
"grad_norm": 1.3178945779800415,
"learning_rate": 0.001,
"loss": 1.6845,
"step": 176300
},
{
"epoch": 57.01357466063349,
"grad_norm": 1.8499455451965332,
"learning_rate": 0.001,
"loss": 1.634,
"step": 176400
},
{
"epoch": 57.0458952811894,
"grad_norm": 1.9029045104980469,
"learning_rate": 0.001,
"loss": 1.5555,
"step": 176500
},
{
"epoch": 57.07821590174532,
"grad_norm": 1.8718750476837158,
"learning_rate": 0.001,
"loss": 1.5629,
"step": 176600
},
{
"epoch": 57.11053652230123,
"grad_norm": 2.0384838581085205,
"learning_rate": 0.001,
"loss": 1.5719,
"step": 176700
},
{
"epoch": 57.142857142857146,
"grad_norm": 1.7239034175872803,
"learning_rate": 0.001,
"loss": 1.5764,
"step": 176800
},
{
"epoch": 57.17517776341306,
"grad_norm": 1.6884541511535645,
"learning_rate": 0.001,
"loss": 1.5866,
"step": 176900
},
{
"epoch": 57.207498383968975,
"grad_norm": 1.2400517463684082,
"learning_rate": 0.001,
"loss": 1.5905,
"step": 177000
},
{
"epoch": 57.23981900452489,
"grad_norm": 2.588717222213745,
"learning_rate": 0.001,
"loss": 1.5864,
"step": 177100
},
{
"epoch": 57.272139625080804,
"grad_norm": 1.7481188774108887,
"learning_rate": 0.001,
"loss": 1.5894,
"step": 177200
},
{
"epoch": 57.30446024563672,
"grad_norm": 2.033562421798706,
"learning_rate": 0.001,
"loss": 1.5931,
"step": 177300
},
{
"epoch": 57.33678086619263,
"grad_norm": 1.4571573734283447,
"learning_rate": 0.001,
"loss": 1.6059,
"step": 177400
},
{
"epoch": 57.36910148674855,
"grad_norm": 1.777367353439331,
"learning_rate": 0.001,
"loss": 1.6077,
"step": 177500
},
{
"epoch": 57.40142210730446,
"grad_norm": 1.3451801538467407,
"learning_rate": 0.001,
"loss": 1.6184,
"step": 177600
},
{
"epoch": 57.43374272786038,
"grad_norm": 1.6739017963409424,
"learning_rate": 0.001,
"loss": 1.6222,
"step": 177700
},
{
"epoch": 57.46606334841629,
"grad_norm": 1.3843618631362915,
"learning_rate": 0.001,
"loss": 1.6129,
"step": 177800
},
{
"epoch": 57.498383968972206,
"grad_norm": 1.6705306768417358,
"learning_rate": 0.001,
"loss": 1.6291,
"step": 177900
},
{
"epoch": 57.53070458952812,
"grad_norm": 1.5358141660690308,
"learning_rate": 0.001,
"loss": 1.6269,
"step": 178000
},
{
"epoch": 57.563025210084035,
"grad_norm": 1.8817569017410278,
"learning_rate": 0.001,
"loss": 1.6297,
"step": 178100
},
{
"epoch": 57.59534583063995,
"grad_norm": 1.316579818725586,
"learning_rate": 0.001,
"loss": 1.622,
"step": 178200
},
{
"epoch": 57.627666451195864,
"grad_norm": 1.7809545993804932,
"learning_rate": 0.001,
"loss": 1.6389,
"step": 178300
},
{
"epoch": 57.65998707175178,
"grad_norm": 1.6181464195251465,
"learning_rate": 0.001,
"loss": 1.6285,
"step": 178400
},
{
"epoch": 57.69230769230769,
"grad_norm": 1.4683103561401367,
"learning_rate": 0.001,
"loss": 1.6487,
"step": 178500
},
{
"epoch": 57.72462831286361,
"grad_norm": 1.442793846130371,
"learning_rate": 0.001,
"loss": 1.637,
"step": 178600
},
{
"epoch": 57.75694893341952,
"grad_norm": 2.156679391860962,
"learning_rate": 0.001,
"loss": 1.6598,
"step": 178700
},
{
"epoch": 57.78926955397544,
"grad_norm": 1.9949902296066284,
"learning_rate": 0.001,
"loss": 1.6546,
"step": 178800
},
{
"epoch": 57.82159017453135,
"grad_norm": 1.6317439079284668,
"learning_rate": 0.001,
"loss": 1.6561,
"step": 178900
},
{
"epoch": 57.853910795087266,
"grad_norm": 1.6811563968658447,
"learning_rate": 0.001,
"loss": 1.6544,
"step": 179000
},
{
"epoch": 57.88623141564318,
"grad_norm": 1.2841479778289795,
"learning_rate": 0.001,
"loss": 1.6608,
"step": 179100
},
{
"epoch": 57.918552036199095,
"grad_norm": 1.2849926948547363,
"learning_rate": 0.001,
"loss": 1.6769,
"step": 179200
},
{
"epoch": 57.95087265675501,
"grad_norm": 1.4529975652694702,
"learning_rate": 0.001,
"loss": 1.6866,
"step": 179300
},
{
"epoch": 57.983193277310924,
"grad_norm": 1.4133366346359253,
"learning_rate": 0.001,
"loss": 1.6887,
"step": 179400
},
{
"epoch": 58.01551389786684,
"grad_norm": 1.697737216949463,
"learning_rate": 0.001,
"loss": 1.5984,
"step": 179500
},
{
"epoch": 58.04783451842275,
"grad_norm": 1.773516058921814,
"learning_rate": 0.001,
"loss": 1.5497,
"step": 179600
},
{
"epoch": 58.08015513897867,
"grad_norm": 2.169342041015625,
"learning_rate": 0.001,
"loss": 1.5616,
"step": 179700
},
{
"epoch": 58.11247575953458,
"grad_norm": 1.8722690343856812,
"learning_rate": 0.001,
"loss": 1.5444,
"step": 179800
},
{
"epoch": 58.1447963800905,
"grad_norm": 1.7068400382995605,
"learning_rate": 0.001,
"loss": 1.5625,
"step": 179900
},
{
"epoch": 58.17711700064641,
"grad_norm": 1.6371480226516724,
"learning_rate": 0.001,
"loss": 1.5506,
"step": 180000
},
{
"epoch": 58.209437621202326,
"grad_norm": 1.5536822080612183,
"learning_rate": 0.001,
"loss": 1.574,
"step": 180100
},
{
"epoch": 58.24175824175824,
"grad_norm": 2.0610296726226807,
"learning_rate": 0.001,
"loss": 1.5961,
"step": 180200
},
{
"epoch": 58.274078862314155,
"grad_norm": 1.963131070137024,
"learning_rate": 0.001,
"loss": 1.5763,
"step": 180300
},
{
"epoch": 58.30639948287007,
"grad_norm": 2.2817301750183105,
"learning_rate": 0.001,
"loss": 1.5851,
"step": 180400
},
{
"epoch": 58.338720103425985,
"grad_norm": 1.568538784980774,
"learning_rate": 0.001,
"loss": 1.6115,
"step": 180500
},
{
"epoch": 58.3710407239819,
"grad_norm": 1.942744493484497,
"learning_rate": 0.001,
"loss": 1.6031,
"step": 180600
},
{
"epoch": 58.403361344537814,
"grad_norm": 1.9709571599960327,
"learning_rate": 0.001,
"loss": 1.5887,
"step": 180700
},
{
"epoch": 58.43568196509373,
"grad_norm": 1.5681718587875366,
"learning_rate": 0.001,
"loss": 1.6019,
"step": 180800
},
{
"epoch": 58.46800258564964,
"grad_norm": 1.5243316888809204,
"learning_rate": 0.001,
"loss": 1.6163,
"step": 180900
},
{
"epoch": 58.50032320620556,
"grad_norm": 1.7645105123519897,
"learning_rate": 0.001,
"loss": 1.6096,
"step": 181000
},
{
"epoch": 58.53264382676147,
"grad_norm": 2.2756471633911133,
"learning_rate": 0.001,
"loss": 1.6097,
"step": 181100
},
{
"epoch": 58.56496444731739,
"grad_norm": 2.1444149017333984,
"learning_rate": 0.001,
"loss": 1.6131,
"step": 181200
},
{
"epoch": 58.5972850678733,
"grad_norm": 1.455983281135559,
"learning_rate": 0.001,
"loss": 1.6317,
"step": 181300
},
{
"epoch": 58.629605688429216,
"grad_norm": 2.3032915592193604,
"learning_rate": 0.001,
"loss": 1.6239,
"step": 181400
},
{
"epoch": 58.66192630898513,
"grad_norm": 2.3108255863189697,
"learning_rate": 0.001,
"loss": 1.6273,
"step": 181500
},
{
"epoch": 58.694246929541045,
"grad_norm": 2.0250985622406006,
"learning_rate": 0.001,
"loss": 1.6292,
"step": 181600
},
{
"epoch": 58.72656755009696,
"grad_norm": 1.6472135782241821,
"learning_rate": 0.001,
"loss": 1.639,
"step": 181700
},
{
"epoch": 58.758888170652874,
"grad_norm": 1.7034873962402344,
"learning_rate": 0.001,
"loss": 1.6434,
"step": 181800
},
{
"epoch": 58.79120879120879,
"grad_norm": 2.028351306915283,
"learning_rate": 0.001,
"loss": 1.6759,
"step": 181900
},
{
"epoch": 58.8235294117647,
"grad_norm": 2.236541986465454,
"learning_rate": 0.001,
"loss": 1.6449,
"step": 182000
},
{
"epoch": 58.85585003232062,
"grad_norm": 1.7271288633346558,
"learning_rate": 0.001,
"loss": 1.64,
"step": 182100
},
{
"epoch": 58.88817065287653,
"grad_norm": 1.748860478401184,
"learning_rate": 0.001,
"loss": 1.6534,
"step": 182200
},
{
"epoch": 58.92049127343245,
"grad_norm": 2.8331689834594727,
"learning_rate": 0.001,
"loss": 1.6617,
"step": 182300
},
{
"epoch": 58.95281189398836,
"grad_norm": 1.4291654825210571,
"learning_rate": 0.001,
"loss": 1.6594,
"step": 182400
},
{
"epoch": 58.985132514544276,
"grad_norm": 2.08453369140625,
"learning_rate": 0.001,
"loss": 1.6599,
"step": 182500
},
{
"epoch": 59.0174531351002,
"grad_norm": 1.8425451517105103,
"learning_rate": 0.001,
"loss": 1.5958,
"step": 182600
},
{
"epoch": 59.04977375565611,
"grad_norm": 2.0206732749938965,
"learning_rate": 0.001,
"loss": 1.542,
"step": 182700
},
{
"epoch": 59.08209437621203,
"grad_norm": 1.8501428365707397,
"learning_rate": 0.001,
"loss": 1.5306,
"step": 182800
},
{
"epoch": 59.11441499676794,
"grad_norm": 1.9489905834197998,
"learning_rate": 0.001,
"loss": 1.5419,
"step": 182900
},
{
"epoch": 59.146735617323856,
"grad_norm": 1.2962005138397217,
"learning_rate": 0.001,
"loss": 1.5349,
"step": 183000
},
{
"epoch": 59.17905623787977,
"grad_norm": 1.915640115737915,
"learning_rate": 0.001,
"loss": 1.5515,
"step": 183100
},
{
"epoch": 59.211376858435685,
"grad_norm": 1.4381942749023438,
"learning_rate": 0.001,
"loss": 1.5804,
"step": 183200
},
{
"epoch": 59.2436974789916,
"grad_norm": 1.5542528629302979,
"learning_rate": 0.001,
"loss": 1.572,
"step": 183300
},
{
"epoch": 59.276018099547514,
"grad_norm": 1.5855448246002197,
"learning_rate": 0.001,
"loss": 1.569,
"step": 183400
},
{
"epoch": 59.30833872010343,
"grad_norm": 1.9453139305114746,
"learning_rate": 0.001,
"loss": 1.5789,
"step": 183500
},
{
"epoch": 59.34065934065934,
"grad_norm": 2.0648319721221924,
"learning_rate": 0.001,
"loss": 1.5983,
"step": 183600
},
{
"epoch": 59.37297996121526,
"grad_norm": 1.8654729127883911,
"learning_rate": 0.001,
"loss": 1.5958,
"step": 183700
},
{
"epoch": 59.40530058177117,
"grad_norm": 1.7432748079299927,
"learning_rate": 0.001,
"loss": 1.5944,
"step": 183800
},
{
"epoch": 59.43762120232709,
"grad_norm": 1.445427417755127,
"learning_rate": 0.001,
"loss": 1.5983,
"step": 183900
},
{
"epoch": 59.469941822883,
"grad_norm": 1.6181834936141968,
"learning_rate": 0.001,
"loss": 1.6008,
"step": 184000
},
{
"epoch": 59.502262443438916,
"grad_norm": 1.9870188236236572,
"learning_rate": 0.001,
"loss": 1.6082,
"step": 184100
},
{
"epoch": 59.53458306399483,
"grad_norm": 1.7534323930740356,
"learning_rate": 0.001,
"loss": 1.6049,
"step": 184200
},
{
"epoch": 59.566903684550745,
"grad_norm": 1.6159641742706299,
"learning_rate": 0.001,
"loss": 1.6105,
"step": 184300
},
{
"epoch": 59.59922430510666,
"grad_norm": 1.5817879438400269,
"learning_rate": 0.001,
"loss": 1.6086,
"step": 184400
},
{
"epoch": 59.631544925662574,
"grad_norm": 1.7402788400650024,
"learning_rate": 0.001,
"loss": 1.6051,
"step": 184500
},
{
"epoch": 59.66386554621849,
"grad_norm": 1.5133683681488037,
"learning_rate": 0.001,
"loss": 1.6258,
"step": 184600
},
{
"epoch": 59.6961861667744,
"grad_norm": 1.5968210697174072,
"learning_rate": 0.001,
"loss": 1.6139,
"step": 184700
},
{
"epoch": 59.72850678733032,
"grad_norm": 1.276124358177185,
"learning_rate": 0.001,
"loss": 1.6218,
"step": 184800
},
{
"epoch": 59.76082740788623,
"grad_norm": 1.579935073852539,
"learning_rate": 0.001,
"loss": 1.6193,
"step": 184900
},
{
"epoch": 59.79314802844215,
"grad_norm": 1.643371820449829,
"learning_rate": 0.001,
"loss": 1.6433,
"step": 185000
},
{
"epoch": 59.82546864899806,
"grad_norm": 1.5976533889770508,
"learning_rate": 0.001,
"loss": 1.6489,
"step": 185100
},
{
"epoch": 59.857789269553976,
"grad_norm": 2.231804847717285,
"learning_rate": 0.001,
"loss": 1.6404,
"step": 185200
},
{
"epoch": 59.89010989010989,
"grad_norm": 1.6676446199417114,
"learning_rate": 0.001,
"loss": 1.6353,
"step": 185300
},
{
"epoch": 59.922430510665805,
"grad_norm": 2.071073055267334,
"learning_rate": 0.001,
"loss": 1.6393,
"step": 185400
},
{
"epoch": 59.95475113122172,
"grad_norm": 1.343000888824463,
"learning_rate": 0.001,
"loss": 1.6471,
"step": 185500
},
{
"epoch": 59.987071751777634,
"grad_norm": 1.4601433277130127,
"learning_rate": 0.001,
"loss": 1.6525,
"step": 185600
},
{
"epoch": 60.01939237233355,
"grad_norm": 1.4083068370819092,
"learning_rate": 0.001,
"loss": 1.5965,
"step": 185700
},
{
"epoch": 60.05171299288946,
"grad_norm": 1.1844639778137207,
"learning_rate": 0.001,
"loss": 1.5373,
"step": 185800
},
{
"epoch": 60.08403361344538,
"grad_norm": 1.5114725828170776,
"learning_rate": 0.001,
"loss": 1.5303,
"step": 185900
},
{
"epoch": 60.11635423400129,
"grad_norm": 1.5674611330032349,
"learning_rate": 0.001,
"loss": 1.5521,
"step": 186000
},
{
"epoch": 60.14867485455721,
"grad_norm": 1.1841026544570923,
"learning_rate": 0.001,
"loss": 1.5435,
"step": 186100
},
{
"epoch": 60.18099547511312,
"grad_norm": 2.0247650146484375,
"learning_rate": 0.001,
"loss": 1.5624,
"step": 186200
},
{
"epoch": 60.213316095669036,
"grad_norm": 0.9949104189872742,
"learning_rate": 0.001,
"loss": 1.5615,
"step": 186300
},
{
"epoch": 60.24563671622495,
"grad_norm": 1.1038603782653809,
"learning_rate": 0.001,
"loss": 1.5493,
"step": 186400
},
{
"epoch": 60.277957336780865,
"grad_norm": 1.5646278858184814,
"learning_rate": 0.001,
"loss": 1.5759,
"step": 186500
},
{
"epoch": 60.31027795733678,
"grad_norm": 1.932615041732788,
"learning_rate": 0.001,
"loss": 1.5586,
"step": 186600
},
{
"epoch": 60.342598577892694,
"grad_norm": 1.3406494855880737,
"learning_rate": 0.001,
"loss": 1.5877,
"step": 186700
},
{
"epoch": 60.37491919844861,
"grad_norm": 1.4349496364593506,
"learning_rate": 0.001,
"loss": 1.5631,
"step": 186800
},
{
"epoch": 60.40723981900452,
"grad_norm": 1.6426575183868408,
"learning_rate": 0.001,
"loss": 1.5677,
"step": 186900
},
{
"epoch": 60.43956043956044,
"grad_norm": 1.5145518779754639,
"learning_rate": 0.001,
"loss": 1.576,
"step": 187000
},
{
"epoch": 60.47188106011635,
"grad_norm": 1.412512183189392,
"learning_rate": 0.001,
"loss": 1.5872,
"step": 187100
},
{
"epoch": 60.50420168067227,
"grad_norm": 1.4870507717132568,
"learning_rate": 0.001,
"loss": 1.5901,
"step": 187200
},
{
"epoch": 60.53652230122818,
"grad_norm": 1.7748714685440063,
"learning_rate": 0.001,
"loss": 1.6159,
"step": 187300
},
{
"epoch": 60.568842921784096,
"grad_norm": 1.033477544784546,
"learning_rate": 0.001,
"loss": 1.603,
"step": 187400
},
{
"epoch": 60.60116354234001,
"grad_norm": 1.6205483675003052,
"learning_rate": 0.001,
"loss": 1.5986,
"step": 187500
},
{
"epoch": 60.633484162895925,
"grad_norm": 1.3940643072128296,
"learning_rate": 0.001,
"loss": 1.5985,
"step": 187600
},
{
"epoch": 60.66580478345184,
"grad_norm": 1.635551929473877,
"learning_rate": 0.001,
"loss": 1.6,
"step": 187700
},
{
"epoch": 60.698125404007754,
"grad_norm": 1.3659719228744507,
"learning_rate": 0.001,
"loss": 1.6206,
"step": 187800
},
{
"epoch": 60.73044602456367,
"grad_norm": 1.8461339473724365,
"learning_rate": 0.001,
"loss": 1.6164,
"step": 187900
},
{
"epoch": 60.762766645119584,
"grad_norm": 1.809069037437439,
"learning_rate": 0.001,
"loss": 1.6324,
"step": 188000
},
{
"epoch": 60.7950872656755,
"grad_norm": 1.3764575719833374,
"learning_rate": 0.001,
"loss": 1.6202,
"step": 188100
},
{
"epoch": 60.82740788623141,
"grad_norm": 1.6674128770828247,
"learning_rate": 0.001,
"loss": 1.6376,
"step": 188200
},
{
"epoch": 60.85972850678733,
"grad_norm": 1.4235379695892334,
"learning_rate": 0.001,
"loss": 1.6172,
"step": 188300
},
{
"epoch": 60.89204912734324,
"grad_norm": 1.6299734115600586,
"learning_rate": 0.001,
"loss": 1.6429,
"step": 188400
},
{
"epoch": 60.924369747899156,
"grad_norm": 1.3563331365585327,
"learning_rate": 0.001,
"loss": 1.6367,
"step": 188500
},
{
"epoch": 60.95669036845507,
"grad_norm": 1.5571542978286743,
"learning_rate": 0.001,
"loss": 1.6252,
"step": 188600
},
{
"epoch": 60.98901098901099,
"grad_norm": 1.4315185546875,
"learning_rate": 0.001,
"loss": 1.6418,
"step": 188700
},
{
"epoch": 61.02133160956691,
"grad_norm": 1.47747802734375,
"learning_rate": 0.001,
"loss": 1.5709,
"step": 188800
},
{
"epoch": 61.05365223012282,
"grad_norm": 1.2938754558563232,
"learning_rate": 0.001,
"loss": 1.5142,
"step": 188900
},
{
"epoch": 61.085972850678736,
"grad_norm": 1.4727519750595093,
"learning_rate": 0.001,
"loss": 1.5282,
"step": 189000
},
{
"epoch": 61.11829347123465,
"grad_norm": 1.3483736515045166,
"learning_rate": 0.001,
"loss": 1.5414,
"step": 189100
},
{
"epoch": 61.150614091790565,
"grad_norm": 1.430799961090088,
"learning_rate": 0.001,
"loss": 1.5445,
"step": 189200
},
{
"epoch": 61.18293471234648,
"grad_norm": 1.290225625038147,
"learning_rate": 0.001,
"loss": 1.5475,
"step": 189300
},
{
"epoch": 61.215255332902395,
"grad_norm": 1.470697283744812,
"learning_rate": 0.001,
"loss": 1.5624,
"step": 189400
},
{
"epoch": 61.24757595345831,
"grad_norm": 1.0034763813018799,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 189500
},
{
"epoch": 61.279896574014224,
"grad_norm": 1.661956548690796,
"learning_rate": 0.001,
"loss": 1.5557,
"step": 189600
},
{
"epoch": 61.31221719457014,
"grad_norm": 1.766558289527893,
"learning_rate": 0.001,
"loss": 1.5837,
"step": 189700
},
{
"epoch": 61.34453781512605,
"grad_norm": 1.3936597108840942,
"learning_rate": 0.001,
"loss": 1.5573,
"step": 189800
},
{
"epoch": 61.37685843568197,
"grad_norm": 1.3482015132904053,
"learning_rate": 0.001,
"loss": 1.571,
"step": 189900
},
{
"epoch": 61.40917905623788,
"grad_norm": 1.3592807054519653,
"learning_rate": 0.001,
"loss": 1.5733,
"step": 190000
},
{
"epoch": 61.441499676793796,
"grad_norm": 1.934733271598816,
"learning_rate": 0.001,
"loss": 1.5876,
"step": 190100
},
{
"epoch": 61.47382029734971,
"grad_norm": 1.3082642555236816,
"learning_rate": 0.001,
"loss": 1.5586,
"step": 190200
},
{
"epoch": 61.506140917905626,
"grad_norm": 1.4099860191345215,
"learning_rate": 0.001,
"loss": 1.5857,
"step": 190300
},
{
"epoch": 61.53846153846154,
"grad_norm": 1.6693425178527832,
"learning_rate": 0.001,
"loss": 1.6016,
"step": 190400
},
{
"epoch": 61.570782159017455,
"grad_norm": 1.2620930671691895,
"learning_rate": 0.001,
"loss": 1.5918,
"step": 190500
},
{
"epoch": 61.60310277957337,
"grad_norm": 1.3933327198028564,
"learning_rate": 0.001,
"loss": 1.5915,
"step": 190600
},
{
"epoch": 61.635423400129284,
"grad_norm": 1.3175150156021118,
"learning_rate": 0.001,
"loss": 1.5881,
"step": 190700
},
{
"epoch": 61.6677440206852,
"grad_norm": 1.3536936044692993,
"learning_rate": 0.001,
"loss": 1.5921,
"step": 190800
},
{
"epoch": 61.70006464124111,
"grad_norm": 1.7247731685638428,
"learning_rate": 0.001,
"loss": 1.5975,
"step": 190900
},
{
"epoch": 61.73238526179703,
"grad_norm": 1.1982550621032715,
"learning_rate": 0.001,
"loss": 1.5834,
"step": 191000
},
{
"epoch": 61.76470588235294,
"grad_norm": 1.3262262344360352,
"learning_rate": 0.001,
"loss": 1.6172,
"step": 191100
},
{
"epoch": 61.79702650290886,
"grad_norm": 1.562690019607544,
"learning_rate": 0.001,
"loss": 1.6095,
"step": 191200
},
{
"epoch": 61.82934712346477,
"grad_norm": 1.2212425470352173,
"learning_rate": 0.001,
"loss": 1.6264,
"step": 191300
},
{
"epoch": 61.861667744020686,
"grad_norm": 1.3418647050857544,
"learning_rate": 0.001,
"loss": 1.6229,
"step": 191400
},
{
"epoch": 61.8939883645766,
"grad_norm": 1.789375901222229,
"learning_rate": 0.001,
"loss": 1.6157,
"step": 191500
},
{
"epoch": 61.926308985132515,
"grad_norm": 1.4057437181472778,
"learning_rate": 0.001,
"loss": 1.6138,
"step": 191600
},
{
"epoch": 61.95862960568843,
"grad_norm": 1.521755576133728,
"learning_rate": 0.001,
"loss": 1.6219,
"step": 191700
},
{
"epoch": 61.990950226244344,
"grad_norm": 1.353166937828064,
"learning_rate": 0.001,
"loss": 1.6395,
"step": 191800
},
{
"epoch": 62.02327084680026,
"grad_norm": 1.3344172239303589,
"learning_rate": 0.001,
"loss": 1.5541,
"step": 191900
},
{
"epoch": 62.05559146735617,
"grad_norm": 1.3394560813903809,
"learning_rate": 0.001,
"loss": 1.516,
"step": 192000
},
{
"epoch": 62.08791208791209,
"grad_norm": 1.1963975429534912,
"learning_rate": 0.001,
"loss": 1.5159,
"step": 192100
},
{
"epoch": 62.120232708468,
"grad_norm": 1.3131014108657837,
"learning_rate": 0.001,
"loss": 1.5267,
"step": 192200
},
{
"epoch": 62.15255332902392,
"grad_norm": 1.1254234313964844,
"learning_rate": 0.001,
"loss": 1.5254,
"step": 192300
},
{
"epoch": 62.18487394957983,
"grad_norm": 1.5363491773605347,
"learning_rate": 0.001,
"loss": 1.5356,
"step": 192400
},
{
"epoch": 62.217194570135746,
"grad_norm": 1.2423816919326782,
"learning_rate": 0.001,
"loss": 1.5334,
"step": 192500
},
{
"epoch": 62.24951519069166,
"grad_norm": 1.5482821464538574,
"learning_rate": 0.001,
"loss": 1.5302,
"step": 192600
},
{
"epoch": 62.281835811247575,
"grad_norm": 1.4487342834472656,
"learning_rate": 0.001,
"loss": 1.543,
"step": 192700
},
{
"epoch": 62.31415643180349,
"grad_norm": 1.3104568719863892,
"learning_rate": 0.001,
"loss": 1.542,
"step": 192800
},
{
"epoch": 62.346477052359404,
"grad_norm": 1.4300580024719238,
"learning_rate": 0.001,
"loss": 1.55,
"step": 192900
},
{
"epoch": 62.37879767291532,
"grad_norm": 2.032675266265869,
"learning_rate": 0.001,
"loss": 1.5562,
"step": 193000
},
{
"epoch": 62.41111829347123,
"grad_norm": 1.2694087028503418,
"learning_rate": 0.001,
"loss": 1.5566,
"step": 193100
},
{
"epoch": 62.44343891402715,
"grad_norm": 1.295148491859436,
"learning_rate": 0.001,
"loss": 1.5688,
"step": 193200
},
{
"epoch": 62.47575953458306,
"grad_norm": 1.4521048069000244,
"learning_rate": 0.001,
"loss": 1.5761,
"step": 193300
},
{
"epoch": 62.50808015513898,
"grad_norm": 1.1412447690963745,
"learning_rate": 0.001,
"loss": 1.5575,
"step": 193400
},
{
"epoch": 62.54040077569489,
"grad_norm": 1.0875858068466187,
"learning_rate": 0.001,
"loss": 1.5761,
"step": 193500
},
{
"epoch": 62.572721396250806,
"grad_norm": 1.2577354907989502,
"learning_rate": 0.001,
"loss": 1.5805,
"step": 193600
},
{
"epoch": 62.60504201680672,
"grad_norm": 1.2637401819229126,
"learning_rate": 0.001,
"loss": 1.5883,
"step": 193700
},
{
"epoch": 62.637362637362635,
"grad_norm": 1.1461881399154663,
"learning_rate": 0.001,
"loss": 1.6009,
"step": 193800
},
{
"epoch": 62.66968325791855,
"grad_norm": 1.2505375146865845,
"learning_rate": 0.001,
"loss": 1.5929,
"step": 193900
},
{
"epoch": 62.702003878474464,
"grad_norm": 1.578495740890503,
"learning_rate": 0.001,
"loss": 1.5837,
"step": 194000
},
{
"epoch": 62.73432449903038,
"grad_norm": 1.4622714519500732,
"learning_rate": 0.001,
"loss": 1.5903,
"step": 194100
},
{
"epoch": 62.76664511958629,
"grad_norm": 1.9387978315353394,
"learning_rate": 0.001,
"loss": 1.6012,
"step": 194200
},
{
"epoch": 62.79896574014221,
"grad_norm": 1.3469456434249878,
"learning_rate": 0.001,
"loss": 1.6006,
"step": 194300
},
{
"epoch": 62.83128636069812,
"grad_norm": 1.1447830200195312,
"learning_rate": 0.001,
"loss": 1.6047,
"step": 194400
},
{
"epoch": 62.86360698125404,
"grad_norm": 1.5646306276321411,
"learning_rate": 0.001,
"loss": 1.6068,
"step": 194500
},
{
"epoch": 62.89592760180995,
"grad_norm": 1.5499316453933716,
"learning_rate": 0.001,
"loss": 1.608,
"step": 194600
},
{
"epoch": 62.928248222365866,
"grad_norm": 1.2377431392669678,
"learning_rate": 0.001,
"loss": 1.6237,
"step": 194700
},
{
"epoch": 62.96056884292178,
"grad_norm": 1.3070772886276245,
"learning_rate": 0.001,
"loss": 1.6235,
"step": 194800
},
{
"epoch": 62.992889463477695,
"grad_norm": 1.2487502098083496,
"learning_rate": 0.001,
"loss": 1.619,
"step": 194900
},
{
"epoch": 63.02521008403362,
"grad_norm": 1.7527961730957031,
"learning_rate": 0.001,
"loss": 1.5414,
"step": 195000
},
{
"epoch": 63.05753070458953,
"grad_norm": 1.3711133003234863,
"learning_rate": 0.001,
"loss": 1.4922,
"step": 195100
},
{
"epoch": 63.089851325145446,
"grad_norm": 1.233734369277954,
"learning_rate": 0.001,
"loss": 1.5233,
"step": 195200
},
{
"epoch": 63.12217194570136,
"grad_norm": 1.3673176765441895,
"learning_rate": 0.001,
"loss": 1.519,
"step": 195300
},
{
"epoch": 63.154492566257275,
"grad_norm": 1.436584711074829,
"learning_rate": 0.001,
"loss": 1.5191,
"step": 195400
},
{
"epoch": 63.18681318681319,
"grad_norm": 1.1519837379455566,
"learning_rate": 0.001,
"loss": 1.5264,
"step": 195500
},
{
"epoch": 63.219133807369104,
"grad_norm": 1.4693679809570312,
"learning_rate": 0.001,
"loss": 1.5199,
"step": 195600
},
{
"epoch": 63.25145442792502,
"grad_norm": 1.6729609966278076,
"learning_rate": 0.001,
"loss": 1.5314,
"step": 195700
},
{
"epoch": 63.28377504848093,
"grad_norm": 1.525174856185913,
"learning_rate": 0.001,
"loss": 1.5369,
"step": 195800
},
{
"epoch": 63.31609566903685,
"grad_norm": 1.814919352531433,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 195900
},
{
"epoch": 63.34841628959276,
"grad_norm": 1.3116391897201538,
"learning_rate": 0.001,
"loss": 1.5378,
"step": 196000
},
{
"epoch": 63.38073691014868,
"grad_norm": 1.5028691291809082,
"learning_rate": 0.001,
"loss": 1.5426,
"step": 196100
},
{
"epoch": 63.41305753070459,
"grad_norm": 1.2397243976593018,
"learning_rate": 0.001,
"loss": 1.5528,
"step": 196200
},
{
"epoch": 63.445378151260506,
"grad_norm": 1.314023733139038,
"learning_rate": 0.001,
"loss": 1.5602,
"step": 196300
},
{
"epoch": 63.47769877181642,
"grad_norm": 1.2115651369094849,
"learning_rate": 0.001,
"loss": 1.5512,
"step": 196400
},
{
"epoch": 63.510019392372335,
"grad_norm": 1.1968430280685425,
"learning_rate": 0.001,
"loss": 1.5558,
"step": 196500
},
{
"epoch": 63.54234001292825,
"grad_norm": 1.373708724975586,
"learning_rate": 0.001,
"loss": 1.569,
"step": 196600
},
{
"epoch": 63.574660633484164,
"grad_norm": 1.7067829370498657,
"learning_rate": 0.001,
"loss": 1.5537,
"step": 196700
},
{
"epoch": 63.60698125404008,
"grad_norm": 1.463901162147522,
"learning_rate": 0.001,
"loss": 1.5788,
"step": 196800
},
{
"epoch": 63.63930187459599,
"grad_norm": 1.5052011013031006,
"learning_rate": 0.001,
"loss": 1.5955,
"step": 196900
},
{
"epoch": 63.67162249515191,
"grad_norm": 1.360450267791748,
"learning_rate": 0.001,
"loss": 1.5795,
"step": 197000
},
{
"epoch": 63.70394311570782,
"grad_norm": 1.6841061115264893,
"learning_rate": 0.001,
"loss": 1.5778,
"step": 197100
},
{
"epoch": 63.73626373626374,
"grad_norm": 1.4469643831253052,
"learning_rate": 0.001,
"loss": 1.5761,
"step": 197200
},
{
"epoch": 63.76858435681965,
"grad_norm": 1.6727570295333862,
"learning_rate": 0.001,
"loss": 1.6075,
"step": 197300
},
{
"epoch": 63.800904977375566,
"grad_norm": 1.2299509048461914,
"learning_rate": 0.001,
"loss": 1.5971,
"step": 197400
},
{
"epoch": 63.83322559793148,
"grad_norm": 1.2899246215820312,
"learning_rate": 0.001,
"loss": 1.5827,
"step": 197500
},
{
"epoch": 63.865546218487395,
"grad_norm": 1.2136132717132568,
"learning_rate": 0.001,
"loss": 1.6126,
"step": 197600
},
{
"epoch": 63.89786683904331,
"grad_norm": 1.9382247924804688,
"learning_rate": 0.001,
"loss": 1.5981,
"step": 197700
},
{
"epoch": 63.930187459599225,
"grad_norm": 1.222715973854065,
"learning_rate": 0.001,
"loss": 1.6005,
"step": 197800
},
{
"epoch": 63.96250808015514,
"grad_norm": 1.4854823350906372,
"learning_rate": 0.001,
"loss": 1.6189,
"step": 197900
},
{
"epoch": 63.994828700711054,
"grad_norm": 1.4821515083312988,
"learning_rate": 0.001,
"loss": 1.6035,
"step": 198000
},
{
"epoch": 64.02714932126698,
"grad_norm": 1.3894232511520386,
"learning_rate": 0.001,
"loss": 1.5266,
"step": 198100
},
{
"epoch": 64.05946994182288,
"grad_norm": 1.1838220357894897,
"learning_rate": 0.001,
"loss": 1.4896,
"step": 198200
},
{
"epoch": 64.0917905623788,
"grad_norm": 1.7862881422042847,
"learning_rate": 0.001,
"loss": 1.5006,
"step": 198300
},
{
"epoch": 64.12411118293471,
"grad_norm": 1.383520483970642,
"learning_rate": 0.001,
"loss": 1.4871,
"step": 198400
},
{
"epoch": 64.15643180349063,
"grad_norm": 1.3534070253372192,
"learning_rate": 0.001,
"loss": 1.5074,
"step": 198500
},
{
"epoch": 64.18875242404654,
"grad_norm": 1.5855531692504883,
"learning_rate": 0.001,
"loss": 1.5113,
"step": 198600
},
{
"epoch": 64.22107304460246,
"grad_norm": 1.326485514640808,
"learning_rate": 0.001,
"loss": 1.5134,
"step": 198700
},
{
"epoch": 64.25339366515837,
"grad_norm": 1.305304765701294,
"learning_rate": 0.001,
"loss": 1.5108,
"step": 198800
},
{
"epoch": 64.28571428571429,
"grad_norm": 1.5865510702133179,
"learning_rate": 0.001,
"loss": 1.5363,
"step": 198900
},
{
"epoch": 64.3180349062702,
"grad_norm": 1.5292768478393555,
"learning_rate": 0.001,
"loss": 1.5366,
"step": 199000
},
{
"epoch": 64.35035552682612,
"grad_norm": 1.3583585023880005,
"learning_rate": 0.001,
"loss": 1.5356,
"step": 199100
},
{
"epoch": 64.38267614738203,
"grad_norm": 1.4706687927246094,
"learning_rate": 0.001,
"loss": 1.528,
"step": 199200
},
{
"epoch": 64.41499676793795,
"grad_norm": 1.8095799684524536,
"learning_rate": 0.001,
"loss": 1.5263,
"step": 199300
},
{
"epoch": 64.44731738849386,
"grad_norm": 1.5559029579162598,
"learning_rate": 0.001,
"loss": 1.5461,
"step": 199400
},
{
"epoch": 64.47963800904978,
"grad_norm": 1.390462040901184,
"learning_rate": 0.001,
"loss": 1.5497,
"step": 199500
},
{
"epoch": 64.51195862960569,
"grad_norm": 1.3401960134506226,
"learning_rate": 0.001,
"loss": 1.5655,
"step": 199600
},
{
"epoch": 64.54427925016161,
"grad_norm": 2.0882558822631836,
"learning_rate": 0.001,
"loss": 1.5622,
"step": 199700
},
{
"epoch": 64.57659987071752,
"grad_norm": 1.2662004232406616,
"learning_rate": 0.001,
"loss": 1.5652,
"step": 199800
},
{
"epoch": 64.60892049127344,
"grad_norm": 1.5303183794021606,
"learning_rate": 0.001,
"loss": 1.5864,
"step": 199900
},
{
"epoch": 64.64124111182934,
"grad_norm": 1.6628869771957397,
"learning_rate": 0.001,
"loss": 1.5721,
"step": 200000
},
{
"epoch": 64.67356173238527,
"grad_norm": 1.6711530685424805,
"learning_rate": 0.001,
"loss": 1.4884,
"step": 200100
},
{
"epoch": 64.70588235294117,
"grad_norm": 1.3670395612716675,
"learning_rate": 0.001,
"loss": 1.4958,
"step": 200200
},
{
"epoch": 64.7382029734971,
"grad_norm": 1.8934025764465332,
"learning_rate": 0.001,
"loss": 1.4823,
"step": 200300
},
{
"epoch": 64.770523594053,
"grad_norm": 1.535398006439209,
"learning_rate": 0.001,
"loss": 1.4938,
"step": 200400
},
{
"epoch": 64.80284421460892,
"grad_norm": 1.779222011566162,
"learning_rate": 0.001,
"loss": 1.5043,
"step": 200500
},
{
"epoch": 64.83516483516483,
"grad_norm": 2.1635754108428955,
"learning_rate": 0.001,
"loss": 1.5106,
"step": 200600
},
{
"epoch": 64.86748545572075,
"grad_norm": 1.879019021987915,
"learning_rate": 0.001,
"loss": 1.4919,
"step": 200700
},
{
"epoch": 64.89980607627666,
"grad_norm": 1.5613999366760254,
"learning_rate": 0.001,
"loss": 1.513,
"step": 200800
},
{
"epoch": 64.93212669683258,
"grad_norm": 1.5712066888809204,
"learning_rate": 0.001,
"loss": 1.5197,
"step": 200900
},
{
"epoch": 64.96444731738849,
"grad_norm": 1.6623351573944092,
"learning_rate": 0.001,
"loss": 1.5177,
"step": 201000
},
{
"epoch": 64.99676793794441,
"grad_norm": 1.5711963176727295,
"learning_rate": 0.001,
"loss": 1.5158,
"step": 201100
},
{
"epoch": 65.02908855850032,
"grad_norm": 1.8463070392608643,
"learning_rate": 0.001,
"loss": 1.5001,
"step": 201200
},
{
"epoch": 65.06140917905624,
"grad_norm": 1.3548403978347778,
"learning_rate": 0.001,
"loss": 1.4754,
"step": 201300
},
{
"epoch": 65.09372979961215,
"grad_norm": 1.5633519887924194,
"learning_rate": 0.001,
"loss": 1.4822,
"step": 201400
},
{
"epoch": 65.12605042016807,
"grad_norm": 2.195394277572632,
"learning_rate": 0.001,
"loss": 1.5201,
"step": 201500
},
{
"epoch": 65.15837104072398,
"grad_norm": 1.3407344818115234,
"learning_rate": 0.001,
"loss": 1.5338,
"step": 201600
},
{
"epoch": 65.1906916612799,
"grad_norm": 1.491759181022644,
"learning_rate": 0.001,
"loss": 1.5135,
"step": 201700
},
{
"epoch": 65.2230122818358,
"grad_norm": 1.965517520904541,
"learning_rate": 0.001,
"loss": 1.5129,
"step": 201800
},
{
"epoch": 65.25533290239173,
"grad_norm": 2.0393636226654053,
"learning_rate": 0.001,
"loss": 1.5265,
"step": 201900
},
{
"epoch": 65.28765352294764,
"grad_norm": 1.3810938596725464,
"learning_rate": 0.001,
"loss": 1.5155,
"step": 202000
},
{
"epoch": 65.31997414350356,
"grad_norm": 1.494787573814392,
"learning_rate": 0.001,
"loss": 1.5327,
"step": 202100
},
{
"epoch": 65.35229476405947,
"grad_norm": 1.9144887924194336,
"learning_rate": 0.001,
"loss": 1.5453,
"step": 202200
},
{
"epoch": 65.38461538461539,
"grad_norm": 2.1711816787719727,
"learning_rate": 0.001,
"loss": 1.5294,
"step": 202300
},
{
"epoch": 65.4169360051713,
"grad_norm": 1.414029836654663,
"learning_rate": 0.001,
"loss": 1.5218,
"step": 202400
},
{
"epoch": 65.44925662572722,
"grad_norm": 1.5755435228347778,
"learning_rate": 0.001,
"loss": 1.5375,
"step": 202500
},
{
"epoch": 65.48157724628312,
"grad_norm": 1.657923698425293,
"learning_rate": 0.001,
"loss": 1.5461,
"step": 202600
},
{
"epoch": 65.51389786683905,
"grad_norm": 1.6497408151626587,
"learning_rate": 0.001,
"loss": 1.562,
"step": 202700
},
{
"epoch": 65.54621848739495,
"grad_norm": 1.772306203842163,
"learning_rate": 0.001,
"loss": 1.5543,
"step": 202800
},
{
"epoch": 65.57853910795087,
"grad_norm": 1.7409076690673828,
"learning_rate": 0.001,
"loss": 1.5639,
"step": 202900
},
{
"epoch": 65.61085972850678,
"grad_norm": 1.824103832244873,
"learning_rate": 0.001,
"loss": 1.5732,
"step": 203000
},
{
"epoch": 65.6431803490627,
"grad_norm": 1.385782241821289,
"learning_rate": 0.001,
"loss": 1.5607,
"step": 203100
},
{
"epoch": 65.67550096961861,
"grad_norm": 1.3521980047225952,
"learning_rate": 0.001,
"loss": 1.5717,
"step": 203200
},
{
"epoch": 65.70782159017453,
"grad_norm": 1.5172401666641235,
"learning_rate": 0.001,
"loss": 1.5736,
"step": 203300
},
{
"epoch": 65.74014221073044,
"grad_norm": 1.758492112159729,
"learning_rate": 0.001,
"loss": 1.5723,
"step": 203400
},
{
"epoch": 65.77246283128636,
"grad_norm": 1.65023934841156,
"learning_rate": 0.001,
"loss": 1.5843,
"step": 203500
},
{
"epoch": 65.80478345184227,
"grad_norm": 1.5059475898742676,
"learning_rate": 0.001,
"loss": 1.5633,
"step": 203600
},
{
"epoch": 65.83710407239819,
"grad_norm": 1.3566142320632935,
"learning_rate": 0.001,
"loss": 1.5806,
"step": 203700
},
{
"epoch": 65.8694246929541,
"grad_norm": 1.3451659679412842,
"learning_rate": 0.001,
"loss": 1.5732,
"step": 203800
},
{
"epoch": 65.90174531351002,
"grad_norm": 1.74308180809021,
"learning_rate": 0.001,
"loss": 1.5913,
"step": 203900
},
{
"epoch": 65.93406593406593,
"grad_norm": 1.441088318824768,
"learning_rate": 0.001,
"loss": 1.5971,
"step": 204000
},
{
"epoch": 65.96638655462185,
"grad_norm": 1.7819565534591675,
"learning_rate": 0.001,
"loss": 1.3976,
"step": 204100
},
{
"epoch": 65.99870717517777,
"grad_norm": 1.3798339366912842,
"learning_rate": 0.001,
"loss": 1.3824,
"step": 204200
},
{
"epoch": 66.03102779573368,
"grad_norm": 1.8737510442733765,
"learning_rate": 0.001,
"loss": 1.4765,
"step": 204300
},
{
"epoch": 66.0633484162896,
"grad_norm": 1.5442428588867188,
"learning_rate": 0.001,
"loss": 1.4773,
"step": 204400
},
{
"epoch": 66.0956690368455,
"grad_norm": 1.6266494989395142,
"learning_rate": 0.001,
"loss": 1.5024,
"step": 204500
},
{
"epoch": 66.12798965740143,
"grad_norm": 1.5699108839035034,
"learning_rate": 0.001,
"loss": 1.4983,
"step": 204600
},
{
"epoch": 66.16031027795734,
"grad_norm": 1.9251000881195068,
"learning_rate": 0.001,
"loss": 1.5087,
"step": 204700
},
{
"epoch": 66.19263089851326,
"grad_norm": 1.989813208580017,
"learning_rate": 0.001,
"loss": 1.5202,
"step": 204800
},
{
"epoch": 66.22495151906917,
"grad_norm": 1.866454005241394,
"learning_rate": 0.001,
"loss": 1.4918,
"step": 204900
},
{
"epoch": 66.25727213962509,
"grad_norm": 1.8392757177352905,
"learning_rate": 0.001,
"loss": 1.5216,
"step": 205000
},
{
"epoch": 66.289592760181,
"grad_norm": 1.8731913566589355,
"learning_rate": 0.001,
"loss": 1.5004,
"step": 205100
},
{
"epoch": 66.32191338073692,
"grad_norm": 2.168161153793335,
"learning_rate": 0.001,
"loss": 1.5282,
"step": 205200
},
{
"epoch": 66.35423400129282,
"grad_norm": 1.7269086837768555,
"learning_rate": 0.001,
"loss": 1.52,
"step": 205300
},
{
"epoch": 66.38655462184875,
"grad_norm": 1.8548578023910522,
"learning_rate": 0.001,
"loss": 1.5159,
"step": 205400
},
{
"epoch": 66.41887524240465,
"grad_norm": 1.6398289203643799,
"learning_rate": 0.001,
"loss": 1.525,
"step": 205500
},
{
"epoch": 66.45119586296057,
"grad_norm": 2.155949592590332,
"learning_rate": 0.001,
"loss": 1.5435,
"step": 205600
},
{
"epoch": 66.48351648351648,
"grad_norm": 1.4126454591751099,
"learning_rate": 0.001,
"loss": 1.5203,
"step": 205700
},
{
"epoch": 66.5158371040724,
"grad_norm": 1.9967403411865234,
"learning_rate": 0.001,
"loss": 1.5325,
"step": 205800
},
{
"epoch": 66.54815772462831,
"grad_norm": 1.7237664461135864,
"learning_rate": 0.001,
"loss": 1.5471,
"step": 205900
},
{
"epoch": 66.58047834518423,
"grad_norm": 2.0278284549713135,
"learning_rate": 0.001,
"loss": 1.5538,
"step": 206000
},
{
"epoch": 66.61279896574014,
"grad_norm": 1.6204816102981567,
"learning_rate": 0.001,
"loss": 1.5529,
"step": 206100
},
{
"epoch": 66.64511958629606,
"grad_norm": 1.772289752960205,
"learning_rate": 0.001,
"loss": 1.5468,
"step": 206200
},
{
"epoch": 66.67744020685197,
"grad_norm": 1.5813918113708496,
"learning_rate": 0.001,
"loss": 1.5616,
"step": 206300
},
{
"epoch": 66.70976082740789,
"grad_norm": 1.7922327518463135,
"learning_rate": 0.001,
"loss": 1.5506,
"step": 206400
},
{
"epoch": 66.7420814479638,
"grad_norm": 1.8799208402633667,
"learning_rate": 0.001,
"loss": 1.5642,
"step": 206500
},
{
"epoch": 66.77440206851972,
"grad_norm": 1.8227005004882812,
"learning_rate": 0.001,
"loss": 1.5639,
"step": 206600
},
{
"epoch": 66.80672268907563,
"grad_norm": 2.0856125354766846,
"learning_rate": 0.001,
"loss": 1.5525,
"step": 206700
},
{
"epoch": 66.83904330963155,
"grad_norm": 1.6521490812301636,
"learning_rate": 0.001,
"loss": 1.566,
"step": 206800
},
{
"epoch": 66.87136393018746,
"grad_norm": 1.7233532667160034,
"learning_rate": 0.001,
"loss": 1.5824,
"step": 206900
},
{
"epoch": 66.90368455074338,
"grad_norm": 1.6359426975250244,
"learning_rate": 0.001,
"loss": 1.5835,
"step": 207000
},
{
"epoch": 66.93600517129929,
"grad_norm": 1.8345274925231934,
"learning_rate": 0.001,
"loss": 1.5703,
"step": 207100
},
{
"epoch": 66.96832579185521,
"grad_norm": 1.5283242464065552,
"learning_rate": 0.001,
"loss": 1.5976,
"step": 207200
},
{
"epoch": 67.00064641241111,
"grad_norm": 2.184720039367676,
"learning_rate": 0.001,
"loss": 1.5708,
"step": 207300
},
{
"epoch": 67.03296703296704,
"grad_norm": 1.640566110610962,
"learning_rate": 0.001,
"loss": 1.4555,
"step": 207400
},
{
"epoch": 67.06528765352294,
"grad_norm": 2.8580191135406494,
"learning_rate": 0.001,
"loss": 1.4653,
"step": 207500
},
{
"epoch": 67.09760827407887,
"grad_norm": 1.787933349609375,
"learning_rate": 0.001,
"loss": 1.4795,
"step": 207600
},
{
"epoch": 67.12992889463477,
"grad_norm": 1.5677714347839355,
"learning_rate": 0.001,
"loss": 1.4791,
"step": 207700
},
{
"epoch": 67.1622495151907,
"grad_norm": 1.6460436582565308,
"learning_rate": 0.001,
"loss": 1.4825,
"step": 207800
},
{
"epoch": 67.1945701357466,
"grad_norm": 2.731843948364258,
"learning_rate": 0.001,
"loss": 1.4921,
"step": 207900
},
{
"epoch": 67.22689075630252,
"grad_norm": 2.03779673576355,
"learning_rate": 0.001,
"loss": 1.5049,
"step": 208000
},
{
"epoch": 67.25921137685843,
"grad_norm": 1.7299399375915527,
"learning_rate": 0.001,
"loss": 1.4893,
"step": 208100
},
{
"epoch": 67.29153199741435,
"grad_norm": 2.380204677581787,
"learning_rate": 0.001,
"loss": 1.4969,
"step": 208200
},
{
"epoch": 67.32385261797026,
"grad_norm": 1.9755852222442627,
"learning_rate": 0.001,
"loss": 1.5002,
"step": 208300
},
{
"epoch": 67.35617323852618,
"grad_norm": 2.0376861095428467,
"learning_rate": 0.001,
"loss": 1.509,
"step": 208400
},
{
"epoch": 67.38849385908209,
"grad_norm": 2.283614158630371,
"learning_rate": 0.001,
"loss": 1.5087,
"step": 208500
},
{
"epoch": 67.42081447963801,
"grad_norm": 2.681487798690796,
"learning_rate": 0.001,
"loss": 1.5321,
"step": 208600
},
{
"epoch": 67.45313510019392,
"grad_norm": 1.8941800594329834,
"learning_rate": 0.001,
"loss": 1.5149,
"step": 208700
},
{
"epoch": 67.48545572074984,
"grad_norm": 2.5822458267211914,
"learning_rate": 0.001,
"loss": 1.5335,
"step": 208800
},
{
"epoch": 67.51777634130575,
"grad_norm": 1.9263173341751099,
"learning_rate": 0.001,
"loss": 1.5415,
"step": 208900
},
{
"epoch": 67.55009696186167,
"grad_norm": 1.4145643711090088,
"learning_rate": 0.001,
"loss": 1.5381,
"step": 209000
},
{
"epoch": 67.58241758241758,
"grad_norm": 2.146472454071045,
"learning_rate": 0.001,
"loss": 1.5523,
"step": 209100
},
{
"epoch": 67.6147382029735,
"grad_norm": 1.6899281740188599,
"learning_rate": 0.001,
"loss": 1.5401,
"step": 209200
},
{
"epoch": 67.6470588235294,
"grad_norm": 1.6108721494674683,
"learning_rate": 0.001,
"loss": 1.534,
"step": 209300
},
{
"epoch": 67.67937944408533,
"grad_norm": 1.9928598403930664,
"learning_rate": 0.001,
"loss": 1.5522,
"step": 209400
},
{
"epoch": 67.71170006464124,
"grad_norm": 1.5368319749832153,
"learning_rate": 0.001,
"loss": 1.5389,
"step": 209500
},
{
"epoch": 67.74402068519716,
"grad_norm": 1.7221187353134155,
"learning_rate": 0.001,
"loss": 1.5469,
"step": 209600
},
{
"epoch": 67.77634130575306,
"grad_norm": 2.3714516162872314,
"learning_rate": 0.001,
"loss": 1.5603,
"step": 209700
},
{
"epoch": 67.80866192630899,
"grad_norm": 1.6930738687515259,
"learning_rate": 0.001,
"loss": 1.5404,
"step": 209800
},
{
"epoch": 67.8409825468649,
"grad_norm": 1.9764783382415771,
"learning_rate": 0.001,
"loss": 1.5577,
"step": 209900
},
{
"epoch": 67.87330316742081,
"grad_norm": 2.144608974456787,
"learning_rate": 0.001,
"loss": 1.5815,
"step": 210000
},
{
"epoch": 67.90562378797672,
"grad_norm": 1.342824935913086,
"learning_rate": 0.001,
"loss": 1.5615,
"step": 210100
},
{
"epoch": 67.93794440853264,
"grad_norm": 1.9291375875473022,
"learning_rate": 0.001,
"loss": 1.5681,
"step": 210200
},
{
"epoch": 67.97026502908855,
"grad_norm": 1.9033061265945435,
"learning_rate": 0.001,
"loss": 1.5818,
"step": 210300
},
{
"epoch": 68.00258564964447,
"grad_norm": 1.3095531463623047,
"learning_rate": 0.001,
"loss": 1.5858,
"step": 210400
},
{
"epoch": 68.0349062702004,
"grad_norm": 1.3913495540618896,
"learning_rate": 0.001,
"loss": 1.4425,
"step": 210500
},
{
"epoch": 68.0672268907563,
"grad_norm": 1.5340535640716553,
"learning_rate": 0.001,
"loss": 1.4683,
"step": 210600
},
{
"epoch": 68.09954751131222,
"grad_norm": 1.2134710550308228,
"learning_rate": 0.001,
"loss": 1.474,
"step": 210700
},
{
"epoch": 68.13186813186813,
"grad_norm": 1.3856017589569092,
"learning_rate": 0.001,
"loss": 1.4976,
"step": 210800
},
{
"epoch": 68.16418875242405,
"grad_norm": 1.045962929725647,
"learning_rate": 0.001,
"loss": 1.4762,
"step": 210900
},
{
"epoch": 68.19650937297996,
"grad_norm": 1.5848826169967651,
"learning_rate": 0.001,
"loss": 1.4904,
"step": 211000
},
{
"epoch": 68.22882999353588,
"grad_norm": 1.4302036762237549,
"learning_rate": 0.001,
"loss": 1.4916,
"step": 211100
},
{
"epoch": 68.26115061409179,
"grad_norm": 1.949602723121643,
"learning_rate": 0.001,
"loss": 1.498,
"step": 211200
},
{
"epoch": 68.29347123464771,
"grad_norm": 1.6095678806304932,
"learning_rate": 0.001,
"loss": 1.4915,
"step": 211300
},
{
"epoch": 68.32579185520362,
"grad_norm": 1.4530701637268066,
"learning_rate": 0.001,
"loss": 1.4801,
"step": 211400
},
{
"epoch": 68.35811247575954,
"grad_norm": 1.9351872205734253,
"learning_rate": 0.001,
"loss": 1.4912,
"step": 211500
},
{
"epoch": 68.39043309631545,
"grad_norm": 1.222960114479065,
"learning_rate": 0.001,
"loss": 1.5062,
"step": 211600
},
{
"epoch": 68.42275371687137,
"grad_norm": 1.8181320428848267,
"learning_rate": 0.001,
"loss": 1.5074,
"step": 211700
},
{
"epoch": 68.45507433742728,
"grad_norm": 1.3645000457763672,
"learning_rate": 0.001,
"loss": 1.5146,
"step": 211800
},
{
"epoch": 68.4873949579832,
"grad_norm": 1.497064232826233,
"learning_rate": 0.001,
"loss": 1.5153,
"step": 211900
},
{
"epoch": 68.5197155785391,
"grad_norm": 1.1851557493209839,
"learning_rate": 0.001,
"loss": 1.5154,
"step": 212000
},
{
"epoch": 68.55203619909503,
"grad_norm": 1.2982760667800903,
"learning_rate": 0.001,
"loss": 1.527,
"step": 212100
},
{
"epoch": 68.58435681965094,
"grad_norm": 2.011042356491089,
"learning_rate": 0.001,
"loss": 1.5124,
"step": 212200
},
{
"epoch": 68.61667744020686,
"grad_norm": 1.7797062397003174,
"learning_rate": 0.001,
"loss": 1.535,
"step": 212300
},
{
"epoch": 68.64899806076276,
"grad_norm": 1.3323163986206055,
"learning_rate": 0.001,
"loss": 1.5241,
"step": 212400
},
{
"epoch": 68.68131868131869,
"grad_norm": 1.4919606447219849,
"learning_rate": 0.001,
"loss": 1.5368,
"step": 212500
},
{
"epoch": 68.7136393018746,
"grad_norm": 1.6354610919952393,
"learning_rate": 0.001,
"loss": 1.5333,
"step": 212600
},
{
"epoch": 68.74595992243052,
"grad_norm": 1.5560182332992554,
"learning_rate": 0.001,
"loss": 1.5401,
"step": 212700
},
{
"epoch": 68.77828054298642,
"grad_norm": 1.115702509880066,
"learning_rate": 0.001,
"loss": 1.5327,
"step": 212800
},
{
"epoch": 68.81060116354234,
"grad_norm": 1.519848108291626,
"learning_rate": 0.001,
"loss": 1.5442,
"step": 212900
},
{
"epoch": 68.84292178409825,
"grad_norm": 1.8405401706695557,
"learning_rate": 0.001,
"loss": 1.5645,
"step": 213000
},
{
"epoch": 68.87524240465417,
"grad_norm": 1.8359476327896118,
"learning_rate": 0.001,
"loss": 1.5686,
"step": 213100
},
{
"epoch": 68.90756302521008,
"grad_norm": 2.003203868865967,
"learning_rate": 0.001,
"loss": 1.5629,
"step": 213200
},
{
"epoch": 68.939883645766,
"grad_norm": 1.3785624504089355,
"learning_rate": 0.001,
"loss": 1.5642,
"step": 213300
},
{
"epoch": 68.97220426632191,
"grad_norm": 1.5034955739974976,
"learning_rate": 0.001,
"loss": 1.5562,
"step": 213400
},
{
"epoch": 69.00452488687783,
"grad_norm": 1.3231661319732666,
"learning_rate": 0.001,
"loss": 1.5538,
"step": 213500
},
{
"epoch": 69.03684550743374,
"grad_norm": 1.3676217794418335,
"learning_rate": 0.001,
"loss": 1.4498,
"step": 213600
},
{
"epoch": 69.06916612798966,
"grad_norm": 1.0177333354949951,
"learning_rate": 0.001,
"loss": 1.4523,
"step": 213700
},
{
"epoch": 69.10148674854557,
"grad_norm": 1.2812739610671997,
"learning_rate": 0.001,
"loss": 1.4668,
"step": 213800
},
{
"epoch": 69.13380736910149,
"grad_norm": 1.3418337106704712,
"learning_rate": 0.001,
"loss": 1.466,
"step": 213900
},
{
"epoch": 69.1661279896574,
"grad_norm": 1.736456274986267,
"learning_rate": 0.001,
"loss": 1.4778,
"step": 214000
},
{
"epoch": 69.19844861021332,
"grad_norm": 1.603867769241333,
"learning_rate": 0.001,
"loss": 1.4582,
"step": 214100
},
{
"epoch": 69.23076923076923,
"grad_norm": 1.2322022914886475,
"learning_rate": 0.001,
"loss": 1.4774,
"step": 214200
},
{
"epoch": 69.26308985132515,
"grad_norm": 1.1912455558776855,
"learning_rate": 0.001,
"loss": 1.4797,
"step": 214300
},
{
"epoch": 69.29541047188106,
"grad_norm": 1.6638360023498535,
"learning_rate": 0.001,
"loss": 1.4747,
"step": 214400
},
{
"epoch": 69.32773109243698,
"grad_norm": 1.4112862348556519,
"learning_rate": 0.001,
"loss": 1.501,
"step": 214500
},
{
"epoch": 69.36005171299288,
"grad_norm": 1.2860196828842163,
"learning_rate": 0.001,
"loss": 1.4834,
"step": 214600
},
{
"epoch": 69.3923723335488,
"grad_norm": 1.7012343406677246,
"learning_rate": 0.001,
"loss": 1.4825,
"step": 214700
},
{
"epoch": 69.42469295410471,
"grad_norm": 1.5980138778686523,
"learning_rate": 0.001,
"loss": 1.4897,
"step": 214800
},
{
"epoch": 69.45701357466064,
"grad_norm": 1.6313142776489258,
"learning_rate": 0.001,
"loss": 1.496,
"step": 214900
},
{
"epoch": 69.48933419521654,
"grad_norm": 1.1970393657684326,
"learning_rate": 0.001,
"loss": 1.5187,
"step": 215000
},
{
"epoch": 69.52165481577246,
"grad_norm": 1.4052519798278809,
"learning_rate": 0.001,
"loss": 1.5155,
"step": 215100
},
{
"epoch": 69.55397543632837,
"grad_norm": 1.7367976903915405,
"learning_rate": 0.001,
"loss": 1.5148,
"step": 215200
},
{
"epoch": 69.5862960568843,
"grad_norm": 1.67872154712677,
"learning_rate": 0.001,
"loss": 1.5208,
"step": 215300
},
{
"epoch": 69.6186166774402,
"grad_norm": 1.4014278650283813,
"learning_rate": 0.001,
"loss": 1.5257,
"step": 215400
},
{
"epoch": 69.65093729799612,
"grad_norm": 1.762558102607727,
"learning_rate": 0.001,
"loss": 1.536,
"step": 215500
},
{
"epoch": 69.68325791855203,
"grad_norm": 1.209141492843628,
"learning_rate": 0.001,
"loss": 1.5288,
"step": 215600
},
{
"epoch": 69.71557853910795,
"grad_norm": 1.2840830087661743,
"learning_rate": 0.001,
"loss": 1.5409,
"step": 215700
},
{
"epoch": 69.74789915966386,
"grad_norm": 1.2475364208221436,
"learning_rate": 0.001,
"loss": 1.5293,
"step": 215800
},
{
"epoch": 69.78021978021978,
"grad_norm": 1.285278081893921,
"learning_rate": 0.001,
"loss": 1.5259,
"step": 215900
},
{
"epoch": 69.81254040077569,
"grad_norm": 1.3252419233322144,
"learning_rate": 0.001,
"loss": 1.5353,
"step": 216000
},
{
"epoch": 69.84486102133161,
"grad_norm": 1.42166006565094,
"learning_rate": 0.001,
"loss": 1.5458,
"step": 216100
},
{
"epoch": 69.87718164188752,
"grad_norm": 1.503132939338684,
"learning_rate": 0.001,
"loss": 1.5286,
"step": 216200
},
{
"epoch": 69.90950226244344,
"grad_norm": 1.2355599403381348,
"learning_rate": 0.001,
"loss": 1.5463,
"step": 216300
},
{
"epoch": 69.94182288299935,
"grad_norm": 1.688442349433899,
"learning_rate": 0.001,
"loss": 1.5628,
"step": 216400
},
{
"epoch": 69.97414350355527,
"grad_norm": 1.7759349346160889,
"learning_rate": 0.001,
"loss": 1.5545,
"step": 216500
},
{
"epoch": 70.00646412411119,
"grad_norm": 1.2405682802200317,
"learning_rate": 0.001,
"loss": 1.5436,
"step": 216600
},
{
"epoch": 70.0387847446671,
"grad_norm": 1.2468485832214355,
"learning_rate": 0.001,
"loss": 1.4279,
"step": 216700
},
{
"epoch": 70.07110536522302,
"grad_norm": 1.375011920928955,
"learning_rate": 0.001,
"loss": 1.436,
"step": 216800
},
{
"epoch": 70.10342598577893,
"grad_norm": 1.234799861907959,
"learning_rate": 0.001,
"loss": 1.4564,
"step": 216900
},
{
"epoch": 70.13574660633485,
"grad_norm": 1.3788397312164307,
"learning_rate": 0.001,
"loss": 1.4562,
"step": 217000
},
{
"epoch": 70.16806722689076,
"grad_norm": 1.3660613298416138,
"learning_rate": 0.001,
"loss": 1.4545,
"step": 217100
},
{
"epoch": 70.20038784744668,
"grad_norm": 1.7291772365570068,
"learning_rate": 0.001,
"loss": 1.4587,
"step": 217200
},
{
"epoch": 70.23270846800258,
"grad_norm": 1.3720569610595703,
"learning_rate": 0.001,
"loss": 1.4719,
"step": 217300
},
{
"epoch": 70.2650290885585,
"grad_norm": 1.24492609500885,
"learning_rate": 0.001,
"loss": 1.4652,
"step": 217400
},
{
"epoch": 70.29734970911441,
"grad_norm": 1.2578587532043457,
"learning_rate": 0.001,
"loss": 1.4792,
"step": 217500
},
{
"epoch": 70.32967032967034,
"grad_norm": 1.7657023668289185,
"learning_rate": 0.001,
"loss": 1.4975,
"step": 217600
},
{
"epoch": 70.36199095022624,
"grad_norm": 1.5366085767745972,
"learning_rate": 0.001,
"loss": 1.4846,
"step": 217700
},
{
"epoch": 70.39431157078216,
"grad_norm": 1.4883160591125488,
"learning_rate": 0.001,
"loss": 1.492,
"step": 217800
},
{
"epoch": 70.42663219133807,
"grad_norm": 1.195124626159668,
"learning_rate": 0.001,
"loss": 1.4911,
"step": 217900
},
{
"epoch": 70.458952811894,
"grad_norm": 1.3525762557983398,
"learning_rate": 0.001,
"loss": 1.4933,
"step": 218000
},
{
"epoch": 70.4912734324499,
"grad_norm": 1.7769513130187988,
"learning_rate": 0.001,
"loss": 1.5039,
"step": 218100
},
{
"epoch": 70.52359405300582,
"grad_norm": 1.4911025762557983,
"learning_rate": 0.001,
"loss": 1.5009,
"step": 218200
},
{
"epoch": 70.55591467356173,
"grad_norm": 1.3646998405456543,
"learning_rate": 0.001,
"loss": 1.5218,
"step": 218300
},
{
"epoch": 70.58823529411765,
"grad_norm": 1.505569338798523,
"learning_rate": 0.001,
"loss": 1.5273,
"step": 218400
},
{
"epoch": 70.62055591467356,
"grad_norm": 1.292610764503479,
"learning_rate": 0.001,
"loss": 1.4906,
"step": 218500
},
{
"epoch": 70.65287653522948,
"grad_norm": 1.2737617492675781,
"learning_rate": 0.001,
"loss": 1.5134,
"step": 218600
},
{
"epoch": 70.68519715578539,
"grad_norm": 1.5007330179214478,
"learning_rate": 0.001,
"loss": 1.5022,
"step": 218700
},
{
"epoch": 70.71751777634131,
"grad_norm": 1.6058465242385864,
"learning_rate": 0.001,
"loss": 1.5234,
"step": 218800
},
{
"epoch": 70.74983839689722,
"grad_norm": 1.323441505432129,
"learning_rate": 0.001,
"loss": 1.5181,
"step": 218900
},
{
"epoch": 70.78215901745314,
"grad_norm": 1.3180164098739624,
"learning_rate": 0.001,
"loss": 1.5223,
"step": 219000
},
{
"epoch": 70.81447963800905,
"grad_norm": 1.2104029655456543,
"learning_rate": 0.001,
"loss": 1.5288,
"step": 219100
},
{
"epoch": 70.84680025856497,
"grad_norm": 1.8017244338989258,
"learning_rate": 0.001,
"loss": 1.5477,
"step": 219200
},
{
"epoch": 70.87912087912088,
"grad_norm": 1.3228713274002075,
"learning_rate": 0.001,
"loss": 1.5343,
"step": 219300
},
{
"epoch": 70.9114414996768,
"grad_norm": 1.1716163158416748,
"learning_rate": 0.001,
"loss": 1.5461,
"step": 219400
},
{
"epoch": 70.9437621202327,
"grad_norm": 1.3336851596832275,
"learning_rate": 0.001,
"loss": 1.5327,
"step": 219500
},
{
"epoch": 70.97608274078863,
"grad_norm": 1.1334633827209473,
"learning_rate": 0.001,
"loss": 1.5371,
"step": 219600
},
{
"epoch": 71.00840336134453,
"grad_norm": 1.48135507106781,
"learning_rate": 0.001,
"loss": 1.5323,
"step": 219700
},
{
"epoch": 71.04072398190046,
"grad_norm": 1.7075530290603638,
"learning_rate": 0.001,
"loss": 1.4132,
"step": 219800
},
{
"epoch": 71.07304460245636,
"grad_norm": 1.3903677463531494,
"learning_rate": 0.001,
"loss": 1.4396,
"step": 219900
},
{
"epoch": 71.10536522301229,
"grad_norm": 1.275710940361023,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 220000
},
{
"epoch": 71.13768584356819,
"grad_norm": 1.2800272703170776,
"learning_rate": 0.001,
"loss": 1.4415,
"step": 220100
},
{
"epoch": 71.17000646412411,
"grad_norm": 1.568637490272522,
"learning_rate": 0.001,
"loss": 1.444,
"step": 220200
},
{
"epoch": 71.20232708468002,
"grad_norm": 1.7946425676345825,
"learning_rate": 0.001,
"loss": 1.4615,
"step": 220300
},
{
"epoch": 71.23464770523594,
"grad_norm": 1.3503689765930176,
"learning_rate": 0.001,
"loss": 1.4686,
"step": 220400
},
{
"epoch": 71.26696832579185,
"grad_norm": 1.2413722276687622,
"learning_rate": 0.001,
"loss": 1.4403,
"step": 220500
},
{
"epoch": 71.29928894634777,
"grad_norm": 1.4415651559829712,
"learning_rate": 0.001,
"loss": 1.4763,
"step": 220600
},
{
"epoch": 71.33160956690368,
"grad_norm": 1.4465628862380981,
"learning_rate": 0.001,
"loss": 1.4753,
"step": 220700
},
{
"epoch": 71.3639301874596,
"grad_norm": 1.5236570835113525,
"learning_rate": 0.001,
"loss": 1.4671,
"step": 220800
},
{
"epoch": 71.39625080801551,
"grad_norm": 1.2270818948745728,
"learning_rate": 0.001,
"loss": 1.4969,
"step": 220900
},
{
"epoch": 71.42857142857143,
"grad_norm": 1.6226907968521118,
"learning_rate": 0.001,
"loss": 1.4899,
"step": 221000
},
{
"epoch": 71.46089204912734,
"grad_norm": 1.3776700496673584,
"learning_rate": 0.001,
"loss": 1.4785,
"step": 221100
},
{
"epoch": 71.49321266968326,
"grad_norm": 1.669882893562317,
"learning_rate": 0.001,
"loss": 1.5002,
"step": 221200
},
{
"epoch": 71.52553329023917,
"grad_norm": 1.332726240158081,
"learning_rate": 0.001,
"loss": 1.4859,
"step": 221300
},
{
"epoch": 71.55785391079509,
"grad_norm": 1.3970365524291992,
"learning_rate": 0.001,
"loss": 1.5053,
"step": 221400
},
{
"epoch": 71.590174531351,
"grad_norm": 1.3187315464019775,
"learning_rate": 0.001,
"loss": 1.504,
"step": 221500
},
{
"epoch": 71.62249515190692,
"grad_norm": 1.5884617567062378,
"learning_rate": 0.001,
"loss": 1.5117,
"step": 221600
},
{
"epoch": 71.65481577246283,
"grad_norm": 2.0635294914245605,
"learning_rate": 0.001,
"loss": 1.5124,
"step": 221700
},
{
"epoch": 71.68713639301875,
"grad_norm": 1.8100731372833252,
"learning_rate": 0.001,
"loss": 1.5019,
"step": 221800
},
{
"epoch": 71.71945701357465,
"grad_norm": 1.8987959623336792,
"learning_rate": 0.001,
"loss": 1.4972,
"step": 221900
},
{
"epoch": 71.75177763413058,
"grad_norm": 1.2269151210784912,
"learning_rate": 0.001,
"loss": 1.5139,
"step": 222000
},
{
"epoch": 71.78409825468648,
"grad_norm": 1.923986792564392,
"learning_rate": 0.001,
"loss": 1.5342,
"step": 222100
},
{
"epoch": 71.8164188752424,
"grad_norm": 1.3730978965759277,
"learning_rate": 0.001,
"loss": 1.5249,
"step": 222200
},
{
"epoch": 71.84873949579831,
"grad_norm": 1.1028276681900024,
"learning_rate": 0.001,
"loss": 1.5397,
"step": 222300
},
{
"epoch": 71.88106011635423,
"grad_norm": 1.7008004188537598,
"learning_rate": 0.001,
"loss": 1.5245,
"step": 222400
},
{
"epoch": 71.91338073691014,
"grad_norm": 1.4056791067123413,
"learning_rate": 0.001,
"loss": 1.5266,
"step": 222500
},
{
"epoch": 71.94570135746606,
"grad_norm": 1.2635778188705444,
"learning_rate": 0.001,
"loss": 1.5311,
"step": 222600
},
{
"epoch": 71.97802197802197,
"grad_norm": 1.1199357509613037,
"learning_rate": 0.001,
"loss": 1.5498,
"step": 222700
},
{
"epoch": 72.01034259857789,
"grad_norm": 1.5272423028945923,
"learning_rate": 0.001,
"loss": 1.489,
"step": 222800
},
{
"epoch": 72.04266321913381,
"grad_norm": 1.735084891319275,
"learning_rate": 0.001,
"loss": 1.4213,
"step": 222900
},
{
"epoch": 72.07498383968972,
"grad_norm": 1.2801902294158936,
"learning_rate": 0.001,
"loss": 1.418,
"step": 223000
},
{
"epoch": 72.10730446024564,
"grad_norm": 1.453973412513733,
"learning_rate": 0.001,
"loss": 1.4238,
"step": 223100
},
{
"epoch": 72.13962508080155,
"grad_norm": 1.2727351188659668,
"learning_rate": 0.001,
"loss": 1.4326,
"step": 223200
},
{
"epoch": 72.17194570135747,
"grad_norm": 1.3244452476501465,
"learning_rate": 0.001,
"loss": 1.4372,
"step": 223300
},
{
"epoch": 72.20426632191338,
"grad_norm": 1.7596348524093628,
"learning_rate": 0.001,
"loss": 1.4515,
"step": 223400
},
{
"epoch": 72.2365869424693,
"grad_norm": 1.57429039478302,
"learning_rate": 0.001,
"loss": 1.4495,
"step": 223500
},
{
"epoch": 72.26890756302521,
"grad_norm": 1.4161080121994019,
"learning_rate": 0.001,
"loss": 1.4574,
"step": 223600
},
{
"epoch": 72.30122818358113,
"grad_norm": 1.1939678192138672,
"learning_rate": 0.001,
"loss": 1.4642,
"step": 223700
},
{
"epoch": 72.33354880413704,
"grad_norm": 1.2868698835372925,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 223800
},
{
"epoch": 72.36586942469296,
"grad_norm": 1.3444679975509644,
"learning_rate": 0.001,
"loss": 1.4681,
"step": 223900
},
{
"epoch": 72.39819004524887,
"grad_norm": 1.895476222038269,
"learning_rate": 0.001,
"loss": 1.4707,
"step": 224000
},
{
"epoch": 72.43051066580479,
"grad_norm": 1.4101219177246094,
"learning_rate": 0.001,
"loss": 1.4822,
"step": 224100
},
{
"epoch": 72.4628312863607,
"grad_norm": 1.52824068069458,
"learning_rate": 0.001,
"loss": 1.494,
"step": 224200
},
{
"epoch": 72.49515190691662,
"grad_norm": 1.4788613319396973,
"learning_rate": 0.001,
"loss": 1.4836,
"step": 224300
},
{
"epoch": 72.52747252747253,
"grad_norm": 1.4013487100601196,
"learning_rate": 0.001,
"loss": 1.4926,
"step": 224400
},
{
"epoch": 72.55979314802845,
"grad_norm": 1.2739520072937012,
"learning_rate": 0.001,
"loss": 1.465,
"step": 224500
},
{
"epoch": 72.59211376858435,
"grad_norm": 1.2549688816070557,
"learning_rate": 0.001,
"loss": 1.4955,
"step": 224600
},
{
"epoch": 72.62443438914028,
"grad_norm": 1.3320270776748657,
"learning_rate": 0.001,
"loss": 1.4994,
"step": 224700
},
{
"epoch": 72.65675500969618,
"grad_norm": 1.5977946519851685,
"learning_rate": 0.001,
"loss": 1.4928,
"step": 224800
},
{
"epoch": 72.6890756302521,
"grad_norm": 1.5983556509017944,
"learning_rate": 0.001,
"loss": 1.5055,
"step": 224900
},
{
"epoch": 72.72139625080801,
"grad_norm": 1.669458270072937,
"learning_rate": 0.001,
"loss": 1.5114,
"step": 225000
},
{
"epoch": 72.75371687136393,
"grad_norm": 1.4556900262832642,
"learning_rate": 0.001,
"loss": 1.5098,
"step": 225100
},
{
"epoch": 72.78603749191984,
"grad_norm": 1.2442396879196167,
"learning_rate": 0.001,
"loss": 1.5158,
"step": 225200
},
{
"epoch": 72.81835811247576,
"grad_norm": 1.9617815017700195,
"learning_rate": 0.001,
"loss": 1.5315,
"step": 225300
},
{
"epoch": 72.85067873303167,
"grad_norm": 1.772283673286438,
"learning_rate": 0.001,
"loss": 1.5213,
"step": 225400
},
{
"epoch": 72.88299935358759,
"grad_norm": 1.5784248113632202,
"learning_rate": 0.001,
"loss": 1.5138,
"step": 225500
},
{
"epoch": 72.9153199741435,
"grad_norm": 1.223120093345642,
"learning_rate": 0.001,
"loss": 1.505,
"step": 225600
},
{
"epoch": 72.94764059469942,
"grad_norm": 1.53673255443573,
"learning_rate": 0.001,
"loss": 1.5254,
"step": 225700
},
{
"epoch": 72.97996121525533,
"grad_norm": 1.6584315299987793,
"learning_rate": 0.001,
"loss": 1.5395,
"step": 225800
},
{
"epoch": 73.01228183581125,
"grad_norm": 1.5241203308105469,
"learning_rate": 0.001,
"loss": 1.4784,
"step": 225900
},
{
"epoch": 73.04460245636716,
"grad_norm": 1.4810051918029785,
"learning_rate": 0.001,
"loss": 1.4017,
"step": 226000
},
{
"epoch": 73.07692307692308,
"grad_norm": 1.470883846282959,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 226100
},
{
"epoch": 73.10924369747899,
"grad_norm": 1.4955079555511475,
"learning_rate": 0.001,
"loss": 1.4219,
"step": 226200
},
{
"epoch": 73.14156431803491,
"grad_norm": 1.5130250453948975,
"learning_rate": 0.001,
"loss": 1.4286,
"step": 226300
},
{
"epoch": 73.17388493859082,
"grad_norm": 1.53693687915802,
"learning_rate": 0.001,
"loss": 1.4288,
"step": 226400
},
{
"epoch": 73.20620555914674,
"grad_norm": 1.5430470705032349,
"learning_rate": 0.001,
"loss": 1.4493,
"step": 226500
},
{
"epoch": 73.23852617970265,
"grad_norm": 1.3034330606460571,
"learning_rate": 0.001,
"loss": 1.4308,
"step": 226600
},
{
"epoch": 73.27084680025857,
"grad_norm": 1.6014649868011475,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 226700
},
{
"epoch": 73.30316742081448,
"grad_norm": 1.783376693725586,
"learning_rate": 0.001,
"loss": 1.4478,
"step": 226800
},
{
"epoch": 73.3354880413704,
"grad_norm": 2.0474274158477783,
"learning_rate": 0.001,
"loss": 1.454,
"step": 226900
},
{
"epoch": 73.3678086619263,
"grad_norm": 1.8938475847244263,
"learning_rate": 0.001,
"loss": 1.4582,
"step": 227000
},
{
"epoch": 73.40012928248223,
"grad_norm": 1.5601502656936646,
"learning_rate": 0.001,
"loss": 1.4664,
"step": 227100
},
{
"epoch": 73.43244990303813,
"grad_norm": 1.4591832160949707,
"learning_rate": 0.001,
"loss": 1.4722,
"step": 227200
},
{
"epoch": 73.46477052359405,
"grad_norm": 1.363129734992981,
"learning_rate": 0.001,
"loss": 1.466,
"step": 227300
},
{
"epoch": 73.49709114414996,
"grad_norm": 2.0992987155914307,
"learning_rate": 0.001,
"loss": 1.4778,
"step": 227400
},
{
"epoch": 73.52941176470588,
"grad_norm": 1.6557480096817017,
"learning_rate": 0.001,
"loss": 1.4882,
"step": 227500
},
{
"epoch": 73.56173238526179,
"grad_norm": 1.4169237613677979,
"learning_rate": 0.001,
"loss": 1.4972,
"step": 227600
},
{
"epoch": 73.59405300581771,
"grad_norm": 1.78652024269104,
"learning_rate": 0.001,
"loss": 1.4891,
"step": 227700
},
{
"epoch": 73.62637362637362,
"grad_norm": 1.6920175552368164,
"learning_rate": 0.001,
"loss": 1.4927,
"step": 227800
},
{
"epoch": 73.65869424692954,
"grad_norm": 1.5663740634918213,
"learning_rate": 0.001,
"loss": 1.4982,
"step": 227900
},
{
"epoch": 73.69101486748545,
"grad_norm": 1.8419337272644043,
"learning_rate": 0.001,
"loss": 1.4864,
"step": 228000
},
{
"epoch": 73.72333548804137,
"grad_norm": 1.727178692817688,
"learning_rate": 0.001,
"loss": 1.5059,
"step": 228100
},
{
"epoch": 73.75565610859728,
"grad_norm": 1.5188404321670532,
"learning_rate": 0.001,
"loss": 1.5047,
"step": 228200
},
{
"epoch": 73.7879767291532,
"grad_norm": 1.3117109537124634,
"learning_rate": 0.001,
"loss": 1.5098,
"step": 228300
},
{
"epoch": 73.82029734970911,
"grad_norm": 2.1388890743255615,
"learning_rate": 0.001,
"loss": 1.5044,
"step": 228400
},
{
"epoch": 73.85261797026503,
"grad_norm": 1.1493655443191528,
"learning_rate": 0.001,
"loss": 1.5153,
"step": 228500
},
{
"epoch": 73.88493859082094,
"grad_norm": 1.6524302959442139,
"learning_rate": 0.001,
"loss": 1.5127,
"step": 228600
},
{
"epoch": 73.91725921137686,
"grad_norm": 1.347652792930603,
"learning_rate": 0.001,
"loss": 1.5169,
"step": 228700
},
{
"epoch": 73.94957983193277,
"grad_norm": 1.7815312147140503,
"learning_rate": 0.001,
"loss": 1.5114,
"step": 228800
},
{
"epoch": 73.98190045248869,
"grad_norm": 1.5086166858673096,
"learning_rate": 0.001,
"loss": 1.525,
"step": 228900
},
{
"epoch": 74.01422107304461,
"grad_norm": 1.7434033155441284,
"learning_rate": 0.001,
"loss": 1.4435,
"step": 229000
},
{
"epoch": 74.04654169360052,
"grad_norm": 1.7308634519577026,
"learning_rate": 0.001,
"loss": 1.3981,
"step": 229100
},
{
"epoch": 74.07886231415644,
"grad_norm": 1.5770615339279175,
"learning_rate": 0.001,
"loss": 1.3953,
"step": 229200
},
{
"epoch": 74.11118293471235,
"grad_norm": 1.534825086593628,
"learning_rate": 0.001,
"loss": 1.4256,
"step": 229300
},
{
"epoch": 74.14350355526827,
"grad_norm": 1.6956117153167725,
"learning_rate": 0.001,
"loss": 1.4186,
"step": 229400
},
{
"epoch": 74.17582417582418,
"grad_norm": 1.602620244026184,
"learning_rate": 0.001,
"loss": 1.4326,
"step": 229500
},
{
"epoch": 74.2081447963801,
"grad_norm": 1.4989371299743652,
"learning_rate": 0.001,
"loss": 1.4343,
"step": 229600
},
{
"epoch": 74.240465416936,
"grad_norm": 1.7642656564712524,
"learning_rate": 0.001,
"loss": 1.4283,
"step": 229700
},
{
"epoch": 74.27278603749193,
"grad_norm": 1.8813875913619995,
"learning_rate": 0.001,
"loss": 1.4426,
"step": 229800
},
{
"epoch": 74.30510665804783,
"grad_norm": 1.7679482698440552,
"learning_rate": 0.001,
"loss": 1.433,
"step": 229900
},
{
"epoch": 74.33742727860376,
"grad_norm": 1.737067699432373,
"learning_rate": 0.001,
"loss": 1.4524,
"step": 230000
},
{
"epoch": 74.36974789915966,
"grad_norm": 1.76546049118042,
"learning_rate": 0.001,
"loss": 1.45,
"step": 230100
},
{
"epoch": 74.40206851971558,
"grad_norm": 2.3247852325439453,
"learning_rate": 0.001,
"loss": 1.4635,
"step": 230200
},
{
"epoch": 74.43438914027149,
"grad_norm": 1.626501202583313,
"learning_rate": 0.001,
"loss": 1.4689,
"step": 230300
},
{
"epoch": 74.46670976082741,
"grad_norm": 1.6809604167938232,
"learning_rate": 0.001,
"loss": 1.4603,
"step": 230400
},
{
"epoch": 74.49903038138332,
"grad_norm": 1.7988243103027344,
"learning_rate": 0.001,
"loss": 1.4745,
"step": 230500
},
{
"epoch": 74.53135100193924,
"grad_norm": 1.522340178489685,
"learning_rate": 0.001,
"loss": 1.4685,
"step": 230600
},
{
"epoch": 74.56367162249515,
"grad_norm": 1.9077357053756714,
"learning_rate": 0.001,
"loss": 1.4749,
"step": 230700
},
{
"epoch": 74.59599224305107,
"grad_norm": 1.8070390224456787,
"learning_rate": 0.001,
"loss": 1.4799,
"step": 230800
},
{
"epoch": 74.62831286360698,
"grad_norm": 1.4802895784378052,
"learning_rate": 0.001,
"loss": 1.4811,
"step": 230900
},
{
"epoch": 74.6606334841629,
"grad_norm": 1.5516326427459717,
"learning_rate": 0.001,
"loss": 1.4832,
"step": 231000
},
{
"epoch": 74.69295410471881,
"grad_norm": 1.808691143989563,
"learning_rate": 0.001,
"loss": 1.4805,
"step": 231100
},
{
"epoch": 74.72527472527473,
"grad_norm": 1.4593627452850342,
"learning_rate": 0.001,
"loss": 1.4918,
"step": 231200
},
{
"epoch": 74.75759534583064,
"grad_norm": 2.0450727939605713,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 231300
},
{
"epoch": 74.78991596638656,
"grad_norm": 1.467057466506958,
"learning_rate": 0.001,
"loss": 1.4979,
"step": 231400
},
{
"epoch": 74.82223658694247,
"grad_norm": 2.430788040161133,
"learning_rate": 0.001,
"loss": 1.5032,
"step": 231500
},
{
"epoch": 74.85455720749839,
"grad_norm": 2.257944107055664,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 231600
},
{
"epoch": 74.8868778280543,
"grad_norm": 1.7993685007095337,
"learning_rate": 0.001,
"loss": 1.5109,
"step": 231700
},
{
"epoch": 74.91919844861022,
"grad_norm": 1.9601980447769165,
"learning_rate": 0.001,
"loss": 1.5183,
"step": 231800
},
{
"epoch": 74.95151906916612,
"grad_norm": 1.4908268451690674,
"learning_rate": 0.001,
"loss": 1.5147,
"step": 231900
},
{
"epoch": 74.98383968972205,
"grad_norm": 1.7062606811523438,
"learning_rate": 0.001,
"loss": 1.5211,
"step": 232000
},
{
"epoch": 75.01616031027795,
"grad_norm": 2.2244179248809814,
"learning_rate": 0.001,
"loss": 1.4202,
"step": 232100
},
{
"epoch": 75.04848093083388,
"grad_norm": 2.967813014984131,
"learning_rate": 0.001,
"loss": 1.3988,
"step": 232200
},
{
"epoch": 75.08080155138978,
"grad_norm": 1.8036996126174927,
"learning_rate": 0.001,
"loss": 1.386,
"step": 232300
},
{
"epoch": 75.1131221719457,
"grad_norm": 2.1683189868927,
"learning_rate": 0.001,
"loss": 1.419,
"step": 232400
},
{
"epoch": 75.14544279250161,
"grad_norm": 2.932931900024414,
"learning_rate": 0.001,
"loss": 1.4245,
"step": 232500
},
{
"epoch": 75.17776341305753,
"grad_norm": 2.6210219860076904,
"learning_rate": 0.001,
"loss": 1.4129,
"step": 232600
},
{
"epoch": 75.21008403361344,
"grad_norm": 3.309596300125122,
"learning_rate": 0.001,
"loss": 1.4139,
"step": 232700
},
{
"epoch": 75.24240465416936,
"grad_norm": 1.9989670515060425,
"learning_rate": 0.001,
"loss": 1.4263,
"step": 232800
},
{
"epoch": 75.27472527472527,
"grad_norm": 1.9537405967712402,
"learning_rate": 0.001,
"loss": 1.4487,
"step": 232900
},
{
"epoch": 75.30704589528119,
"grad_norm": 2.422278642654419,
"learning_rate": 0.001,
"loss": 1.4465,
"step": 233000
},
{
"epoch": 75.3393665158371,
"grad_norm": 2.3491227626800537,
"learning_rate": 0.001,
"loss": 1.4383,
"step": 233100
},
{
"epoch": 75.37168713639302,
"grad_norm": 2.1019949913024902,
"learning_rate": 0.001,
"loss": 1.4362,
"step": 233200
},
{
"epoch": 75.40400775694893,
"grad_norm": 2.695218801498413,
"learning_rate": 0.001,
"loss": 1.4497,
"step": 233300
},
{
"epoch": 75.43632837750485,
"grad_norm": 1.9660553932189941,
"learning_rate": 0.001,
"loss": 1.458,
"step": 233400
},
{
"epoch": 75.46864899806076,
"grad_norm": 1.967105746269226,
"learning_rate": 0.001,
"loss": 1.4643,
"step": 233500
},
{
"epoch": 75.50096961861668,
"grad_norm": 2.1649856567382812,
"learning_rate": 0.001,
"loss": 1.4456,
"step": 233600
},
{
"epoch": 75.53329023917259,
"grad_norm": 1.8373686075210571,
"learning_rate": 0.001,
"loss": 1.4674,
"step": 233700
},
{
"epoch": 75.56561085972851,
"grad_norm": 2.12976336479187,
"learning_rate": 0.001,
"loss": 1.4769,
"step": 233800
},
{
"epoch": 75.59793148028442,
"grad_norm": 2.144550085067749,
"learning_rate": 0.001,
"loss": 1.4602,
"step": 233900
},
{
"epoch": 75.63025210084034,
"grad_norm": 3.1020848751068115,
"learning_rate": 0.001,
"loss": 1.4685,
"step": 234000
},
{
"epoch": 75.66257272139624,
"grad_norm": 2.4213857650756836,
"learning_rate": 0.001,
"loss": 1.4697,
"step": 234100
},
{
"epoch": 75.69489334195217,
"grad_norm": 2.1722898483276367,
"learning_rate": 0.001,
"loss": 1.4678,
"step": 234200
},
{
"epoch": 75.72721396250807,
"grad_norm": 1.684613585472107,
"learning_rate": 0.001,
"loss": 1.4831,
"step": 234300
},
{
"epoch": 75.759534583064,
"grad_norm": 2.1862945556640625,
"learning_rate": 0.001,
"loss": 1.5017,
"step": 234400
},
{
"epoch": 75.7918552036199,
"grad_norm": 2.201857805252075,
"learning_rate": 0.001,
"loss": 1.4851,
"step": 234500
},
{
"epoch": 75.82417582417582,
"grad_norm": 2.5875611305236816,
"learning_rate": 0.001,
"loss": 1.4841,
"step": 234600
},
{
"epoch": 75.85649644473173,
"grad_norm": 1.9884369373321533,
"learning_rate": 0.001,
"loss": 1.4974,
"step": 234700
},
{
"epoch": 75.88881706528765,
"grad_norm": 3.2347846031188965,
"learning_rate": 0.001,
"loss": 1.5078,
"step": 234800
},
{
"epoch": 75.92113768584356,
"grad_norm": 2.0802674293518066,
"learning_rate": 0.001,
"loss": 1.5003,
"step": 234900
},
{
"epoch": 75.95345830639948,
"grad_norm": 2.8741228580474854,
"learning_rate": 0.001,
"loss": 1.5101,
"step": 235000
},
{
"epoch": 75.98577892695539,
"grad_norm": 2.135455846786499,
"learning_rate": 0.001,
"loss": 1.5075,
"step": 235100
},
{
"epoch": 76.01809954751131,
"grad_norm": 1.8048343658447266,
"learning_rate": 0.001,
"loss": 1.4565,
"step": 235200
},
{
"epoch": 76.05042016806723,
"grad_norm": 1.3127402067184448,
"learning_rate": 0.001,
"loss": 1.3891,
"step": 235300
},
{
"epoch": 76.08274078862314,
"grad_norm": 1.7419129610061646,
"learning_rate": 0.001,
"loss": 1.3925,
"step": 235400
},
{
"epoch": 76.11506140917906,
"grad_norm": 2.105221748352051,
"learning_rate": 0.001,
"loss": 1.3929,
"step": 235500
},
{
"epoch": 76.14738202973497,
"grad_norm": 1.7963032722473145,
"learning_rate": 0.001,
"loss": 1.3977,
"step": 235600
},
{
"epoch": 76.17970265029089,
"grad_norm": 1.381103754043579,
"learning_rate": 0.001,
"loss": 1.4133,
"step": 235700
},
{
"epoch": 76.2120232708468,
"grad_norm": 1.445009708404541,
"learning_rate": 0.001,
"loss": 1.4154,
"step": 235800
},
{
"epoch": 76.24434389140272,
"grad_norm": 1.4311461448669434,
"learning_rate": 0.001,
"loss": 1.4155,
"step": 235900
},
{
"epoch": 76.27666451195863,
"grad_norm": 2.1381478309631348,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 236000
},
{
"epoch": 76.30898513251455,
"grad_norm": 1.51384437084198,
"learning_rate": 0.001,
"loss": 1.4256,
"step": 236100
},
{
"epoch": 76.34130575307046,
"grad_norm": 1.7977445125579834,
"learning_rate": 0.001,
"loss": 1.4269,
"step": 236200
},
{
"epoch": 76.37362637362638,
"grad_norm": 1.336828351020813,
"learning_rate": 0.001,
"loss": 1.4484,
"step": 236300
},
{
"epoch": 76.40594699418229,
"grad_norm": 1.1708205938339233,
"learning_rate": 0.001,
"loss": 1.4448,
"step": 236400
},
{
"epoch": 76.43826761473821,
"grad_norm": 1.705637812614441,
"learning_rate": 0.001,
"loss": 1.4505,
"step": 236500
},
{
"epoch": 76.47058823529412,
"grad_norm": 1.7289289236068726,
"learning_rate": 0.001,
"loss": 1.4456,
"step": 236600
},
{
"epoch": 76.50290885585004,
"grad_norm": 1.8329416513442993,
"learning_rate": 0.001,
"loss": 1.4425,
"step": 236700
},
{
"epoch": 76.53522947640595,
"grad_norm": 1.7150590419769287,
"learning_rate": 0.001,
"loss": 1.4523,
"step": 236800
},
{
"epoch": 76.56755009696187,
"grad_norm": 1.7819806337356567,
"learning_rate": 0.001,
"loss": 1.473,
"step": 236900
},
{
"epoch": 76.59987071751777,
"grad_norm": 1.5425524711608887,
"learning_rate": 0.001,
"loss": 1.4757,
"step": 237000
},
{
"epoch": 76.6321913380737,
"grad_norm": 1.1247155666351318,
"learning_rate": 0.001,
"loss": 1.4697,
"step": 237100
},
{
"epoch": 76.6645119586296,
"grad_norm": 1.6578909158706665,
"learning_rate": 0.001,
"loss": 1.4583,
"step": 237200
},
{
"epoch": 76.69683257918552,
"grad_norm": 1.2041404247283936,
"learning_rate": 0.001,
"loss": 1.4745,
"step": 237300
},
{
"epoch": 76.72915319974143,
"grad_norm": 1.705149531364441,
"learning_rate": 0.001,
"loss": 1.4875,
"step": 237400
},
{
"epoch": 76.76147382029735,
"grad_norm": 1.6393407583236694,
"learning_rate": 0.001,
"loss": 1.4586,
"step": 237500
},
{
"epoch": 76.79379444085326,
"grad_norm": 1.6923236846923828,
"learning_rate": 0.001,
"loss": 1.4897,
"step": 237600
},
{
"epoch": 76.82611506140918,
"grad_norm": 1.614906907081604,
"learning_rate": 0.001,
"loss": 1.4746,
"step": 237700
},
{
"epoch": 76.85843568196509,
"grad_norm": 2.1430628299713135,
"learning_rate": 0.001,
"loss": 1.4956,
"step": 237800
},
{
"epoch": 76.89075630252101,
"grad_norm": 1.7753914594650269,
"learning_rate": 0.001,
"loss": 1.486,
"step": 237900
},
{
"epoch": 76.92307692307692,
"grad_norm": 2.0059542655944824,
"learning_rate": 0.001,
"loss": 1.4898,
"step": 238000
},
{
"epoch": 76.95539754363284,
"grad_norm": 1.5522229671478271,
"learning_rate": 0.001,
"loss": 1.5014,
"step": 238100
},
{
"epoch": 76.98771816418875,
"grad_norm": 1.8205959796905518,
"learning_rate": 0.001,
"loss": 1.4776,
"step": 238200
},
{
"epoch": 77.02003878474467,
"grad_norm": 1.3158133029937744,
"learning_rate": 0.001,
"loss": 1.4418,
"step": 238300
},
{
"epoch": 77.05235940530058,
"grad_norm": 1.4433218240737915,
"learning_rate": 0.001,
"loss": 1.382,
"step": 238400
},
{
"epoch": 77.0846800258565,
"grad_norm": 1.2018743753433228,
"learning_rate": 0.001,
"loss": 1.3971,
"step": 238500
},
{
"epoch": 77.11700064641241,
"grad_norm": 1.389436960220337,
"learning_rate": 0.001,
"loss": 1.4006,
"step": 238600
},
{
"epoch": 77.14932126696833,
"grad_norm": 1.3059638738632202,
"learning_rate": 0.001,
"loss": 1.4039,
"step": 238700
},
{
"epoch": 77.18164188752424,
"grad_norm": 1.8309720754623413,
"learning_rate": 0.001,
"loss": 1.4085,
"step": 238800
},
{
"epoch": 77.21396250808016,
"grad_norm": 1.0015380382537842,
"learning_rate": 0.001,
"loss": 1.4163,
"step": 238900
},
{
"epoch": 77.24628312863607,
"grad_norm": 1.542875051498413,
"learning_rate": 0.001,
"loss": 1.4042,
"step": 239000
},
{
"epoch": 77.27860374919199,
"grad_norm": 1.3617465496063232,
"learning_rate": 0.001,
"loss": 1.4172,
"step": 239100
},
{
"epoch": 77.3109243697479,
"grad_norm": 1.180490255355835,
"learning_rate": 0.001,
"loss": 1.4358,
"step": 239200
},
{
"epoch": 77.34324499030382,
"grad_norm": 1.5970758199691772,
"learning_rate": 0.001,
"loss": 1.414,
"step": 239300
},
{
"epoch": 77.37556561085972,
"grad_norm": 1.1102935075759888,
"learning_rate": 0.001,
"loss": 1.4356,
"step": 239400
},
{
"epoch": 77.40788623141565,
"grad_norm": 1.6913446187973022,
"learning_rate": 0.001,
"loss": 1.4347,
"step": 239500
},
{
"epoch": 77.44020685197155,
"grad_norm": 1.8373421430587769,
"learning_rate": 0.001,
"loss": 1.44,
"step": 239600
},
{
"epoch": 77.47252747252747,
"grad_norm": 1.3673992156982422,
"learning_rate": 0.001,
"loss": 1.4487,
"step": 239700
},
{
"epoch": 77.50484809308338,
"grad_norm": 1.2254825830459595,
"learning_rate": 0.001,
"loss": 1.4322,
"step": 239800
},
{
"epoch": 77.5371687136393,
"grad_norm": 1.5254865884780884,
"learning_rate": 0.001,
"loss": 1.4477,
"step": 239900
},
{
"epoch": 77.56948933419521,
"grad_norm": 1.1001614332199097,
"learning_rate": 0.001,
"loss": 1.4487,
"step": 240000
},
{
"epoch": 77.60180995475113,
"grad_norm": 1.5636470317840576,
"learning_rate": 0.001,
"loss": 1.4462,
"step": 240100
},
{
"epoch": 77.63413057530704,
"grad_norm": 1.3541321754455566,
"learning_rate": 0.001,
"loss": 1.4686,
"step": 240200
},
{
"epoch": 77.66645119586296,
"grad_norm": 1.5194716453552246,
"learning_rate": 0.001,
"loss": 1.4682,
"step": 240300
},
{
"epoch": 77.69877181641887,
"grad_norm": 1.3677459955215454,
"learning_rate": 0.001,
"loss": 1.4641,
"step": 240400
},
{
"epoch": 77.73109243697479,
"grad_norm": 1.2505600452423096,
"learning_rate": 0.001,
"loss": 1.4731,
"step": 240500
},
{
"epoch": 77.7634130575307,
"grad_norm": 1.9318885803222656,
"learning_rate": 0.001,
"loss": 1.4637,
"step": 240600
},
{
"epoch": 77.79573367808662,
"grad_norm": 1.281347393989563,
"learning_rate": 0.001,
"loss": 1.4755,
"step": 240700
},
{
"epoch": 77.82805429864253,
"grad_norm": 1.1534476280212402,
"learning_rate": 0.001,
"loss": 1.4858,
"step": 240800
},
{
"epoch": 77.86037491919845,
"grad_norm": 1.7815430164337158,
"learning_rate": 0.001,
"loss": 1.4736,
"step": 240900
},
{
"epoch": 77.89269553975436,
"grad_norm": 1.4169514179229736,
"learning_rate": 0.001,
"loss": 1.4705,
"step": 241000
},
{
"epoch": 77.92501616031028,
"grad_norm": 1.6190577745437622,
"learning_rate": 0.001,
"loss": 1.4857,
"step": 241100
},
{
"epoch": 77.95733678086619,
"grad_norm": 1.2750232219696045,
"learning_rate": 0.001,
"loss": 1.4855,
"step": 241200
},
{
"epoch": 77.98965740142211,
"grad_norm": 1.9691710472106934,
"learning_rate": 0.001,
"loss": 1.4847,
"step": 241300
},
{
"epoch": 78.02197802197803,
"grad_norm": 1.8732845783233643,
"learning_rate": 0.001,
"loss": 1.4216,
"step": 241400
},
{
"epoch": 78.05429864253394,
"grad_norm": 1.4587857723236084,
"learning_rate": 0.001,
"loss": 1.375,
"step": 241500
},
{
"epoch": 78.08661926308986,
"grad_norm": 1.6787368059158325,
"learning_rate": 0.001,
"loss": 1.3748,
"step": 241600
},
{
"epoch": 78.11893988364577,
"grad_norm": 1.9325082302093506,
"learning_rate": 0.001,
"loss": 1.3799,
"step": 241700
},
{
"epoch": 78.15126050420169,
"grad_norm": 1.581204891204834,
"learning_rate": 0.001,
"loss": 1.3954,
"step": 241800
},
{
"epoch": 78.1835811247576,
"grad_norm": 1.268236517906189,
"learning_rate": 0.001,
"loss": 1.4081,
"step": 241900
},
{
"epoch": 78.21590174531352,
"grad_norm": 1.1501832008361816,
"learning_rate": 0.001,
"loss": 1.3864,
"step": 242000
},
{
"epoch": 78.24822236586942,
"grad_norm": 1.5832775831222534,
"learning_rate": 0.001,
"loss": 1.4037,
"step": 242100
},
{
"epoch": 78.28054298642535,
"grad_norm": 1.183275580406189,
"learning_rate": 0.001,
"loss": 1.4192,
"step": 242200
},
{
"epoch": 78.31286360698125,
"grad_norm": 1.3821377754211426,
"learning_rate": 0.001,
"loss": 1.4132,
"step": 242300
},
{
"epoch": 78.34518422753717,
"grad_norm": 1.8374381065368652,
"learning_rate": 0.001,
"loss": 1.4166,
"step": 242400
},
{
"epoch": 78.37750484809308,
"grad_norm": 1.411993145942688,
"learning_rate": 0.001,
"loss": 1.4339,
"step": 242500
},
{
"epoch": 78.409825468649,
"grad_norm": 1.4325950145721436,
"learning_rate": 0.001,
"loss": 1.4164,
"step": 242600
},
{
"epoch": 78.44214608920491,
"grad_norm": 1.267930030822754,
"learning_rate": 0.001,
"loss": 1.4269,
"step": 242700
},
{
"epoch": 78.47446670976083,
"grad_norm": 1.6772940158843994,
"learning_rate": 0.001,
"loss": 1.4347,
"step": 242800
},
{
"epoch": 78.50678733031674,
"grad_norm": 1.1386587619781494,
"learning_rate": 0.001,
"loss": 1.4254,
"step": 242900
},
{
"epoch": 78.53910795087266,
"grad_norm": 1.543060302734375,
"learning_rate": 0.001,
"loss": 1.4602,
"step": 243000
},
{
"epoch": 78.57142857142857,
"grad_norm": 1.7481422424316406,
"learning_rate": 0.001,
"loss": 1.4502,
"step": 243100
},
{
"epoch": 78.60374919198449,
"grad_norm": 1.3415696620941162,
"learning_rate": 0.001,
"loss": 1.4378,
"step": 243200
},
{
"epoch": 78.6360698125404,
"grad_norm": 1.9283690452575684,
"learning_rate": 0.001,
"loss": 1.4477,
"step": 243300
},
{
"epoch": 78.66839043309632,
"grad_norm": 1.751171350479126,
"learning_rate": 0.001,
"loss": 1.4509,
"step": 243400
},
{
"epoch": 78.70071105365223,
"grad_norm": 1.3747504949569702,
"learning_rate": 0.001,
"loss": 1.457,
"step": 243500
},
{
"epoch": 78.73303167420815,
"grad_norm": 1.1686395406723022,
"learning_rate": 0.001,
"loss": 1.463,
"step": 243600
},
{
"epoch": 78.76535229476406,
"grad_norm": 1.3300254344940186,
"learning_rate": 0.001,
"loss": 1.466,
"step": 243700
},
{
"epoch": 78.79767291531998,
"grad_norm": 1.3648974895477295,
"learning_rate": 0.001,
"loss": 1.4751,
"step": 243800
},
{
"epoch": 78.82999353587589,
"grad_norm": 1.5902628898620605,
"learning_rate": 0.001,
"loss": 1.4767,
"step": 243900
},
{
"epoch": 78.86231415643181,
"grad_norm": 1.3859655857086182,
"learning_rate": 0.001,
"loss": 1.4816,
"step": 244000
},
{
"epoch": 78.89463477698771,
"grad_norm": 1.2946065664291382,
"learning_rate": 0.001,
"loss": 1.4648,
"step": 244100
},
{
"epoch": 78.92695539754364,
"grad_norm": 1.4596238136291504,
"learning_rate": 0.001,
"loss": 1.4797,
"step": 244200
},
{
"epoch": 78.95927601809954,
"grad_norm": 1.8877861499786377,
"learning_rate": 0.001,
"loss": 1.4949,
"step": 244300
},
{
"epoch": 78.99159663865547,
"grad_norm": 1.3379746675491333,
"learning_rate": 0.001,
"loss": 1.4773,
"step": 244400
},
{
"epoch": 79.02391725921137,
"grad_norm": 1.222055196762085,
"learning_rate": 0.001,
"loss": 1.4021,
"step": 244500
},
{
"epoch": 79.0562378797673,
"grad_norm": 1.5215052366256714,
"learning_rate": 0.001,
"loss": 1.3588,
"step": 244600
},
{
"epoch": 79.0885585003232,
"grad_norm": 1.3643509149551392,
"learning_rate": 0.001,
"loss": 1.3556,
"step": 244700
},
{
"epoch": 79.12087912087912,
"grad_norm": 1.1437932252883911,
"learning_rate": 0.001,
"loss": 1.373,
"step": 244800
},
{
"epoch": 79.15319974143503,
"grad_norm": 1.728434681892395,
"learning_rate": 0.001,
"loss": 1.3982,
"step": 244900
},
{
"epoch": 79.18552036199095,
"grad_norm": 1.548836588859558,
"learning_rate": 0.001,
"loss": 1.3995,
"step": 245000
},
{
"epoch": 79.21784098254686,
"grad_norm": 1.6287895441055298,
"learning_rate": 0.001,
"loss": 1.4102,
"step": 245100
},
{
"epoch": 79.25016160310278,
"grad_norm": 1.41975736618042,
"learning_rate": 0.001,
"loss": 1.3977,
"step": 245200
},
{
"epoch": 79.28248222365869,
"grad_norm": 1.7041893005371094,
"learning_rate": 0.001,
"loss": 1.4033,
"step": 245300
},
{
"epoch": 79.31480284421461,
"grad_norm": 1.2006598711013794,
"learning_rate": 0.001,
"loss": 1.3992,
"step": 245400
},
{
"epoch": 79.34712346477052,
"grad_norm": 1.4916554689407349,
"learning_rate": 0.001,
"loss": 1.4173,
"step": 245500
},
{
"epoch": 79.37944408532644,
"grad_norm": 1.2780492305755615,
"learning_rate": 0.001,
"loss": 1.411,
"step": 245600
},
{
"epoch": 79.41176470588235,
"grad_norm": 1.5021244287490845,
"learning_rate": 0.001,
"loss": 1.4245,
"step": 245700
},
{
"epoch": 79.44408532643827,
"grad_norm": 1.749388575553894,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 245800
},
{
"epoch": 79.47640594699418,
"grad_norm": 1.5458804368972778,
"learning_rate": 0.001,
"loss": 1.4227,
"step": 245900
},
{
"epoch": 79.5087265675501,
"grad_norm": 1.1557130813598633,
"learning_rate": 0.001,
"loss": 1.4187,
"step": 246000
},
{
"epoch": 79.541047188106,
"grad_norm": 1.2262699604034424,
"learning_rate": 0.001,
"loss": 1.4303,
"step": 246100
},
{
"epoch": 79.57336780866193,
"grad_norm": 1.551119327545166,
"learning_rate": 0.001,
"loss": 1.4337,
"step": 246200
},
{
"epoch": 79.60568842921784,
"grad_norm": 1.3149312734603882,
"learning_rate": 0.001,
"loss": 1.4397,
"step": 246300
},
{
"epoch": 79.63800904977376,
"grad_norm": 1.6271798610687256,
"learning_rate": 0.001,
"loss": 1.4455,
"step": 246400
},
{
"epoch": 79.67032967032966,
"grad_norm": 1.2145382165908813,
"learning_rate": 0.001,
"loss": 1.4507,
"step": 246500
},
{
"epoch": 79.70265029088559,
"grad_norm": 1.9027819633483887,
"learning_rate": 0.001,
"loss": 1.4427,
"step": 246600
},
{
"epoch": 79.7349709114415,
"grad_norm": 1.2835354804992676,
"learning_rate": 0.001,
"loss": 1.449,
"step": 246700
},
{
"epoch": 79.76729153199742,
"grad_norm": 1.7002805471420288,
"learning_rate": 0.001,
"loss": 1.4463,
"step": 246800
},
{
"epoch": 79.79961215255332,
"grad_norm": 1.5688732862472534,
"learning_rate": 0.001,
"loss": 1.469,
"step": 246900
},
{
"epoch": 79.83193277310924,
"grad_norm": 1.4354397058486938,
"learning_rate": 0.001,
"loss": 1.4698,
"step": 247000
},
{
"epoch": 79.86425339366515,
"grad_norm": 1.341257095336914,
"learning_rate": 0.001,
"loss": 1.4717,
"step": 247100
},
{
"epoch": 79.89657401422107,
"grad_norm": 1.5465447902679443,
"learning_rate": 0.001,
"loss": 1.4684,
"step": 247200
},
{
"epoch": 79.92889463477698,
"grad_norm": 1.3759920597076416,
"learning_rate": 0.001,
"loss": 1.4786,
"step": 247300
},
{
"epoch": 79.9612152553329,
"grad_norm": 1.4622517824172974,
"learning_rate": 0.001,
"loss": 1.4871,
"step": 247400
},
{
"epoch": 79.99353587588882,
"grad_norm": 1.8062729835510254,
"learning_rate": 0.001,
"loss": 1.4742,
"step": 247500
},
{
"epoch": 80.02585649644473,
"grad_norm": 1.5291780233383179,
"learning_rate": 0.001,
"loss": 1.3891,
"step": 247600
},
{
"epoch": 80.05817711700065,
"grad_norm": 1.3742475509643555,
"learning_rate": 0.001,
"loss": 1.3608,
"step": 247700
},
{
"epoch": 80.09049773755656,
"grad_norm": 1.9539769887924194,
"learning_rate": 0.001,
"loss": 1.3556,
"step": 247800
},
{
"epoch": 80.12281835811248,
"grad_norm": 1.1654654741287231,
"learning_rate": 0.001,
"loss": 1.3742,
"step": 247900
},
{
"epoch": 80.15513897866839,
"grad_norm": 1.399845838546753,
"learning_rate": 0.001,
"loss": 1.3858,
"step": 248000
},
{
"epoch": 80.18745959922431,
"grad_norm": 1.262036919593811,
"learning_rate": 0.001,
"loss": 1.3776,
"step": 248100
},
{
"epoch": 80.21978021978022,
"grad_norm": 1.5510685443878174,
"learning_rate": 0.001,
"loss": 1.389,
"step": 248200
},
{
"epoch": 80.25210084033614,
"grad_norm": 1.46701180934906,
"learning_rate": 0.001,
"loss": 1.3855,
"step": 248300
},
{
"epoch": 80.28442146089205,
"grad_norm": 1.2954543828964233,
"learning_rate": 0.001,
"loss": 1.4089,
"step": 248400
},
{
"epoch": 80.31674208144797,
"grad_norm": 1.3932864665985107,
"learning_rate": 0.001,
"loss": 1.397,
"step": 248500
},
{
"epoch": 80.34906270200388,
"grad_norm": 1.9922958612442017,
"learning_rate": 0.001,
"loss": 1.4035,
"step": 248600
},
{
"epoch": 80.3813833225598,
"grad_norm": 1.5023183822631836,
"learning_rate": 0.001,
"loss": 1.4206,
"step": 248700
},
{
"epoch": 80.4137039431157,
"grad_norm": 1.24849271774292,
"learning_rate": 0.001,
"loss": 1.4199,
"step": 248800
},
{
"epoch": 80.44602456367163,
"grad_norm": 1.5220226049423218,
"learning_rate": 0.001,
"loss": 1.4132,
"step": 248900
},
{
"epoch": 80.47834518422754,
"grad_norm": 1.7662017345428467,
"learning_rate": 0.001,
"loss": 1.4207,
"step": 249000
},
{
"epoch": 80.51066580478346,
"grad_norm": 1.3429661989212036,
"learning_rate": 0.001,
"loss": 1.4198,
"step": 249100
},
{
"epoch": 80.54298642533936,
"grad_norm": 1.4158062934875488,
"learning_rate": 0.001,
"loss": 1.4209,
"step": 249200
},
{
"epoch": 80.57530704589529,
"grad_norm": 1.7861323356628418,
"learning_rate": 0.001,
"loss": 1.4366,
"step": 249300
},
{
"epoch": 80.6076276664512,
"grad_norm": 1.6265937089920044,
"learning_rate": 0.001,
"loss": 1.4373,
"step": 249400
},
{
"epoch": 80.63994828700712,
"grad_norm": 1.6524442434310913,
"learning_rate": 0.001,
"loss": 1.4231,
"step": 249500
},
{
"epoch": 80.67226890756302,
"grad_norm": 1.2062203884124756,
"learning_rate": 0.001,
"loss": 1.4376,
"step": 249600
},
{
"epoch": 80.70458952811894,
"grad_norm": 1.3392670154571533,
"learning_rate": 0.001,
"loss": 1.4349,
"step": 249700
},
{
"epoch": 80.73691014867485,
"grad_norm": 1.2870832681655884,
"learning_rate": 0.001,
"loss": 1.4465,
"step": 249800
},
{
"epoch": 80.76923076923077,
"grad_norm": 1.8183658123016357,
"learning_rate": 0.001,
"loss": 1.4394,
"step": 249900
},
{
"epoch": 80.80155138978668,
"grad_norm": 1.2293808460235596,
"learning_rate": 0.001,
"loss": 1.4462,
"step": 250000
},
{
"epoch": 80.8338720103426,
"grad_norm": 1.3528752326965332,
"learning_rate": 0.001,
"loss": 1.4494,
"step": 250100
},
{
"epoch": 80.86619263089851,
"grad_norm": 1.4483890533447266,
"learning_rate": 0.001,
"loss": 1.4601,
"step": 250200
},
{
"epoch": 80.89851325145443,
"grad_norm": 1.7228915691375732,
"learning_rate": 0.001,
"loss": 1.4732,
"step": 250300
},
{
"epoch": 80.93083387201034,
"grad_norm": 1.433459997177124,
"learning_rate": 0.001,
"loss": 1.4635,
"step": 250400
},
{
"epoch": 80.96315449256626,
"grad_norm": 1.4736968278884888,
"learning_rate": 0.001,
"loss": 1.4726,
"step": 250500
},
{
"epoch": 80.99547511312217,
"grad_norm": 1.5733822584152222,
"learning_rate": 0.001,
"loss": 1.477,
"step": 250600
},
{
"epoch": 81.02779573367809,
"grad_norm": 1.2300752401351929,
"learning_rate": 0.001,
"loss": 1.3528,
"step": 250700
},
{
"epoch": 81.060116354234,
"grad_norm": 1.4760788679122925,
"learning_rate": 0.001,
"loss": 1.3679,
"step": 250800
},
{
"epoch": 81.09243697478992,
"grad_norm": 1.3421930074691772,
"learning_rate": 0.001,
"loss": 1.3545,
"step": 250900
},
{
"epoch": 81.12475759534583,
"grad_norm": 1.601110816001892,
"learning_rate": 0.001,
"loss": 1.3735,
"step": 251000
},
{
"epoch": 81.15707821590175,
"grad_norm": 1.5888081789016724,
"learning_rate": 0.001,
"loss": 1.3718,
"step": 251100
},
{
"epoch": 81.18939883645766,
"grad_norm": 1.6814547777175903,
"learning_rate": 0.001,
"loss": 1.3788,
"step": 251200
},
{
"epoch": 81.22171945701358,
"grad_norm": 1.3557676076889038,
"learning_rate": 0.001,
"loss": 1.3679,
"step": 251300
},
{
"epoch": 81.25404007756948,
"grad_norm": 1.371376872062683,
"learning_rate": 0.001,
"loss": 1.3757,
"step": 251400
},
{
"epoch": 81.2863606981254,
"grad_norm": 1.612276554107666,
"learning_rate": 0.001,
"loss": 1.4043,
"step": 251500
},
{
"epoch": 81.31868131868131,
"grad_norm": 1.2297378778457642,
"learning_rate": 0.001,
"loss": 1.4041,
"step": 251600
},
{
"epoch": 81.35100193923724,
"grad_norm": 1.345565915107727,
"learning_rate": 0.001,
"loss": 1.3923,
"step": 251700
},
{
"epoch": 81.38332255979314,
"grad_norm": 1.4563668966293335,
"learning_rate": 0.001,
"loss": 1.3974,
"step": 251800
},
{
"epoch": 81.41564318034906,
"grad_norm": 1.5582501888275146,
"learning_rate": 0.001,
"loss": 1.3899,
"step": 251900
},
{
"epoch": 81.44796380090497,
"grad_norm": 1.8762598037719727,
"learning_rate": 0.001,
"loss": 1.4286,
"step": 252000
},
{
"epoch": 81.4802844214609,
"grad_norm": 1.5365785360336304,
"learning_rate": 0.001,
"loss": 1.3621,
"step": 252100
},
{
"epoch": 81.5126050420168,
"grad_norm": 1.6336952447891235,
"learning_rate": 0.001,
"loss": 1.3729,
"step": 252200
},
{
"epoch": 81.54492566257272,
"grad_norm": 1.3633315563201904,
"learning_rate": 0.001,
"loss": 1.3715,
"step": 252300
},
{
"epoch": 81.57724628312863,
"grad_norm": 1.5147680044174194,
"learning_rate": 0.001,
"loss": 1.3741,
"step": 252400
},
{
"epoch": 81.60956690368455,
"grad_norm": 1.6999508142471313,
"learning_rate": 0.001,
"loss": 1.3838,
"step": 252500
},
{
"epoch": 81.64188752424046,
"grad_norm": 1.7111316919326782,
"learning_rate": 0.001,
"loss": 1.3809,
"step": 252600
},
{
"epoch": 81.67420814479638,
"grad_norm": 1.6138224601745605,
"learning_rate": 0.001,
"loss": 1.3821,
"step": 252700
},
{
"epoch": 81.70652876535229,
"grad_norm": 1.4667565822601318,
"learning_rate": 0.001,
"loss": 1.3884,
"step": 252800
},
{
"epoch": 81.73884938590821,
"grad_norm": 1.8227312564849854,
"learning_rate": 0.001,
"loss": 1.3706,
"step": 252900
},
{
"epoch": 81.77117000646412,
"grad_norm": 1.9187544584274292,
"learning_rate": 0.001,
"loss": 1.3792,
"step": 253000
},
{
"epoch": 81.80349062702004,
"grad_norm": 1.924202799797058,
"learning_rate": 0.001,
"loss": 1.3939,
"step": 253100
},
{
"epoch": 81.83581124757595,
"grad_norm": 1.2514687776565552,
"learning_rate": 0.001,
"loss": 1.3865,
"step": 253200
},
{
"epoch": 81.86813186813187,
"grad_norm": 1.6029008626937866,
"learning_rate": 0.001,
"loss": 1.3787,
"step": 253300
},
{
"epoch": 81.90045248868778,
"grad_norm": 1.713050127029419,
"learning_rate": 0.001,
"loss": 1.3912,
"step": 253400
},
{
"epoch": 81.9327731092437,
"grad_norm": 1.6208316087722778,
"learning_rate": 0.001,
"loss": 1.4037,
"step": 253500
},
{
"epoch": 81.9650937297996,
"grad_norm": 1.3602346181869507,
"learning_rate": 0.001,
"loss": 1.383,
"step": 253600
},
{
"epoch": 81.99741435035553,
"grad_norm": 1.491992473602295,
"learning_rate": 0.001,
"loss": 1.3741,
"step": 253700
},
{
"epoch": 82.02973497091145,
"grad_norm": 1.610087275505066,
"learning_rate": 0.001,
"loss": 1.3552,
"step": 253800
},
{
"epoch": 82.06205559146736,
"grad_norm": 1.4379409551620483,
"learning_rate": 0.001,
"loss": 1.3534,
"step": 253900
},
{
"epoch": 82.09437621202328,
"grad_norm": 1.615574836730957,
"learning_rate": 0.001,
"loss": 1.3662,
"step": 254000
},
{
"epoch": 82.12669683257919,
"grad_norm": 1.4320043325424194,
"learning_rate": 0.001,
"loss": 1.3639,
"step": 254100
},
{
"epoch": 82.1590174531351,
"grad_norm": 1.6757969856262207,
"learning_rate": 0.001,
"loss": 1.3757,
"step": 254200
},
{
"epoch": 82.19133807369101,
"grad_norm": 1.4462916851043701,
"learning_rate": 0.001,
"loss": 1.384,
"step": 254300
},
{
"epoch": 82.22365869424694,
"grad_norm": 1.493852138519287,
"learning_rate": 0.001,
"loss": 1.3761,
"step": 254400
},
{
"epoch": 82.25597931480284,
"grad_norm": 1.9881926774978638,
"learning_rate": 0.001,
"loss": 1.3756,
"step": 254500
},
{
"epoch": 82.28829993535876,
"grad_norm": 1.6097559928894043,
"learning_rate": 0.001,
"loss": 1.3852,
"step": 254600
},
{
"epoch": 82.32062055591467,
"grad_norm": 1.6144851446151733,
"learning_rate": 0.001,
"loss": 1.3873,
"step": 254700
},
{
"epoch": 82.3529411764706,
"grad_norm": 1.5550802946090698,
"learning_rate": 0.001,
"loss": 1.4031,
"step": 254800
},
{
"epoch": 82.3852617970265,
"grad_norm": 1.4737643003463745,
"learning_rate": 0.001,
"loss": 1.4082,
"step": 254900
},
{
"epoch": 82.41758241758242,
"grad_norm": 1.783548355102539,
"learning_rate": 0.001,
"loss": 1.4012,
"step": 255000
},
{
"epoch": 82.44990303813833,
"grad_norm": 2.3701467514038086,
"learning_rate": 0.001,
"loss": 1.3943,
"step": 255100
},
{
"epoch": 82.48222365869425,
"grad_norm": 1.7233704328536987,
"learning_rate": 0.001,
"loss": 1.4078,
"step": 255200
},
{
"epoch": 82.51454427925016,
"grad_norm": 1.7663028240203857,
"learning_rate": 0.001,
"loss": 1.4284,
"step": 255300
},
{
"epoch": 82.54686489980608,
"grad_norm": 1.783216953277588,
"learning_rate": 0.001,
"loss": 1.4185,
"step": 255400
},
{
"epoch": 82.57918552036199,
"grad_norm": 1.5642690658569336,
"learning_rate": 0.001,
"loss": 1.4163,
"step": 255500
},
{
"epoch": 82.61150614091791,
"grad_norm": 1.4378348588943481,
"learning_rate": 0.001,
"loss": 1.4309,
"step": 255600
},
{
"epoch": 82.64382676147382,
"grad_norm": 1.5882807970046997,
"learning_rate": 0.001,
"loss": 1.4192,
"step": 255700
},
{
"epoch": 82.67614738202974,
"grad_norm": 1.7995021343231201,
"learning_rate": 0.001,
"loss": 1.4283,
"step": 255800
},
{
"epoch": 82.70846800258565,
"grad_norm": 1.6401058435440063,
"learning_rate": 0.001,
"loss": 1.45,
"step": 255900
},
{
"epoch": 82.74078862314157,
"grad_norm": 1.5894571542739868,
"learning_rate": 0.001,
"loss": 1.4369,
"step": 256000
},
{
"epoch": 82.77310924369748,
"grad_norm": 1.8430721759796143,
"learning_rate": 0.001,
"loss": 1.4388,
"step": 256100
},
{
"epoch": 82.8054298642534,
"grad_norm": 1.5611604452133179,
"learning_rate": 0.001,
"loss": 1.4412,
"step": 256200
},
{
"epoch": 82.8377504848093,
"grad_norm": 2.033487319946289,
"learning_rate": 0.001,
"loss": 1.4436,
"step": 256300
},
{
"epoch": 82.87007110536523,
"grad_norm": 1.4882752895355225,
"learning_rate": 0.001,
"loss": 1.4413,
"step": 256400
},
{
"epoch": 82.90239172592113,
"grad_norm": 1.4577480554580688,
"learning_rate": 0.001,
"loss": 1.4621,
"step": 256500
},
{
"epoch": 82.93471234647706,
"grad_norm": 1.5735620260238647,
"learning_rate": 0.001,
"loss": 1.4514,
"step": 256600
},
{
"epoch": 82.96703296703296,
"grad_norm": 1.4674712419509888,
"learning_rate": 0.001,
"loss": 1.463,
"step": 256700
},
{
"epoch": 82.99935358758889,
"grad_norm": 1.8225946426391602,
"learning_rate": 0.001,
"loss": 1.4465,
"step": 256800
},
{
"epoch": 83.03167420814479,
"grad_norm": 1.5670487880706787,
"learning_rate": 0.001,
"loss": 1.3387,
"step": 256900
},
{
"epoch": 83.06399482870071,
"grad_norm": 1.5646357536315918,
"learning_rate": 0.001,
"loss": 1.3405,
"step": 257000
},
{
"epoch": 83.09631544925662,
"grad_norm": 2.9608230590820312,
"learning_rate": 0.001,
"loss": 1.3511,
"step": 257100
},
{
"epoch": 83.12863606981254,
"grad_norm": 1.7706575393676758,
"learning_rate": 0.001,
"loss": 1.3575,
"step": 257200
},
{
"epoch": 83.16095669036845,
"grad_norm": 1.9454774856567383,
"learning_rate": 0.001,
"loss": 1.3653,
"step": 257300
},
{
"epoch": 83.19327731092437,
"grad_norm": 2.1506412029266357,
"learning_rate": 0.001,
"loss": 1.365,
"step": 257400
},
{
"epoch": 83.22559793148028,
"grad_norm": 1.6990796327590942,
"learning_rate": 0.001,
"loss": 1.3607,
"step": 257500
},
{
"epoch": 83.2579185520362,
"grad_norm": 1.5187764167785645,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 257600
},
{
"epoch": 83.29023917259211,
"grad_norm": 2.0596764087677,
"learning_rate": 0.001,
"loss": 1.3851,
"step": 257700
},
{
"epoch": 83.32255979314803,
"grad_norm": 2.375506639480591,
"learning_rate": 0.001,
"loss": 1.37,
"step": 257800
},
{
"epoch": 83.35488041370394,
"grad_norm": 1.7246047258377075,
"learning_rate": 0.001,
"loss": 1.3808,
"step": 257900
},
{
"epoch": 83.38720103425986,
"grad_norm": 1.8666071891784668,
"learning_rate": 0.001,
"loss": 1.3807,
"step": 258000
},
{
"epoch": 83.41952165481577,
"grad_norm": 2.562657117843628,
"learning_rate": 0.001,
"loss": 1.3899,
"step": 258100
},
{
"epoch": 83.45184227537169,
"grad_norm": 1.9849448204040527,
"learning_rate": 0.001,
"loss": 1.4007,
"step": 258200
},
{
"epoch": 83.4841628959276,
"grad_norm": 1.7674652338027954,
"learning_rate": 0.001,
"loss": 1.4047,
"step": 258300
},
{
"epoch": 83.51648351648352,
"grad_norm": 1.463085412979126,
"learning_rate": 0.001,
"loss": 1.4104,
"step": 258400
},
{
"epoch": 83.54880413703943,
"grad_norm": 1.8578214645385742,
"learning_rate": 0.001,
"loss": 1.3986,
"step": 258500
},
{
"epoch": 83.58112475759535,
"grad_norm": 1.5158841609954834,
"learning_rate": 0.001,
"loss": 1.4092,
"step": 258600
},
{
"epoch": 83.61344537815125,
"grad_norm": 1.6846269369125366,
"learning_rate": 0.001,
"loss": 1.4267,
"step": 258700
},
{
"epoch": 83.64576599870718,
"grad_norm": 1.8053178787231445,
"learning_rate": 0.001,
"loss": 1.4272,
"step": 258800
},
{
"epoch": 83.67808661926308,
"grad_norm": 1.9504740238189697,
"learning_rate": 0.001,
"loss": 1.4189,
"step": 258900
},
{
"epoch": 83.710407239819,
"grad_norm": 1.8589684963226318,
"learning_rate": 0.001,
"loss": 1.4227,
"step": 259000
},
{
"epoch": 83.74272786037491,
"grad_norm": 1.7385740280151367,
"learning_rate": 0.001,
"loss": 1.4319,
"step": 259100
},
{
"epoch": 83.77504848093083,
"grad_norm": 2.115756034851074,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 259200
},
{
"epoch": 83.80736910148674,
"grad_norm": 2.0501482486724854,
"learning_rate": 0.001,
"loss": 1.4405,
"step": 259300
},
{
"epoch": 83.83968972204266,
"grad_norm": 2.1972477436065674,
"learning_rate": 0.001,
"loss": 1.4478,
"step": 259400
},
{
"epoch": 83.87201034259857,
"grad_norm": 1.725358486175537,
"learning_rate": 0.001,
"loss": 1.4471,
"step": 259500
},
{
"epoch": 83.9043309631545,
"grad_norm": 2.9141781330108643,
"learning_rate": 0.001,
"loss": 1.4524,
"step": 259600
},
{
"epoch": 83.9366515837104,
"grad_norm": 2.246513843536377,
"learning_rate": 0.001,
"loss": 1.4489,
"step": 259700
},
{
"epoch": 83.96897220426632,
"grad_norm": 2.0187714099884033,
"learning_rate": 0.001,
"loss": 1.4628,
"step": 259800
},
{
"epoch": 84.00129282482224,
"grad_norm": 1.463592290878296,
"learning_rate": 0.001,
"loss": 1.4409,
"step": 259900
},
{
"epoch": 84.03361344537815,
"grad_norm": 1.7122960090637207,
"learning_rate": 0.001,
"loss": 1.3349,
"step": 260000
},
{
"epoch": 84.06593406593407,
"grad_norm": 2.092839241027832,
"learning_rate": 0.001,
"loss": 1.344,
"step": 260100
},
{
"epoch": 84.09825468648998,
"grad_norm": 1.5537153482437134,
"learning_rate": 0.001,
"loss": 1.3284,
"step": 260200
},
{
"epoch": 84.1305753070459,
"grad_norm": 2.3516452312469482,
"learning_rate": 0.001,
"loss": 1.3623,
"step": 260300
},
{
"epoch": 84.16289592760181,
"grad_norm": 1.9590219259262085,
"learning_rate": 0.001,
"loss": 1.3563,
"step": 260400
},
{
"epoch": 84.19521654815773,
"grad_norm": 2.0436458587646484,
"learning_rate": 0.001,
"loss": 1.3603,
"step": 260500
},
{
"epoch": 84.22753716871364,
"grad_norm": 1.8916879892349243,
"learning_rate": 0.001,
"loss": 1.3727,
"step": 260600
},
{
"epoch": 84.25985778926956,
"grad_norm": 1.6515244245529175,
"learning_rate": 0.001,
"loss": 1.3543,
"step": 260700
},
{
"epoch": 84.29217840982547,
"grad_norm": 1.9168320894241333,
"learning_rate": 0.001,
"loss": 1.359,
"step": 260800
},
{
"epoch": 84.32449903038139,
"grad_norm": 1.362146258354187,
"learning_rate": 0.001,
"loss": 1.3704,
"step": 260900
},
{
"epoch": 84.3568196509373,
"grad_norm": 1.4027669429779053,
"learning_rate": 0.001,
"loss": 1.3927,
"step": 261000
},
{
"epoch": 84.38914027149322,
"grad_norm": 2.048327684402466,
"learning_rate": 0.001,
"loss": 1.3778,
"step": 261100
},
{
"epoch": 84.42146089204913,
"grad_norm": 1.6923198699951172,
"learning_rate": 0.001,
"loss": 1.3958,
"step": 261200
},
{
"epoch": 84.45378151260505,
"grad_norm": 1.995417833328247,
"learning_rate": 0.001,
"loss": 1.3905,
"step": 261300
},
{
"epoch": 84.48610213316095,
"grad_norm": 1.4216890335083008,
"learning_rate": 0.001,
"loss": 1.3911,
"step": 261400
},
{
"epoch": 84.51842275371688,
"grad_norm": 1.6288414001464844,
"learning_rate": 0.001,
"loss": 1.3948,
"step": 261500
},
{
"epoch": 84.55074337427278,
"grad_norm": 1.6087497472763062,
"learning_rate": 0.001,
"loss": 1.4058,
"step": 261600
},
{
"epoch": 84.5830639948287,
"grad_norm": 1.9099173545837402,
"learning_rate": 0.001,
"loss": 1.4095,
"step": 261700
},
{
"epoch": 84.61538461538461,
"grad_norm": 1.6727123260498047,
"learning_rate": 0.001,
"loss": 1.4036,
"step": 261800
},
{
"epoch": 84.64770523594053,
"grad_norm": 1.634691596031189,
"learning_rate": 0.001,
"loss": 1.3952,
"step": 261900
},
{
"epoch": 84.68002585649644,
"grad_norm": 1.4719901084899902,
"learning_rate": 0.001,
"loss": 1.4121,
"step": 262000
},
{
"epoch": 84.71234647705236,
"grad_norm": 1.393352746963501,
"learning_rate": 0.001,
"loss": 1.4265,
"step": 262100
},
{
"epoch": 84.74466709760827,
"grad_norm": 2.0315215587615967,
"learning_rate": 0.001,
"loss": 1.4141,
"step": 262200
},
{
"epoch": 84.7769877181642,
"grad_norm": 1.6670891046524048,
"learning_rate": 0.001,
"loss": 1.4297,
"step": 262300
},
{
"epoch": 84.8093083387201,
"grad_norm": 1.483593225479126,
"learning_rate": 0.001,
"loss": 1.4321,
"step": 262400
},
{
"epoch": 84.84162895927602,
"grad_norm": 1.6042678356170654,
"learning_rate": 0.001,
"loss": 1.4326,
"step": 262500
},
{
"epoch": 84.87394957983193,
"grad_norm": 2.126124382019043,
"learning_rate": 0.001,
"loss": 1.4334,
"step": 262600
},
{
"epoch": 84.90627020038785,
"grad_norm": 2.080559015274048,
"learning_rate": 0.001,
"loss": 1.4407,
"step": 262700
},
{
"epoch": 84.93859082094376,
"grad_norm": 1.4968855381011963,
"learning_rate": 0.001,
"loss": 1.4497,
"step": 262800
},
{
"epoch": 84.97091144149968,
"grad_norm": 1.5889743566513062,
"learning_rate": 0.001,
"loss": 1.4473,
"step": 262900
},
{
"epoch": 85.00323206205559,
"grad_norm": 1.3979679346084595,
"learning_rate": 0.001,
"loss": 1.4448,
"step": 263000
},
{
"epoch": 85.03555268261151,
"grad_norm": 1.3192113637924194,
"learning_rate": 0.001,
"loss": 1.3214,
"step": 263100
},
{
"epoch": 85.06787330316742,
"grad_norm": 1.7462128400802612,
"learning_rate": 0.001,
"loss": 1.3282,
"step": 263200
},
{
"epoch": 85.10019392372334,
"grad_norm": 1.6237643957138062,
"learning_rate": 0.001,
"loss": 1.3365,
"step": 263300
},
{
"epoch": 85.13251454427925,
"grad_norm": 1.1121832132339478,
"learning_rate": 0.001,
"loss": 1.3521,
"step": 263400
},
{
"epoch": 85.16483516483517,
"grad_norm": 1.6434276103973389,
"learning_rate": 0.001,
"loss": 1.3479,
"step": 263500
},
{
"epoch": 85.19715578539108,
"grad_norm": 1.5122075080871582,
"learning_rate": 0.001,
"loss": 1.3442,
"step": 263600
},
{
"epoch": 85.229476405947,
"grad_norm": 1.3585845232009888,
"learning_rate": 0.001,
"loss": 1.3416,
"step": 263700
},
{
"epoch": 85.2617970265029,
"grad_norm": 1.9766629934310913,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 263800
},
{
"epoch": 85.29411764705883,
"grad_norm": 2.283902645111084,
"learning_rate": 0.001,
"loss": 1.3539,
"step": 263900
},
{
"epoch": 85.32643826761473,
"grad_norm": 1.3756550550460815,
"learning_rate": 0.001,
"loss": 1.369,
"step": 264000
},
{
"epoch": 85.35875888817066,
"grad_norm": 1.5105441808700562,
"learning_rate": 0.001,
"loss": 1.3812,
"step": 264100
},
{
"epoch": 85.39107950872656,
"grad_norm": 1.454532265663147,
"learning_rate": 0.001,
"loss": 1.3693,
"step": 264200
},
{
"epoch": 85.42340012928248,
"grad_norm": 1.5482038259506226,
"learning_rate": 0.001,
"loss": 1.3863,
"step": 264300
},
{
"epoch": 85.45572074983839,
"grad_norm": 1.2381619215011597,
"learning_rate": 0.001,
"loss": 1.3806,
"step": 264400
},
{
"epoch": 85.48804137039431,
"grad_norm": 1.6195075511932373,
"learning_rate": 0.001,
"loss": 1.3876,
"step": 264500
},
{
"epoch": 85.52036199095022,
"grad_norm": 1.5150704383850098,
"learning_rate": 0.001,
"loss": 1.3783,
"step": 264600
},
{
"epoch": 85.55268261150614,
"grad_norm": 1.4213989973068237,
"learning_rate": 0.001,
"loss": 1.4106,
"step": 264700
},
{
"epoch": 85.58500323206205,
"grad_norm": 1.2862284183502197,
"learning_rate": 0.001,
"loss": 1.4004,
"step": 264800
},
{
"epoch": 85.61732385261797,
"grad_norm": 1.1904622316360474,
"learning_rate": 0.001,
"loss": 1.4059,
"step": 264900
},
{
"epoch": 85.64964447317388,
"grad_norm": 1.8824856281280518,
"learning_rate": 0.001,
"loss": 1.4107,
"step": 265000
},
{
"epoch": 85.6819650937298,
"grad_norm": 1.5507681369781494,
"learning_rate": 0.001,
"loss": 1.4064,
"step": 265100
},
{
"epoch": 85.71428571428571,
"grad_norm": 1.504534363746643,
"learning_rate": 0.001,
"loss": 1.418,
"step": 265200
},
{
"epoch": 85.74660633484163,
"grad_norm": 1.5033137798309326,
"learning_rate": 0.001,
"loss": 1.4111,
"step": 265300
},
{
"epoch": 85.77892695539754,
"grad_norm": 1.2780414819717407,
"learning_rate": 0.001,
"loss": 1.4097,
"step": 265400
},
{
"epoch": 85.81124757595346,
"grad_norm": 1.5234276056289673,
"learning_rate": 0.001,
"loss": 1.434,
"step": 265500
},
{
"epoch": 85.84356819650937,
"grad_norm": 1.3871499300003052,
"learning_rate": 0.001,
"loss": 1.4078,
"step": 265600
},
{
"epoch": 85.87588881706529,
"grad_norm": 1.79600191116333,
"learning_rate": 0.001,
"loss": 1.4202,
"step": 265700
},
{
"epoch": 85.9082094376212,
"grad_norm": 1.615275263786316,
"learning_rate": 0.001,
"loss": 1.4334,
"step": 265800
},
{
"epoch": 85.94053005817712,
"grad_norm": 1.3826439380645752,
"learning_rate": 0.001,
"loss": 1.4434,
"step": 265900
},
{
"epoch": 85.97285067873302,
"grad_norm": 1.8904972076416016,
"learning_rate": 0.001,
"loss": 1.435,
"step": 266000
},
{
"epoch": 86.00517129928895,
"grad_norm": 1.3957210779190063,
"learning_rate": 0.001,
"loss": 1.4284,
"step": 266100
},
{
"epoch": 86.03749191984487,
"grad_norm": 1.2213150262832642,
"learning_rate": 0.001,
"loss": 1.3186,
"step": 266200
},
{
"epoch": 86.06981254040078,
"grad_norm": 1.178726315498352,
"learning_rate": 0.001,
"loss": 1.3325,
"step": 266300
},
{
"epoch": 86.1021331609567,
"grad_norm": 1.5732803344726562,
"learning_rate": 0.001,
"loss": 1.3244,
"step": 266400
},
{
"epoch": 86.1344537815126,
"grad_norm": 1.2888745069503784,
"learning_rate": 0.001,
"loss": 1.3385,
"step": 266500
},
{
"epoch": 86.16677440206853,
"grad_norm": 1.5115408897399902,
"learning_rate": 0.001,
"loss": 1.3398,
"step": 266600
},
{
"epoch": 86.19909502262443,
"grad_norm": 2.0756070613861084,
"learning_rate": 0.001,
"loss": 1.3203,
"step": 266700
},
{
"epoch": 86.23141564318036,
"grad_norm": 1.7549015283584595,
"learning_rate": 0.001,
"loss": 1.3565,
"step": 266800
},
{
"epoch": 86.26373626373626,
"grad_norm": 1.3304147720336914,
"learning_rate": 0.001,
"loss": 1.362,
"step": 266900
},
{
"epoch": 86.29605688429218,
"grad_norm": 1.5407073497772217,
"learning_rate": 0.001,
"loss": 1.3648,
"step": 267000
},
{
"epoch": 86.32837750484809,
"grad_norm": 1.5653663873672485,
"learning_rate": 0.001,
"loss": 1.3639,
"step": 267100
},
{
"epoch": 86.36069812540401,
"grad_norm": 1.2268964052200317,
"learning_rate": 0.001,
"loss": 1.359,
"step": 267200
},
{
"epoch": 86.39301874595992,
"grad_norm": 1.3261157274246216,
"learning_rate": 0.001,
"loss": 1.3758,
"step": 267300
},
{
"epoch": 86.42533936651584,
"grad_norm": 1.8261868953704834,
"learning_rate": 0.001,
"loss": 1.374,
"step": 267400
},
{
"epoch": 86.45765998707175,
"grad_norm": 1.3722460269927979,
"learning_rate": 0.001,
"loss": 1.3656,
"step": 267500
},
{
"epoch": 86.48998060762767,
"grad_norm": 1.5559968948364258,
"learning_rate": 0.001,
"loss": 1.388,
"step": 267600
},
{
"epoch": 86.52230122818358,
"grad_norm": 1.9659545421600342,
"learning_rate": 0.001,
"loss": 1.3898,
"step": 267700
},
{
"epoch": 86.5546218487395,
"grad_norm": 1.393257737159729,
"learning_rate": 0.001,
"loss": 1.3928,
"step": 267800
},
{
"epoch": 86.58694246929541,
"grad_norm": 1.4295693635940552,
"learning_rate": 0.001,
"loss": 1.3957,
"step": 267900
},
{
"epoch": 86.61926308985133,
"grad_norm": 1.5853294134140015,
"learning_rate": 0.001,
"loss": 1.3883,
"step": 268000
},
{
"epoch": 86.65158371040724,
"grad_norm": 1.360254168510437,
"learning_rate": 0.001,
"loss": 1.4047,
"step": 268100
},
{
"epoch": 86.68390433096316,
"grad_norm": 1.257365107536316,
"learning_rate": 0.001,
"loss": 1.4012,
"step": 268200
},
{
"epoch": 86.71622495151907,
"grad_norm": 1.6614432334899902,
"learning_rate": 0.001,
"loss": 1.4063,
"step": 268300
},
{
"epoch": 86.74854557207499,
"grad_norm": 1.2073627710342407,
"learning_rate": 0.001,
"loss": 1.4084,
"step": 268400
},
{
"epoch": 86.7808661926309,
"grad_norm": 1.3388640880584717,
"learning_rate": 0.001,
"loss": 1.4041,
"step": 268500
},
{
"epoch": 86.81318681318682,
"grad_norm": 1.7202292680740356,
"learning_rate": 0.001,
"loss": 1.424,
"step": 268600
},
{
"epoch": 86.84550743374272,
"grad_norm": 1.631931185722351,
"learning_rate": 0.001,
"loss": 1.4005,
"step": 268700
},
{
"epoch": 86.87782805429865,
"grad_norm": 1.8990880250930786,
"learning_rate": 0.001,
"loss": 1.4113,
"step": 268800
},
{
"epoch": 86.91014867485455,
"grad_norm": 1.2895227670669556,
"learning_rate": 0.001,
"loss": 1.4209,
"step": 268900
},
{
"epoch": 86.94246929541048,
"grad_norm": 1.7723702192306519,
"learning_rate": 0.001,
"loss": 1.4159,
"step": 269000
},
{
"epoch": 86.97478991596638,
"grad_norm": 1.6865254640579224,
"learning_rate": 0.001,
"loss": 1.4282,
"step": 269100
},
{
"epoch": 87.0071105365223,
"grad_norm": 1.446299433708191,
"learning_rate": 0.001,
"loss": 1.4039,
"step": 269200
},
{
"epoch": 87.03943115707821,
"grad_norm": 1.504185438156128,
"learning_rate": 0.001,
"loss": 1.3048,
"step": 269300
},
{
"epoch": 87.07175177763413,
"grad_norm": 1.6202774047851562,
"learning_rate": 0.001,
"loss": 1.3274,
"step": 269400
},
{
"epoch": 87.10407239819004,
"grad_norm": 1.397711992263794,
"learning_rate": 0.001,
"loss": 1.332,
"step": 269500
},
{
"epoch": 87.13639301874596,
"grad_norm": 1.3761570453643799,
"learning_rate": 0.001,
"loss": 1.3404,
"step": 269600
},
{
"epoch": 87.16871363930187,
"grad_norm": 1.325406551361084,
"learning_rate": 0.001,
"loss": 1.3239,
"step": 269700
},
{
"epoch": 87.20103425985779,
"grad_norm": 1.437890648841858,
"learning_rate": 0.001,
"loss": 1.339,
"step": 269800
},
{
"epoch": 87.2333548804137,
"grad_norm": 1.7336199283599854,
"learning_rate": 0.001,
"loss": 1.3359,
"step": 269900
},
{
"epoch": 87.26567550096962,
"grad_norm": 1.6384884119033813,
"learning_rate": 0.001,
"loss": 1.3537,
"step": 270000
},
{
"epoch": 87.29799612152553,
"grad_norm": 1.107170581817627,
"learning_rate": 0.001,
"loss": 1.3564,
"step": 270100
},
{
"epoch": 87.33031674208145,
"grad_norm": 1.309773325920105,
"learning_rate": 0.001,
"loss": 1.3473,
"step": 270200
},
{
"epoch": 87.36263736263736,
"grad_norm": 1.4358161687850952,
"learning_rate": 0.001,
"loss": 1.3514,
"step": 270300
},
{
"epoch": 87.39495798319328,
"grad_norm": 1.4592732191085815,
"learning_rate": 0.001,
"loss": 1.3612,
"step": 270400
},
{
"epoch": 87.42727860374919,
"grad_norm": 1.3932117223739624,
"learning_rate": 0.001,
"loss": 1.3652,
"step": 270500
},
{
"epoch": 87.45959922430511,
"grad_norm": 1.40018892288208,
"learning_rate": 0.001,
"loss": 1.3729,
"step": 270600
},
{
"epoch": 87.49191984486102,
"grad_norm": 1.3737910985946655,
"learning_rate": 0.001,
"loss": 1.3881,
"step": 270700
},
{
"epoch": 87.52424046541694,
"grad_norm": 1.7012450695037842,
"learning_rate": 0.001,
"loss": 1.3594,
"step": 270800
},
{
"epoch": 87.55656108597285,
"grad_norm": 1.3596134185791016,
"learning_rate": 0.001,
"loss": 1.3786,
"step": 270900
},
{
"epoch": 87.58888170652877,
"grad_norm": 1.5316884517669678,
"learning_rate": 0.001,
"loss": 1.3865,
"step": 271000
},
{
"epoch": 87.62120232708467,
"grad_norm": 1.3480607271194458,
"learning_rate": 0.001,
"loss": 1.394,
"step": 271100
},
{
"epoch": 87.6535229476406,
"grad_norm": 2.001589775085449,
"learning_rate": 0.001,
"loss": 1.3797,
"step": 271200
},
{
"epoch": 87.6858435681965,
"grad_norm": 1.6091517210006714,
"learning_rate": 0.001,
"loss": 1.3991,
"step": 271300
},
{
"epoch": 87.71816418875243,
"grad_norm": 1.4985905885696411,
"learning_rate": 0.001,
"loss": 1.3981,
"step": 271400
},
{
"epoch": 87.75048480930833,
"grad_norm": 1.3070586919784546,
"learning_rate": 0.001,
"loss": 1.3882,
"step": 271500
},
{
"epoch": 87.78280542986425,
"grad_norm": 1.2479877471923828,
"learning_rate": 0.001,
"loss": 1.404,
"step": 271600
},
{
"epoch": 87.81512605042016,
"grad_norm": 1.6475880146026611,
"learning_rate": 0.001,
"loss": 1.4079,
"step": 271700
},
{
"epoch": 87.84744667097608,
"grad_norm": 1.3737822771072388,
"learning_rate": 0.001,
"loss": 1.4075,
"step": 271800
},
{
"epoch": 87.87976729153199,
"grad_norm": 1.7142956256866455,
"learning_rate": 0.001,
"loss": 1.4215,
"step": 271900
},
{
"epoch": 87.91208791208791,
"grad_norm": 1.502916693687439,
"learning_rate": 0.001,
"loss": 1.4263,
"step": 272000
},
{
"epoch": 87.94440853264382,
"grad_norm": 1.7450618743896484,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 272100
},
{
"epoch": 87.97672915319974,
"grad_norm": 1.3914827108383179,
"learning_rate": 0.001,
"loss": 1.4373,
"step": 272200
},
{
"epoch": 88.00904977375566,
"grad_norm": 1.648630142211914,
"learning_rate": 0.001,
"loss": 1.3958,
"step": 272300
},
{
"epoch": 88.04137039431157,
"grad_norm": 2.1626803874969482,
"learning_rate": 0.001,
"loss": 1.3087,
"step": 272400
},
{
"epoch": 88.07369101486749,
"grad_norm": 1.2503234148025513,
"learning_rate": 0.001,
"loss": 1.3049,
"step": 272500
},
{
"epoch": 88.1060116354234,
"grad_norm": 1.269621729850769,
"learning_rate": 0.001,
"loss": 1.3133,
"step": 272600
},
{
"epoch": 88.13833225597932,
"grad_norm": 1.31803560256958,
"learning_rate": 0.001,
"loss": 1.3127,
"step": 272700
},
{
"epoch": 88.17065287653523,
"grad_norm": 1.381618857383728,
"learning_rate": 0.001,
"loss": 1.3277,
"step": 272800
},
{
"epoch": 88.20297349709115,
"grad_norm": 1.1675587892532349,
"learning_rate": 0.001,
"loss": 1.3239,
"step": 272900
},
{
"epoch": 88.23529411764706,
"grad_norm": 1.4538698196411133,
"learning_rate": 0.001,
"loss": 1.3337,
"step": 273000
},
{
"epoch": 88.26761473820298,
"grad_norm": 1.5730509757995605,
"learning_rate": 0.001,
"loss": 1.346,
"step": 273100
},
{
"epoch": 88.29993535875889,
"grad_norm": 1.3985658884048462,
"learning_rate": 0.001,
"loss": 1.3592,
"step": 273200
},
{
"epoch": 88.33225597931481,
"grad_norm": 1.5334653854370117,
"learning_rate": 0.001,
"loss": 1.3554,
"step": 273300
},
{
"epoch": 88.36457659987072,
"grad_norm": 1.9616245031356812,
"learning_rate": 0.001,
"loss": 1.3547,
"step": 273400
},
{
"epoch": 88.39689722042664,
"grad_norm": 1.3973840475082397,
"learning_rate": 0.001,
"loss": 1.3593,
"step": 273500
},
{
"epoch": 88.42921784098255,
"grad_norm": 1.1953452825546265,
"learning_rate": 0.001,
"loss": 1.3572,
"step": 273600
},
{
"epoch": 88.46153846153847,
"grad_norm": 1.8578697443008423,
"learning_rate": 0.001,
"loss": 1.3583,
"step": 273700
},
{
"epoch": 88.49385908209437,
"grad_norm": 1.4416879415512085,
"learning_rate": 0.001,
"loss": 1.3729,
"step": 273800
},
{
"epoch": 88.5261797026503,
"grad_norm": 2.0024890899658203,
"learning_rate": 0.001,
"loss": 1.3853,
"step": 273900
},
{
"epoch": 88.5585003232062,
"grad_norm": 1.6199134588241577,
"learning_rate": 0.001,
"loss": 1.373,
"step": 274000
},
{
"epoch": 88.59082094376213,
"grad_norm": 1.5681766271591187,
"learning_rate": 0.001,
"loss": 1.3819,
"step": 274100
},
{
"epoch": 88.62314156431803,
"grad_norm": 1.6268113851547241,
"learning_rate": 0.001,
"loss": 1.3839,
"step": 274200
},
{
"epoch": 88.65546218487395,
"grad_norm": 1.4910027980804443,
"learning_rate": 0.001,
"loss": 1.3858,
"step": 274300
},
{
"epoch": 88.68778280542986,
"grad_norm": 1.4038636684417725,
"learning_rate": 0.001,
"loss": 1.3966,
"step": 274400
},
{
"epoch": 88.72010342598578,
"grad_norm": 1.312757968902588,
"learning_rate": 0.001,
"loss": 1.3955,
"step": 274500
},
{
"epoch": 88.75242404654169,
"grad_norm": 1.3066078424453735,
"learning_rate": 0.001,
"loss": 1.385,
"step": 274600
},
{
"epoch": 88.78474466709761,
"grad_norm": 1.2062005996704102,
"learning_rate": 0.001,
"loss": 1.3955,
"step": 274700
},
{
"epoch": 88.81706528765352,
"grad_norm": 1.2288823127746582,
"learning_rate": 0.001,
"loss": 1.4011,
"step": 274800
},
{
"epoch": 88.84938590820944,
"grad_norm": 2.141970157623291,
"learning_rate": 0.001,
"loss": 1.4041,
"step": 274900
},
{
"epoch": 88.88170652876535,
"grad_norm": 1.2106239795684814,
"learning_rate": 0.001,
"loss": 1.4121,
"step": 275000
},
{
"epoch": 88.91402714932127,
"grad_norm": 1.3385686874389648,
"learning_rate": 0.001,
"loss": 1.4053,
"step": 275100
},
{
"epoch": 88.94634776987718,
"grad_norm": 1.5660736560821533,
"learning_rate": 0.001,
"loss": 1.4016,
"step": 275200
},
{
"epoch": 88.9786683904331,
"grad_norm": 1.3535319566726685,
"learning_rate": 0.001,
"loss": 1.4102,
"step": 275300
},
{
"epoch": 89.01098901098901,
"grad_norm": 1.3070379495620728,
"learning_rate": 0.001,
"loss": 1.38,
"step": 275400
},
{
"epoch": 89.04330963154493,
"grad_norm": 1.358125925064087,
"learning_rate": 0.001,
"loss": 1.3073,
"step": 275500
},
{
"epoch": 89.07563025210084,
"grad_norm": 1.6103869676589966,
"learning_rate": 0.001,
"loss": 1.3029,
"step": 275600
},
{
"epoch": 89.10795087265676,
"grad_norm": 1.5382152795791626,
"learning_rate": 0.001,
"loss": 1.3233,
"step": 275700
},
{
"epoch": 89.14027149321267,
"grad_norm": 1.531084418296814,
"learning_rate": 0.001,
"loss": 1.3215,
"step": 275800
},
{
"epoch": 89.17259211376859,
"grad_norm": 1.5296151638031006,
"learning_rate": 0.001,
"loss": 1.3305,
"step": 275900
},
{
"epoch": 89.2049127343245,
"grad_norm": 1.82796049118042,
"learning_rate": 0.001,
"loss": 1.3347,
"step": 276000
},
{
"epoch": 89.23723335488042,
"grad_norm": 1.1894311904907227,
"learning_rate": 0.001,
"loss": 1.3383,
"step": 276100
},
{
"epoch": 89.26955397543632,
"grad_norm": 1.527662992477417,
"learning_rate": 0.001,
"loss": 1.3355,
"step": 276200
},
{
"epoch": 89.30187459599225,
"grad_norm": 1.5246412754058838,
"learning_rate": 0.001,
"loss": 1.3265,
"step": 276300
},
{
"epoch": 89.33419521654815,
"grad_norm": 1.4023152589797974,
"learning_rate": 0.001,
"loss": 1.3486,
"step": 276400
},
{
"epoch": 89.36651583710407,
"grad_norm": 1.3282690048217773,
"learning_rate": 0.001,
"loss": 1.3665,
"step": 276500
},
{
"epoch": 89.39883645765998,
"grad_norm": 2.154405117034912,
"learning_rate": 0.001,
"loss": 1.3574,
"step": 276600
},
{
"epoch": 89.4311570782159,
"grad_norm": 1.7010021209716797,
"learning_rate": 0.001,
"loss": 1.3414,
"step": 276700
},
{
"epoch": 89.46347769877181,
"grad_norm": 1.6260164976119995,
"learning_rate": 0.001,
"loss": 1.358,
"step": 276800
},
{
"epoch": 89.49579831932773,
"grad_norm": 1.5198596715927124,
"learning_rate": 0.001,
"loss": 1.3693,
"step": 276900
},
{
"epoch": 89.52811893988364,
"grad_norm": 1.574953317642212,
"learning_rate": 0.001,
"loss": 1.3618,
"step": 277000
},
{
"epoch": 89.56043956043956,
"grad_norm": 1.2591584920883179,
"learning_rate": 0.001,
"loss": 1.3681,
"step": 277100
},
{
"epoch": 89.59276018099547,
"grad_norm": 1.8173555135726929,
"learning_rate": 0.001,
"loss": 1.3669,
"step": 277200
},
{
"epoch": 89.62508080155139,
"grad_norm": 1.726732850074768,
"learning_rate": 0.001,
"loss": 1.3834,
"step": 277300
},
{
"epoch": 89.6574014221073,
"grad_norm": 1.7921017408370972,
"learning_rate": 0.001,
"loss": 1.3829,
"step": 277400
},
{
"epoch": 89.68972204266322,
"grad_norm": 1.6196974515914917,
"learning_rate": 0.001,
"loss": 1.3706,
"step": 277500
},
{
"epoch": 89.72204266321913,
"grad_norm": 1.81668221950531,
"learning_rate": 0.001,
"loss": 1.392,
"step": 277600
},
{
"epoch": 89.75436328377505,
"grad_norm": 1.2870680093765259,
"learning_rate": 0.001,
"loss": 1.3807,
"step": 277700
},
{
"epoch": 89.78668390433096,
"grad_norm": 1.490224003791809,
"learning_rate": 0.001,
"loss": 1.3803,
"step": 277800
},
{
"epoch": 89.81900452488688,
"grad_norm": 1.2039817571640015,
"learning_rate": 0.001,
"loss": 1.4091,
"step": 277900
},
{
"epoch": 89.85132514544279,
"grad_norm": 1.740922212600708,
"learning_rate": 0.001,
"loss": 1.3826,
"step": 278000
},
{
"epoch": 89.88364576599871,
"grad_norm": 1.3540980815887451,
"learning_rate": 0.001,
"loss": 1.4017,
"step": 278100
},
{
"epoch": 89.91596638655462,
"grad_norm": 1.3727006912231445,
"learning_rate": 0.001,
"loss": 1.3957,
"step": 278200
},
{
"epoch": 89.94828700711054,
"grad_norm": 1.3438173532485962,
"learning_rate": 0.001,
"loss": 1.4151,
"step": 278300
},
{
"epoch": 89.98060762766644,
"grad_norm": 1.239018201828003,
"learning_rate": 0.001,
"loss": 1.3964,
"step": 278400
},
{
"epoch": 90.01292824822237,
"grad_norm": 1.4747555255889893,
"learning_rate": 0.001,
"loss": 1.3498,
"step": 278500
},
{
"epoch": 90.04524886877829,
"grad_norm": 1.7845624685287476,
"learning_rate": 0.001,
"loss": 1.3009,
"step": 278600
},
{
"epoch": 90.0775694893342,
"grad_norm": 1.762743353843689,
"learning_rate": 0.001,
"loss": 1.3042,
"step": 278700
},
{
"epoch": 90.10989010989012,
"grad_norm": 1.501381278038025,
"learning_rate": 0.001,
"loss": 1.3031,
"step": 278800
},
{
"epoch": 90.14221073044602,
"grad_norm": 1.4839171171188354,
"learning_rate": 0.001,
"loss": 1.3144,
"step": 278900
},
{
"epoch": 90.17453135100195,
"grad_norm": 1.9532594680786133,
"learning_rate": 0.001,
"loss": 1.307,
"step": 279000
},
{
"epoch": 90.20685197155785,
"grad_norm": 1.5692918300628662,
"learning_rate": 0.001,
"loss": 1.3059,
"step": 279100
},
{
"epoch": 90.23917259211377,
"grad_norm": 1.5409700870513916,
"learning_rate": 0.001,
"loss": 1.3446,
"step": 279200
},
{
"epoch": 90.27149321266968,
"grad_norm": 1.4734153747558594,
"learning_rate": 0.001,
"loss": 1.3298,
"step": 279300
},
{
"epoch": 90.3038138332256,
"grad_norm": 2.3025665283203125,
"learning_rate": 0.001,
"loss": 1.323,
"step": 279400
},
{
"epoch": 90.33613445378151,
"grad_norm": 1.684672236442566,
"learning_rate": 0.001,
"loss": 1.3369,
"step": 279500
},
{
"epoch": 90.36845507433743,
"grad_norm": 1.2520039081573486,
"learning_rate": 0.001,
"loss": 1.3374,
"step": 279600
},
{
"epoch": 90.40077569489334,
"grad_norm": 1.5118873119354248,
"learning_rate": 0.001,
"loss": 1.3459,
"step": 279700
},
{
"epoch": 90.43309631544926,
"grad_norm": 1.4605162143707275,
"learning_rate": 0.001,
"loss": 1.3351,
"step": 279800
},
{
"epoch": 90.46541693600517,
"grad_norm": 1.6042977571487427,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 279900
},
{
"epoch": 90.49773755656109,
"grad_norm": 1.2875202894210815,
"learning_rate": 0.001,
"loss": 1.3567,
"step": 280000
},
{
"epoch": 90.530058177117,
"grad_norm": 1.8810664415359497,
"learning_rate": 0.001,
"loss": 1.3597,
"step": 280100
},
{
"epoch": 90.56237879767292,
"grad_norm": 1.9839472770690918,
"learning_rate": 0.001,
"loss": 1.3621,
"step": 280200
},
{
"epoch": 90.59469941822883,
"grad_norm": 1.4278959035873413,
"learning_rate": 0.001,
"loss": 1.3671,
"step": 280300
},
{
"epoch": 90.62702003878475,
"grad_norm": 2.0551869869232178,
"learning_rate": 0.001,
"loss": 1.3681,
"step": 280400
},
{
"epoch": 90.65934065934066,
"grad_norm": 2.150033950805664,
"learning_rate": 0.001,
"loss": 1.3537,
"step": 280500
},
{
"epoch": 90.69166127989658,
"grad_norm": 1.7056057453155518,
"learning_rate": 0.001,
"loss": 1.3803,
"step": 280600
},
{
"epoch": 90.72398190045249,
"grad_norm": 1.5661425590515137,
"learning_rate": 0.001,
"loss": 1.3802,
"step": 280700
},
{
"epoch": 90.75630252100841,
"grad_norm": 1.3944406509399414,
"learning_rate": 0.001,
"loss": 1.3771,
"step": 280800
},
{
"epoch": 90.78862314156432,
"grad_norm": 1.9906152486801147,
"learning_rate": 0.001,
"loss": 1.3777,
"step": 280900
},
{
"epoch": 90.82094376212024,
"grad_norm": 1.5358821153640747,
"learning_rate": 0.001,
"loss": 1.397,
"step": 281000
},
{
"epoch": 90.85326438267614,
"grad_norm": 1.3231483697891235,
"learning_rate": 0.001,
"loss": 1.4156,
"step": 281100
},
{
"epoch": 90.88558500323207,
"grad_norm": 1.4134284257888794,
"learning_rate": 0.001,
"loss": 1.4063,
"step": 281200
},
{
"epoch": 90.91790562378797,
"grad_norm": 1.4724249839782715,
"learning_rate": 0.001,
"loss": 1.4094,
"step": 281300
},
{
"epoch": 90.9502262443439,
"grad_norm": 1.496097207069397,
"learning_rate": 0.001,
"loss": 1.3932,
"step": 281400
},
{
"epoch": 90.9825468648998,
"grad_norm": 1.5194125175476074,
"learning_rate": 0.001,
"loss": 1.4018,
"step": 281500
},
{
"epoch": 91.01486748545572,
"grad_norm": 1.8468122482299805,
"learning_rate": 0.001,
"loss": 1.3315,
"step": 281600
},
{
"epoch": 91.04718810601163,
"grad_norm": 1.8446292877197266,
"learning_rate": 0.001,
"loss": 1.2891,
"step": 281700
},
{
"epoch": 91.07950872656755,
"grad_norm": 1.7370030879974365,
"learning_rate": 0.001,
"loss": 1.2877,
"step": 281800
},
{
"epoch": 91.11182934712346,
"grad_norm": 1.6517528295516968,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 281900
},
{
"epoch": 91.14414996767938,
"grad_norm": 2.1205289363861084,
"learning_rate": 0.001,
"loss": 1.3125,
"step": 282000
},
{
"epoch": 91.17647058823529,
"grad_norm": 2.161062717437744,
"learning_rate": 0.001,
"loss": 1.3193,
"step": 282100
},
{
"epoch": 91.20879120879121,
"grad_norm": 1.7916630506515503,
"learning_rate": 0.001,
"loss": 1.3146,
"step": 282200
},
{
"epoch": 91.24111182934712,
"grad_norm": 1.6187920570373535,
"learning_rate": 0.001,
"loss": 1.3213,
"step": 282300
},
{
"epoch": 91.27343244990304,
"grad_norm": 1.746664047241211,
"learning_rate": 0.001,
"loss": 1.3202,
"step": 282400
},
{
"epoch": 91.30575307045895,
"grad_norm": 1.3970807790756226,
"learning_rate": 0.001,
"loss": 1.3343,
"step": 282500
},
{
"epoch": 91.33807369101487,
"grad_norm": 1.7289519309997559,
"learning_rate": 0.001,
"loss": 1.3327,
"step": 282600
},
{
"epoch": 91.37039431157078,
"grad_norm": 1.9510860443115234,
"learning_rate": 0.001,
"loss": 1.338,
"step": 282700
},
{
"epoch": 91.4027149321267,
"grad_norm": 1.3857265710830688,
"learning_rate": 0.001,
"loss": 1.346,
"step": 282800
},
{
"epoch": 91.4350355526826,
"grad_norm": 1.7828974723815918,
"learning_rate": 0.001,
"loss": 1.3388,
"step": 282900
},
{
"epoch": 91.46735617323853,
"grad_norm": 1.7658045291900635,
"learning_rate": 0.001,
"loss": 1.3423,
"step": 283000
},
{
"epoch": 91.49967679379444,
"grad_norm": 1.5482114553451538,
"learning_rate": 0.001,
"loss": 1.3551,
"step": 283100
},
{
"epoch": 91.53199741435036,
"grad_norm": 1.4986413717269897,
"learning_rate": 0.001,
"loss": 1.3514,
"step": 283200
},
{
"epoch": 91.56431803490626,
"grad_norm": 1.9350371360778809,
"learning_rate": 0.001,
"loss": 1.3518,
"step": 283300
},
{
"epoch": 91.59663865546219,
"grad_norm": 2.14019775390625,
"learning_rate": 0.001,
"loss": 1.3653,
"step": 283400
},
{
"epoch": 91.6289592760181,
"grad_norm": 1.604882836341858,
"learning_rate": 0.001,
"loss": 1.3631,
"step": 283500
},
{
"epoch": 91.66127989657402,
"grad_norm": 1.8683799505233765,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 283600
},
{
"epoch": 91.69360051712992,
"grad_norm": 1.643358588218689,
"learning_rate": 0.001,
"loss": 1.3727,
"step": 283700
},
{
"epoch": 91.72592113768584,
"grad_norm": 1.5758355855941772,
"learning_rate": 0.001,
"loss": 1.3723,
"step": 283800
},
{
"epoch": 91.75824175824175,
"grad_norm": 1.7453467845916748,
"learning_rate": 0.001,
"loss": 1.3727,
"step": 283900
},
{
"epoch": 91.79056237879767,
"grad_norm": 2.0909583568573,
"learning_rate": 0.001,
"loss": 1.3846,
"step": 284000
},
{
"epoch": 91.82288299935358,
"grad_norm": 1.766341209411621,
"learning_rate": 0.001,
"loss": 1.3861,
"step": 284100
},
{
"epoch": 91.8552036199095,
"grad_norm": 1.5741945505142212,
"learning_rate": 0.001,
"loss": 1.3976,
"step": 284200
},
{
"epoch": 91.88752424046541,
"grad_norm": 1.8239383697509766,
"learning_rate": 0.001,
"loss": 1.386,
"step": 284300
},
{
"epoch": 91.91984486102133,
"grad_norm": 2.422290563583374,
"learning_rate": 0.001,
"loss": 1.3837,
"step": 284400
},
{
"epoch": 91.95216548157724,
"grad_norm": 2.4389803409576416,
"learning_rate": 0.001,
"loss": 1.394,
"step": 284500
},
{
"epoch": 91.98448610213316,
"grad_norm": 1.7738986015319824,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 284600
},
{
"epoch": 92.01680672268908,
"grad_norm": 2.3665008544921875,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 284700
},
{
"epoch": 92.04912734324499,
"grad_norm": 3.186936140060425,
"learning_rate": 0.001,
"loss": 1.2915,
"step": 284800
},
{
"epoch": 92.08144796380091,
"grad_norm": 2.5595545768737793,
"learning_rate": 0.001,
"loss": 1.2999,
"step": 284900
},
{
"epoch": 92.11376858435682,
"grad_norm": 3.0340330600738525,
"learning_rate": 0.001,
"loss": 1.3,
"step": 285000
},
{
"epoch": 92.14608920491274,
"grad_norm": 1.8480430841445923,
"learning_rate": 0.001,
"loss": 1.2998,
"step": 285100
},
{
"epoch": 92.17840982546865,
"grad_norm": 2.4935553073883057,
"learning_rate": 0.001,
"loss": 1.3171,
"step": 285200
},
{
"epoch": 92.21073044602457,
"grad_norm": 2.521172046661377,
"learning_rate": 0.001,
"loss": 1.3087,
"step": 285300
},
{
"epoch": 92.24305106658048,
"grad_norm": 2.152034044265747,
"learning_rate": 0.001,
"loss": 1.3142,
"step": 285400
},
{
"epoch": 92.2753716871364,
"grad_norm": 2.193843126296997,
"learning_rate": 0.001,
"loss": 1.3315,
"step": 285500
},
{
"epoch": 92.3076923076923,
"grad_norm": 1.6960961818695068,
"learning_rate": 0.001,
"loss": 1.3246,
"step": 285600
},
{
"epoch": 92.34001292824823,
"grad_norm": 2.940481185913086,
"learning_rate": 0.001,
"loss": 1.3284,
"step": 285700
},
{
"epoch": 92.37233354880414,
"grad_norm": 2.660595178604126,
"learning_rate": 0.001,
"loss": 1.3291,
"step": 285800
},
{
"epoch": 92.40465416936006,
"grad_norm": 2.349214792251587,
"learning_rate": 0.001,
"loss": 1.3288,
"step": 285900
},
{
"epoch": 92.43697478991596,
"grad_norm": 1.7059687376022339,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 286000
},
{
"epoch": 92.46929541047189,
"grad_norm": 1.6006224155426025,
"learning_rate": 0.001,
"loss": 1.3443,
"step": 286100
},
{
"epoch": 92.5016160310278,
"grad_norm": 1.9527106285095215,
"learning_rate": 0.001,
"loss": 1.3383,
"step": 286200
},
{
"epoch": 92.53393665158372,
"grad_norm": 2.163050413131714,
"learning_rate": 0.001,
"loss": 1.3555,
"step": 286300
},
{
"epoch": 92.56625727213962,
"grad_norm": 2.1168384552001953,
"learning_rate": 0.001,
"loss": 1.3498,
"step": 286400
},
{
"epoch": 92.59857789269554,
"grad_norm": 2.3591318130493164,
"learning_rate": 0.001,
"loss": 1.3501,
"step": 286500
},
{
"epoch": 92.63089851325145,
"grad_norm": 2.360686779022217,
"learning_rate": 0.001,
"loss": 1.3589,
"step": 286600
},
{
"epoch": 92.66321913380737,
"grad_norm": 2.009597063064575,
"learning_rate": 0.001,
"loss": 1.3676,
"step": 286700
},
{
"epoch": 92.69553975436328,
"grad_norm": 2.3390748500823975,
"learning_rate": 0.001,
"loss": 1.3599,
"step": 286800
},
{
"epoch": 92.7278603749192,
"grad_norm": 1.2635425329208374,
"learning_rate": 0.001,
"loss": 1.3601,
"step": 286900
},
{
"epoch": 92.76018099547511,
"grad_norm": 1.7672971487045288,
"learning_rate": 0.001,
"loss": 1.3781,
"step": 287000
},
{
"epoch": 92.79250161603103,
"grad_norm": 1.7889213562011719,
"learning_rate": 0.001,
"loss": 1.3591,
"step": 287100
},
{
"epoch": 92.82482223658694,
"grad_norm": 2.1574652194976807,
"learning_rate": 0.001,
"loss": 1.3865,
"step": 287200
},
{
"epoch": 92.85714285714286,
"grad_norm": 1.9692292213439941,
"learning_rate": 0.001,
"loss": 1.3787,
"step": 287300
},
{
"epoch": 92.88946347769877,
"grad_norm": 2.3051788806915283,
"learning_rate": 0.001,
"loss": 1.3668,
"step": 287400
},
{
"epoch": 92.92178409825469,
"grad_norm": 1.5948925018310547,
"learning_rate": 0.001,
"loss": 1.3725,
"step": 287500
},
{
"epoch": 92.9541047188106,
"grad_norm": 1.6683539152145386,
"learning_rate": 0.001,
"loss": 1.3765,
"step": 287600
},
{
"epoch": 92.98642533936652,
"grad_norm": 2.239222288131714,
"learning_rate": 0.001,
"loss": 1.3927,
"step": 287700
},
{
"epoch": 93.01874595992243,
"grad_norm": 1.803403615951538,
"learning_rate": 0.001,
"loss": 1.322,
"step": 287800
},
{
"epoch": 93.05106658047835,
"grad_norm": 1.5085774660110474,
"learning_rate": 0.001,
"loss": 1.2678,
"step": 287900
},
{
"epoch": 93.08338720103426,
"grad_norm": 1.6019845008850098,
"learning_rate": 0.001,
"loss": 1.2806,
"step": 288000
},
{
"epoch": 93.11570782159018,
"grad_norm": 1.068983554840088,
"learning_rate": 0.001,
"loss": 1.2895,
"step": 288100
},
{
"epoch": 93.14802844214609,
"grad_norm": 2.3375728130340576,
"learning_rate": 0.001,
"loss": 1.2959,
"step": 288200
},
{
"epoch": 93.180349062702,
"grad_norm": 1.768635630607605,
"learning_rate": 0.001,
"loss": 1.2912,
"step": 288300
},
{
"epoch": 93.21266968325791,
"grad_norm": 1.4623817205429077,
"learning_rate": 0.001,
"loss": 1.2952,
"step": 288400
},
{
"epoch": 93.24499030381384,
"grad_norm": 1.6178747415542603,
"learning_rate": 0.001,
"loss": 1.2921,
"step": 288500
},
{
"epoch": 93.27731092436974,
"grad_norm": 1.7693052291870117,
"learning_rate": 0.001,
"loss": 1.3118,
"step": 288600
},
{
"epoch": 93.30963154492567,
"grad_norm": 1.5522078275680542,
"learning_rate": 0.001,
"loss": 1.3093,
"step": 288700
},
{
"epoch": 93.34195216548157,
"grad_norm": 1.8236924409866333,
"learning_rate": 0.001,
"loss": 1.3198,
"step": 288800
},
{
"epoch": 93.3742727860375,
"grad_norm": 2.227855682373047,
"learning_rate": 0.001,
"loss": 1.3177,
"step": 288900
},
{
"epoch": 93.4065934065934,
"grad_norm": 1.994999885559082,
"learning_rate": 0.001,
"loss": 1.3189,
"step": 289000
},
{
"epoch": 93.43891402714932,
"grad_norm": 1.5782331228256226,
"learning_rate": 0.001,
"loss": 1.3391,
"step": 289100
},
{
"epoch": 93.47123464770523,
"grad_norm": 2.008796215057373,
"learning_rate": 0.001,
"loss": 1.3323,
"step": 289200
},
{
"epoch": 93.50355526826115,
"grad_norm": 1.780877709388733,
"learning_rate": 0.001,
"loss": 1.3351,
"step": 289300
},
{
"epoch": 93.53587588881706,
"grad_norm": 2.309138059616089,
"learning_rate": 0.001,
"loss": 1.3436,
"step": 289400
},
{
"epoch": 93.56819650937298,
"grad_norm": 1.6557698249816895,
"learning_rate": 0.001,
"loss": 1.3627,
"step": 289500
},
{
"epoch": 93.60051712992889,
"grad_norm": 1.8269431591033936,
"learning_rate": 0.001,
"loss": 1.3427,
"step": 289600
},
{
"epoch": 93.63283775048481,
"grad_norm": 1.5965728759765625,
"learning_rate": 0.001,
"loss": 1.3409,
"step": 289700
},
{
"epoch": 93.66515837104072,
"grad_norm": 1.6863973140716553,
"learning_rate": 0.001,
"loss": 1.355,
"step": 289800
},
{
"epoch": 93.69747899159664,
"grad_norm": 1.3488801717758179,
"learning_rate": 0.001,
"loss": 1.3732,
"step": 289900
},
{
"epoch": 93.72979961215255,
"grad_norm": 1.7931402921676636,
"learning_rate": 0.001,
"loss": 1.3582,
"step": 290000
},
{
"epoch": 93.76212023270847,
"grad_norm": 1.497398853302002,
"learning_rate": 0.001,
"loss": 1.366,
"step": 290100
},
{
"epoch": 93.79444085326438,
"grad_norm": 1.2366420030593872,
"learning_rate": 0.001,
"loss": 1.3797,
"step": 290200
},
{
"epoch": 93.8267614738203,
"grad_norm": 2.0114481449127197,
"learning_rate": 0.001,
"loss": 1.3792,
"step": 290300
},
{
"epoch": 93.8590820943762,
"grad_norm": 1.6956149339675903,
"learning_rate": 0.001,
"loss": 1.3652,
"step": 290400
},
{
"epoch": 93.89140271493213,
"grad_norm": 1.594327688217163,
"learning_rate": 0.001,
"loss": 1.3796,
"step": 290500
},
{
"epoch": 93.92372333548803,
"grad_norm": 2.3135488033294678,
"learning_rate": 0.001,
"loss": 1.3669,
"step": 290600
},
{
"epoch": 93.95604395604396,
"grad_norm": 1.8480385541915894,
"learning_rate": 0.001,
"loss": 1.3994,
"step": 290700
},
{
"epoch": 93.98836457659988,
"grad_norm": 1.6887590885162354,
"learning_rate": 0.001,
"loss": 1.3761,
"step": 290800
},
{
"epoch": 94.02068519715579,
"grad_norm": 2.628980875015259,
"learning_rate": 0.001,
"loss": 1.3282,
"step": 290900
},
{
"epoch": 94.0530058177117,
"grad_norm": 1.4327553510665894,
"learning_rate": 0.001,
"loss": 1.2685,
"step": 291000
},
{
"epoch": 94.08532643826761,
"grad_norm": 1.4887218475341797,
"learning_rate": 0.001,
"loss": 1.2829,
"step": 291100
},
{
"epoch": 94.11764705882354,
"grad_norm": 1.1396749019622803,
"learning_rate": 0.001,
"loss": 1.2873,
"step": 291200
},
{
"epoch": 94.14996767937944,
"grad_norm": 2.3619840145111084,
"learning_rate": 0.001,
"loss": 1.2925,
"step": 291300
},
{
"epoch": 94.18228829993537,
"grad_norm": 2.0263519287109375,
"learning_rate": 0.001,
"loss": 1.2792,
"step": 291400
},
{
"epoch": 94.21460892049127,
"grad_norm": 1.562637448310852,
"learning_rate": 0.001,
"loss": 1.298,
"step": 291500
},
{
"epoch": 94.2469295410472,
"grad_norm": 1.2812947034835815,
"learning_rate": 0.001,
"loss": 1.2939,
"step": 291600
},
{
"epoch": 94.2792501616031,
"grad_norm": 1.9622937440872192,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 291700
},
{
"epoch": 94.31157078215902,
"grad_norm": 1.60319983959198,
"learning_rate": 0.001,
"loss": 1.3061,
"step": 291800
},
{
"epoch": 94.34389140271493,
"grad_norm": 2.189262866973877,
"learning_rate": 0.001,
"loss": 1.3139,
"step": 291900
},
{
"epoch": 94.37621202327085,
"grad_norm": 1.7461422681808472,
"learning_rate": 0.001,
"loss": 1.3061,
"step": 292000
},
{
"epoch": 94.40853264382676,
"grad_norm": 1.0402483940124512,
"learning_rate": 0.001,
"loss": 1.3211,
"step": 292100
},
{
"epoch": 94.44085326438268,
"grad_norm": 2.0994131565093994,
"learning_rate": 0.001,
"loss": 1.3321,
"step": 292200
},
{
"epoch": 94.47317388493859,
"grad_norm": 1.897391676902771,
"learning_rate": 0.001,
"loss": 1.3356,
"step": 292300
},
{
"epoch": 94.50549450549451,
"grad_norm": 1.9235081672668457,
"learning_rate": 0.001,
"loss": 1.3465,
"step": 292400
},
{
"epoch": 94.53781512605042,
"grad_norm": 1.3842601776123047,
"learning_rate": 0.001,
"loss": 1.3398,
"step": 292500
},
{
"epoch": 94.57013574660634,
"grad_norm": 1.4253175258636475,
"learning_rate": 0.001,
"loss": 1.323,
"step": 292600
},
{
"epoch": 94.60245636716225,
"grad_norm": 1.9249699115753174,
"learning_rate": 0.001,
"loss": 1.3413,
"step": 292700
},
{
"epoch": 94.63477698771817,
"grad_norm": 1.3795268535614014,
"learning_rate": 0.001,
"loss": 1.3444,
"step": 292800
},
{
"epoch": 94.66709760827408,
"grad_norm": 1.1008977890014648,
"learning_rate": 0.001,
"loss": 1.35,
"step": 292900
},
{
"epoch": 94.69941822883,
"grad_norm": 1.3548905849456787,
"learning_rate": 0.001,
"loss": 1.3545,
"step": 293000
},
{
"epoch": 94.7317388493859,
"grad_norm": 1.6496766805648804,
"learning_rate": 0.001,
"loss": 1.3644,
"step": 293100
},
{
"epoch": 94.76405946994183,
"grad_norm": 1.6453754901885986,
"learning_rate": 0.001,
"loss": 1.369,
"step": 293200
},
{
"epoch": 94.79638009049773,
"grad_norm": 1.339321494102478,
"learning_rate": 0.001,
"loss": 1.3799,
"step": 293300
},
{
"epoch": 94.82870071105366,
"grad_norm": 1.661134123802185,
"learning_rate": 0.001,
"loss": 1.3595,
"step": 293400
},
{
"epoch": 94.86102133160956,
"grad_norm": 1.2583048343658447,
"learning_rate": 0.001,
"loss": 1.3511,
"step": 293500
},
{
"epoch": 94.89334195216549,
"grad_norm": 2.5929088592529297,
"learning_rate": 0.001,
"loss": 1.3687,
"step": 293600
},
{
"epoch": 94.9256625727214,
"grad_norm": 1.5293428897857666,
"learning_rate": 0.001,
"loss": 1.3756,
"step": 293700
},
{
"epoch": 94.95798319327731,
"grad_norm": 1.5770586729049683,
"learning_rate": 0.001,
"loss": 1.376,
"step": 293800
},
{
"epoch": 94.99030381383322,
"grad_norm": 1.988012671470642,
"learning_rate": 0.001,
"loss": 1.3761,
"step": 293900
},
{
"epoch": 95.02262443438914,
"grad_norm": 1.481789469718933,
"learning_rate": 0.001,
"loss": 1.3097,
"step": 294000
},
{
"epoch": 95.05494505494505,
"grad_norm": 1.075170874595642,
"learning_rate": 0.001,
"loss": 1.2704,
"step": 294100
},
{
"epoch": 95.08726567550097,
"grad_norm": 1.4342702627182007,
"learning_rate": 0.001,
"loss": 1.272,
"step": 294200
},
{
"epoch": 95.11958629605688,
"grad_norm": 2.060530424118042,
"learning_rate": 0.001,
"loss": 1.2735,
"step": 294300
},
{
"epoch": 95.1519069166128,
"grad_norm": 1.5602858066558838,
"learning_rate": 0.001,
"loss": 1.297,
"step": 294400
},
{
"epoch": 95.18422753716871,
"grad_norm": 1.2675756216049194,
"learning_rate": 0.001,
"loss": 1.2855,
"step": 294500
},
{
"epoch": 95.21654815772463,
"grad_norm": 1.542402744293213,
"learning_rate": 0.001,
"loss": 1.2923,
"step": 294600
},
{
"epoch": 95.24886877828054,
"grad_norm": 1.4929137229919434,
"learning_rate": 0.001,
"loss": 1.2877,
"step": 294700
},
{
"epoch": 95.28118939883646,
"grad_norm": 1.457393765449524,
"learning_rate": 0.001,
"loss": 1.299,
"step": 294800
},
{
"epoch": 95.31351001939237,
"grad_norm": 1.5335299968719482,
"learning_rate": 0.001,
"loss": 1.3107,
"step": 294900
},
{
"epoch": 95.34583063994829,
"grad_norm": 1.3645907640457153,
"learning_rate": 0.001,
"loss": 1.3076,
"step": 295000
},
{
"epoch": 95.3781512605042,
"grad_norm": 1.1843338012695312,
"learning_rate": 0.001,
"loss": 1.325,
"step": 295100
},
{
"epoch": 95.41047188106012,
"grad_norm": 1.1182949542999268,
"learning_rate": 0.001,
"loss": 1.3131,
"step": 295200
},
{
"epoch": 95.44279250161603,
"grad_norm": 1.6379339694976807,
"learning_rate": 0.001,
"loss": 1.3221,
"step": 295300
},
{
"epoch": 95.47511312217195,
"grad_norm": 1.287140965461731,
"learning_rate": 0.001,
"loss": 1.3242,
"step": 295400
},
{
"epoch": 95.50743374272786,
"grad_norm": 1.493634819984436,
"learning_rate": 0.001,
"loss": 1.3275,
"step": 295500
},
{
"epoch": 95.53975436328378,
"grad_norm": 1.4656271934509277,
"learning_rate": 0.001,
"loss": 1.3395,
"step": 295600
},
{
"epoch": 95.57207498383968,
"grad_norm": 1.4360830783843994,
"learning_rate": 0.001,
"loss": 1.3297,
"step": 295700
},
{
"epoch": 95.6043956043956,
"grad_norm": 1.2041678428649902,
"learning_rate": 0.001,
"loss": 1.3359,
"step": 295800
},
{
"epoch": 95.63671622495151,
"grad_norm": 1.5483204126358032,
"learning_rate": 0.001,
"loss": 1.3443,
"step": 295900
},
{
"epoch": 95.66903684550743,
"grad_norm": 1.6295415163040161,
"learning_rate": 0.001,
"loss": 1.3356,
"step": 296000
},
{
"epoch": 95.70135746606334,
"grad_norm": 1.411381483078003,
"learning_rate": 0.001,
"loss": 1.3448,
"step": 296100
},
{
"epoch": 95.73367808661926,
"grad_norm": 1.4215126037597656,
"learning_rate": 0.001,
"loss": 1.3629,
"step": 296200
},
{
"epoch": 95.76599870717517,
"grad_norm": 1.3287146091461182,
"learning_rate": 0.001,
"loss": 1.3309,
"step": 296300
},
{
"epoch": 95.7983193277311,
"grad_norm": 1.434643268585205,
"learning_rate": 0.001,
"loss": 1.3654,
"step": 296400
},
{
"epoch": 95.830639948287,
"grad_norm": 1.254253625869751,
"learning_rate": 0.001,
"loss": 1.3602,
"step": 296500
},
{
"epoch": 95.86296056884292,
"grad_norm": 1.2696986198425293,
"learning_rate": 0.001,
"loss": 1.3635,
"step": 296600
},
{
"epoch": 95.89528118939883,
"grad_norm": 1.7204337120056152,
"learning_rate": 0.001,
"loss": 1.358,
"step": 296700
},
{
"epoch": 95.92760180995475,
"grad_norm": 1.567694902420044,
"learning_rate": 0.001,
"loss": 1.3732,
"step": 296800
},
{
"epoch": 95.95992243051066,
"grad_norm": 1.5819817781448364,
"learning_rate": 0.001,
"loss": 1.3611,
"step": 296900
},
{
"epoch": 95.99224305106658,
"grad_norm": 1.3724982738494873,
"learning_rate": 0.001,
"loss": 1.3735,
"step": 297000
},
{
"epoch": 96.0245636716225,
"grad_norm": 1.350100040435791,
"learning_rate": 0.001,
"loss": 1.3103,
"step": 297100
},
{
"epoch": 96.05688429217841,
"grad_norm": 1.3620035648345947,
"learning_rate": 0.001,
"loss": 1.2624,
"step": 297200
},
{
"epoch": 96.08920491273433,
"grad_norm": 1.5321812629699707,
"learning_rate": 0.001,
"loss": 1.2719,
"step": 297300
},
{
"epoch": 96.12152553329024,
"grad_norm": 1.5273820161819458,
"learning_rate": 0.001,
"loss": 1.2647,
"step": 297400
},
{
"epoch": 96.15384615384616,
"grad_norm": 1.4025139808654785,
"learning_rate": 0.001,
"loss": 1.2806,
"step": 297500
},
{
"epoch": 96.18616677440207,
"grad_norm": 1.4742810726165771,
"learning_rate": 0.001,
"loss": 1.2827,
"step": 297600
},
{
"epoch": 96.21848739495799,
"grad_norm": 2.5691189765930176,
"learning_rate": 0.001,
"loss": 1.2955,
"step": 297700
},
{
"epoch": 96.2508080155139,
"grad_norm": 1.4609074592590332,
"learning_rate": 0.001,
"loss": 1.279,
"step": 297800
},
{
"epoch": 96.28312863606982,
"grad_norm": 1.2360470294952393,
"learning_rate": 0.001,
"loss": 1.2769,
"step": 297900
},
{
"epoch": 96.31544925662573,
"grad_norm": 1.4440158605575562,
"learning_rate": 0.001,
"loss": 1.3036,
"step": 298000
},
{
"epoch": 96.34776987718165,
"grad_norm": 1.631263256072998,
"learning_rate": 0.001,
"loss": 1.3134,
"step": 298100
},
{
"epoch": 96.38009049773756,
"grad_norm": 1.2250239849090576,
"learning_rate": 0.001,
"loss": 1.3279,
"step": 298200
},
{
"epoch": 96.41241111829348,
"grad_norm": 2.013558864593506,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 298300
},
{
"epoch": 96.44473173884938,
"grad_norm": 1.4350364208221436,
"learning_rate": 0.001,
"loss": 1.3067,
"step": 298400
},
{
"epoch": 96.4770523594053,
"grad_norm": 1.2664413452148438,
"learning_rate": 0.001,
"loss": 1.3204,
"step": 298500
},
{
"epoch": 96.50937297996121,
"grad_norm": 1.6226825714111328,
"learning_rate": 0.001,
"loss": 1.3254,
"step": 298600
},
{
"epoch": 96.54169360051714,
"grad_norm": 1.538313627243042,
"learning_rate": 0.001,
"loss": 1.3281,
"step": 298700
},
{
"epoch": 96.57401422107304,
"grad_norm": 1.3091726303100586,
"learning_rate": 0.001,
"loss": 1.3295,
"step": 298800
},
{
"epoch": 96.60633484162896,
"grad_norm": 1.818063497543335,
"learning_rate": 0.001,
"loss": 1.3339,
"step": 298900
},
{
"epoch": 96.63865546218487,
"grad_norm": 1.6741087436676025,
"learning_rate": 0.001,
"loss": 1.3173,
"step": 299000
},
{
"epoch": 96.6709760827408,
"grad_norm": 1.7017314434051514,
"learning_rate": 0.001,
"loss": 1.3337,
"step": 299100
},
{
"epoch": 96.7032967032967,
"grad_norm": 1.8388307094573975,
"learning_rate": 0.001,
"loss": 1.3484,
"step": 299200
},
{
"epoch": 96.73561732385262,
"grad_norm": 1.7868112325668335,
"learning_rate": 0.001,
"loss": 1.3395,
"step": 299300
},
{
"epoch": 96.76793794440853,
"grad_norm": 1.7302414178848267,
"learning_rate": 0.001,
"loss": 1.3587,
"step": 299400
},
{
"epoch": 96.80025856496445,
"grad_norm": 1.8578855991363525,
"learning_rate": 0.001,
"loss": 1.3487,
"step": 299500
},
{
"epoch": 96.83257918552036,
"grad_norm": 1.5445219278335571,
"learning_rate": 0.001,
"loss": 1.3575,
"step": 299600
},
{
"epoch": 96.86489980607628,
"grad_norm": 1.4697306156158447,
"learning_rate": 0.001,
"loss": 1.3564,
"step": 299700
},
{
"epoch": 96.89722042663219,
"grad_norm": 1.7938448190689087,
"learning_rate": 0.001,
"loss": 1.3585,
"step": 299800
},
{
"epoch": 96.92954104718811,
"grad_norm": 1.2490466833114624,
"learning_rate": 0.001,
"loss": 1.3491,
"step": 299900
},
{
"epoch": 96.96186166774402,
"grad_norm": 1.7172377109527588,
"learning_rate": 0.001,
"loss": 1.3683,
"step": 300000
},
{
"epoch": 96.99418228829994,
"grad_norm": 1.2884808778762817,
"learning_rate": 0.001,
"loss": 1.3553,
"step": 300100
},
{
"epoch": 97.02650290885585,
"grad_norm": 1.6161185503005981,
"learning_rate": 0.001,
"loss": 1.2776,
"step": 300200
},
{
"epoch": 97.05882352941177,
"grad_norm": 1.2775593996047974,
"learning_rate": 0.001,
"loss": 1.2645,
"step": 300300
},
{
"epoch": 97.09114414996768,
"grad_norm": 1.9834895133972168,
"learning_rate": 0.001,
"loss": 1.2643,
"step": 300400
},
{
"epoch": 97.1234647705236,
"grad_norm": 1.7059547901153564,
"learning_rate": 0.001,
"loss": 1.265,
"step": 300500
},
{
"epoch": 97.1557853910795,
"grad_norm": 1.3417140245437622,
"learning_rate": 0.001,
"loss": 1.2757,
"step": 300600
},
{
"epoch": 97.18810601163543,
"grad_norm": 1.8753329515457153,
"learning_rate": 0.001,
"loss": 1.2701,
"step": 300700
},
{
"epoch": 97.22042663219133,
"grad_norm": 2.8010098934173584,
"learning_rate": 0.001,
"loss": 1.2742,
"step": 300800
},
{
"epoch": 97.25274725274726,
"grad_norm": 1.9707610607147217,
"learning_rate": 0.001,
"loss": 1.2872,
"step": 300900
},
{
"epoch": 97.28506787330316,
"grad_norm": 1.6114462614059448,
"learning_rate": 0.001,
"loss": 1.2982,
"step": 301000
},
{
"epoch": 97.31738849385908,
"grad_norm": 2.399348735809326,
"learning_rate": 0.001,
"loss": 1.305,
"step": 301100
},
{
"epoch": 97.34970911441499,
"grad_norm": 1.589941143989563,
"learning_rate": 0.001,
"loss": 1.2997,
"step": 301200
},
{
"epoch": 97.38202973497091,
"grad_norm": 1.3192157745361328,
"learning_rate": 0.001,
"loss": 1.3076,
"step": 301300
},
{
"epoch": 97.41435035552682,
"grad_norm": 1.3781229257583618,
"learning_rate": 0.001,
"loss": 1.2986,
"step": 301400
},
{
"epoch": 97.44667097608274,
"grad_norm": 1.4433748722076416,
"learning_rate": 0.001,
"loss": 1.3069,
"step": 301500
},
{
"epoch": 97.47899159663865,
"grad_norm": 1.935983657836914,
"learning_rate": 0.001,
"loss": 1.3077,
"step": 301600
},
{
"epoch": 97.51131221719457,
"grad_norm": 1.6531956195831299,
"learning_rate": 0.001,
"loss": 1.3318,
"step": 301700
},
{
"epoch": 97.54363283775048,
"grad_norm": 1.5503556728363037,
"learning_rate": 0.001,
"loss": 1.3257,
"step": 301800
},
{
"epoch": 97.5759534583064,
"grad_norm": 1.3275309801101685,
"learning_rate": 0.001,
"loss": 1.3209,
"step": 301900
},
{
"epoch": 97.60827407886231,
"grad_norm": 1.3888849020004272,
"learning_rate": 0.001,
"loss": 1.337,
"step": 302000
},
{
"epoch": 97.64059469941823,
"grad_norm": 1.237357497215271,
"learning_rate": 0.001,
"loss": 1.3319,
"step": 302100
},
{
"epoch": 97.67291531997414,
"grad_norm": 2.0060534477233887,
"learning_rate": 0.001,
"loss": 1.3208,
"step": 302200
},
{
"epoch": 97.70523594053006,
"grad_norm": 1.7788901329040527,
"learning_rate": 0.001,
"loss": 1.3274,
"step": 302300
},
{
"epoch": 97.73755656108597,
"grad_norm": 1.6663439273834229,
"learning_rate": 0.001,
"loss": 1.3295,
"step": 302400
},
{
"epoch": 97.76987718164189,
"grad_norm": 1.3143705129623413,
"learning_rate": 0.001,
"loss": 1.3469,
"step": 302500
},
{
"epoch": 97.8021978021978,
"grad_norm": 1.6444023847579956,
"learning_rate": 0.001,
"loss": 1.3575,
"step": 302600
},
{
"epoch": 97.83451842275372,
"grad_norm": 1.390714168548584,
"learning_rate": 0.001,
"loss": 1.3473,
"step": 302700
},
{
"epoch": 97.86683904330962,
"grad_norm": 1.401758074760437,
"learning_rate": 0.001,
"loss": 1.3507,
"step": 302800
},
{
"epoch": 97.89915966386555,
"grad_norm": 1.391494631767273,
"learning_rate": 0.001,
"loss": 1.3494,
"step": 302900
},
{
"epoch": 97.93148028442145,
"grad_norm": 1.4199986457824707,
"learning_rate": 0.001,
"loss": 1.3634,
"step": 303000
},
{
"epoch": 97.96380090497738,
"grad_norm": 1.9891318082809448,
"learning_rate": 0.001,
"loss": 1.3637,
"step": 303100
},
{
"epoch": 97.99612152553328,
"grad_norm": 1.6727330684661865,
"learning_rate": 0.001,
"loss": 1.3514,
"step": 303200
},
{
"epoch": 98.0284421460892,
"grad_norm": 1.7826147079467773,
"learning_rate": 0.001,
"loss": 1.2465,
"step": 303300
},
{
"epoch": 98.06076276664513,
"grad_norm": 1.378698706626892,
"learning_rate": 0.001,
"loss": 1.2539,
"step": 303400
},
{
"epoch": 98.09308338720103,
"grad_norm": 2.177722930908203,
"learning_rate": 0.001,
"loss": 1.2587,
"step": 303500
},
{
"epoch": 98.12540400775696,
"grad_norm": 1.6054373979568481,
"learning_rate": 0.001,
"loss": 1.2646,
"step": 303600
},
{
"epoch": 98.15772462831286,
"grad_norm": 1.6011995077133179,
"learning_rate": 0.001,
"loss": 1.2693,
"step": 303700
},
{
"epoch": 98.19004524886878,
"grad_norm": 1.323742151260376,
"learning_rate": 0.001,
"loss": 1.2759,
"step": 303800
},
{
"epoch": 98.22236586942469,
"grad_norm": 1.7112693786621094,
"learning_rate": 0.001,
"loss": 1.2837,
"step": 303900
},
{
"epoch": 98.25468648998061,
"grad_norm": 1.6237554550170898,
"learning_rate": 0.001,
"loss": 1.2882,
"step": 304000
},
{
"epoch": 98.28700711053652,
"grad_norm": 1.648337960243225,
"learning_rate": 0.001,
"loss": 1.2927,
"step": 304100
},
{
"epoch": 98.31932773109244,
"grad_norm": 1.1869858503341675,
"learning_rate": 0.001,
"loss": 1.2889,
"step": 304200
},
{
"epoch": 98.35164835164835,
"grad_norm": 1.593783974647522,
"learning_rate": 0.001,
"loss": 1.2865,
"step": 304300
},
{
"epoch": 98.38396897220427,
"grad_norm": 1.4546592235565186,
"learning_rate": 0.001,
"loss": 1.2997,
"step": 304400
},
{
"epoch": 98.41628959276018,
"grad_norm": 1.6142128705978394,
"learning_rate": 0.001,
"loss": 1.3054,
"step": 304500
},
{
"epoch": 98.4486102133161,
"grad_norm": 1.240814447402954,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 304600
},
{
"epoch": 98.48093083387201,
"grad_norm": 1.5925500392913818,
"learning_rate": 0.001,
"loss": 1.2954,
"step": 304700
},
{
"epoch": 98.51325145442793,
"grad_norm": 1.2763153314590454,
"learning_rate": 0.001,
"loss": 1.306,
"step": 304800
},
{
"epoch": 98.54557207498384,
"grad_norm": 1.9138118028640747,
"learning_rate": 0.001,
"loss": 1.3358,
"step": 304900
},
{
"epoch": 98.57789269553976,
"grad_norm": 1.1177512407302856,
"learning_rate": 0.001,
"loss": 1.3252,
"step": 305000
},
{
"epoch": 98.61021331609567,
"grad_norm": 1.306567668914795,
"learning_rate": 0.001,
"loss": 1.3218,
"step": 305100
},
{
"epoch": 98.64253393665159,
"grad_norm": 1.3440361022949219,
"learning_rate": 0.001,
"loss": 1.3306,
"step": 305200
},
{
"epoch": 98.6748545572075,
"grad_norm": 1.7852712869644165,
"learning_rate": 0.001,
"loss": 1.3287,
"step": 305300
},
{
"epoch": 98.70717517776342,
"grad_norm": 1.544251799583435,
"learning_rate": 0.001,
"loss": 1.3272,
"step": 305400
},
{
"epoch": 98.73949579831933,
"grad_norm": 1.6691638231277466,
"learning_rate": 0.001,
"loss": 1.3318,
"step": 305500
},
{
"epoch": 98.77181641887525,
"grad_norm": 1.2872183322906494,
"learning_rate": 0.001,
"loss": 1.3438,
"step": 305600
},
{
"epoch": 98.80413703943115,
"grad_norm": 1.4782793521881104,
"learning_rate": 0.001,
"loss": 1.3327,
"step": 305700
},
{
"epoch": 98.83645765998708,
"grad_norm": 1.3585878610610962,
"learning_rate": 0.001,
"loss": 1.3379,
"step": 305800
},
{
"epoch": 98.86877828054298,
"grad_norm": 1.5928707122802734,
"learning_rate": 0.001,
"loss": 1.3404,
"step": 305900
},
{
"epoch": 98.9010989010989,
"grad_norm": 1.4327731132507324,
"learning_rate": 0.001,
"loss": 1.3412,
"step": 306000
},
{
"epoch": 98.93341952165481,
"grad_norm": 1.3819307088851929,
"learning_rate": 0.001,
"loss": 1.3494,
"step": 306100
},
{
"epoch": 98.96574014221073,
"grad_norm": 1.9289947748184204,
"learning_rate": 0.001,
"loss": 1.357,
"step": 306200
},
{
"epoch": 98.99806076276664,
"grad_norm": 1.3652219772338867,
"learning_rate": 0.001,
"loss": 1.3472,
"step": 306300
},
{
"epoch": 99.03038138332256,
"grad_norm": 1.3954764604568481,
"learning_rate": 0.001,
"loss": 1.2461,
"step": 306400
},
{
"epoch": 99.06270200387847,
"grad_norm": 1.5966042280197144,
"learning_rate": 0.001,
"loss": 1.2396,
"step": 306500
},
{
"epoch": 99.09502262443439,
"grad_norm": 2.1443591117858887,
"learning_rate": 0.001,
"loss": 1.2605,
"step": 306600
},
{
"epoch": 99.1273432449903,
"grad_norm": 2.153681993484497,
"learning_rate": 0.001,
"loss": 1.2622,
"step": 306700
},
{
"epoch": 99.15966386554622,
"grad_norm": 1.8812059164047241,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 306800
},
{
"epoch": 99.19198448610213,
"grad_norm": 2.035917043685913,
"learning_rate": 0.001,
"loss": 1.2719,
"step": 306900
},
{
"epoch": 99.22430510665805,
"grad_norm": 1.74689519405365,
"learning_rate": 0.001,
"loss": 1.2686,
"step": 307000
},
{
"epoch": 99.25662572721396,
"grad_norm": 1.7956931591033936,
"learning_rate": 0.001,
"loss": 1.2769,
"step": 307100
},
{
"epoch": 99.28894634776988,
"grad_norm": 1.8091312646865845,
"learning_rate": 0.001,
"loss": 1.2945,
"step": 307200
},
{
"epoch": 99.32126696832579,
"grad_norm": 1.6559561491012573,
"learning_rate": 0.001,
"loss": 1.2896,
"step": 307300
},
{
"epoch": 99.35358758888171,
"grad_norm": 1.7476838827133179,
"learning_rate": 0.001,
"loss": 1.2869,
"step": 307400
},
{
"epoch": 99.38590820943762,
"grad_norm": 1.8559825420379639,
"learning_rate": 0.001,
"loss": 1.2964,
"step": 307500
},
{
"epoch": 99.41822882999354,
"grad_norm": 1.685070514678955,
"learning_rate": 0.001,
"loss": 1.2959,
"step": 307600
},
{
"epoch": 99.45054945054945,
"grad_norm": 1.797820806503296,
"learning_rate": 0.001,
"loss": 1.2983,
"step": 307700
},
{
"epoch": 99.48287007110537,
"grad_norm": 2.027237892150879,
"learning_rate": 0.001,
"loss": 1.2979,
"step": 307800
},
{
"epoch": 99.51519069166127,
"grad_norm": 1.6880706548690796,
"learning_rate": 0.001,
"loss": 1.3203,
"step": 307900
},
{
"epoch": 99.5475113122172,
"grad_norm": 1.6182732582092285,
"learning_rate": 0.001,
"loss": 1.3138,
"step": 308000
},
{
"epoch": 99.5798319327731,
"grad_norm": 1.6047966480255127,
"learning_rate": 0.001,
"loss": 1.3283,
"step": 308100
},
{
"epoch": 99.61215255332903,
"grad_norm": 1.5363796949386597,
"learning_rate": 0.001,
"loss": 1.3223,
"step": 308200
},
{
"epoch": 99.64447317388493,
"grad_norm": 1.6853973865509033,
"learning_rate": 0.001,
"loss": 1.3131,
"step": 308300
},
{
"epoch": 99.67679379444085,
"grad_norm": 1.4846042394638062,
"learning_rate": 0.001,
"loss": 1.3217,
"step": 308400
},
{
"epoch": 99.70911441499676,
"grad_norm": 1.318103313446045,
"learning_rate": 0.001,
"loss": 1.3254,
"step": 308500
},
{
"epoch": 99.74143503555268,
"grad_norm": 1.6862074136734009,
"learning_rate": 0.001,
"loss": 1.3293,
"step": 308600
},
{
"epoch": 99.77375565610859,
"grad_norm": 1.6789426803588867,
"learning_rate": 0.001,
"loss": 1.3265,
"step": 308700
},
{
"epoch": 99.80607627666451,
"grad_norm": 1.5269800424575806,
"learning_rate": 0.001,
"loss": 1.3194,
"step": 308800
},
{
"epoch": 99.83839689722042,
"grad_norm": 1.6079939603805542,
"learning_rate": 0.001,
"loss": 1.3328,
"step": 308900
},
{
"epoch": 99.87071751777634,
"grad_norm": 2.0882716178894043,
"learning_rate": 0.001,
"loss": 1.3442,
"step": 309000
},
{
"epoch": 99.90303813833225,
"grad_norm": 1.3886873722076416,
"learning_rate": 0.001,
"loss": 1.3455,
"step": 309100
},
{
"epoch": 99.93535875888817,
"grad_norm": 1.6168712377548218,
"learning_rate": 0.001,
"loss": 1.3428,
"step": 309200
},
{
"epoch": 99.96767937944408,
"grad_norm": 1.3099786043167114,
"learning_rate": 0.001,
"loss": 1.3482,
"step": 309300
},
{
"epoch": 100.0,
"grad_norm": 2.370666742324829,
"learning_rate": 0.001,
"loss": 1.3159,
"step": 309400
},
{
"epoch": 100.0,
"step": 309400,
"total_flos": 1.7026842083328e+18,
"train_loss": 0.2512842774098353,
"train_runtime": 5679.3413,
"train_samples_per_second": 1743.16,
"train_steps_per_second": 54.478
}
],
"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": true
},
"attributes": {}
}
},
"total_flos": 1.7026842083328e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}