SD14_100K_A_Corr / trainer_state.json
AusmitM's picture
Upload 31 files
d3c2030 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": 13.095431327819824,
"learning_rate": 9.900000000000002e-06,
"loss": 4.2463,
"step": 100
},
{
"epoch": 0.06464124111182935,
"grad_norm": 27.387174606323242,
"learning_rate": 1.9900000000000003e-05,
"loss": 4.0764,
"step": 200
},
{
"epoch": 0.09696186166774402,
"grad_norm": 8.1353759765625,
"learning_rate": 2.9900000000000002e-05,
"loss": 4.0007,
"step": 300
},
{
"epoch": 0.1292824822236587,
"grad_norm": 11.949505805969238,
"learning_rate": 3.99e-05,
"loss": 3.8759,
"step": 400
},
{
"epoch": 0.16160310277957338,
"grad_norm": 10.47620677947998,
"learning_rate": 4.99e-05,
"loss": 3.6782,
"step": 500
},
{
"epoch": 0.19392372333548805,
"grad_norm": 18.38490104675293,
"learning_rate": 5.9900000000000006e-05,
"loss": 3.491,
"step": 600
},
{
"epoch": 0.22624434389140272,
"grad_norm": 53.37411117553711,
"learning_rate": 6.99e-05,
"loss": 3.2757,
"step": 700
},
{
"epoch": 0.2585649644473174,
"grad_norm": 17.150442123413086,
"learning_rate": 7.99e-05,
"loss": 3.1093,
"step": 800
},
{
"epoch": 0.2908855850032321,
"grad_norm": 5.678936004638672,
"learning_rate": 8.989999999999999e-05,
"loss": 3.022,
"step": 900
},
{
"epoch": 0.32320620555914675,
"grad_norm": 6.282465934753418,
"learning_rate": 9.99e-05,
"loss": 2.957,
"step": 1000
},
{
"epoch": 0.3555268261150614,
"grad_norm": 5.60117244720459,
"learning_rate": 0.0001099,
"loss": 2.8868,
"step": 1100
},
{
"epoch": 0.3878474466709761,
"grad_norm": 4.572321891784668,
"learning_rate": 0.00011990000000000001,
"loss": 2.8396,
"step": 1200
},
{
"epoch": 0.42016806722689076,
"grad_norm": 3.3965022563934326,
"learning_rate": 0.00012989999999999999,
"loss": 2.7941,
"step": 1300
},
{
"epoch": 0.45248868778280543,
"grad_norm": 16.95421028137207,
"learning_rate": 0.0001399,
"loss": 2.7427,
"step": 1400
},
{
"epoch": 0.4848093083387201,
"grad_norm": 6.816357612609863,
"learning_rate": 0.0001499,
"loss": 2.7324,
"step": 1500
},
{
"epoch": 0.5171299288946348,
"grad_norm": 2.1214568614959717,
"learning_rate": 0.00015989999999999998,
"loss": 2.7426,
"step": 1600
},
{
"epoch": 0.5494505494505495,
"grad_norm": 3.358969211578369,
"learning_rate": 0.0001699,
"loss": 2.715,
"step": 1700
},
{
"epoch": 0.5817711700064642,
"grad_norm": 2.701083183288574,
"learning_rate": 0.0001799,
"loss": 2.6622,
"step": 1800
},
{
"epoch": 0.6140917905623788,
"grad_norm": 1.8063291311264038,
"learning_rate": 0.0001899,
"loss": 2.6827,
"step": 1900
},
{
"epoch": 0.6464124111182935,
"grad_norm": 1.8204689025878906,
"learning_rate": 0.0001999,
"loss": 2.6492,
"step": 2000
},
{
"epoch": 0.6787330316742082,
"grad_norm": 1.6437947750091553,
"learning_rate": 0.0002099,
"loss": 2.6523,
"step": 2100
},
{
"epoch": 0.7110536522301228,
"grad_norm": 2.960686445236206,
"learning_rate": 0.0002199,
"loss": 2.6393,
"step": 2200
},
{
"epoch": 0.7433742727860375,
"grad_norm": 1.3150275945663452,
"learning_rate": 0.0002299,
"loss": 2.5826,
"step": 2300
},
{
"epoch": 0.7756948933419522,
"grad_norm": 2.5215647220611572,
"learning_rate": 0.0002399,
"loss": 2.612,
"step": 2400
},
{
"epoch": 0.8080155138978669,
"grad_norm": 1.298682451248169,
"learning_rate": 0.0002499,
"loss": 2.5858,
"step": 2500
},
{
"epoch": 0.8403361344537815,
"grad_norm": 1.9009894132614136,
"learning_rate": 0.00025990000000000003,
"loss": 2.6145,
"step": 2600
},
{
"epoch": 0.8726567550096962,
"grad_norm": 1.924782395362854,
"learning_rate": 0.0002699,
"loss": 2.5587,
"step": 2700
},
{
"epoch": 0.9049773755656109,
"grad_norm": 1.8015832901000977,
"learning_rate": 0.0002799,
"loss": 2.554,
"step": 2800
},
{
"epoch": 0.9372979961215255,
"grad_norm": 1.7930667400360107,
"learning_rate": 0.0002899,
"loss": 2.5686,
"step": 2900
},
{
"epoch": 0.9696186166774402,
"grad_norm": 1.5759750604629517,
"learning_rate": 0.0002999,
"loss": 2.54,
"step": 3000
},
{
"epoch": 1.0019392372333549,
"grad_norm": 1.3244950771331787,
"learning_rate": 0.0003099,
"loss": 2.5637,
"step": 3100
},
{
"epoch": 1.0342598577892697,
"grad_norm": 1.9222184419631958,
"learning_rate": 0.0003199,
"loss": 2.5038,
"step": 3200
},
{
"epoch": 1.0665804783451842,
"grad_norm": 1.4104559421539307,
"learning_rate": 0.00032990000000000005,
"loss": 2.5322,
"step": 3300
},
{
"epoch": 1.098901098901099,
"grad_norm": 2.111445188522339,
"learning_rate": 0.00033989999999999997,
"loss": 2.4966,
"step": 3400
},
{
"epoch": 1.1312217194570136,
"grad_norm": 1.793367624282837,
"learning_rate": 0.0003499,
"loss": 2.5059,
"step": 3500
},
{
"epoch": 1.1635423400129283,
"grad_norm": 1.2840224504470825,
"learning_rate": 0.0003599,
"loss": 2.4945,
"step": 3600
},
{
"epoch": 1.195862960568843,
"grad_norm": 1.9581950902938843,
"learning_rate": 0.0003699,
"loss": 2.4998,
"step": 3700
},
{
"epoch": 1.2281835811247577,
"grad_norm": 1.221644639968872,
"learning_rate": 0.0003799,
"loss": 2.518,
"step": 3800
},
{
"epoch": 1.2605042016806722,
"grad_norm": 1.7324455976486206,
"learning_rate": 0.00038990000000000004,
"loss": 2.5482,
"step": 3900
},
{
"epoch": 1.292824822236587,
"grad_norm": 3.273299217224121,
"learning_rate": 0.00039989999999999996,
"loss": 2.509,
"step": 4000
},
{
"epoch": 1.3251454427925016,
"grad_norm": 1.7108508348464966,
"learning_rate": 0.0004099,
"loss": 2.514,
"step": 4100
},
{
"epoch": 1.3574660633484164,
"grad_norm": 1.50692617893219,
"learning_rate": 0.0004199,
"loss": 2.5092,
"step": 4200
},
{
"epoch": 1.389786683904331,
"grad_norm": 1.5326952934265137,
"learning_rate": 0.0004299,
"loss": 2.5145,
"step": 4300
},
{
"epoch": 1.4221073044602457,
"grad_norm": 1.7152796983718872,
"learning_rate": 0.0004399,
"loss": 2.5145,
"step": 4400
},
{
"epoch": 1.4544279250161603,
"grad_norm": 1.7978283166885376,
"learning_rate": 0.00044990000000000004,
"loss": 2.515,
"step": 4500
},
{
"epoch": 1.486748545572075,
"grad_norm": 0.933148205280304,
"learning_rate": 0.0004599,
"loss": 2.5079,
"step": 4600
},
{
"epoch": 1.5190691661279896,
"grad_norm": 1.4725111722946167,
"learning_rate": 0.0004699,
"loss": 2.501,
"step": 4700
},
{
"epoch": 1.5513897866839044,
"grad_norm": 1.3290444612503052,
"learning_rate": 0.0004799,
"loss": 2.4955,
"step": 4800
},
{
"epoch": 1.5837104072398192,
"grad_norm": 1.7420613765716553,
"learning_rate": 0.0004899,
"loss": 2.4826,
"step": 4900
},
{
"epoch": 1.6160310277957337,
"grad_norm": 2.0405023097991943,
"learning_rate": 0.0004999000000000001,
"loss": 2.5143,
"step": 5000
},
{
"epoch": 1.6483516483516483,
"grad_norm": 3.256430149078369,
"learning_rate": 0.0005099,
"loss": 2.505,
"step": 5100
},
{
"epoch": 1.680672268907563,
"grad_norm": 1.9231915473937988,
"learning_rate": 0.0005199,
"loss": 2.4882,
"step": 5200
},
{
"epoch": 1.7129928894634778,
"grad_norm": 1.8035612106323242,
"learning_rate": 0.0005299,
"loss": 2.47,
"step": 5300
},
{
"epoch": 1.7453135100193924,
"grad_norm": 1.3306243419647217,
"learning_rate": 0.0005399000000000001,
"loss": 2.5006,
"step": 5400
},
{
"epoch": 1.777634130575307,
"grad_norm": 1.123727560043335,
"learning_rate": 0.0005499000000000001,
"loss": 2.4856,
"step": 5500
},
{
"epoch": 1.8099547511312217,
"grad_norm": 1.8601796627044678,
"learning_rate": 0.0005599,
"loss": 2.4775,
"step": 5600
},
{
"epoch": 1.8422753716871365,
"grad_norm": 1.7166662216186523,
"learning_rate": 0.0005698999999999999,
"loss": 2.4893,
"step": 5700
},
{
"epoch": 1.874595992243051,
"grad_norm": 1.2359657287597656,
"learning_rate": 0.0005799,
"loss": 2.505,
"step": 5800
},
{
"epoch": 1.9069166127989656,
"grad_norm": 1.1526943445205688,
"learning_rate": 0.0005899,
"loss": 2.4919,
"step": 5900
},
{
"epoch": 1.9392372333548804,
"grad_norm": 1.1822352409362793,
"learning_rate": 0.0005999,
"loss": 2.5048,
"step": 6000
},
{
"epoch": 1.9715578539107952,
"grad_norm": 1.2245287895202637,
"learning_rate": 0.0006099,
"loss": 2.482,
"step": 6100
},
{
"epoch": 2.0038784744667097,
"grad_norm": 1.1707814931869507,
"learning_rate": 0.0006199,
"loss": 2.5006,
"step": 6200
},
{
"epoch": 2.0361990950226243,
"grad_norm": 1.1745480298995972,
"learning_rate": 0.0006299000000000001,
"loss": 2.4224,
"step": 6300
},
{
"epoch": 2.0685197155785393,
"grad_norm": 1.0892423391342163,
"learning_rate": 0.0006399,
"loss": 2.4111,
"step": 6400
},
{
"epoch": 2.100840336134454,
"grad_norm": 1.217278242111206,
"learning_rate": 0.0006499,
"loss": 2.4131,
"step": 6500
},
{
"epoch": 2.1331609566903684,
"grad_norm": 1.2329050302505493,
"learning_rate": 0.0006599,
"loss": 2.4312,
"step": 6600
},
{
"epoch": 2.165481577246283,
"grad_norm": 1.2498464584350586,
"learning_rate": 0.0006699000000000001,
"loss": 2.4321,
"step": 6700
},
{
"epoch": 2.197802197802198,
"grad_norm": 1.7037843465805054,
"learning_rate": 0.0006799,
"loss": 2.3998,
"step": 6800
},
{
"epoch": 2.2301228183581125,
"grad_norm": 1.5184770822525024,
"learning_rate": 0.0006899,
"loss": 2.4153,
"step": 6900
},
{
"epoch": 2.262443438914027,
"grad_norm": 1.695165991783142,
"learning_rate": 0.0006998999999999999,
"loss": 2.4218,
"step": 7000
},
{
"epoch": 2.2947640594699417,
"grad_norm": 1.3623040914535522,
"learning_rate": 0.0007099,
"loss": 2.4172,
"step": 7100
},
{
"epoch": 2.3270846800258567,
"grad_norm": 1.1840554475784302,
"learning_rate": 0.0007199,
"loss": 2.4598,
"step": 7200
},
{
"epoch": 2.3594053005817712,
"grad_norm": 1.2055315971374512,
"learning_rate": 0.0007299,
"loss": 2.4048,
"step": 7300
},
{
"epoch": 2.391725921137686,
"grad_norm": 1.4990323781967163,
"learning_rate": 0.0007399,
"loss": 2.4364,
"step": 7400
},
{
"epoch": 2.4240465416936003,
"grad_norm": 1.1268545389175415,
"learning_rate": 0.0007499000000000001,
"loss": 2.4543,
"step": 7500
},
{
"epoch": 2.4563671622495153,
"grad_norm": 1.3047798871994019,
"learning_rate": 0.0007599,
"loss": 2.4456,
"step": 7600
},
{
"epoch": 2.48868778280543,
"grad_norm": 1.170345425605774,
"learning_rate": 0.0007699,
"loss": 2.4349,
"step": 7700
},
{
"epoch": 2.5210084033613445,
"grad_norm": 0.9993893504142761,
"learning_rate": 0.0007799,
"loss": 2.3841,
"step": 7800
},
{
"epoch": 2.553329023917259,
"grad_norm": 0.9757175445556641,
"learning_rate": 0.0007899000000000001,
"loss": 2.399,
"step": 7900
},
{
"epoch": 2.585649644473174,
"grad_norm": 1.2529906034469604,
"learning_rate": 0.0007999000000000001,
"loss": 2.4359,
"step": 8000
},
{
"epoch": 2.6179702650290886,
"grad_norm": 1.183002233505249,
"learning_rate": 0.0008099,
"loss": 2.4184,
"step": 8100
},
{
"epoch": 2.650290885585003,
"grad_norm": 1.2000353336334229,
"learning_rate": 0.0008198999999999999,
"loss": 2.4102,
"step": 8200
},
{
"epoch": 2.682611506140918,
"grad_norm": 1.0429084300994873,
"learning_rate": 0.0008299,
"loss": 2.4061,
"step": 8300
},
{
"epoch": 2.7149321266968327,
"grad_norm": 1.2502565383911133,
"learning_rate": 0.0008399,
"loss": 2.4636,
"step": 8400
},
{
"epoch": 2.7472527472527473,
"grad_norm": 1.2110387086868286,
"learning_rate": 0.0008499,
"loss": 2.4175,
"step": 8500
},
{
"epoch": 2.779573367808662,
"grad_norm": 1.1316890716552734,
"learning_rate": 0.0008599,
"loss": 2.4055,
"step": 8600
},
{
"epoch": 2.8118939883645764,
"grad_norm": 1.0335315465927124,
"learning_rate": 0.0008699000000000001,
"loss": 2.4026,
"step": 8700
},
{
"epoch": 2.8442146089204914,
"grad_norm": 1.1183189153671265,
"learning_rate": 0.0008799000000000001,
"loss": 2.4466,
"step": 8800
},
{
"epoch": 2.876535229476406,
"grad_norm": 1.2400306463241577,
"learning_rate": 0.0008899,
"loss": 2.4484,
"step": 8900
},
{
"epoch": 2.9088558500323205,
"grad_norm": 1.0672909021377563,
"learning_rate": 0.0008999,
"loss": 2.3913,
"step": 9000
},
{
"epoch": 2.9411764705882355,
"grad_norm": 1.250365138053894,
"learning_rate": 0.0009099,
"loss": 2.4352,
"step": 9100
},
{
"epoch": 2.97349709114415,
"grad_norm": 1.1961923837661743,
"learning_rate": 0.0009199000000000001,
"loss": 2.438,
"step": 9200
},
{
"epoch": 3.0058177117000646,
"grad_norm": 1.6543998718261719,
"learning_rate": 0.0009299,
"loss": 2.4589,
"step": 9300
},
{
"epoch": 3.038138332255979,
"grad_norm": 1.4493199586868286,
"learning_rate": 0.0009399,
"loss": 2.3203,
"step": 9400
},
{
"epoch": 3.070458952811894,
"grad_norm": 1.0506430864334106,
"learning_rate": 0.0009498999999999999,
"loss": 2.3324,
"step": 9500
},
{
"epoch": 3.1027795733678087,
"grad_norm": 1.1909950971603394,
"learning_rate": 0.0009599,
"loss": 2.3315,
"step": 9600
},
{
"epoch": 3.1351001939237233,
"grad_norm": 1.7181323766708374,
"learning_rate": 0.0009699,
"loss": 2.3593,
"step": 9700
},
{
"epoch": 3.167420814479638,
"grad_norm": 1.3096699714660645,
"learning_rate": 0.0009799,
"loss": 2.3457,
"step": 9800
},
{
"epoch": 3.199741435035553,
"grad_norm": 1.4763398170471191,
"learning_rate": 0.0009899,
"loss": 2.3232,
"step": 9900
},
{
"epoch": 3.2320620555914674,
"grad_norm": 1.4502891302108765,
"learning_rate": 0.0009999,
"loss": 2.3432,
"step": 10000
},
{
"epoch": 3.264382676147382,
"grad_norm": 1.4951964616775513,
"learning_rate": 0.001,
"loss": 2.3599,
"step": 10100
},
{
"epoch": 3.2967032967032965,
"grad_norm": 1.3563085794448853,
"learning_rate": 0.001,
"loss": 2.3509,
"step": 10200
},
{
"epoch": 3.3290239172592115,
"grad_norm": 1.1576285362243652,
"learning_rate": 0.001,
"loss": 2.3666,
"step": 10300
},
{
"epoch": 3.361344537815126,
"grad_norm": 1.0529580116271973,
"learning_rate": 0.001,
"loss": 2.3372,
"step": 10400
},
{
"epoch": 3.3936651583710407,
"grad_norm": 1.1564232110977173,
"learning_rate": 0.001,
"loss": 2.3488,
"step": 10500
},
{
"epoch": 3.425985778926955,
"grad_norm": 1.1305166482925415,
"learning_rate": 0.001,
"loss": 2.3936,
"step": 10600
},
{
"epoch": 3.45830639948287,
"grad_norm": 1.464630126953125,
"learning_rate": 0.001,
"loss": 2.3706,
"step": 10700
},
{
"epoch": 3.490627020038785,
"grad_norm": 1.1218030452728271,
"learning_rate": 0.001,
"loss": 2.4045,
"step": 10800
},
{
"epoch": 3.5229476405946993,
"grad_norm": 1.1466821432113647,
"learning_rate": 0.001,
"loss": 2.3568,
"step": 10900
},
{
"epoch": 3.555268261150614,
"grad_norm": 0.9143931269645691,
"learning_rate": 0.001,
"loss": 2.395,
"step": 11000
},
{
"epoch": 3.587588881706529,
"grad_norm": 1.290088176727295,
"learning_rate": 0.001,
"loss": 2.3372,
"step": 11100
},
{
"epoch": 3.6199095022624435,
"grad_norm": 1.271133542060852,
"learning_rate": 0.001,
"loss": 2.3731,
"step": 11200
},
{
"epoch": 3.652230122818358,
"grad_norm": 2.0801732540130615,
"learning_rate": 0.001,
"loss": 2.3895,
"step": 11300
},
{
"epoch": 3.684550743374273,
"grad_norm": 1.0689164400100708,
"learning_rate": 0.001,
"loss": 2.3635,
"step": 11400
},
{
"epoch": 3.7168713639301876,
"grad_norm": 1.1849030256271362,
"learning_rate": 0.001,
"loss": 2.3524,
"step": 11500
},
{
"epoch": 3.749191984486102,
"grad_norm": 1.3510334491729736,
"learning_rate": 0.001,
"loss": 2.3691,
"step": 11600
},
{
"epoch": 3.7815126050420167,
"grad_norm": 0.9980648159980774,
"learning_rate": 0.001,
"loss": 2.3792,
"step": 11700
},
{
"epoch": 3.8138332255979313,
"grad_norm": 1.0921436548233032,
"learning_rate": 0.001,
"loss": 2.3765,
"step": 11800
},
{
"epoch": 3.8461538461538463,
"grad_norm": 1.3819752931594849,
"learning_rate": 0.001,
"loss": 2.359,
"step": 11900
},
{
"epoch": 3.878474466709761,
"grad_norm": 1.0975066423416138,
"learning_rate": 0.001,
"loss": 2.365,
"step": 12000
},
{
"epoch": 3.9107950872656754,
"grad_norm": 1.1619728803634644,
"learning_rate": 0.001,
"loss": 2.3778,
"step": 12100
},
{
"epoch": 3.9431157078215904,
"grad_norm": 1.3813674449920654,
"learning_rate": 0.001,
"loss": 2.3733,
"step": 12200
},
{
"epoch": 3.975436328377505,
"grad_norm": 1.6252094507217407,
"learning_rate": 0.001,
"loss": 2.3926,
"step": 12300
},
{
"epoch": 4.0077569489334195,
"grad_norm": 1.207340121269226,
"learning_rate": 0.001,
"loss": 2.3659,
"step": 12400
},
{
"epoch": 4.040077569489334,
"grad_norm": 1.151745319366455,
"learning_rate": 0.001,
"loss": 2.2764,
"step": 12500
},
{
"epoch": 4.072398190045249,
"grad_norm": 1.115214467048645,
"learning_rate": 0.001,
"loss": 2.2529,
"step": 12600
},
{
"epoch": 4.104718810601163,
"grad_norm": 0.8362165689468384,
"learning_rate": 0.001,
"loss": 2.286,
"step": 12700
},
{
"epoch": 4.137039431157079,
"grad_norm": 1.1398061513900757,
"learning_rate": 0.001,
"loss": 2.2718,
"step": 12800
},
{
"epoch": 4.169360051712993,
"grad_norm": 1.1849474906921387,
"learning_rate": 0.001,
"loss": 2.2881,
"step": 12900
},
{
"epoch": 4.201680672268908,
"grad_norm": 0.7654844522476196,
"learning_rate": 0.001,
"loss": 2.2998,
"step": 13000
},
{
"epoch": 4.234001292824822,
"grad_norm": 0.9867515563964844,
"learning_rate": 0.001,
"loss": 2.2835,
"step": 13100
},
{
"epoch": 4.266321913380737,
"grad_norm": 1.1559300422668457,
"learning_rate": 0.001,
"loss": 2.2413,
"step": 13200
},
{
"epoch": 4.298642533936651,
"grad_norm": 0.9872420430183411,
"learning_rate": 0.001,
"loss": 2.2769,
"step": 13300
},
{
"epoch": 4.330963154492566,
"grad_norm": 1.1229034662246704,
"learning_rate": 0.001,
"loss": 2.2509,
"step": 13400
},
{
"epoch": 4.3632837750484805,
"grad_norm": 1.0585553646087646,
"learning_rate": 0.001,
"loss": 2.3067,
"step": 13500
},
{
"epoch": 4.395604395604396,
"grad_norm": 1.0657356977462769,
"learning_rate": 0.001,
"loss": 2.2857,
"step": 13600
},
{
"epoch": 4.4279250161603105,
"grad_norm": 1.3249043226242065,
"learning_rate": 0.001,
"loss": 2.3003,
"step": 13700
},
{
"epoch": 4.460245636716225,
"grad_norm": 1.142127275466919,
"learning_rate": 0.001,
"loss": 2.3132,
"step": 13800
},
{
"epoch": 4.49256625727214,
"grad_norm": 1.1387176513671875,
"learning_rate": 0.001,
"loss": 2.3045,
"step": 13900
},
{
"epoch": 4.524886877828054,
"grad_norm": 1.2085492610931396,
"learning_rate": 0.001,
"loss": 2.2948,
"step": 14000
},
{
"epoch": 4.557207498383969,
"grad_norm": 0.9948108792304993,
"learning_rate": 0.001,
"loss": 2.2626,
"step": 14100
},
{
"epoch": 4.589528118939883,
"grad_norm": 1.3945204019546509,
"learning_rate": 0.001,
"loss": 2.2821,
"step": 14200
},
{
"epoch": 4.621848739495798,
"grad_norm": 1.0511122941970825,
"learning_rate": 0.001,
"loss": 2.2967,
"step": 14300
},
{
"epoch": 4.654169360051713,
"grad_norm": 0.8755561113357544,
"learning_rate": 0.001,
"loss": 2.2845,
"step": 14400
},
{
"epoch": 4.686489980607628,
"grad_norm": 0.8116604685783386,
"learning_rate": 0.001,
"loss": 2.3187,
"step": 14500
},
{
"epoch": 4.7188106011635425,
"grad_norm": 1.3444281816482544,
"learning_rate": 0.001,
"loss": 2.3013,
"step": 14600
},
{
"epoch": 4.751131221719457,
"grad_norm": 1.0668421983718872,
"learning_rate": 0.001,
"loss": 2.2815,
"step": 14700
},
{
"epoch": 4.783451842275372,
"grad_norm": 1.055733561515808,
"learning_rate": 0.001,
"loss": 2.3033,
"step": 14800
},
{
"epoch": 4.815772462831286,
"grad_norm": 0.9193568825721741,
"learning_rate": 0.001,
"loss": 2.2927,
"step": 14900
},
{
"epoch": 4.848093083387201,
"grad_norm": 1.006926417350769,
"learning_rate": 0.001,
"loss": 2.2657,
"step": 15000
},
{
"epoch": 4.880413703943116,
"grad_norm": 0.8166502714157104,
"learning_rate": 0.001,
"loss": 2.276,
"step": 15100
},
{
"epoch": 4.912734324499031,
"grad_norm": 1.0166735649108887,
"learning_rate": 0.001,
"loss": 2.2926,
"step": 15200
},
{
"epoch": 4.945054945054945,
"grad_norm": 1.075845718383789,
"learning_rate": 0.001,
"loss": 2.2833,
"step": 15300
},
{
"epoch": 4.97737556561086,
"grad_norm": 0.8848145604133606,
"learning_rate": 0.001,
"loss": 2.3029,
"step": 15400
},
{
"epoch": 5.009696186166774,
"grad_norm": 0.8544314503669739,
"learning_rate": 0.001,
"loss": 2.2695,
"step": 15500
},
{
"epoch": 5.042016806722689,
"grad_norm": 1.4420232772827148,
"learning_rate": 0.001,
"loss": 2.1842,
"step": 15600
},
{
"epoch": 5.0743374272786035,
"grad_norm": 1.000048041343689,
"learning_rate": 0.001,
"loss": 2.1832,
"step": 15700
},
{
"epoch": 5.106658047834518,
"grad_norm": 1.0307416915893555,
"learning_rate": 0.001,
"loss": 2.2021,
"step": 15800
},
{
"epoch": 5.1389786683904335,
"grad_norm": 1.2275282144546509,
"learning_rate": 0.001,
"loss": 2.1989,
"step": 15900
},
{
"epoch": 5.171299288946348,
"grad_norm": 1.0649629831314087,
"learning_rate": 0.001,
"loss": 2.2266,
"step": 16000
},
{
"epoch": 5.203619909502263,
"grad_norm": 1.2316499948501587,
"learning_rate": 0.001,
"loss": 2.1996,
"step": 16100
},
{
"epoch": 5.235940530058177,
"grad_norm": 1.3741744756698608,
"learning_rate": 0.001,
"loss": 2.1916,
"step": 16200
},
{
"epoch": 5.268261150614092,
"grad_norm": 1.0122957229614258,
"learning_rate": 0.001,
"loss": 2.2525,
"step": 16300
},
{
"epoch": 5.300581771170006,
"grad_norm": 0.8483036160469055,
"learning_rate": 0.001,
"loss": 2.184,
"step": 16400
},
{
"epoch": 5.332902391725921,
"grad_norm": 0.8776037096977234,
"learning_rate": 0.001,
"loss": 2.2341,
"step": 16500
},
{
"epoch": 5.365223012281835,
"grad_norm": 0.9180140495300293,
"learning_rate": 0.001,
"loss": 2.2265,
"step": 16600
},
{
"epoch": 5.397543632837751,
"grad_norm": 1.063555359840393,
"learning_rate": 0.001,
"loss": 2.1839,
"step": 16700
},
{
"epoch": 5.429864253393665,
"grad_norm": 1.938733458518982,
"learning_rate": 0.001,
"loss": 2.2229,
"step": 16800
},
{
"epoch": 5.46218487394958,
"grad_norm": 1.2461786270141602,
"learning_rate": 0.001,
"loss": 2.2305,
"step": 16900
},
{
"epoch": 5.4945054945054945,
"grad_norm": 0.8778449296951294,
"learning_rate": 0.001,
"loss": 2.2053,
"step": 17000
},
{
"epoch": 5.526826115061409,
"grad_norm": 0.9430673718452454,
"learning_rate": 0.001,
"loss": 2.2403,
"step": 17100
},
{
"epoch": 5.559146735617324,
"grad_norm": 0.7641558051109314,
"learning_rate": 0.001,
"loss": 2.2149,
"step": 17200
},
{
"epoch": 5.591467356173238,
"grad_norm": 1.0826328992843628,
"learning_rate": 0.001,
"loss": 2.2042,
"step": 17300
},
{
"epoch": 5.623787976729153,
"grad_norm": 1.050527811050415,
"learning_rate": 0.001,
"loss": 2.2511,
"step": 17400
},
{
"epoch": 5.656108597285068,
"grad_norm": 0.9223970174789429,
"learning_rate": 0.001,
"loss": 2.2315,
"step": 17500
},
{
"epoch": 5.688429217840983,
"grad_norm": 1.0374590158462524,
"learning_rate": 0.001,
"loss": 2.243,
"step": 17600
},
{
"epoch": 5.720749838396897,
"grad_norm": 1.022735834121704,
"learning_rate": 0.001,
"loss": 2.2103,
"step": 17700
},
{
"epoch": 5.753070458952812,
"grad_norm": 1.0346524715423584,
"learning_rate": 0.001,
"loss": 2.2325,
"step": 17800
},
{
"epoch": 5.785391079508726,
"grad_norm": 0.9487394094467163,
"learning_rate": 0.001,
"loss": 2.2501,
"step": 17900
},
{
"epoch": 5.817711700064641,
"grad_norm": 0.923183798789978,
"learning_rate": 0.001,
"loss": 2.2162,
"step": 18000
},
{
"epoch": 5.850032320620556,
"grad_norm": 1.3014050722122192,
"learning_rate": 0.001,
"loss": 2.2293,
"step": 18100
},
{
"epoch": 5.882352941176471,
"grad_norm": 1.249762773513794,
"learning_rate": 0.001,
"loss": 2.2131,
"step": 18200
},
{
"epoch": 5.914673561732386,
"grad_norm": 0.8731359243392944,
"learning_rate": 0.001,
"loss": 2.253,
"step": 18300
},
{
"epoch": 5.9469941822883,
"grad_norm": 0.896523118019104,
"learning_rate": 0.001,
"loss": 2.265,
"step": 18400
},
{
"epoch": 5.979314802844215,
"grad_norm": 0.9707561731338501,
"learning_rate": 0.001,
"loss": 2.2056,
"step": 18500
},
{
"epoch": 6.011635423400129,
"grad_norm": 1.1334419250488281,
"learning_rate": 0.001,
"loss": 2.1646,
"step": 18600
},
{
"epoch": 6.043956043956044,
"grad_norm": 1.1426136493682861,
"learning_rate": 0.001,
"loss": 2.1307,
"step": 18700
},
{
"epoch": 6.076276664511958,
"grad_norm": 1.0500696897506714,
"learning_rate": 0.001,
"loss": 2.1347,
"step": 18800
},
{
"epoch": 6.108597285067873,
"grad_norm": 0.8252379894256592,
"learning_rate": 0.001,
"loss": 2.1297,
"step": 18900
},
{
"epoch": 6.140917905623788,
"grad_norm": 1.1823996305465698,
"learning_rate": 0.001,
"loss": 2.1171,
"step": 19000
},
{
"epoch": 6.173238526179703,
"grad_norm": 0.8585313558578491,
"learning_rate": 0.001,
"loss": 2.1368,
"step": 19100
},
{
"epoch": 6.2055591467356175,
"grad_norm": 0.9864341616630554,
"learning_rate": 0.001,
"loss": 2.145,
"step": 19200
},
{
"epoch": 6.237879767291532,
"grad_norm": 1.1134485006332397,
"learning_rate": 0.001,
"loss": 2.1667,
"step": 19300
},
{
"epoch": 6.270200387847447,
"grad_norm": 1.1856287717819214,
"learning_rate": 0.001,
"loss": 2.1582,
"step": 19400
},
{
"epoch": 6.302521008403361,
"grad_norm": 1.1392621994018555,
"learning_rate": 0.001,
"loss": 2.165,
"step": 19500
},
{
"epoch": 6.334841628959276,
"grad_norm": 0.9821904897689819,
"learning_rate": 0.001,
"loss": 2.1773,
"step": 19600
},
{
"epoch": 6.36716224951519,
"grad_norm": 1.0998677015304565,
"learning_rate": 0.001,
"loss": 2.1583,
"step": 19700
},
{
"epoch": 6.399482870071106,
"grad_norm": 1.2467516660690308,
"learning_rate": 0.001,
"loss": 2.1471,
"step": 19800
},
{
"epoch": 6.43180349062702,
"grad_norm": 1.0288869142532349,
"learning_rate": 0.001,
"loss": 2.1588,
"step": 19900
},
{
"epoch": 6.464124111182935,
"grad_norm": 0.7433337569236755,
"learning_rate": 0.001,
"loss": 2.1435,
"step": 20000
},
{
"epoch": 6.496444731738849,
"grad_norm": 0.9702023863792419,
"learning_rate": 0.001,
"loss": 2.1831,
"step": 20100
},
{
"epoch": 6.528765352294764,
"grad_norm": 1.1337813138961792,
"learning_rate": 0.001,
"loss": 2.1695,
"step": 20200
},
{
"epoch": 6.5610859728506785,
"grad_norm": 1.0064531564712524,
"learning_rate": 0.001,
"loss": 2.1348,
"step": 20300
},
{
"epoch": 6.593406593406593,
"grad_norm": 1.3694676160812378,
"learning_rate": 0.001,
"loss": 2.1895,
"step": 20400
},
{
"epoch": 6.625727213962508,
"grad_norm": 0.8968884348869324,
"learning_rate": 0.001,
"loss": 2.1538,
"step": 20500
},
{
"epoch": 6.658047834518423,
"grad_norm": 1.3198908567428589,
"learning_rate": 0.001,
"loss": 2.1787,
"step": 20600
},
{
"epoch": 6.690368455074338,
"grad_norm": 1.086137294769287,
"learning_rate": 0.001,
"loss": 2.1851,
"step": 20700
},
{
"epoch": 6.722689075630252,
"grad_norm": 0.8597367405891418,
"learning_rate": 0.001,
"loss": 2.1799,
"step": 20800
},
{
"epoch": 6.755009696186167,
"grad_norm": 1.2422113418579102,
"learning_rate": 0.001,
"loss": 2.1767,
"step": 20900
},
{
"epoch": 6.787330316742081,
"grad_norm": 0.9910422563552856,
"learning_rate": 0.001,
"loss": 2.1606,
"step": 21000
},
{
"epoch": 6.819650937297996,
"grad_norm": 1.27292799949646,
"learning_rate": 0.001,
"loss": 2.1678,
"step": 21100
},
{
"epoch": 6.85197155785391,
"grad_norm": 0.9695398807525635,
"learning_rate": 0.001,
"loss": 2.182,
"step": 21200
},
{
"epoch": 6.884292178409826,
"grad_norm": 1.1529979705810547,
"learning_rate": 0.001,
"loss": 2.1706,
"step": 21300
},
{
"epoch": 6.91661279896574,
"grad_norm": 1.3126720190048218,
"learning_rate": 0.001,
"loss": 2.1573,
"step": 21400
},
{
"epoch": 6.948933419521655,
"grad_norm": 0.9190894961357117,
"learning_rate": 0.001,
"loss": 2.1563,
"step": 21500
},
{
"epoch": 6.98125404007757,
"grad_norm": 1.3218880891799927,
"learning_rate": 0.001,
"loss": 2.1972,
"step": 21600
},
{
"epoch": 7.013574660633484,
"grad_norm": 0.8746786117553711,
"learning_rate": 0.001,
"loss": 2.1198,
"step": 21700
},
{
"epoch": 7.045895281189399,
"grad_norm": 0.9623684287071228,
"learning_rate": 0.001,
"loss": 2.0991,
"step": 21800
},
{
"epoch": 7.078215901745313,
"grad_norm": 0.9384899735450745,
"learning_rate": 0.001,
"loss": 2.0808,
"step": 21900
},
{
"epoch": 7.110536522301228,
"grad_norm": 1.1331701278686523,
"learning_rate": 0.001,
"loss": 2.0982,
"step": 22000
},
{
"epoch": 7.142857142857143,
"grad_norm": 1.1500688791275024,
"learning_rate": 0.001,
"loss": 2.0825,
"step": 22100
},
{
"epoch": 7.175177763413058,
"grad_norm": 1.047314167022705,
"learning_rate": 0.001,
"loss": 2.0834,
"step": 22200
},
{
"epoch": 7.207498383968972,
"grad_norm": 1.0547418594360352,
"learning_rate": 0.001,
"loss": 2.071,
"step": 22300
},
{
"epoch": 7.239819004524887,
"grad_norm": 1.3998901844024658,
"learning_rate": 0.001,
"loss": 2.1047,
"step": 22400
},
{
"epoch": 7.2721396250808015,
"grad_norm": 0.9676983952522278,
"learning_rate": 0.001,
"loss": 2.0935,
"step": 22500
},
{
"epoch": 7.304460245636716,
"grad_norm": 0.9831491708755493,
"learning_rate": 0.001,
"loss": 2.0842,
"step": 22600
},
{
"epoch": 7.336780866192631,
"grad_norm": 0.971725344657898,
"learning_rate": 0.001,
"loss": 2.1248,
"step": 22700
},
{
"epoch": 7.369101486748546,
"grad_norm": 0.9703531861305237,
"learning_rate": 0.001,
"loss": 2.0816,
"step": 22800
},
{
"epoch": 7.401422107304461,
"grad_norm": 0.8289443850517273,
"learning_rate": 0.001,
"loss": 2.0674,
"step": 22900
},
{
"epoch": 7.433742727860375,
"grad_norm": 1.0580024719238281,
"learning_rate": 0.001,
"loss": 2.1269,
"step": 23000
},
{
"epoch": 7.46606334841629,
"grad_norm": 1.3143529891967773,
"learning_rate": 0.001,
"loss": 2.1089,
"step": 23100
},
{
"epoch": 7.498383968972204,
"grad_norm": 1.0106147527694702,
"learning_rate": 0.001,
"loss": 2.1267,
"step": 23200
},
{
"epoch": 7.530704589528119,
"grad_norm": 1.0150158405303955,
"learning_rate": 0.001,
"loss": 2.1166,
"step": 23300
},
{
"epoch": 7.563025210084033,
"grad_norm": 1.2520421743392944,
"learning_rate": 0.001,
"loss": 2.1166,
"step": 23400
},
{
"epoch": 7.595345830639948,
"grad_norm": 1.198043704032898,
"learning_rate": 0.001,
"loss": 2.1213,
"step": 23500
},
{
"epoch": 7.6276664511958625,
"grad_norm": 1.2207905054092407,
"learning_rate": 0.001,
"loss": 2.1108,
"step": 23600
},
{
"epoch": 7.659987071751778,
"grad_norm": 0.9647412300109863,
"learning_rate": 0.001,
"loss": 2.1268,
"step": 23700
},
{
"epoch": 7.6923076923076925,
"grad_norm": 0.9915140271186829,
"learning_rate": 0.001,
"loss": 2.1438,
"step": 23800
},
{
"epoch": 7.724628312863607,
"grad_norm": 1.1954071521759033,
"learning_rate": 0.001,
"loss": 2.0931,
"step": 23900
},
{
"epoch": 7.756948933419522,
"grad_norm": 0.9521498084068298,
"learning_rate": 0.001,
"loss": 2.1314,
"step": 24000
},
{
"epoch": 7.789269553975436,
"grad_norm": 0.9276489615440369,
"learning_rate": 0.001,
"loss": 2.107,
"step": 24100
},
{
"epoch": 7.821590174531351,
"grad_norm": 1.0093352794647217,
"learning_rate": 0.001,
"loss": 2.1245,
"step": 24200
},
{
"epoch": 7.853910795087265,
"grad_norm": 0.842779278755188,
"learning_rate": 0.001,
"loss": 2.1334,
"step": 24300
},
{
"epoch": 7.886231415643181,
"grad_norm": 1.007067322731018,
"learning_rate": 0.001,
"loss": 2.1178,
"step": 24400
},
{
"epoch": 7.918552036199095,
"grad_norm": 1.259667158126831,
"learning_rate": 0.001,
"loss": 2.1335,
"step": 24500
},
{
"epoch": 7.95087265675501,
"grad_norm": 1.6199828386306763,
"learning_rate": 0.001,
"loss": 2.1201,
"step": 24600
},
{
"epoch": 7.983193277310924,
"grad_norm": 1.0307115316390991,
"learning_rate": 0.001,
"loss": 2.1404,
"step": 24700
},
{
"epoch": 8.015513897866839,
"grad_norm": 3.100094795227051,
"learning_rate": 0.001,
"loss": 2.0168,
"step": 24800
},
{
"epoch": 8.047834518422754,
"grad_norm": 1.230669379234314,
"learning_rate": 0.001,
"loss": 2.0227,
"step": 24900
},
{
"epoch": 8.080155138978668,
"grad_norm": 1.1248514652252197,
"learning_rate": 0.001,
"loss": 2.0109,
"step": 25000
},
{
"epoch": 8.112475759534583,
"grad_norm": 1.1911290884017944,
"learning_rate": 0.001,
"loss": 2.0383,
"step": 25100
},
{
"epoch": 8.144796380090497,
"grad_norm": 0.9947310090065002,
"learning_rate": 0.001,
"loss": 2.0398,
"step": 25200
},
{
"epoch": 8.177117000646412,
"grad_norm": 1.1442896127700806,
"learning_rate": 0.001,
"loss": 2.0483,
"step": 25300
},
{
"epoch": 8.209437621202326,
"grad_norm": 0.847846508026123,
"learning_rate": 0.001,
"loss": 2.0151,
"step": 25400
},
{
"epoch": 8.241758241758241,
"grad_norm": 1.1538875102996826,
"learning_rate": 0.001,
"loss": 2.0274,
"step": 25500
},
{
"epoch": 8.274078862314157,
"grad_norm": 0.9806484580039978,
"learning_rate": 0.001,
"loss": 2.0477,
"step": 25600
},
{
"epoch": 8.306399482870072,
"grad_norm": 0.9989823698997498,
"learning_rate": 0.001,
"loss": 2.0475,
"step": 25700
},
{
"epoch": 8.338720103425986,
"grad_norm": 1.151330828666687,
"learning_rate": 0.001,
"loss": 2.0621,
"step": 25800
},
{
"epoch": 8.371040723981901,
"grad_norm": 1.0853585004806519,
"learning_rate": 0.001,
"loss": 2.0313,
"step": 25900
},
{
"epoch": 8.403361344537815,
"grad_norm": 1.1173166036605835,
"learning_rate": 0.001,
"loss": 2.0538,
"step": 26000
},
{
"epoch": 8.43568196509373,
"grad_norm": 1.0314850807189941,
"learning_rate": 0.001,
"loss": 2.0368,
"step": 26100
},
{
"epoch": 8.468002585649645,
"grad_norm": 1.2047348022460938,
"learning_rate": 0.001,
"loss": 2.0552,
"step": 26200
},
{
"epoch": 8.50032320620556,
"grad_norm": 0.9672854542732239,
"learning_rate": 0.001,
"loss": 2.1015,
"step": 26300
},
{
"epoch": 8.532643826761474,
"grad_norm": 0.9105637073516846,
"learning_rate": 0.001,
"loss": 2.0828,
"step": 26400
},
{
"epoch": 8.564964447317388,
"grad_norm": 1.086616039276123,
"learning_rate": 0.001,
"loss": 2.0289,
"step": 26500
},
{
"epoch": 8.597285067873303,
"grad_norm": 0.9836768507957458,
"learning_rate": 0.001,
"loss": 2.0851,
"step": 26600
},
{
"epoch": 8.629605688429217,
"grad_norm": 1.150235891342163,
"learning_rate": 0.001,
"loss": 2.0542,
"step": 26700
},
{
"epoch": 8.661926308985132,
"grad_norm": 1.1072510480880737,
"learning_rate": 0.001,
"loss": 2.0797,
"step": 26800
},
{
"epoch": 8.694246929541046,
"grad_norm": 1.2430179119110107,
"learning_rate": 0.001,
"loss": 2.0976,
"step": 26900
},
{
"epoch": 8.726567550096961,
"grad_norm": 0.9480435848236084,
"learning_rate": 0.001,
"loss": 2.0835,
"step": 27000
},
{
"epoch": 8.758888170652877,
"grad_norm": 1.1376099586486816,
"learning_rate": 0.001,
"loss": 2.0695,
"step": 27100
},
{
"epoch": 8.791208791208792,
"grad_norm": 1.1211824417114258,
"learning_rate": 0.001,
"loss": 2.0724,
"step": 27200
},
{
"epoch": 8.823529411764707,
"grad_norm": 0.8404349684715271,
"learning_rate": 0.001,
"loss": 2.0693,
"step": 27300
},
{
"epoch": 8.855850032320621,
"grad_norm": 1.0358692407608032,
"learning_rate": 0.001,
"loss": 2.0874,
"step": 27400
},
{
"epoch": 8.888170652876536,
"grad_norm": 0.7218955755233765,
"learning_rate": 0.001,
"loss": 2.0887,
"step": 27500
},
{
"epoch": 8.92049127343245,
"grad_norm": 1.412729024887085,
"learning_rate": 0.001,
"loss": 2.0919,
"step": 27600
},
{
"epoch": 8.952811893988365,
"grad_norm": 0.8630658984184265,
"learning_rate": 0.001,
"loss": 2.1076,
"step": 27700
},
{
"epoch": 8.98513251454428,
"grad_norm": 0.9641337990760803,
"learning_rate": 0.001,
"loss": 2.0819,
"step": 27800
},
{
"epoch": 9.017453135100194,
"grad_norm": 1.5362216234207153,
"learning_rate": 0.001,
"loss": 1.9746,
"step": 27900
},
{
"epoch": 9.049773755656108,
"grad_norm": 1.55039381980896,
"learning_rate": 0.001,
"loss": 1.9523,
"step": 28000
},
{
"epoch": 9.082094376212023,
"grad_norm": 1.211641550064087,
"learning_rate": 0.001,
"loss": 1.9809,
"step": 28100
},
{
"epoch": 9.114414996767938,
"grad_norm": 1.4977613687515259,
"learning_rate": 0.001,
"loss": 1.9573,
"step": 28200
},
{
"epoch": 9.146735617323852,
"grad_norm": 1.6715325117111206,
"learning_rate": 0.001,
"loss": 2.0006,
"step": 28300
},
{
"epoch": 9.179056237879767,
"grad_norm": 1.270294189453125,
"learning_rate": 0.001,
"loss": 1.9826,
"step": 28400
},
{
"epoch": 9.211376858435681,
"grad_norm": 1.3731482028961182,
"learning_rate": 0.001,
"loss": 1.9976,
"step": 28500
},
{
"epoch": 9.243697478991596,
"grad_norm": 1.2195913791656494,
"learning_rate": 0.001,
"loss": 1.9656,
"step": 28600
},
{
"epoch": 9.276018099547512,
"grad_norm": 1.3323032855987549,
"learning_rate": 0.001,
"loss": 2.0285,
"step": 28700
},
{
"epoch": 9.308338720103427,
"grad_norm": 1.1946659088134766,
"learning_rate": 0.001,
"loss": 1.9931,
"step": 28800
},
{
"epoch": 9.340659340659341,
"grad_norm": 1.231381893157959,
"learning_rate": 0.001,
"loss": 1.9987,
"step": 28900
},
{
"epoch": 9.372979961215256,
"grad_norm": 1.3914852142333984,
"learning_rate": 0.001,
"loss": 2.0162,
"step": 29000
},
{
"epoch": 9.40530058177117,
"grad_norm": 1.6865482330322266,
"learning_rate": 0.001,
"loss": 1.9987,
"step": 29100
},
{
"epoch": 9.437621202327085,
"grad_norm": 1.3092128038406372,
"learning_rate": 0.001,
"loss": 2.0095,
"step": 29200
},
{
"epoch": 9.469941822883,
"grad_norm": 1.674773097038269,
"learning_rate": 0.001,
"loss": 2.0127,
"step": 29300
},
{
"epoch": 9.502262443438914,
"grad_norm": 1.4838887453079224,
"learning_rate": 0.001,
"loss": 2.0138,
"step": 29400
},
{
"epoch": 9.534583063994829,
"grad_norm": 1.4192705154418945,
"learning_rate": 0.001,
"loss": 2.0291,
"step": 29500
},
{
"epoch": 9.566903684550743,
"grad_norm": 1.3291783332824707,
"learning_rate": 0.001,
"loss": 2.0384,
"step": 29600
},
{
"epoch": 9.599224305106658,
"grad_norm": 1.399062156677246,
"learning_rate": 0.001,
"loss": 2.0315,
"step": 29700
},
{
"epoch": 9.631544925662572,
"grad_norm": 1.213178038597107,
"learning_rate": 0.001,
"loss": 2.0544,
"step": 29800
},
{
"epoch": 9.663865546218487,
"grad_norm": 1.2992242574691772,
"learning_rate": 0.001,
"loss": 2.0378,
"step": 29900
},
{
"epoch": 9.696186166774401,
"grad_norm": 1.3114560842514038,
"learning_rate": 0.001,
"loss": 2.0445,
"step": 30000
},
{
"epoch": 9.728506787330316,
"grad_norm": 1.130974531173706,
"learning_rate": 0.001,
"loss": 2.0377,
"step": 30100
},
{
"epoch": 9.760827407886232,
"grad_norm": 1.3303745985031128,
"learning_rate": 0.001,
"loss": 2.0535,
"step": 30200
},
{
"epoch": 9.793148028442147,
"grad_norm": 1.8423625230789185,
"learning_rate": 0.001,
"loss": 2.0334,
"step": 30300
},
{
"epoch": 9.825468648998061,
"grad_norm": 1.193260669708252,
"learning_rate": 0.001,
"loss": 2.026,
"step": 30400
},
{
"epoch": 9.857789269553976,
"grad_norm": 1.9798153638839722,
"learning_rate": 0.001,
"loss": 2.0475,
"step": 30500
},
{
"epoch": 9.89010989010989,
"grad_norm": 1.1298526525497437,
"learning_rate": 0.001,
"loss": 2.044,
"step": 30600
},
{
"epoch": 9.922430510665805,
"grad_norm": 1.5975474119186401,
"learning_rate": 0.001,
"loss": 2.0417,
"step": 30700
},
{
"epoch": 9.95475113122172,
"grad_norm": 1.2891868352890015,
"learning_rate": 0.001,
"loss": 2.0445,
"step": 30800
},
{
"epoch": 9.987071751777634,
"grad_norm": 1.7390249967575073,
"learning_rate": 0.001,
"loss": 2.0436,
"step": 30900
},
{
"epoch": 10.019392372333549,
"grad_norm": 0.9371219277381897,
"learning_rate": 0.001,
"loss": 1.9984,
"step": 31000
},
{
"epoch": 10.051712992889463,
"grad_norm": 1.240126132965088,
"learning_rate": 0.001,
"loss": 1.9411,
"step": 31100
},
{
"epoch": 10.084033613445378,
"grad_norm": 1.129967451095581,
"learning_rate": 0.001,
"loss": 1.9458,
"step": 31200
},
{
"epoch": 10.116354234001292,
"grad_norm": 1.056111216545105,
"learning_rate": 0.001,
"loss": 1.951,
"step": 31300
},
{
"epoch": 10.148674854557207,
"grad_norm": 1.1332019567489624,
"learning_rate": 0.001,
"loss": 1.9352,
"step": 31400
},
{
"epoch": 10.180995475113122,
"grad_norm": 1.170355200767517,
"learning_rate": 0.001,
"loss": 1.9528,
"step": 31500
},
{
"epoch": 10.213316095669036,
"grad_norm": 1.212427020072937,
"learning_rate": 0.001,
"loss": 1.9544,
"step": 31600
},
{
"epoch": 10.24563671622495,
"grad_norm": 1.2170082330703735,
"learning_rate": 0.001,
"loss": 1.9603,
"step": 31700
},
{
"epoch": 10.277957336780867,
"grad_norm": 1.2054773569107056,
"learning_rate": 0.001,
"loss": 1.9755,
"step": 31800
},
{
"epoch": 10.310277957336782,
"grad_norm": 1.0994369983673096,
"learning_rate": 0.001,
"loss": 1.9709,
"step": 31900
},
{
"epoch": 10.342598577892696,
"grad_norm": 1.1936355829238892,
"learning_rate": 0.001,
"loss": 1.9493,
"step": 32000
},
{
"epoch": 10.37491919844861,
"grad_norm": 1.029075026512146,
"learning_rate": 0.001,
"loss": 1.9632,
"step": 32100
},
{
"epoch": 10.407239819004525,
"grad_norm": 0.9728875756263733,
"learning_rate": 0.001,
"loss": 1.9951,
"step": 32200
},
{
"epoch": 10.43956043956044,
"grad_norm": 1.2093101739883423,
"learning_rate": 0.001,
"loss": 1.9654,
"step": 32300
},
{
"epoch": 10.471881060116354,
"grad_norm": 1.241963267326355,
"learning_rate": 0.001,
"loss": 1.9874,
"step": 32400
},
{
"epoch": 10.504201680672269,
"grad_norm": 1.0996415615081787,
"learning_rate": 0.001,
"loss": 1.9809,
"step": 32500
},
{
"epoch": 10.536522301228183,
"grad_norm": 1.0012050867080688,
"learning_rate": 0.001,
"loss": 1.9745,
"step": 32600
},
{
"epoch": 10.568842921784098,
"grad_norm": 1.1308503150939941,
"learning_rate": 0.001,
"loss": 1.9829,
"step": 32700
},
{
"epoch": 10.601163542340013,
"grad_norm": 1.2286540269851685,
"learning_rate": 0.001,
"loss": 1.9493,
"step": 32800
},
{
"epoch": 10.633484162895927,
"grad_norm": 0.9340251684188843,
"learning_rate": 0.001,
"loss": 2.0062,
"step": 32900
},
{
"epoch": 10.665804783451842,
"grad_norm": 1.0742608308792114,
"learning_rate": 0.001,
"loss": 1.974,
"step": 33000
},
{
"epoch": 10.698125404007756,
"grad_norm": 1.1647969484329224,
"learning_rate": 0.001,
"loss": 1.9751,
"step": 33100
},
{
"epoch": 10.73044602456367,
"grad_norm": 1.2428866624832153,
"learning_rate": 0.001,
"loss": 2.0049,
"step": 33200
},
{
"epoch": 10.762766645119587,
"grad_norm": 1.4069253206253052,
"learning_rate": 0.001,
"loss": 2.0109,
"step": 33300
},
{
"epoch": 10.795087265675502,
"grad_norm": 1.2430641651153564,
"learning_rate": 0.001,
"loss": 1.9871,
"step": 33400
},
{
"epoch": 10.827407886231416,
"grad_norm": 1.2892264127731323,
"learning_rate": 0.001,
"loss": 2.0063,
"step": 33500
},
{
"epoch": 10.85972850678733,
"grad_norm": 1.044710397720337,
"learning_rate": 0.001,
"loss": 2.0003,
"step": 33600
},
{
"epoch": 10.892049127343245,
"grad_norm": 1.239604115486145,
"learning_rate": 0.001,
"loss": 1.9966,
"step": 33700
},
{
"epoch": 10.92436974789916,
"grad_norm": 1.3418306112289429,
"learning_rate": 0.001,
"loss": 1.9852,
"step": 33800
},
{
"epoch": 10.956690368455074,
"grad_norm": 1.0788873434066772,
"learning_rate": 0.001,
"loss": 1.9987,
"step": 33900
},
{
"epoch": 10.989010989010989,
"grad_norm": 1.139272928237915,
"learning_rate": 0.001,
"loss": 2.0007,
"step": 34000
},
{
"epoch": 11.021331609566904,
"grad_norm": 1.0919997692108154,
"learning_rate": 0.001,
"loss": 1.9835,
"step": 34100
},
{
"epoch": 11.053652230122818,
"grad_norm": 1.0319784879684448,
"learning_rate": 0.001,
"loss": 1.8933,
"step": 34200
},
{
"epoch": 11.085972850678733,
"grad_norm": 1.3085379600524902,
"learning_rate": 0.001,
"loss": 1.9088,
"step": 34300
},
{
"epoch": 11.118293471234647,
"grad_norm": 1.3764021396636963,
"learning_rate": 0.001,
"loss": 1.912,
"step": 34400
},
{
"epoch": 11.150614091790562,
"grad_norm": 1.1798242330551147,
"learning_rate": 0.001,
"loss": 1.9078,
"step": 34500
},
{
"epoch": 11.182934712346476,
"grad_norm": 1.325455904006958,
"learning_rate": 0.001,
"loss": 1.9255,
"step": 34600
},
{
"epoch": 11.215255332902391,
"grad_norm": 1.0184882879257202,
"learning_rate": 0.001,
"loss": 1.9248,
"step": 34700
},
{
"epoch": 11.247575953458306,
"grad_norm": 0.9781094193458557,
"learning_rate": 0.001,
"loss": 1.922,
"step": 34800
},
{
"epoch": 11.279896574014222,
"grad_norm": 1.0457218885421753,
"learning_rate": 0.001,
"loss": 1.8872,
"step": 34900
},
{
"epoch": 11.312217194570136,
"grad_norm": 1.1817752122879028,
"learning_rate": 0.001,
"loss": 1.9091,
"step": 35000
},
{
"epoch": 11.344537815126051,
"grad_norm": 1.059022307395935,
"learning_rate": 0.001,
"loss": 1.9213,
"step": 35100
},
{
"epoch": 11.376858435681966,
"grad_norm": 1.2279465198516846,
"learning_rate": 0.001,
"loss": 1.9173,
"step": 35200
},
{
"epoch": 11.40917905623788,
"grad_norm": 1.0559117794036865,
"learning_rate": 0.001,
"loss": 1.9462,
"step": 35300
},
{
"epoch": 11.441499676793795,
"grad_norm": 1.155202031135559,
"learning_rate": 0.001,
"loss": 1.9219,
"step": 35400
},
{
"epoch": 11.47382029734971,
"grad_norm": 0.9445791244506836,
"learning_rate": 0.001,
"loss": 1.9351,
"step": 35500
},
{
"epoch": 11.506140917905624,
"grad_norm": 1.164797306060791,
"learning_rate": 0.001,
"loss": 1.9704,
"step": 35600
},
{
"epoch": 11.538461538461538,
"grad_norm": 1.1770981550216675,
"learning_rate": 0.001,
"loss": 1.9473,
"step": 35700
},
{
"epoch": 11.570782159017453,
"grad_norm": 1.2111883163452148,
"learning_rate": 0.001,
"loss": 1.9358,
"step": 35800
},
{
"epoch": 11.603102779573367,
"grad_norm": 1.2625328302383423,
"learning_rate": 0.001,
"loss": 1.9276,
"step": 35900
},
{
"epoch": 11.635423400129282,
"grad_norm": 1.070054054260254,
"learning_rate": 0.001,
"loss": 1.9501,
"step": 36000
},
{
"epoch": 11.667744020685197,
"grad_norm": 1.6539897918701172,
"learning_rate": 0.001,
"loss": 1.9679,
"step": 36100
},
{
"epoch": 11.700064641241111,
"grad_norm": 0.9244352579116821,
"learning_rate": 0.001,
"loss": 1.9429,
"step": 36200
},
{
"epoch": 11.732385261797026,
"grad_norm": 1.3357621431350708,
"learning_rate": 0.001,
"loss": 1.9378,
"step": 36300
},
{
"epoch": 11.764705882352942,
"grad_norm": 1.4329630136489868,
"learning_rate": 0.001,
"loss": 1.9545,
"step": 36400
},
{
"epoch": 11.797026502908857,
"grad_norm": 1.078933835029602,
"learning_rate": 0.001,
"loss": 1.9275,
"step": 36500
},
{
"epoch": 11.829347123464771,
"grad_norm": 0.9876687526702881,
"learning_rate": 0.001,
"loss": 1.9787,
"step": 36600
},
{
"epoch": 11.861667744020686,
"grad_norm": 1.1422828435897827,
"learning_rate": 0.001,
"loss": 1.9797,
"step": 36700
},
{
"epoch": 11.8939883645766,
"grad_norm": 1.062204360961914,
"learning_rate": 0.001,
"loss": 1.9526,
"step": 36800
},
{
"epoch": 11.926308985132515,
"grad_norm": 1.3464689254760742,
"learning_rate": 0.001,
"loss": 1.9804,
"step": 36900
},
{
"epoch": 11.95862960568843,
"grad_norm": 0.9646903276443481,
"learning_rate": 0.001,
"loss": 1.9566,
"step": 37000
},
{
"epoch": 11.990950226244344,
"grad_norm": 1.1506516933441162,
"learning_rate": 0.001,
"loss": 1.9728,
"step": 37100
},
{
"epoch": 12.023270846800258,
"grad_norm": 1.369059681892395,
"learning_rate": 0.001,
"loss": 1.8877,
"step": 37200
},
{
"epoch": 12.055591467356173,
"grad_norm": 1.1158785820007324,
"learning_rate": 0.001,
"loss": 1.8268,
"step": 37300
},
{
"epoch": 12.087912087912088,
"grad_norm": 1.0913095474243164,
"learning_rate": 0.001,
"loss": 1.8475,
"step": 37400
},
{
"epoch": 12.120232708468002,
"grad_norm": 1.241456151008606,
"learning_rate": 0.001,
"loss": 1.8694,
"step": 37500
},
{
"epoch": 12.152553329023917,
"grad_norm": 1.1001811027526855,
"learning_rate": 0.001,
"loss": 1.8699,
"step": 37600
},
{
"epoch": 12.184873949579831,
"grad_norm": 1.1026263236999512,
"learning_rate": 0.001,
"loss": 1.8862,
"step": 37700
},
{
"epoch": 12.217194570135746,
"grad_norm": 1.1104148626327515,
"learning_rate": 0.001,
"loss": 1.885,
"step": 37800
},
{
"epoch": 12.24951519069166,
"grad_norm": 1.0062155723571777,
"learning_rate": 0.001,
"loss": 1.8719,
"step": 37900
},
{
"epoch": 12.281835811247577,
"grad_norm": 1.2362042665481567,
"learning_rate": 0.001,
"loss": 1.8963,
"step": 38000
},
{
"epoch": 12.314156431803491,
"grad_norm": 1.0042673349380493,
"learning_rate": 0.001,
"loss": 1.9027,
"step": 38100
},
{
"epoch": 12.346477052359406,
"grad_norm": 1.2307231426239014,
"learning_rate": 0.001,
"loss": 1.8827,
"step": 38200
},
{
"epoch": 12.37879767291532,
"grad_norm": 0.9624270796775818,
"learning_rate": 0.001,
"loss": 1.8745,
"step": 38300
},
{
"epoch": 12.411118293471235,
"grad_norm": 0.9763580560684204,
"learning_rate": 0.001,
"loss": 1.878,
"step": 38400
},
{
"epoch": 12.44343891402715,
"grad_norm": 1.371140480041504,
"learning_rate": 0.001,
"loss": 1.9079,
"step": 38500
},
{
"epoch": 12.475759534583064,
"grad_norm": 1.0613393783569336,
"learning_rate": 0.001,
"loss": 1.8867,
"step": 38600
},
{
"epoch": 12.508080155138979,
"grad_norm": 1.1260520219802856,
"learning_rate": 0.001,
"loss": 1.8916,
"step": 38700
},
{
"epoch": 12.540400775694893,
"grad_norm": 1.4606720209121704,
"learning_rate": 0.001,
"loss": 1.9039,
"step": 38800
},
{
"epoch": 12.572721396250808,
"grad_norm": 1.113877773284912,
"learning_rate": 0.001,
"loss": 1.9194,
"step": 38900
},
{
"epoch": 12.605042016806722,
"grad_norm": 1.2178810834884644,
"learning_rate": 0.001,
"loss": 1.9354,
"step": 39000
},
{
"epoch": 12.637362637362637,
"grad_norm": 1.0655773878097534,
"learning_rate": 0.001,
"loss": 1.9348,
"step": 39100
},
{
"epoch": 12.669683257918551,
"grad_norm": 1.2934608459472656,
"learning_rate": 0.001,
"loss": 1.9038,
"step": 39200
},
{
"epoch": 12.702003878474466,
"grad_norm": 0.9688911437988281,
"learning_rate": 0.001,
"loss": 1.9057,
"step": 39300
},
{
"epoch": 12.73432449903038,
"grad_norm": 1.1328833103179932,
"learning_rate": 0.001,
"loss": 1.9568,
"step": 39400
},
{
"epoch": 12.766645119586297,
"grad_norm": 1.2693794965744019,
"learning_rate": 0.001,
"loss": 1.9161,
"step": 39500
},
{
"epoch": 12.798965740142211,
"grad_norm": 1.2547452449798584,
"learning_rate": 0.001,
"loss": 1.9138,
"step": 39600
},
{
"epoch": 12.831286360698126,
"grad_norm": 1.2441757917404175,
"learning_rate": 0.001,
"loss": 1.9496,
"step": 39700
},
{
"epoch": 12.86360698125404,
"grad_norm": 0.9224479794502258,
"learning_rate": 0.001,
"loss": 1.9181,
"step": 39800
},
{
"epoch": 12.895927601809955,
"grad_norm": 1.2042218446731567,
"learning_rate": 0.001,
"loss": 1.9154,
"step": 39900
},
{
"epoch": 12.92824822236587,
"grad_norm": 1.195986032485962,
"learning_rate": 0.001,
"loss": 1.92,
"step": 40000
},
{
"epoch": 12.960568842921784,
"grad_norm": 1.5201036930084229,
"learning_rate": 0.001,
"loss": 1.9273,
"step": 40100
},
{
"epoch": 12.992889463477699,
"grad_norm": 1.1624977588653564,
"learning_rate": 0.001,
"loss": 1.9191,
"step": 40200
},
{
"epoch": 13.025210084033613,
"grad_norm": 0.8809813261032104,
"learning_rate": 0.001,
"loss": 1.8388,
"step": 40300
},
{
"epoch": 13.057530704589528,
"grad_norm": 1.2070914506912231,
"learning_rate": 0.001,
"loss": 1.8258,
"step": 40400
},
{
"epoch": 13.089851325145442,
"grad_norm": 1.0898829698562622,
"learning_rate": 0.001,
"loss": 1.864,
"step": 40500
},
{
"epoch": 13.122171945701357,
"grad_norm": 1.113163709640503,
"learning_rate": 0.001,
"loss": 1.8202,
"step": 40600
},
{
"epoch": 13.154492566257272,
"grad_norm": 1.1067817211151123,
"learning_rate": 0.001,
"loss": 1.8332,
"step": 40700
},
{
"epoch": 13.186813186813186,
"grad_norm": 1.3998067378997803,
"learning_rate": 0.001,
"loss": 1.8404,
"step": 40800
},
{
"epoch": 13.2191338073691,
"grad_norm": 1.395437479019165,
"learning_rate": 0.001,
"loss": 1.8519,
"step": 40900
},
{
"epoch": 13.251454427925015,
"grad_norm": 1.0215939283370972,
"learning_rate": 0.001,
"loss": 1.8144,
"step": 41000
},
{
"epoch": 13.283775048480932,
"grad_norm": 1.2167524099349976,
"learning_rate": 0.001,
"loss": 1.8414,
"step": 41100
},
{
"epoch": 13.316095669036846,
"grad_norm": 1.1152799129486084,
"learning_rate": 0.001,
"loss": 1.8569,
"step": 41200
},
{
"epoch": 13.34841628959276,
"grad_norm": 1.06120765209198,
"learning_rate": 0.001,
"loss": 1.8496,
"step": 41300
},
{
"epoch": 13.380736910148675,
"grad_norm": 1.1193475723266602,
"learning_rate": 0.001,
"loss": 1.8679,
"step": 41400
},
{
"epoch": 13.41305753070459,
"grad_norm": 1.1804929971694946,
"learning_rate": 0.001,
"loss": 1.8768,
"step": 41500
},
{
"epoch": 13.445378151260504,
"grad_norm": 0.9798558354377747,
"learning_rate": 0.001,
"loss": 1.8611,
"step": 41600
},
{
"epoch": 13.477698771816419,
"grad_norm": 1.071977138519287,
"learning_rate": 0.001,
"loss": 1.8651,
"step": 41700
},
{
"epoch": 13.510019392372334,
"grad_norm": 1.1536909341812134,
"learning_rate": 0.001,
"loss": 1.8736,
"step": 41800
},
{
"epoch": 13.542340012928248,
"grad_norm": 1.042947769165039,
"learning_rate": 0.001,
"loss": 1.8417,
"step": 41900
},
{
"epoch": 13.574660633484163,
"grad_norm": 1.2445282936096191,
"learning_rate": 0.001,
"loss": 1.8717,
"step": 42000
},
{
"epoch": 13.606981254040077,
"grad_norm": 1.1522380113601685,
"learning_rate": 0.001,
"loss": 1.8796,
"step": 42100
},
{
"epoch": 13.639301874595992,
"grad_norm": 0.980404257774353,
"learning_rate": 0.001,
"loss": 1.8773,
"step": 42200
},
{
"epoch": 13.671622495151906,
"grad_norm": 1.2718123197555542,
"learning_rate": 0.001,
"loss": 1.8546,
"step": 42300
},
{
"epoch": 13.70394311570782,
"grad_norm": 1.400966763496399,
"learning_rate": 0.001,
"loss": 1.8677,
"step": 42400
},
{
"epoch": 13.736263736263737,
"grad_norm": 1.1766020059585571,
"learning_rate": 0.001,
"loss": 1.8764,
"step": 42500
},
{
"epoch": 13.768584356819652,
"grad_norm": 1.259521484375,
"learning_rate": 0.001,
"loss": 1.8938,
"step": 42600
},
{
"epoch": 13.800904977375566,
"grad_norm": 1.1797550916671753,
"learning_rate": 0.001,
"loss": 1.8738,
"step": 42700
},
{
"epoch": 13.83322559793148,
"grad_norm": 0.9888365268707275,
"learning_rate": 0.001,
"loss": 1.9153,
"step": 42800
},
{
"epoch": 13.865546218487395,
"grad_norm": 1.108959436416626,
"learning_rate": 0.001,
"loss": 1.8916,
"step": 42900
},
{
"epoch": 13.89786683904331,
"grad_norm": 1.1953341960906982,
"learning_rate": 0.001,
"loss": 1.9091,
"step": 43000
},
{
"epoch": 13.930187459599225,
"grad_norm": 0.9718614816665649,
"learning_rate": 0.001,
"loss": 1.8896,
"step": 43100
},
{
"epoch": 13.96250808015514,
"grad_norm": 1.102332592010498,
"learning_rate": 0.001,
"loss": 1.8836,
"step": 43200
},
{
"epoch": 13.994828700711054,
"grad_norm": 0.9656371474266052,
"learning_rate": 0.001,
"loss": 1.8902,
"step": 43300
},
{
"epoch": 14.027149321266968,
"grad_norm": 0.9415162801742554,
"learning_rate": 0.001,
"loss": 1.8123,
"step": 43400
},
{
"epoch": 14.059469941822883,
"grad_norm": 1.268556833267212,
"learning_rate": 0.001,
"loss": 1.7736,
"step": 43500
},
{
"epoch": 14.091790562378797,
"grad_norm": 1.369551658630371,
"learning_rate": 0.001,
"loss": 1.7756,
"step": 43600
},
{
"epoch": 14.124111182934712,
"grad_norm": 1.0541764497756958,
"learning_rate": 0.001,
"loss": 1.8139,
"step": 43700
},
{
"epoch": 14.156431803490626,
"grad_norm": 1.0418062210083008,
"learning_rate": 0.001,
"loss": 1.7946,
"step": 43800
},
{
"epoch": 14.188752424046541,
"grad_norm": 1.2742817401885986,
"learning_rate": 0.001,
"loss": 1.8125,
"step": 43900
},
{
"epoch": 14.221073044602456,
"grad_norm": 1.1431653499603271,
"learning_rate": 0.001,
"loss": 1.8091,
"step": 44000
},
{
"epoch": 14.25339366515837,
"grad_norm": 1.1454989910125732,
"learning_rate": 0.001,
"loss": 1.8023,
"step": 44100
},
{
"epoch": 14.285714285714286,
"grad_norm": 1.1051077842712402,
"learning_rate": 0.001,
"loss": 1.787,
"step": 44200
},
{
"epoch": 14.318034906270201,
"grad_norm": 1.0722583532333374,
"learning_rate": 0.001,
"loss": 1.8183,
"step": 44300
},
{
"epoch": 14.350355526826116,
"grad_norm": 0.8678897619247437,
"learning_rate": 0.001,
"loss": 1.8147,
"step": 44400
},
{
"epoch": 14.38267614738203,
"grad_norm": 0.9575488567352295,
"learning_rate": 0.001,
"loss": 1.8209,
"step": 44500
},
{
"epoch": 14.414996767937945,
"grad_norm": 1.1212340593338013,
"learning_rate": 0.001,
"loss": 1.8389,
"step": 44600
},
{
"epoch": 14.44731738849386,
"grad_norm": 3.612347364425659,
"learning_rate": 0.001,
"loss": 1.8051,
"step": 44700
},
{
"epoch": 14.479638009049774,
"grad_norm": 1.1123112440109253,
"learning_rate": 0.001,
"loss": 1.8191,
"step": 44800
},
{
"epoch": 14.511958629605688,
"grad_norm": 1.093819260597229,
"learning_rate": 0.001,
"loss": 1.8469,
"step": 44900
},
{
"epoch": 14.544279250161603,
"grad_norm": 1.1778242588043213,
"learning_rate": 0.001,
"loss": 1.8386,
"step": 45000
},
{
"epoch": 14.576599870717518,
"grad_norm": 1.2398154735565186,
"learning_rate": 0.001,
"loss": 1.8262,
"step": 45100
},
{
"epoch": 14.608920491273432,
"grad_norm": 1.4576246738433838,
"learning_rate": 0.001,
"loss": 1.8475,
"step": 45200
},
{
"epoch": 14.641241111829347,
"grad_norm": 1.1424200534820557,
"learning_rate": 0.001,
"loss": 1.8487,
"step": 45300
},
{
"epoch": 14.673561732385261,
"grad_norm": 1.0567972660064697,
"learning_rate": 0.001,
"loss": 1.8465,
"step": 45400
},
{
"epoch": 14.705882352941176,
"grad_norm": 1.2708767652511597,
"learning_rate": 0.001,
"loss": 1.8401,
"step": 45500
},
{
"epoch": 14.738202973497092,
"grad_norm": 1.1260626316070557,
"learning_rate": 0.001,
"loss": 1.8524,
"step": 45600
},
{
"epoch": 14.770523594053007,
"grad_norm": 1.2315834760665894,
"learning_rate": 0.001,
"loss": 1.8688,
"step": 45700
},
{
"epoch": 14.802844214608921,
"grad_norm": 1.148695945739746,
"learning_rate": 0.001,
"loss": 1.858,
"step": 45800
},
{
"epoch": 14.835164835164836,
"grad_norm": 1.004560112953186,
"learning_rate": 0.001,
"loss": 1.8637,
"step": 45900
},
{
"epoch": 14.86748545572075,
"grad_norm": 1.2938129901885986,
"learning_rate": 0.001,
"loss": 1.8759,
"step": 46000
},
{
"epoch": 14.899806076276665,
"grad_norm": 0.936123251914978,
"learning_rate": 0.001,
"loss": 1.8408,
"step": 46100
},
{
"epoch": 14.93212669683258,
"grad_norm": 1.173580288887024,
"learning_rate": 0.001,
"loss": 1.8871,
"step": 46200
},
{
"epoch": 14.964447317388494,
"grad_norm": 1.340844988822937,
"learning_rate": 0.001,
"loss": 1.8669,
"step": 46300
},
{
"epoch": 14.996767937944409,
"grad_norm": 1.3199015855789185,
"learning_rate": 0.001,
"loss": 1.875,
"step": 46400
},
{
"epoch": 15.029088558500323,
"grad_norm": 1.3935327529907227,
"learning_rate": 0.001,
"loss": 1.7577,
"step": 46500
},
{
"epoch": 15.061409179056238,
"grad_norm": 1.2479134798049927,
"learning_rate": 0.001,
"loss": 1.7524,
"step": 46600
},
{
"epoch": 15.093729799612152,
"grad_norm": 1.3339827060699463,
"learning_rate": 0.001,
"loss": 1.778,
"step": 46700
},
{
"epoch": 15.126050420168067,
"grad_norm": 1.185196042060852,
"learning_rate": 0.001,
"loss": 1.7368,
"step": 46800
},
{
"epoch": 15.158371040723981,
"grad_norm": 1.2535922527313232,
"learning_rate": 0.001,
"loss": 1.7511,
"step": 46900
},
{
"epoch": 15.190691661279896,
"grad_norm": 1.3402800559997559,
"learning_rate": 0.001,
"loss": 1.7662,
"step": 47000
},
{
"epoch": 15.22301228183581,
"grad_norm": 1.272594928741455,
"learning_rate": 0.001,
"loss": 1.7792,
"step": 47100
},
{
"epoch": 15.255332902391725,
"grad_norm": 1.1950360536575317,
"learning_rate": 0.001,
"loss": 1.8125,
"step": 47200
},
{
"epoch": 15.287653522947641,
"grad_norm": 1.2349659204483032,
"learning_rate": 0.001,
"loss": 1.7832,
"step": 47300
},
{
"epoch": 15.319974143503556,
"grad_norm": 1.061517357826233,
"learning_rate": 0.001,
"loss": 1.7894,
"step": 47400
},
{
"epoch": 15.35229476405947,
"grad_norm": 1.3945839405059814,
"learning_rate": 0.001,
"loss": 1.809,
"step": 47500
},
{
"epoch": 15.384615384615385,
"grad_norm": 1.3774470090866089,
"learning_rate": 0.001,
"loss": 1.7642,
"step": 47600
},
{
"epoch": 15.4169360051713,
"grad_norm": 1.2366855144500732,
"learning_rate": 0.001,
"loss": 1.7812,
"step": 47700
},
{
"epoch": 15.449256625727214,
"grad_norm": 1.4002078771591187,
"learning_rate": 0.001,
"loss": 1.8015,
"step": 47800
},
{
"epoch": 15.481577246283129,
"grad_norm": 1.4305744171142578,
"learning_rate": 0.001,
"loss": 1.7981,
"step": 47900
},
{
"epoch": 15.513897866839043,
"grad_norm": 1.2977144718170166,
"learning_rate": 0.001,
"loss": 1.8124,
"step": 48000
},
{
"epoch": 15.546218487394958,
"grad_norm": 1.0210585594177246,
"learning_rate": 0.001,
"loss": 1.8203,
"step": 48100
},
{
"epoch": 15.578539107950872,
"grad_norm": 1.2214510440826416,
"learning_rate": 0.001,
"loss": 1.8164,
"step": 48200
},
{
"epoch": 15.610859728506787,
"grad_norm": 1.0851165056228638,
"learning_rate": 0.001,
"loss": 1.7905,
"step": 48300
},
{
"epoch": 15.643180349062701,
"grad_norm": 2.2630977630615234,
"learning_rate": 0.001,
"loss": 1.8129,
"step": 48400
},
{
"epoch": 15.675500969618616,
"grad_norm": 1.0727578401565552,
"learning_rate": 0.001,
"loss": 1.8309,
"step": 48500
},
{
"epoch": 15.70782159017453,
"grad_norm": 1.122301459312439,
"learning_rate": 0.001,
"loss": 1.8025,
"step": 48600
},
{
"epoch": 15.740142210730447,
"grad_norm": 1.1632217168807983,
"learning_rate": 0.001,
"loss": 1.8057,
"step": 48700
},
{
"epoch": 15.772462831286362,
"grad_norm": 1.2424322366714478,
"learning_rate": 0.001,
"loss": 1.8345,
"step": 48800
},
{
"epoch": 15.804783451842276,
"grad_norm": 0.9887394309043884,
"learning_rate": 0.001,
"loss": 1.8425,
"step": 48900
},
{
"epoch": 15.83710407239819,
"grad_norm": 1.1400974988937378,
"learning_rate": 0.001,
"loss": 1.8399,
"step": 49000
},
{
"epoch": 15.869424692954105,
"grad_norm": 1.1223629713058472,
"learning_rate": 0.001,
"loss": 1.813,
"step": 49100
},
{
"epoch": 15.90174531351002,
"grad_norm": 1.370543122291565,
"learning_rate": 0.001,
"loss": 1.831,
"step": 49200
},
{
"epoch": 15.934065934065934,
"grad_norm": 1.1231876611709595,
"learning_rate": 0.001,
"loss": 1.8377,
"step": 49300
},
{
"epoch": 15.966386554621849,
"grad_norm": 1.1462494134902954,
"learning_rate": 0.001,
"loss": 1.843,
"step": 49400
},
{
"epoch": 15.998707175177763,
"grad_norm": 1.1465389728546143,
"learning_rate": 0.001,
"loss": 1.7893,
"step": 49500
},
{
"epoch": 16.031027795733678,
"grad_norm": 1.4136383533477783,
"learning_rate": 0.001,
"loss": 1.7237,
"step": 49600
},
{
"epoch": 16.063348416289593,
"grad_norm": 1.3214054107666016,
"learning_rate": 0.001,
"loss": 1.7424,
"step": 49700
},
{
"epoch": 16.095669036845507,
"grad_norm": 1.1296522617340088,
"learning_rate": 0.001,
"loss": 1.7382,
"step": 49800
},
{
"epoch": 16.12798965740142,
"grad_norm": 1.1522901058197021,
"learning_rate": 0.001,
"loss": 1.7455,
"step": 49900
},
{
"epoch": 16.160310277957336,
"grad_norm": 1.4975008964538574,
"learning_rate": 0.001,
"loss": 1.7403,
"step": 50000
},
{
"epoch": 16.19263089851325,
"grad_norm": 1.11152982711792,
"learning_rate": 0.001,
"loss": 1.7476,
"step": 50100
},
{
"epoch": 16.224951519069165,
"grad_norm": 1.299704670906067,
"learning_rate": 0.001,
"loss": 1.739,
"step": 50200
},
{
"epoch": 16.25727213962508,
"grad_norm": 1.0670026540756226,
"learning_rate": 0.001,
"loss": 1.7473,
"step": 50300
},
{
"epoch": 16.289592760180994,
"grad_norm": 1.260867714881897,
"learning_rate": 0.001,
"loss": 1.7428,
"step": 50400
},
{
"epoch": 16.32191338073691,
"grad_norm": 1.5163524150848389,
"learning_rate": 0.001,
"loss": 1.7542,
"step": 50500
},
{
"epoch": 16.354234001292824,
"grad_norm": 1.0129752159118652,
"learning_rate": 0.001,
"loss": 1.7653,
"step": 50600
},
{
"epoch": 16.386554621848738,
"grad_norm": 1.3281183242797852,
"learning_rate": 0.001,
"loss": 1.7604,
"step": 50700
},
{
"epoch": 16.418875242404653,
"grad_norm": 1.205254077911377,
"learning_rate": 0.001,
"loss": 1.7695,
"step": 50800
},
{
"epoch": 16.451195862960567,
"grad_norm": 1.1382715702056885,
"learning_rate": 0.001,
"loss": 1.7736,
"step": 50900
},
{
"epoch": 16.483516483516482,
"grad_norm": 1.2368077039718628,
"learning_rate": 0.001,
"loss": 1.7838,
"step": 51000
},
{
"epoch": 16.5158371040724,
"grad_norm": 1.3391265869140625,
"learning_rate": 0.001,
"loss": 1.7735,
"step": 51100
},
{
"epoch": 16.548157724628314,
"grad_norm": 1.1239650249481201,
"learning_rate": 0.001,
"loss": 1.7567,
"step": 51200
},
{
"epoch": 16.58047834518423,
"grad_norm": 1.0369645357131958,
"learning_rate": 0.001,
"loss": 1.7963,
"step": 51300
},
{
"epoch": 16.612798965740144,
"grad_norm": 1.4317647218704224,
"learning_rate": 0.001,
"loss": 1.781,
"step": 51400
},
{
"epoch": 16.645119586296058,
"grad_norm": 1.2334084510803223,
"learning_rate": 0.001,
"loss": 1.7876,
"step": 51500
},
{
"epoch": 16.677440206851973,
"grad_norm": 1.2364823818206787,
"learning_rate": 0.001,
"loss": 1.7724,
"step": 51600
},
{
"epoch": 16.709760827407887,
"grad_norm": 1.032331943511963,
"learning_rate": 0.001,
"loss": 1.7978,
"step": 51700
},
{
"epoch": 16.742081447963802,
"grad_norm": 1.4971531629562378,
"learning_rate": 0.001,
"loss": 1.7797,
"step": 51800
},
{
"epoch": 16.774402068519716,
"grad_norm": 1.1399532556533813,
"learning_rate": 0.001,
"loss": 1.7998,
"step": 51900
},
{
"epoch": 16.80672268907563,
"grad_norm": 4.143343925476074,
"learning_rate": 0.001,
"loss": 1.8055,
"step": 52000
},
{
"epoch": 16.839043309631545,
"grad_norm": 1.2093541622161865,
"learning_rate": 0.001,
"loss": 1.785,
"step": 52100
},
{
"epoch": 16.87136393018746,
"grad_norm": 1.04611337184906,
"learning_rate": 0.001,
"loss": 1.7681,
"step": 52200
},
{
"epoch": 16.903684550743375,
"grad_norm": 1.2969337701797485,
"learning_rate": 0.001,
"loss": 1.7753,
"step": 52300
},
{
"epoch": 16.93600517129929,
"grad_norm": 0.9581918716430664,
"learning_rate": 0.001,
"loss": 1.7911,
"step": 52400
},
{
"epoch": 16.968325791855204,
"grad_norm": 1.0658535957336426,
"learning_rate": 0.001,
"loss": 1.8223,
"step": 52500
},
{
"epoch": 17.00064641241112,
"grad_norm": 1.322858214378357,
"learning_rate": 0.001,
"loss": 1.7659,
"step": 52600
},
{
"epoch": 17.032967032967033,
"grad_norm": 1.4845845699310303,
"learning_rate": 0.001,
"loss": 1.6901,
"step": 52700
},
{
"epoch": 17.065287653522947,
"grad_norm": 1.4620217084884644,
"learning_rate": 0.001,
"loss": 1.6839,
"step": 52800
},
{
"epoch": 17.097608274078862,
"grad_norm": 1.4743014574050903,
"learning_rate": 0.001,
"loss": 1.6718,
"step": 52900
},
{
"epoch": 17.129928894634777,
"grad_norm": 1.7777990102767944,
"learning_rate": 0.001,
"loss": 1.6903,
"step": 53000
},
{
"epoch": 17.16224951519069,
"grad_norm": 1.8911314010620117,
"learning_rate": 0.001,
"loss": 1.7087,
"step": 53100
},
{
"epoch": 17.194570135746606,
"grad_norm": 1.5767875909805298,
"learning_rate": 0.001,
"loss": 1.7232,
"step": 53200
},
{
"epoch": 17.22689075630252,
"grad_norm": 1.7391340732574463,
"learning_rate": 0.001,
"loss": 1.7499,
"step": 53300
},
{
"epoch": 17.259211376858435,
"grad_norm": 1.4621258974075317,
"learning_rate": 0.001,
"loss": 1.7194,
"step": 53400
},
{
"epoch": 17.29153199741435,
"grad_norm": 2.076847553253174,
"learning_rate": 0.001,
"loss": 1.7175,
"step": 53500
},
{
"epoch": 17.323852617970264,
"grad_norm": 1.7916910648345947,
"learning_rate": 0.001,
"loss": 1.7091,
"step": 53600
},
{
"epoch": 17.35617323852618,
"grad_norm": 1.8870574235916138,
"learning_rate": 0.001,
"loss": 1.741,
"step": 53700
},
{
"epoch": 17.388493859082093,
"grad_norm": 1.267464518547058,
"learning_rate": 0.001,
"loss": 1.7362,
"step": 53800
},
{
"epoch": 17.420814479638008,
"grad_norm": 1.4546117782592773,
"learning_rate": 0.001,
"loss": 1.7316,
"step": 53900
},
{
"epoch": 17.453135100193922,
"grad_norm": 1.8361940383911133,
"learning_rate": 0.001,
"loss": 1.7163,
"step": 54000
},
{
"epoch": 17.485455720749837,
"grad_norm": 1.8543264865875244,
"learning_rate": 0.001,
"loss": 1.7469,
"step": 54100
},
{
"epoch": 17.517776341305755,
"grad_norm": 1.6080793142318726,
"learning_rate": 0.001,
"loss": 1.7247,
"step": 54200
},
{
"epoch": 17.55009696186167,
"grad_norm": 1.7696819305419922,
"learning_rate": 0.001,
"loss": 1.7532,
"step": 54300
},
{
"epoch": 17.582417582417584,
"grad_norm": 1.9339938163757324,
"learning_rate": 0.001,
"loss": 1.7803,
"step": 54400
},
{
"epoch": 17.6147382029735,
"grad_norm": 1.6124098300933838,
"learning_rate": 0.001,
"loss": 1.7597,
"step": 54500
},
{
"epoch": 17.647058823529413,
"grad_norm": 1.9301472902297974,
"learning_rate": 0.001,
"loss": 1.7487,
"step": 54600
},
{
"epoch": 17.679379444085328,
"grad_norm": 1.3814643621444702,
"learning_rate": 0.001,
"loss": 1.7435,
"step": 54700
},
{
"epoch": 17.711700064641242,
"grad_norm": 2.3363003730773926,
"learning_rate": 0.001,
"loss": 1.7477,
"step": 54800
},
{
"epoch": 17.744020685197157,
"grad_norm": 1.5960017442703247,
"learning_rate": 0.001,
"loss": 1.7944,
"step": 54900
},
{
"epoch": 17.77634130575307,
"grad_norm": 1.4182417392730713,
"learning_rate": 0.001,
"loss": 1.757,
"step": 55000
},
{
"epoch": 17.808661926308986,
"grad_norm": 1.5050570964813232,
"learning_rate": 0.001,
"loss": 1.7788,
"step": 55100
},
{
"epoch": 17.8409825468649,
"grad_norm": 1.7510799169540405,
"learning_rate": 0.001,
"loss": 1.7685,
"step": 55200
},
{
"epoch": 17.873303167420815,
"grad_norm": 1.8426209688186646,
"learning_rate": 0.001,
"loss": 1.7701,
"step": 55300
},
{
"epoch": 17.90562378797673,
"grad_norm": 1.778796672821045,
"learning_rate": 0.001,
"loss": 1.7546,
"step": 55400
},
{
"epoch": 17.937944408532644,
"grad_norm": 1.843400001525879,
"learning_rate": 0.001,
"loss": 1.7567,
"step": 55500
},
{
"epoch": 17.97026502908856,
"grad_norm": 1.376536250114441,
"learning_rate": 0.001,
"loss": 1.7876,
"step": 55600
},
{
"epoch": 18.002585649644473,
"grad_norm": 1.459142804145813,
"learning_rate": 0.001,
"loss": 1.7969,
"step": 55700
},
{
"epoch": 18.034906270200388,
"grad_norm": 1.385663628578186,
"learning_rate": 0.001,
"loss": 1.6557,
"step": 55800
},
{
"epoch": 18.067226890756302,
"grad_norm": 1.577804446220398,
"learning_rate": 0.001,
"loss": 1.651,
"step": 55900
},
{
"epoch": 18.099547511312217,
"grad_norm": 0.9823166131973267,
"learning_rate": 0.001,
"loss": 1.6503,
"step": 56000
},
{
"epoch": 18.13186813186813,
"grad_norm": 1.701475739479065,
"learning_rate": 0.001,
"loss": 1.6771,
"step": 56100
},
{
"epoch": 18.164188752424046,
"grad_norm": 1.289601445198059,
"learning_rate": 0.001,
"loss": 1.6993,
"step": 56200
},
{
"epoch": 18.19650937297996,
"grad_norm": 2.680657386779785,
"learning_rate": 0.001,
"loss": 1.692,
"step": 56300
},
{
"epoch": 18.228829993535875,
"grad_norm": 1.2981613874435425,
"learning_rate": 0.001,
"loss": 1.7194,
"step": 56400
},
{
"epoch": 18.26115061409179,
"grad_norm": 1.715312123298645,
"learning_rate": 0.001,
"loss": 1.6828,
"step": 56500
},
{
"epoch": 18.293471234647704,
"grad_norm": 1.151294469833374,
"learning_rate": 0.001,
"loss": 1.6828,
"step": 56600
},
{
"epoch": 18.32579185520362,
"grad_norm": 1.3789284229278564,
"learning_rate": 0.001,
"loss": 1.718,
"step": 56700
},
{
"epoch": 18.358112475759533,
"grad_norm": 1.1405805349349976,
"learning_rate": 0.001,
"loss": 1.65,
"step": 56800
},
{
"epoch": 18.390433096315448,
"grad_norm": 1.230582594871521,
"learning_rate": 0.001,
"loss": 1.7174,
"step": 56900
},
{
"epoch": 18.422753716871362,
"grad_norm": 1.3198965787887573,
"learning_rate": 0.001,
"loss": 1.7173,
"step": 57000
},
{
"epoch": 18.455074337427277,
"grad_norm": 1.647703766822815,
"learning_rate": 0.001,
"loss": 1.7111,
"step": 57100
},
{
"epoch": 18.48739495798319,
"grad_norm": 1.202138066291809,
"learning_rate": 0.001,
"loss": 1.6879,
"step": 57200
},
{
"epoch": 18.51971557853911,
"grad_norm": 1.23879075050354,
"learning_rate": 0.001,
"loss": 1.734,
"step": 57300
},
{
"epoch": 18.552036199095024,
"grad_norm": 1.201494574546814,
"learning_rate": 0.001,
"loss": 1.7566,
"step": 57400
},
{
"epoch": 18.58435681965094,
"grad_norm": 1.2573801279067993,
"learning_rate": 0.001,
"loss": 1.7185,
"step": 57500
},
{
"epoch": 18.616677440206853,
"grad_norm": 1.2716763019561768,
"learning_rate": 0.001,
"loss": 1.7072,
"step": 57600
},
{
"epoch": 18.648998060762768,
"grad_norm": 1.058627724647522,
"learning_rate": 0.001,
"loss": 1.7288,
"step": 57700
},
{
"epoch": 18.681318681318682,
"grad_norm": 1.353332281112671,
"learning_rate": 0.001,
"loss": 1.7155,
"step": 57800
},
{
"epoch": 18.713639301874597,
"grad_norm": 1.175697684288025,
"learning_rate": 0.001,
"loss": 1.6966,
"step": 57900
},
{
"epoch": 18.74595992243051,
"grad_norm": 1.0386375188827515,
"learning_rate": 0.001,
"loss": 1.7585,
"step": 58000
},
{
"epoch": 18.778280542986426,
"grad_norm": 1.2818589210510254,
"learning_rate": 0.001,
"loss": 1.7699,
"step": 58100
},
{
"epoch": 18.81060116354234,
"grad_norm": 1.291780710220337,
"learning_rate": 0.001,
"loss": 1.742,
"step": 58200
},
{
"epoch": 18.842921784098255,
"grad_norm": 1.204770803451538,
"learning_rate": 0.001,
"loss": 1.7425,
"step": 58300
},
{
"epoch": 18.87524240465417,
"grad_norm": 1.2708287239074707,
"learning_rate": 0.001,
"loss": 1.7588,
"step": 58400
},
{
"epoch": 18.907563025210084,
"grad_norm": 1.3189966678619385,
"learning_rate": 0.001,
"loss": 1.7107,
"step": 58500
},
{
"epoch": 18.939883645766,
"grad_norm": 1.2411608695983887,
"learning_rate": 0.001,
"loss": 1.7359,
"step": 58600
},
{
"epoch": 18.972204266321913,
"grad_norm": 1.3548706769943237,
"learning_rate": 0.001,
"loss": 1.7459,
"step": 58700
},
{
"epoch": 19.004524886877828,
"grad_norm": 1.1609103679656982,
"learning_rate": 0.001,
"loss": 1.7467,
"step": 58800
},
{
"epoch": 19.036845507433743,
"grad_norm": 1.4791889190673828,
"learning_rate": 0.001,
"loss": 1.6426,
"step": 58900
},
{
"epoch": 19.069166127989657,
"grad_norm": 1.2185018062591553,
"learning_rate": 0.001,
"loss": 1.6269,
"step": 59000
},
{
"epoch": 19.10148674854557,
"grad_norm": 1.4684048891067505,
"learning_rate": 0.001,
"loss": 1.6587,
"step": 59100
},
{
"epoch": 19.133807369101486,
"grad_norm": 1.2556350231170654,
"learning_rate": 0.001,
"loss": 1.667,
"step": 59200
},
{
"epoch": 19.1661279896574,
"grad_norm": 1.3185192346572876,
"learning_rate": 0.001,
"loss": 1.6482,
"step": 59300
},
{
"epoch": 19.198448610213315,
"grad_norm": 1.174451470375061,
"learning_rate": 0.001,
"loss": 1.6581,
"step": 59400
},
{
"epoch": 19.23076923076923,
"grad_norm": 1.5593920946121216,
"learning_rate": 0.001,
"loss": 1.6536,
"step": 59500
},
{
"epoch": 19.263089851325145,
"grad_norm": 1.0820014476776123,
"learning_rate": 0.001,
"loss": 1.6585,
"step": 59600
},
{
"epoch": 19.29541047188106,
"grad_norm": 1.4829204082489014,
"learning_rate": 0.001,
"loss": 1.6528,
"step": 59700
},
{
"epoch": 19.327731092436974,
"grad_norm": 1.426621437072754,
"learning_rate": 0.001,
"loss": 1.6563,
"step": 59800
},
{
"epoch": 19.360051712992888,
"grad_norm": 1.2473586797714233,
"learning_rate": 0.001,
"loss": 1.6806,
"step": 59900
},
{
"epoch": 19.392372333548803,
"grad_norm": 1.1561217308044434,
"learning_rate": 0.001,
"loss": 1.6744,
"step": 60000
},
{
"epoch": 19.424692954104717,
"grad_norm": 1.4965547323226929,
"learning_rate": 0.001,
"loss": 1.6999,
"step": 60100
},
{
"epoch": 19.457013574660632,
"grad_norm": 1.284369707107544,
"learning_rate": 0.001,
"loss": 1.6816,
"step": 60200
},
{
"epoch": 19.489334195216546,
"grad_norm": 1.2189420461654663,
"learning_rate": 0.001,
"loss": 1.6687,
"step": 60300
},
{
"epoch": 19.521654815772465,
"grad_norm": 1.1701105833053589,
"learning_rate": 0.001,
"loss": 1.6865,
"step": 60400
},
{
"epoch": 19.55397543632838,
"grad_norm": 1.2209968566894531,
"learning_rate": 0.001,
"loss": 1.6981,
"step": 60500
},
{
"epoch": 19.586296056884294,
"grad_norm": 0.8396730422973633,
"learning_rate": 0.001,
"loss": 1.6924,
"step": 60600
},
{
"epoch": 19.618616677440208,
"grad_norm": 1.237061858177185,
"learning_rate": 0.001,
"loss": 1.6996,
"step": 60700
},
{
"epoch": 19.650937297996123,
"grad_norm": 1.1365478038787842,
"learning_rate": 0.001,
"loss": 1.6981,
"step": 60800
},
{
"epoch": 19.683257918552037,
"grad_norm": 1.2317266464233398,
"learning_rate": 0.001,
"loss": 1.6784,
"step": 60900
},
{
"epoch": 19.715578539107952,
"grad_norm": 1.3470901250839233,
"learning_rate": 0.001,
"loss": 1.6913,
"step": 61000
},
{
"epoch": 19.747899159663866,
"grad_norm": 1.2688798904418945,
"learning_rate": 0.001,
"loss": 1.6841,
"step": 61100
},
{
"epoch": 19.78021978021978,
"grad_norm": 1.069466471672058,
"learning_rate": 0.001,
"loss": 1.7036,
"step": 61200
},
{
"epoch": 19.812540400775696,
"grad_norm": 1.6772282123565674,
"learning_rate": 0.001,
"loss": 1.7298,
"step": 61300
},
{
"epoch": 19.84486102133161,
"grad_norm": 1.3225016593933105,
"learning_rate": 0.001,
"loss": 1.711,
"step": 61400
},
{
"epoch": 19.877181641887525,
"grad_norm": 1.3533378839492798,
"learning_rate": 0.001,
"loss": 1.7338,
"step": 61500
},
{
"epoch": 19.90950226244344,
"grad_norm": 1.0688648223876953,
"learning_rate": 0.001,
"loss": 1.7381,
"step": 61600
},
{
"epoch": 19.941822882999354,
"grad_norm": 1.357704758644104,
"learning_rate": 0.001,
"loss": 1.7413,
"step": 61700
},
{
"epoch": 19.97414350355527,
"grad_norm": 1.1575063467025757,
"learning_rate": 0.001,
"loss": 1.7246,
"step": 61800
},
{
"epoch": 20.006464124111183,
"grad_norm": 1.2985217571258545,
"learning_rate": 0.001,
"loss": 1.721,
"step": 61900
},
{
"epoch": 20.038784744667097,
"grad_norm": 1.1972153186798096,
"learning_rate": 0.001,
"loss": 1.6357,
"step": 62000
},
{
"epoch": 20.071105365223012,
"grad_norm": 1.1683275699615479,
"learning_rate": 0.001,
"loss": 1.5913,
"step": 62100
},
{
"epoch": 20.103425985778927,
"grad_norm": 1.3947675228118896,
"learning_rate": 0.001,
"loss": 1.614,
"step": 62200
},
{
"epoch": 20.13574660633484,
"grad_norm": 1.3403626680374146,
"learning_rate": 0.001,
"loss": 1.606,
"step": 62300
},
{
"epoch": 20.168067226890756,
"grad_norm": 1.252703309059143,
"learning_rate": 0.001,
"loss": 1.6326,
"step": 62400
},
{
"epoch": 20.20038784744667,
"grad_norm": 1.4022725820541382,
"learning_rate": 0.001,
"loss": 1.6434,
"step": 62500
},
{
"epoch": 20.232708468002585,
"grad_norm": 1.2247436046600342,
"learning_rate": 0.001,
"loss": 1.6279,
"step": 62600
},
{
"epoch": 20.2650290885585,
"grad_norm": 1.0815917253494263,
"learning_rate": 0.001,
"loss": 1.6277,
"step": 62700
},
{
"epoch": 20.297349709114414,
"grad_norm": 1.089803695678711,
"learning_rate": 0.001,
"loss": 1.6229,
"step": 62800
},
{
"epoch": 20.32967032967033,
"grad_norm": 1.2179733514785767,
"learning_rate": 0.001,
"loss": 1.6327,
"step": 62900
},
{
"epoch": 20.361990950226243,
"grad_norm": 1.2288011312484741,
"learning_rate": 0.001,
"loss": 1.647,
"step": 63000
},
{
"epoch": 20.394311570782158,
"grad_norm": 1.0536634922027588,
"learning_rate": 0.001,
"loss": 1.6695,
"step": 63100
},
{
"epoch": 20.426632191338072,
"grad_norm": 1.0153943300247192,
"learning_rate": 0.001,
"loss": 1.664,
"step": 63200
},
{
"epoch": 20.458952811893987,
"grad_norm": 1.130033016204834,
"learning_rate": 0.001,
"loss": 1.6725,
"step": 63300
},
{
"epoch": 20.4912734324499,
"grad_norm": 1.515398621559143,
"learning_rate": 0.001,
"loss": 1.6572,
"step": 63400
},
{
"epoch": 20.52359405300582,
"grad_norm": 1.2673180103302002,
"learning_rate": 0.001,
"loss": 1.665,
"step": 63500
},
{
"epoch": 20.555914673561734,
"grad_norm": 1.1285369396209717,
"learning_rate": 0.001,
"loss": 1.661,
"step": 63600
},
{
"epoch": 20.58823529411765,
"grad_norm": 1.2637361288070679,
"learning_rate": 0.001,
"loss": 1.67,
"step": 63700
},
{
"epoch": 20.620555914673563,
"grad_norm": 1.3248004913330078,
"learning_rate": 0.001,
"loss": 1.6913,
"step": 63800
},
{
"epoch": 20.652876535229478,
"grad_norm": 1.1528964042663574,
"learning_rate": 0.001,
"loss": 1.6517,
"step": 63900
},
{
"epoch": 20.685197155785392,
"grad_norm": 1.1508557796478271,
"learning_rate": 0.001,
"loss": 1.6593,
"step": 64000
},
{
"epoch": 20.717517776341307,
"grad_norm": 1.1110694408416748,
"learning_rate": 0.001,
"loss": 1.6933,
"step": 64100
},
{
"epoch": 20.74983839689722,
"grad_norm": 1.089123010635376,
"learning_rate": 0.001,
"loss": 1.6657,
"step": 64200
},
{
"epoch": 20.782159017453136,
"grad_norm": 1.448936939239502,
"learning_rate": 0.001,
"loss": 1.7123,
"step": 64300
},
{
"epoch": 20.81447963800905,
"grad_norm": 1.3776893615722656,
"learning_rate": 0.001,
"loss": 1.7107,
"step": 64400
},
{
"epoch": 20.846800258564965,
"grad_norm": 1.4579541683197021,
"learning_rate": 0.001,
"loss": 1.6942,
"step": 64500
},
{
"epoch": 20.87912087912088,
"grad_norm": 1.0205689668655396,
"learning_rate": 0.001,
"loss": 1.6835,
"step": 64600
},
{
"epoch": 20.911441499676794,
"grad_norm": 1.1773357391357422,
"learning_rate": 0.001,
"loss": 1.6896,
"step": 64700
},
{
"epoch": 20.94376212023271,
"grad_norm": 1.2993382215499878,
"learning_rate": 0.001,
"loss": 1.6996,
"step": 64800
},
{
"epoch": 20.976082740788623,
"grad_norm": 0.9848776459693909,
"learning_rate": 0.001,
"loss": 1.6891,
"step": 64900
},
{
"epoch": 21.008403361344538,
"grad_norm": 1.2597471475601196,
"learning_rate": 0.001,
"loss": 1.6802,
"step": 65000
},
{
"epoch": 21.040723981900452,
"grad_norm": 1.2813159227371216,
"learning_rate": 0.001,
"loss": 1.5808,
"step": 65100
},
{
"epoch": 21.073044602456367,
"grad_norm": 1.2395695447921753,
"learning_rate": 0.001,
"loss": 1.5823,
"step": 65200
},
{
"epoch": 21.10536522301228,
"grad_norm": 1.3502123355865479,
"learning_rate": 0.001,
"loss": 1.5888,
"step": 65300
},
{
"epoch": 21.137685843568196,
"grad_norm": 1.1473206281661987,
"learning_rate": 0.001,
"loss": 1.6125,
"step": 65400
},
{
"epoch": 21.17000646412411,
"grad_norm": 1.0330958366394043,
"learning_rate": 0.001,
"loss": 1.6036,
"step": 65500
},
{
"epoch": 21.202327084680025,
"grad_norm": 1.2344865798950195,
"learning_rate": 0.001,
"loss": 1.6114,
"step": 65600
},
{
"epoch": 21.23464770523594,
"grad_norm": 1.0328295230865479,
"learning_rate": 0.001,
"loss": 1.612,
"step": 65700
},
{
"epoch": 21.266968325791854,
"grad_norm": 1.1887953281402588,
"learning_rate": 0.001,
"loss": 1.6174,
"step": 65800
},
{
"epoch": 21.29928894634777,
"grad_norm": 3.9064321517944336,
"learning_rate": 0.001,
"loss": 1.625,
"step": 65900
},
{
"epoch": 21.331609566903683,
"grad_norm": 1.2309563159942627,
"learning_rate": 0.001,
"loss": 1.632,
"step": 66000
},
{
"epoch": 21.363930187459598,
"grad_norm": 1.387037992477417,
"learning_rate": 0.001,
"loss": 1.62,
"step": 66100
},
{
"epoch": 21.396250808015512,
"grad_norm": 1.321847677230835,
"learning_rate": 0.001,
"loss": 1.6048,
"step": 66200
},
{
"epoch": 21.428571428571427,
"grad_norm": 1.0067797899246216,
"learning_rate": 0.001,
"loss": 1.6096,
"step": 66300
},
{
"epoch": 21.46089204912734,
"grad_norm": 1.2617535591125488,
"learning_rate": 0.001,
"loss": 1.6441,
"step": 66400
},
{
"epoch": 21.49321266968326,
"grad_norm": 1.4177860021591187,
"learning_rate": 0.001,
"loss": 1.6144,
"step": 66500
},
{
"epoch": 21.525533290239174,
"grad_norm": 1.5710135698318481,
"learning_rate": 0.001,
"loss": 1.6221,
"step": 66600
},
{
"epoch": 21.55785391079509,
"grad_norm": 1.2180819511413574,
"learning_rate": 0.001,
"loss": 1.645,
"step": 66700
},
{
"epoch": 21.590174531351003,
"grad_norm": 1.226199984550476,
"learning_rate": 0.001,
"loss": 1.6302,
"step": 66800
},
{
"epoch": 21.622495151906918,
"grad_norm": 1.044456958770752,
"learning_rate": 0.001,
"loss": 1.6276,
"step": 66900
},
{
"epoch": 21.654815772462833,
"grad_norm": 1.2039384841918945,
"learning_rate": 0.001,
"loss": 1.6742,
"step": 67000
},
{
"epoch": 21.687136393018747,
"grad_norm": 1.0783025026321411,
"learning_rate": 0.001,
"loss": 1.6594,
"step": 67100
},
{
"epoch": 21.71945701357466,
"grad_norm": 1.4424864053726196,
"learning_rate": 0.001,
"loss": 1.6641,
"step": 67200
},
{
"epoch": 21.751777634130576,
"grad_norm": 1.9445148706436157,
"learning_rate": 0.001,
"loss": 1.6619,
"step": 67300
},
{
"epoch": 21.78409825468649,
"grad_norm": 1.016932487487793,
"learning_rate": 0.001,
"loss": 1.6451,
"step": 67400
},
{
"epoch": 21.816418875242405,
"grad_norm": 1.1570857763290405,
"learning_rate": 0.001,
"loss": 1.6723,
"step": 67500
},
{
"epoch": 21.84873949579832,
"grad_norm": 1.112880825996399,
"learning_rate": 0.001,
"loss": 1.6544,
"step": 67600
},
{
"epoch": 21.881060116354234,
"grad_norm": 1.3634724617004395,
"learning_rate": 0.001,
"loss": 1.6492,
"step": 67700
},
{
"epoch": 21.91338073691015,
"grad_norm": 1.1221792697906494,
"learning_rate": 0.001,
"loss": 1.686,
"step": 67800
},
{
"epoch": 21.945701357466064,
"grad_norm": 1.2058967351913452,
"learning_rate": 0.001,
"loss": 1.6735,
"step": 67900
},
{
"epoch": 21.978021978021978,
"grad_norm": 1.2786732912063599,
"learning_rate": 0.001,
"loss": 1.7036,
"step": 68000
},
{
"epoch": 22.010342598577893,
"grad_norm": 1.4422574043273926,
"learning_rate": 0.001,
"loss": 1.6345,
"step": 68100
},
{
"epoch": 22.042663219133807,
"grad_norm": 1.175736904144287,
"learning_rate": 0.001,
"loss": 1.5827,
"step": 68200
},
{
"epoch": 22.07498383968972,
"grad_norm": 1.1469517946243286,
"learning_rate": 0.001,
"loss": 1.571,
"step": 68300
},
{
"epoch": 22.107304460245636,
"grad_norm": 1.5781559944152832,
"learning_rate": 0.001,
"loss": 1.5555,
"step": 68400
},
{
"epoch": 22.13962508080155,
"grad_norm": 1.413362979888916,
"learning_rate": 0.001,
"loss": 1.5768,
"step": 68500
},
{
"epoch": 22.171945701357465,
"grad_norm": 1.2031275033950806,
"learning_rate": 0.001,
"loss": 1.5876,
"step": 68600
},
{
"epoch": 22.20426632191338,
"grad_norm": 1.3570070266723633,
"learning_rate": 0.001,
"loss": 1.597,
"step": 68700
},
{
"epoch": 22.236586942469295,
"grad_norm": 1.4896953105926514,
"learning_rate": 0.001,
"loss": 1.5844,
"step": 68800
},
{
"epoch": 22.26890756302521,
"grad_norm": 1.5189759731292725,
"learning_rate": 0.001,
"loss": 1.5878,
"step": 68900
},
{
"epoch": 22.301228183581124,
"grad_norm": 1.5635501146316528,
"learning_rate": 0.001,
"loss": 1.6088,
"step": 69000
},
{
"epoch": 22.33354880413704,
"grad_norm": 1.3236138820648193,
"learning_rate": 0.001,
"loss": 1.5919,
"step": 69100
},
{
"epoch": 22.365869424692953,
"grad_norm": 1.1415072679519653,
"learning_rate": 0.001,
"loss": 1.6074,
"step": 69200
},
{
"epoch": 22.398190045248867,
"grad_norm": 1.212083101272583,
"learning_rate": 0.001,
"loss": 1.5884,
"step": 69300
},
{
"epoch": 22.430510665804782,
"grad_norm": 1.3305213451385498,
"learning_rate": 0.001,
"loss": 1.5937,
"step": 69400
},
{
"epoch": 22.462831286360696,
"grad_norm": 1.476341724395752,
"learning_rate": 0.001,
"loss": 1.5918,
"step": 69500
},
{
"epoch": 22.49515190691661,
"grad_norm": 1.1090694665908813,
"learning_rate": 0.001,
"loss": 1.589,
"step": 69600
},
{
"epoch": 22.52747252747253,
"grad_norm": 1.5420668125152588,
"learning_rate": 0.001,
"loss": 1.616,
"step": 69700
},
{
"epoch": 22.559793148028444,
"grad_norm": 1.281031847000122,
"learning_rate": 0.001,
"loss": 1.6115,
"step": 69800
},
{
"epoch": 22.59211376858436,
"grad_norm": 1.1826776266098022,
"learning_rate": 0.001,
"loss": 1.6089,
"step": 69900
},
{
"epoch": 22.624434389140273,
"grad_norm": 1.137690544128418,
"learning_rate": 0.001,
"loss": 1.6299,
"step": 70000
},
{
"epoch": 22.656755009696187,
"grad_norm": 1.1811046600341797,
"learning_rate": 0.001,
"loss": 1.6457,
"step": 70100
},
{
"epoch": 22.689075630252102,
"grad_norm": 1.142443060874939,
"learning_rate": 0.001,
"loss": 1.6284,
"step": 70200
},
{
"epoch": 22.721396250808017,
"grad_norm": 1.719937801361084,
"learning_rate": 0.001,
"loss": 1.6254,
"step": 70300
},
{
"epoch": 22.75371687136393,
"grad_norm": 1.6573597192764282,
"learning_rate": 0.001,
"loss": 1.6294,
"step": 70400
},
{
"epoch": 22.786037491919846,
"grad_norm": 1.1464518308639526,
"learning_rate": 0.001,
"loss": 1.6272,
"step": 70500
},
{
"epoch": 22.81835811247576,
"grad_norm": 1.2429357767105103,
"learning_rate": 0.001,
"loss": 1.6429,
"step": 70600
},
{
"epoch": 22.850678733031675,
"grad_norm": 1.3033969402313232,
"learning_rate": 0.001,
"loss": 1.6559,
"step": 70700
},
{
"epoch": 22.88299935358759,
"grad_norm": 1.6557401418685913,
"learning_rate": 0.001,
"loss": 1.6476,
"step": 70800
},
{
"epoch": 22.915319974143504,
"grad_norm": 1.163193702697754,
"learning_rate": 0.001,
"loss": 1.6383,
"step": 70900
},
{
"epoch": 22.94764059469942,
"grad_norm": 1.5962588787078857,
"learning_rate": 0.001,
"loss": 1.6528,
"step": 71000
},
{
"epoch": 22.979961215255333,
"grad_norm": 1.1483979225158691,
"learning_rate": 0.001,
"loss": 1.6624,
"step": 71100
},
{
"epoch": 23.012281835811248,
"grad_norm": 1.3430731296539307,
"learning_rate": 0.001,
"loss": 1.5737,
"step": 71200
},
{
"epoch": 23.044602456367162,
"grad_norm": 1.3681293725967407,
"learning_rate": 0.001,
"loss": 1.5217,
"step": 71300
},
{
"epoch": 23.076923076923077,
"grad_norm": 1.325210690498352,
"learning_rate": 0.001,
"loss": 1.5488,
"step": 71400
},
{
"epoch": 23.10924369747899,
"grad_norm": 1.750307559967041,
"learning_rate": 0.001,
"loss": 1.5457,
"step": 71500
},
{
"epoch": 23.141564318034906,
"grad_norm": 1.3093291521072388,
"learning_rate": 0.001,
"loss": 1.5522,
"step": 71600
},
{
"epoch": 23.17388493859082,
"grad_norm": 1.2910124063491821,
"learning_rate": 0.001,
"loss": 1.5526,
"step": 71700
},
{
"epoch": 23.206205559146735,
"grad_norm": 1.2840170860290527,
"learning_rate": 0.001,
"loss": 1.5682,
"step": 71800
},
{
"epoch": 23.23852617970265,
"grad_norm": 1.2806519269943237,
"learning_rate": 0.001,
"loss": 1.5885,
"step": 71900
},
{
"epoch": 23.270846800258564,
"grad_norm": 1.2111592292785645,
"learning_rate": 0.001,
"loss": 1.5734,
"step": 72000
},
{
"epoch": 23.30316742081448,
"grad_norm": 1.3556594848632812,
"learning_rate": 0.001,
"loss": 1.5792,
"step": 72100
},
{
"epoch": 23.335488041370393,
"grad_norm": 1.2785731554031372,
"learning_rate": 0.001,
"loss": 1.5682,
"step": 72200
},
{
"epoch": 23.367808661926308,
"grad_norm": 1.3190655708312988,
"learning_rate": 0.001,
"loss": 1.5695,
"step": 72300
},
{
"epoch": 23.400129282482222,
"grad_norm": 1.1459965705871582,
"learning_rate": 0.001,
"loss": 1.5899,
"step": 72400
},
{
"epoch": 23.432449903038137,
"grad_norm": 1.3239068984985352,
"learning_rate": 0.001,
"loss": 1.5818,
"step": 72500
},
{
"epoch": 23.46477052359405,
"grad_norm": 1.6447606086730957,
"learning_rate": 0.001,
"loss": 1.5798,
"step": 72600
},
{
"epoch": 23.49709114414997,
"grad_norm": 1.2553894519805908,
"learning_rate": 0.001,
"loss": 1.5897,
"step": 72700
},
{
"epoch": 23.529411764705884,
"grad_norm": 1.5192135572433472,
"learning_rate": 0.001,
"loss": 1.5671,
"step": 72800
},
{
"epoch": 23.5617323852618,
"grad_norm": 1.1816911697387695,
"learning_rate": 0.001,
"loss": 1.5876,
"step": 72900
},
{
"epoch": 23.594053005817713,
"grad_norm": 1.2005702257156372,
"learning_rate": 0.001,
"loss": 1.5963,
"step": 73000
},
{
"epoch": 23.626373626373628,
"grad_norm": 1.5252586603164673,
"learning_rate": 0.001,
"loss": 1.5819,
"step": 73100
},
{
"epoch": 23.658694246929542,
"grad_norm": 1.3027052879333496,
"learning_rate": 0.001,
"loss": 1.5909,
"step": 73200
},
{
"epoch": 23.691014867485457,
"grad_norm": 1.3421530723571777,
"learning_rate": 0.001,
"loss": 1.602,
"step": 73300
},
{
"epoch": 23.72333548804137,
"grad_norm": 1.4905993938446045,
"learning_rate": 0.001,
"loss": 1.625,
"step": 73400
},
{
"epoch": 23.755656108597286,
"grad_norm": 1.4335204362869263,
"learning_rate": 0.001,
"loss": 1.5941,
"step": 73500
},
{
"epoch": 23.7879767291532,
"grad_norm": 1.2546992301940918,
"learning_rate": 0.001,
"loss": 1.6238,
"step": 73600
},
{
"epoch": 23.820297349709115,
"grad_norm": 1.2537373304367065,
"learning_rate": 0.001,
"loss": 1.6303,
"step": 73700
},
{
"epoch": 23.85261797026503,
"grad_norm": 2.2106618881225586,
"learning_rate": 0.001,
"loss": 1.6129,
"step": 73800
},
{
"epoch": 23.884938590820944,
"grad_norm": 1.3618272542953491,
"learning_rate": 0.001,
"loss": 1.6093,
"step": 73900
},
{
"epoch": 23.91725921137686,
"grad_norm": 1.2944997549057007,
"learning_rate": 0.001,
"loss": 1.6103,
"step": 74000
},
{
"epoch": 23.949579831932773,
"grad_norm": 1.2308852672576904,
"learning_rate": 0.001,
"loss": 1.6456,
"step": 74100
},
{
"epoch": 23.981900452488688,
"grad_norm": 1.3738027811050415,
"learning_rate": 0.001,
"loss": 1.6603,
"step": 74200
},
{
"epoch": 24.014221073044602,
"grad_norm": 2.0176162719726562,
"learning_rate": 0.001,
"loss": 1.5526,
"step": 74300
},
{
"epoch": 24.046541693600517,
"grad_norm": 1.2242016792297363,
"learning_rate": 0.001,
"loss": 1.4932,
"step": 74400
},
{
"epoch": 24.07886231415643,
"grad_norm": 1.212552785873413,
"learning_rate": 0.001,
"loss": 1.5145,
"step": 74500
},
{
"epoch": 24.111182934712346,
"grad_norm": 1.89344322681427,
"learning_rate": 0.001,
"loss": 1.5289,
"step": 74600
},
{
"epoch": 24.14350355526826,
"grad_norm": 1.278113603591919,
"learning_rate": 0.001,
"loss": 1.5372,
"step": 74700
},
{
"epoch": 24.175824175824175,
"grad_norm": 1.6612149477005005,
"learning_rate": 0.001,
"loss": 1.5326,
"step": 74800
},
{
"epoch": 24.20814479638009,
"grad_norm": 1.2389644384384155,
"learning_rate": 0.001,
"loss": 1.5162,
"step": 74900
},
{
"epoch": 24.240465416936004,
"grad_norm": 1.5246014595031738,
"learning_rate": 0.001,
"loss": 1.5539,
"step": 75000
},
{
"epoch": 24.27278603749192,
"grad_norm": 1.496571660041809,
"learning_rate": 0.001,
"loss": 1.5298,
"step": 75100
},
{
"epoch": 24.305106658047833,
"grad_norm": 1.1371721029281616,
"learning_rate": 0.001,
"loss": 1.5627,
"step": 75200
},
{
"epoch": 24.337427278603748,
"grad_norm": 1.488763689994812,
"learning_rate": 0.001,
"loss": 1.5464,
"step": 75300
},
{
"epoch": 24.369747899159663,
"grad_norm": 2.095946788787842,
"learning_rate": 0.001,
"loss": 1.5717,
"step": 75400
},
{
"epoch": 24.402068519715577,
"grad_norm": 1.4641380310058594,
"learning_rate": 0.001,
"loss": 1.5728,
"step": 75500
},
{
"epoch": 24.43438914027149,
"grad_norm": 1.5061386823654175,
"learning_rate": 0.001,
"loss": 1.5399,
"step": 75600
},
{
"epoch": 24.466709760827406,
"grad_norm": 1.5154993534088135,
"learning_rate": 0.001,
"loss": 1.5743,
"step": 75700
},
{
"epoch": 24.49903038138332,
"grad_norm": 1.573975682258606,
"learning_rate": 0.001,
"loss": 1.5653,
"step": 75800
},
{
"epoch": 24.53135100193924,
"grad_norm": 1.5374081134796143,
"learning_rate": 0.001,
"loss": 1.572,
"step": 75900
},
{
"epoch": 24.563671622495153,
"grad_norm": 1.331465482711792,
"learning_rate": 0.001,
"loss": 1.5671,
"step": 76000
},
{
"epoch": 24.595992243051068,
"grad_norm": 1.6944043636322021,
"learning_rate": 0.001,
"loss": 1.569,
"step": 76100
},
{
"epoch": 24.628312863606983,
"grad_norm": 1.5414694547653198,
"learning_rate": 0.001,
"loss": 1.5945,
"step": 76200
},
{
"epoch": 24.660633484162897,
"grad_norm": 1.74053156375885,
"learning_rate": 0.001,
"loss": 1.5894,
"step": 76300
},
{
"epoch": 24.69295410471881,
"grad_norm": 1.4306658506393433,
"learning_rate": 0.001,
"loss": 1.5921,
"step": 76400
},
{
"epoch": 24.725274725274726,
"grad_norm": 1.250169038772583,
"learning_rate": 0.001,
"loss": 1.5789,
"step": 76500
},
{
"epoch": 24.75759534583064,
"grad_norm": 1.2672016620635986,
"learning_rate": 0.001,
"loss": 1.5937,
"step": 76600
},
{
"epoch": 24.789915966386555,
"grad_norm": 1.3557816743850708,
"learning_rate": 0.001,
"loss": 1.592,
"step": 76700
},
{
"epoch": 24.82223658694247,
"grad_norm": 1.330208420753479,
"learning_rate": 0.001,
"loss": 1.5801,
"step": 76800
},
{
"epoch": 24.854557207498384,
"grad_norm": 1.3479870557785034,
"learning_rate": 0.001,
"loss": 1.5849,
"step": 76900
},
{
"epoch": 24.8868778280543,
"grad_norm": 1.4242347478866577,
"learning_rate": 0.001,
"loss": 1.5957,
"step": 77000
},
{
"epoch": 24.919198448610214,
"grad_norm": 1.527997374534607,
"learning_rate": 0.001,
"loss": 1.6222,
"step": 77100
},
{
"epoch": 24.951519069166128,
"grad_norm": 1.1563555002212524,
"learning_rate": 0.001,
"loss": 1.6155,
"step": 77200
},
{
"epoch": 24.983839689722043,
"grad_norm": 1.1758341789245605,
"learning_rate": 0.001,
"loss": 1.6147,
"step": 77300
},
{
"epoch": 25.016160310277957,
"grad_norm": 0.742283284664154,
"learning_rate": 0.001,
"loss": 1.4905,
"step": 77400
},
{
"epoch": 25.048480930833872,
"grad_norm": 0.9083603620529175,
"learning_rate": 0.001,
"loss": 1.4952,
"step": 77500
},
{
"epoch": 25.080801551389786,
"grad_norm": 0.7578412294387817,
"learning_rate": 0.001,
"loss": 1.4976,
"step": 77600
},
{
"epoch": 25.1131221719457,
"grad_norm": 0.7569114565849304,
"learning_rate": 0.001,
"loss": 1.5075,
"step": 77700
},
{
"epoch": 25.145442792501616,
"grad_norm": 0.7160850763320923,
"learning_rate": 0.001,
"loss": 1.524,
"step": 77800
},
{
"epoch": 25.17776341305753,
"grad_norm": 0.8957599997520447,
"learning_rate": 0.001,
"loss": 1.5142,
"step": 77900
},
{
"epoch": 25.210084033613445,
"grad_norm": 0.9639133810997009,
"learning_rate": 0.001,
"loss": 1.4995,
"step": 78000
},
{
"epoch": 25.24240465416936,
"grad_norm": 0.7975667119026184,
"learning_rate": 0.001,
"loss": 1.5078,
"step": 78100
},
{
"epoch": 25.274725274725274,
"grad_norm": 0.6604834794998169,
"learning_rate": 0.001,
"loss": 1.5515,
"step": 78200
},
{
"epoch": 25.30704589528119,
"grad_norm": 1.7004810571670532,
"learning_rate": 0.001,
"loss": 1.5241,
"step": 78300
},
{
"epoch": 25.339366515837103,
"grad_norm": 0.6456866264343262,
"learning_rate": 0.001,
"loss": 1.5068,
"step": 78400
},
{
"epoch": 25.371687136393017,
"grad_norm": 0.7533178329467773,
"learning_rate": 0.001,
"loss": 1.5126,
"step": 78500
},
{
"epoch": 25.404007756948932,
"grad_norm": 1.254917860031128,
"learning_rate": 0.001,
"loss": 1.5386,
"step": 78600
},
{
"epoch": 25.436328377504847,
"grad_norm": 0.8305799961090088,
"learning_rate": 0.001,
"loss": 1.5495,
"step": 78700
},
{
"epoch": 25.46864899806076,
"grad_norm": 0.5019985437393188,
"learning_rate": 0.001,
"loss": 1.5522,
"step": 78800
},
{
"epoch": 25.50096961861668,
"grad_norm": 1.266822338104248,
"learning_rate": 0.001,
"loss": 1.5586,
"step": 78900
},
{
"epoch": 25.533290239172594,
"grad_norm": 1.1242650747299194,
"learning_rate": 0.001,
"loss": 1.548,
"step": 79000
},
{
"epoch": 25.56561085972851,
"grad_norm": 1.264953374862671,
"learning_rate": 0.001,
"loss": 1.5599,
"step": 79100
},
{
"epoch": 25.597931480284423,
"grad_norm": 0.8981252908706665,
"learning_rate": 0.001,
"loss": 1.5602,
"step": 79200
},
{
"epoch": 25.630252100840337,
"grad_norm": 0.9501145482063293,
"learning_rate": 0.001,
"loss": 1.5501,
"step": 79300
},
{
"epoch": 25.662572721396252,
"grad_norm": 0.9055042266845703,
"learning_rate": 0.001,
"loss": 1.5692,
"step": 79400
},
{
"epoch": 25.694893341952167,
"grad_norm": 0.6795365810394287,
"learning_rate": 0.001,
"loss": 1.5801,
"step": 79500
},
{
"epoch": 25.72721396250808,
"grad_norm": 0.8403079509735107,
"learning_rate": 0.001,
"loss": 1.574,
"step": 79600
},
{
"epoch": 25.759534583063996,
"grad_norm": 0.8397261500358582,
"learning_rate": 0.001,
"loss": 1.5633,
"step": 79700
},
{
"epoch": 25.79185520361991,
"grad_norm": 1.120841383934021,
"learning_rate": 0.001,
"loss": 1.5868,
"step": 79800
},
{
"epoch": 25.824175824175825,
"grad_norm": 0.8841111660003662,
"learning_rate": 0.001,
"loss": 1.5438,
"step": 79900
},
{
"epoch": 25.85649644473174,
"grad_norm": 1.7660539150238037,
"learning_rate": 0.001,
"loss": 1.5841,
"step": 80000
},
{
"epoch": 25.888817065287654,
"grad_norm": 0.6177299618721008,
"learning_rate": 0.001,
"loss": 1.5619,
"step": 80100
},
{
"epoch": 25.92113768584357,
"grad_norm": 0.9627276659011841,
"learning_rate": 0.001,
"loss": 1.5964,
"step": 80200
},
{
"epoch": 25.953458306399483,
"grad_norm": 1.162492036819458,
"learning_rate": 0.001,
"loss": 1.5836,
"step": 80300
},
{
"epoch": 25.985778926955398,
"grad_norm": 0.6118494868278503,
"learning_rate": 0.001,
"loss": 1.5949,
"step": 80400
},
{
"epoch": 26.018099547511312,
"grad_norm": 1.2347427606582642,
"learning_rate": 0.001,
"loss": 1.5292,
"step": 80500
},
{
"epoch": 26.050420168067227,
"grad_norm": 1.450631856918335,
"learning_rate": 0.001,
"loss": 1.4459,
"step": 80600
},
{
"epoch": 26.08274078862314,
"grad_norm": 1.2779650688171387,
"learning_rate": 0.001,
"loss": 1.4594,
"step": 80700
},
{
"epoch": 26.115061409179056,
"grad_norm": 1.435947299003601,
"learning_rate": 0.001,
"loss": 1.4688,
"step": 80800
},
{
"epoch": 26.14738202973497,
"grad_norm": 1.2987701892852783,
"learning_rate": 0.001,
"loss": 1.5039,
"step": 80900
},
{
"epoch": 26.179702650290885,
"grad_norm": 1.240431547164917,
"learning_rate": 0.001,
"loss": 1.5013,
"step": 81000
},
{
"epoch": 26.2120232708468,
"grad_norm": 1.5258148908615112,
"learning_rate": 0.001,
"loss": 1.4991,
"step": 81100
},
{
"epoch": 26.244343891402714,
"grad_norm": 1.3428306579589844,
"learning_rate": 0.001,
"loss": 1.504,
"step": 81200
},
{
"epoch": 26.27666451195863,
"grad_norm": 1.8016172647476196,
"learning_rate": 0.001,
"loss": 1.5192,
"step": 81300
},
{
"epoch": 26.308985132514543,
"grad_norm": 1.2530312538146973,
"learning_rate": 0.001,
"loss": 1.5187,
"step": 81400
},
{
"epoch": 26.341305753070458,
"grad_norm": 1.2320021390914917,
"learning_rate": 0.001,
"loss": 1.5256,
"step": 81500
},
{
"epoch": 26.373626373626372,
"grad_norm": 1.5778974294662476,
"learning_rate": 0.001,
"loss": 1.503,
"step": 81600
},
{
"epoch": 26.405946994182287,
"grad_norm": 1.6406939029693604,
"learning_rate": 0.001,
"loss": 1.5297,
"step": 81700
},
{
"epoch": 26.4382676147382,
"grad_norm": 1.3841694593429565,
"learning_rate": 0.001,
"loss": 1.5295,
"step": 81800
},
{
"epoch": 26.470588235294116,
"grad_norm": 1.5115548372268677,
"learning_rate": 0.001,
"loss": 1.5216,
"step": 81900
},
{
"epoch": 26.50290885585003,
"grad_norm": 1.463423252105713,
"learning_rate": 0.001,
"loss": 1.522,
"step": 82000
},
{
"epoch": 26.53522947640595,
"grad_norm": 1.2583633661270142,
"learning_rate": 0.001,
"loss": 1.5367,
"step": 82100
},
{
"epoch": 26.567550096961863,
"grad_norm": 1.482848048210144,
"learning_rate": 0.001,
"loss": 1.5252,
"step": 82200
},
{
"epoch": 26.599870717517778,
"grad_norm": 1.3020163774490356,
"learning_rate": 0.001,
"loss": 1.5197,
"step": 82300
},
{
"epoch": 26.632191338073692,
"grad_norm": 1.1481016874313354,
"learning_rate": 0.001,
"loss": 1.5356,
"step": 82400
},
{
"epoch": 26.664511958629607,
"grad_norm": 1.3201098442077637,
"learning_rate": 0.001,
"loss": 1.5355,
"step": 82500
},
{
"epoch": 26.69683257918552,
"grad_norm": 1.2914501428604126,
"learning_rate": 0.001,
"loss": 1.5497,
"step": 82600
},
{
"epoch": 26.729153199741436,
"grad_norm": 1.4507853984832764,
"learning_rate": 0.001,
"loss": 1.5417,
"step": 82700
},
{
"epoch": 26.76147382029735,
"grad_norm": 1.4626699686050415,
"learning_rate": 0.001,
"loss": 1.5646,
"step": 82800
},
{
"epoch": 26.793794440853265,
"grad_norm": 1.5812556743621826,
"learning_rate": 0.001,
"loss": 1.5514,
"step": 82900
},
{
"epoch": 26.82611506140918,
"grad_norm": 1.2999969720840454,
"learning_rate": 0.001,
"loss": 1.5597,
"step": 83000
},
{
"epoch": 26.858435681965094,
"grad_norm": 1.7724205255508423,
"learning_rate": 0.001,
"loss": 1.5526,
"step": 83100
},
{
"epoch": 26.89075630252101,
"grad_norm": 1.5374313592910767,
"learning_rate": 0.001,
"loss": 1.5583,
"step": 83200
},
{
"epoch": 26.923076923076923,
"grad_norm": 1.34683358669281,
"learning_rate": 0.001,
"loss": 1.5715,
"step": 83300
},
{
"epoch": 26.955397543632838,
"grad_norm": 1.9225049018859863,
"learning_rate": 0.001,
"loss": 1.5586,
"step": 83400
},
{
"epoch": 26.987718164188752,
"grad_norm": 1.2125778198242188,
"learning_rate": 0.001,
"loss": 1.5621,
"step": 83500
},
{
"epoch": 27.020038784744667,
"grad_norm": 1.5823465585708618,
"learning_rate": 0.001,
"loss": 1.5058,
"step": 83600
},
{
"epoch": 27.05235940530058,
"grad_norm": 1.2081438302993774,
"learning_rate": 0.001,
"loss": 1.4497,
"step": 83700
},
{
"epoch": 27.084680025856496,
"grad_norm": 1.27534019947052,
"learning_rate": 0.001,
"loss": 1.4587,
"step": 83800
},
{
"epoch": 27.11700064641241,
"grad_norm": 1.159661054611206,
"learning_rate": 0.001,
"loss": 1.4757,
"step": 83900
},
{
"epoch": 27.149321266968325,
"grad_norm": 1.3436905145645142,
"learning_rate": 0.001,
"loss": 1.4684,
"step": 84000
},
{
"epoch": 27.18164188752424,
"grad_norm": 1.2977607250213623,
"learning_rate": 0.001,
"loss": 1.5002,
"step": 84100
},
{
"epoch": 27.213962508080154,
"grad_norm": 1.5422593355178833,
"learning_rate": 0.001,
"loss": 1.4569,
"step": 84200
},
{
"epoch": 27.24628312863607,
"grad_norm": 1.330285906791687,
"learning_rate": 0.001,
"loss": 1.478,
"step": 84300
},
{
"epoch": 27.278603749191983,
"grad_norm": 1.6561344861984253,
"learning_rate": 0.001,
"loss": 1.4716,
"step": 84400
},
{
"epoch": 27.310924369747898,
"grad_norm": 1.6572483777999878,
"learning_rate": 0.001,
"loss": 1.4789,
"step": 84500
},
{
"epoch": 27.343244990303813,
"grad_norm": 1.295915961265564,
"learning_rate": 0.001,
"loss": 1.5091,
"step": 84600
},
{
"epoch": 27.375565610859727,
"grad_norm": 1.2116016149520874,
"learning_rate": 0.001,
"loss": 1.503,
"step": 84700
},
{
"epoch": 27.40788623141564,
"grad_norm": 1.413427472114563,
"learning_rate": 0.001,
"loss": 1.4932,
"step": 84800
},
{
"epoch": 27.440206851971556,
"grad_norm": 1.242362380027771,
"learning_rate": 0.001,
"loss": 1.5037,
"step": 84900
},
{
"epoch": 27.47252747252747,
"grad_norm": 1.4885907173156738,
"learning_rate": 0.001,
"loss": 1.4834,
"step": 85000
},
{
"epoch": 27.50484809308339,
"grad_norm": 1.2446436882019043,
"learning_rate": 0.001,
"loss": 1.5187,
"step": 85100
},
{
"epoch": 27.537168713639304,
"grad_norm": 1.154815435409546,
"learning_rate": 0.001,
"loss": 1.5187,
"step": 85200
},
{
"epoch": 27.569489334195218,
"grad_norm": 1.386157751083374,
"learning_rate": 0.001,
"loss": 1.5252,
"step": 85300
},
{
"epoch": 27.601809954751133,
"grad_norm": 1.07121741771698,
"learning_rate": 0.001,
"loss": 1.5063,
"step": 85400
},
{
"epoch": 27.634130575307047,
"grad_norm": 1.6424086093902588,
"learning_rate": 0.001,
"loss": 1.5084,
"step": 85500
},
{
"epoch": 27.66645119586296,
"grad_norm": 1.3184316158294678,
"learning_rate": 0.001,
"loss": 1.5178,
"step": 85600
},
{
"epoch": 27.698771816418876,
"grad_norm": 1.3914283514022827,
"learning_rate": 0.001,
"loss": 1.5375,
"step": 85700
},
{
"epoch": 27.73109243697479,
"grad_norm": 1.3977998495101929,
"learning_rate": 0.001,
"loss": 1.5232,
"step": 85800
},
{
"epoch": 27.763413057530705,
"grad_norm": 1.423851728439331,
"learning_rate": 0.001,
"loss": 1.5201,
"step": 85900
},
{
"epoch": 27.79573367808662,
"grad_norm": 1.258535623550415,
"learning_rate": 0.001,
"loss": 1.5244,
"step": 86000
},
{
"epoch": 27.828054298642535,
"grad_norm": 1.1898850202560425,
"learning_rate": 0.001,
"loss": 1.524,
"step": 86100
},
{
"epoch": 27.86037491919845,
"grad_norm": 1.3646154403686523,
"learning_rate": 0.001,
"loss": 1.5416,
"step": 86200
},
{
"epoch": 27.892695539754364,
"grad_norm": 1.1520329713821411,
"learning_rate": 0.001,
"loss": 1.5248,
"step": 86300
},
{
"epoch": 27.92501616031028,
"grad_norm": 1.2999958992004395,
"learning_rate": 0.001,
"loss": 1.5227,
"step": 86400
},
{
"epoch": 27.957336780866193,
"grad_norm": 1.1944336891174316,
"learning_rate": 0.001,
"loss": 1.5457,
"step": 86500
},
{
"epoch": 27.989657401422107,
"grad_norm": 1.6716809272766113,
"learning_rate": 0.001,
"loss": 1.5364,
"step": 86600
},
{
"epoch": 28.021978021978022,
"grad_norm": 1.3939216136932373,
"learning_rate": 0.001,
"loss": 1.4927,
"step": 86700
},
{
"epoch": 28.054298642533936,
"grad_norm": 1.2728296518325806,
"learning_rate": 0.001,
"loss": 1.4457,
"step": 86800
},
{
"epoch": 28.08661926308985,
"grad_norm": 1.3491313457489014,
"learning_rate": 0.001,
"loss": 1.4111,
"step": 86900
},
{
"epoch": 28.118939883645766,
"grad_norm": 1.2990353107452393,
"learning_rate": 0.001,
"loss": 1.45,
"step": 87000
},
{
"epoch": 28.15126050420168,
"grad_norm": 1.213227391242981,
"learning_rate": 0.001,
"loss": 1.4374,
"step": 87100
},
{
"epoch": 28.183581124757595,
"grad_norm": 1.2896127700805664,
"learning_rate": 0.001,
"loss": 1.4648,
"step": 87200
},
{
"epoch": 28.21590174531351,
"grad_norm": 1.5600861310958862,
"learning_rate": 0.001,
"loss": 1.4622,
"step": 87300
},
{
"epoch": 28.248222365869424,
"grad_norm": 1.3157583475112915,
"learning_rate": 0.001,
"loss": 1.4648,
"step": 87400
},
{
"epoch": 28.28054298642534,
"grad_norm": 1.2662124633789062,
"learning_rate": 0.001,
"loss": 1.4799,
"step": 87500
},
{
"epoch": 28.312863606981253,
"grad_norm": 1.4469647407531738,
"learning_rate": 0.001,
"loss": 1.4426,
"step": 87600
},
{
"epoch": 28.345184227537167,
"grad_norm": 1.5201246738433838,
"learning_rate": 0.001,
"loss": 1.4455,
"step": 87700
},
{
"epoch": 28.377504848093082,
"grad_norm": 1.735161542892456,
"learning_rate": 0.001,
"loss": 1.4936,
"step": 87800
},
{
"epoch": 28.409825468648997,
"grad_norm": 1.433800458908081,
"learning_rate": 0.001,
"loss": 1.4776,
"step": 87900
},
{
"epoch": 28.44214608920491,
"grad_norm": 1.529483675956726,
"learning_rate": 0.001,
"loss": 1.4762,
"step": 88000
},
{
"epoch": 28.474466709760826,
"grad_norm": 1.7155773639678955,
"learning_rate": 0.001,
"loss": 1.4661,
"step": 88100
},
{
"epoch": 28.50678733031674,
"grad_norm": 1.5531314611434937,
"learning_rate": 0.001,
"loss": 1.4836,
"step": 88200
},
{
"epoch": 28.53910795087266,
"grad_norm": 1.2565120458602905,
"learning_rate": 0.001,
"loss": 1.4938,
"step": 88300
},
{
"epoch": 28.571428571428573,
"grad_norm": 1.0435504913330078,
"learning_rate": 0.001,
"loss": 1.4919,
"step": 88400
},
{
"epoch": 28.603749191984488,
"grad_norm": 1.5444915294647217,
"learning_rate": 0.001,
"loss": 1.5075,
"step": 88500
},
{
"epoch": 28.636069812540402,
"grad_norm": 2.226731300354004,
"learning_rate": 0.001,
"loss": 1.5065,
"step": 88600
},
{
"epoch": 28.668390433096317,
"grad_norm": 1.4304759502410889,
"learning_rate": 0.001,
"loss": 1.506,
"step": 88700
},
{
"epoch": 28.70071105365223,
"grad_norm": 1.0051342248916626,
"learning_rate": 0.001,
"loss": 1.5066,
"step": 88800
},
{
"epoch": 28.733031674208146,
"grad_norm": 1.3363932371139526,
"learning_rate": 0.001,
"loss": 1.5029,
"step": 88900
},
{
"epoch": 28.76535229476406,
"grad_norm": 4.342236042022705,
"learning_rate": 0.001,
"loss": 1.5065,
"step": 89000
},
{
"epoch": 28.797672915319975,
"grad_norm": 1.1559197902679443,
"learning_rate": 0.001,
"loss": 1.5142,
"step": 89100
},
{
"epoch": 28.82999353587589,
"grad_norm": 1.5334970951080322,
"learning_rate": 0.001,
"loss": 1.5089,
"step": 89200
},
{
"epoch": 28.862314156431804,
"grad_norm": 1.2204309701919556,
"learning_rate": 0.001,
"loss": 1.5117,
"step": 89300
},
{
"epoch": 28.89463477698772,
"grad_norm": 1.5689836740493774,
"learning_rate": 0.001,
"loss": 1.5135,
"step": 89400
},
{
"epoch": 28.926955397543633,
"grad_norm": 1.6976715326309204,
"learning_rate": 0.001,
"loss": 1.5148,
"step": 89500
},
{
"epoch": 28.959276018099548,
"grad_norm": 1.4873236417770386,
"learning_rate": 0.001,
"loss": 1.5158,
"step": 89600
},
{
"epoch": 28.991596638655462,
"grad_norm": 1.242458701133728,
"learning_rate": 0.001,
"loss": 1.5225,
"step": 89700
},
{
"epoch": 29.023917259211377,
"grad_norm": 1.1908669471740723,
"learning_rate": 0.001,
"loss": 1.4639,
"step": 89800
},
{
"epoch": 29.05623787976729,
"grad_norm": 1.4540772438049316,
"learning_rate": 0.001,
"loss": 1.4076,
"step": 89900
},
{
"epoch": 29.088558500323206,
"grad_norm": 1.3074499368667603,
"learning_rate": 0.001,
"loss": 1.4015,
"step": 90000
},
{
"epoch": 29.12087912087912,
"grad_norm": 2.1484622955322266,
"learning_rate": 0.001,
"loss": 1.425,
"step": 90100
},
{
"epoch": 29.153199741435035,
"grad_norm": 1.9532856941223145,
"learning_rate": 0.001,
"loss": 1.4144,
"step": 90200
},
{
"epoch": 29.18552036199095,
"grad_norm": 1.4356775283813477,
"learning_rate": 0.001,
"loss": 1.4342,
"step": 90300
},
{
"epoch": 29.217840982546864,
"grad_norm": 1.7443933486938477,
"learning_rate": 0.001,
"loss": 1.431,
"step": 90400
},
{
"epoch": 29.25016160310278,
"grad_norm": 1.4974429607391357,
"learning_rate": 0.001,
"loss": 1.424,
"step": 90500
},
{
"epoch": 29.282482223658693,
"grad_norm": 1.5972236394882202,
"learning_rate": 0.001,
"loss": 1.4474,
"step": 90600
},
{
"epoch": 29.314802844214608,
"grad_norm": 1.5421713590621948,
"learning_rate": 0.001,
"loss": 1.4521,
"step": 90700
},
{
"epoch": 29.347123464770522,
"grad_norm": 1.3467155694961548,
"learning_rate": 0.001,
"loss": 1.4596,
"step": 90800
},
{
"epoch": 29.379444085326437,
"grad_norm": 1.146463394165039,
"learning_rate": 0.001,
"loss": 1.4456,
"step": 90900
},
{
"epoch": 29.41176470588235,
"grad_norm": 1.3262990713119507,
"learning_rate": 0.001,
"loss": 1.4791,
"step": 91000
},
{
"epoch": 29.444085326438266,
"grad_norm": 1.5677353143692017,
"learning_rate": 0.001,
"loss": 1.4612,
"step": 91100
},
{
"epoch": 29.47640594699418,
"grad_norm": 1.2217344045639038,
"learning_rate": 0.001,
"loss": 1.4476,
"step": 91200
},
{
"epoch": 29.5087265675501,
"grad_norm": 1.3716522455215454,
"learning_rate": 0.001,
"loss": 1.4687,
"step": 91300
},
{
"epoch": 29.541047188106013,
"grad_norm": 1.5650326013565063,
"learning_rate": 0.001,
"loss": 1.4932,
"step": 91400
},
{
"epoch": 29.573367808661928,
"grad_norm": 1.4391878843307495,
"learning_rate": 0.001,
"loss": 1.4665,
"step": 91500
},
{
"epoch": 29.605688429217842,
"grad_norm": 1.2627538442611694,
"learning_rate": 0.001,
"loss": 1.4443,
"step": 91600
},
{
"epoch": 29.638009049773757,
"grad_norm": 1.4286977052688599,
"learning_rate": 0.001,
"loss": 1.4829,
"step": 91700
},
{
"epoch": 29.67032967032967,
"grad_norm": 1.380190372467041,
"learning_rate": 0.001,
"loss": 1.4936,
"step": 91800
},
{
"epoch": 29.702650290885586,
"grad_norm": 1.3603507280349731,
"learning_rate": 0.001,
"loss": 1.4793,
"step": 91900
},
{
"epoch": 29.7349709114415,
"grad_norm": 1.2642713785171509,
"learning_rate": 0.001,
"loss": 1.4848,
"step": 92000
},
{
"epoch": 29.767291531997415,
"grad_norm": 1.3949614763259888,
"learning_rate": 0.001,
"loss": 1.4902,
"step": 92100
},
{
"epoch": 29.79961215255333,
"grad_norm": 1.346086025238037,
"learning_rate": 0.001,
"loss": 1.5032,
"step": 92200
},
{
"epoch": 29.831932773109244,
"grad_norm": 1.2533193826675415,
"learning_rate": 0.001,
"loss": 1.4987,
"step": 92300
},
{
"epoch": 29.86425339366516,
"grad_norm": 1.5642977952957153,
"learning_rate": 0.001,
"loss": 1.4958,
"step": 92400
},
{
"epoch": 29.896574014221073,
"grad_norm": 1.1529120206832886,
"learning_rate": 0.001,
"loss": 1.504,
"step": 92500
},
{
"epoch": 29.928894634776988,
"grad_norm": 1.5595033168792725,
"learning_rate": 0.001,
"loss": 1.5043,
"step": 92600
},
{
"epoch": 29.961215255332903,
"grad_norm": 1.3069615364074707,
"learning_rate": 0.001,
"loss": 1.4877,
"step": 92700
},
{
"epoch": 29.993535875888817,
"grad_norm": 1.4206944704055786,
"learning_rate": 0.001,
"loss": 1.5047,
"step": 92800
},
{
"epoch": 30.02585649644473,
"grad_norm": 1.6076793670654297,
"learning_rate": 0.001,
"loss": 1.4092,
"step": 92900
},
{
"epoch": 30.058177117000646,
"grad_norm": 1.626736044883728,
"learning_rate": 0.001,
"loss": 1.3785,
"step": 93000
},
{
"epoch": 30.09049773755656,
"grad_norm": 1.416502833366394,
"learning_rate": 0.001,
"loss": 1.3829,
"step": 93100
},
{
"epoch": 30.122818358112475,
"grad_norm": 1.39216947555542,
"learning_rate": 0.001,
"loss": 1.3948,
"step": 93200
},
{
"epoch": 30.15513897866839,
"grad_norm": 1.3596535921096802,
"learning_rate": 0.001,
"loss": 1.4033,
"step": 93300
},
{
"epoch": 30.187459599224304,
"grad_norm": 1.544952630996704,
"learning_rate": 0.001,
"loss": 1.4145,
"step": 93400
},
{
"epoch": 30.21978021978022,
"grad_norm": 1.465815782546997,
"learning_rate": 0.001,
"loss": 1.4243,
"step": 93500
},
{
"epoch": 30.252100840336134,
"grad_norm": 1.2440087795257568,
"learning_rate": 0.001,
"loss": 1.434,
"step": 93600
},
{
"epoch": 30.284421460892048,
"grad_norm": 1.3168740272521973,
"learning_rate": 0.001,
"loss": 1.4139,
"step": 93700
},
{
"epoch": 30.316742081447963,
"grad_norm": 1.6248412132263184,
"learning_rate": 0.001,
"loss": 1.4096,
"step": 93800
},
{
"epoch": 30.349062702003877,
"grad_norm": 1.1371262073516846,
"learning_rate": 0.001,
"loss": 1.4351,
"step": 93900
},
{
"epoch": 30.381383322559792,
"grad_norm": 1.4596260786056519,
"learning_rate": 0.001,
"loss": 1.4342,
"step": 94000
},
{
"epoch": 30.413703943115706,
"grad_norm": 1.284958004951477,
"learning_rate": 0.001,
"loss": 1.4516,
"step": 94100
},
{
"epoch": 30.44602456367162,
"grad_norm": 1.0324749946594238,
"learning_rate": 0.001,
"loss": 1.4382,
"step": 94200
},
{
"epoch": 30.478345184227535,
"grad_norm": 1.3807328939437866,
"learning_rate": 0.001,
"loss": 1.4617,
"step": 94300
},
{
"epoch": 30.51066580478345,
"grad_norm": 1.4561772346496582,
"learning_rate": 0.001,
"loss": 1.4732,
"step": 94400
},
{
"epoch": 30.542986425339368,
"grad_norm": 1.1771188974380493,
"learning_rate": 0.001,
"loss": 1.4479,
"step": 94500
},
{
"epoch": 30.575307045895283,
"grad_norm": 1.7424471378326416,
"learning_rate": 0.001,
"loss": 1.4705,
"step": 94600
},
{
"epoch": 30.607627666451197,
"grad_norm": 1.817940354347229,
"learning_rate": 0.001,
"loss": 1.462,
"step": 94700
},
{
"epoch": 30.639948287007112,
"grad_norm": 1.380663275718689,
"learning_rate": 0.001,
"loss": 1.4239,
"step": 94800
},
{
"epoch": 30.672268907563026,
"grad_norm": 1.4679547548294067,
"learning_rate": 0.001,
"loss": 1.4532,
"step": 94900
},
{
"epoch": 30.70458952811894,
"grad_norm": 1.6584173440933228,
"learning_rate": 0.001,
"loss": 1.4572,
"step": 95000
},
{
"epoch": 30.736910148674855,
"grad_norm": 1.5099067687988281,
"learning_rate": 0.001,
"loss": 1.4836,
"step": 95100
},
{
"epoch": 30.76923076923077,
"grad_norm": 1.4512988328933716,
"learning_rate": 0.001,
"loss": 1.4611,
"step": 95200
},
{
"epoch": 30.801551389786685,
"grad_norm": 1.297056794166565,
"learning_rate": 0.001,
"loss": 1.462,
"step": 95300
},
{
"epoch": 30.8338720103426,
"grad_norm": 1.4180004596710205,
"learning_rate": 0.001,
"loss": 1.4916,
"step": 95400
},
{
"epoch": 30.866192630898514,
"grad_norm": 1.7167716026306152,
"learning_rate": 0.001,
"loss": 1.4824,
"step": 95500
},
{
"epoch": 30.89851325145443,
"grad_norm": 1.228950023651123,
"learning_rate": 0.001,
"loss": 1.4697,
"step": 95600
},
{
"epoch": 30.930833872010343,
"grad_norm": 1.4081132411956787,
"learning_rate": 0.001,
"loss": 1.5114,
"step": 95700
},
{
"epoch": 30.963154492566257,
"grad_norm": 1.2866050004959106,
"learning_rate": 0.001,
"loss": 1.484,
"step": 95800
},
{
"epoch": 30.995475113122172,
"grad_norm": 1.4979358911514282,
"learning_rate": 0.001,
"loss": 1.4559,
"step": 95900
},
{
"epoch": 31.027795733678087,
"grad_norm": 1.3178406953811646,
"learning_rate": 0.001,
"loss": 1.3567,
"step": 96000
},
{
"epoch": 31.060116354234,
"grad_norm": 1.5758633613586426,
"learning_rate": 0.001,
"loss": 1.3671,
"step": 96100
},
{
"epoch": 31.092436974789916,
"grad_norm": 1.5738314390182495,
"learning_rate": 0.001,
"loss": 1.3945,
"step": 96200
},
{
"epoch": 31.12475759534583,
"grad_norm": 1.34505295753479,
"learning_rate": 0.001,
"loss": 1.4101,
"step": 96300
},
{
"epoch": 31.157078215901745,
"grad_norm": 1.502578616142273,
"learning_rate": 0.001,
"loss": 1.3946,
"step": 96400
},
{
"epoch": 31.18939883645766,
"grad_norm": 1.2991825342178345,
"learning_rate": 0.001,
"loss": 1.3664,
"step": 96500
},
{
"epoch": 31.221719457013574,
"grad_norm": 1.2653495073318481,
"learning_rate": 0.001,
"loss": 1.416,
"step": 96600
},
{
"epoch": 31.25404007756949,
"grad_norm": 1.8182306289672852,
"learning_rate": 0.001,
"loss": 1.3977,
"step": 96700
},
{
"epoch": 31.286360698125403,
"grad_norm": 1.8426049947738647,
"learning_rate": 0.001,
"loss": 1.4165,
"step": 96800
},
{
"epoch": 31.318681318681318,
"grad_norm": 1.3844722509384155,
"learning_rate": 0.001,
"loss": 1.4174,
"step": 96900
},
{
"epoch": 31.351001939237232,
"grad_norm": 1.6439272165298462,
"learning_rate": 0.001,
"loss": 1.4053,
"step": 97000
},
{
"epoch": 31.383322559793147,
"grad_norm": 1.369141936302185,
"learning_rate": 0.001,
"loss": 1.4124,
"step": 97100
},
{
"epoch": 31.41564318034906,
"grad_norm": 1.6224870681762695,
"learning_rate": 0.001,
"loss": 1.4294,
"step": 97200
},
{
"epoch": 31.447963800904976,
"grad_norm": 1.3477481603622437,
"learning_rate": 0.001,
"loss": 1.4287,
"step": 97300
},
{
"epoch": 31.48028442146089,
"grad_norm": 1.3420053720474243,
"learning_rate": 0.001,
"loss": 1.4152,
"step": 97400
},
{
"epoch": 31.51260504201681,
"grad_norm": 1.5057765245437622,
"learning_rate": 0.001,
"loss": 1.4473,
"step": 97500
},
{
"epoch": 31.544925662572723,
"grad_norm": 1.417724609375,
"learning_rate": 0.001,
"loss": 1.4357,
"step": 97600
},
{
"epoch": 31.577246283128638,
"grad_norm": 1.4711930751800537,
"learning_rate": 0.001,
"loss": 1.4333,
"step": 97700
},
{
"epoch": 31.609566903684552,
"grad_norm": 1.3188661336898804,
"learning_rate": 0.001,
"loss": 1.4402,
"step": 97800
},
{
"epoch": 31.641887524240467,
"grad_norm": 1.7794547080993652,
"learning_rate": 0.001,
"loss": 1.4465,
"step": 97900
},
{
"epoch": 31.67420814479638,
"grad_norm": 1.7140164375305176,
"learning_rate": 0.001,
"loss": 1.4473,
"step": 98000
},
{
"epoch": 31.706528765352296,
"grad_norm": 1.1292674541473389,
"learning_rate": 0.001,
"loss": 1.4553,
"step": 98100
},
{
"epoch": 31.73884938590821,
"grad_norm": 1.4880545139312744,
"learning_rate": 0.001,
"loss": 1.4568,
"step": 98200
},
{
"epoch": 31.771170006464125,
"grad_norm": 1.5296339988708496,
"learning_rate": 0.001,
"loss": 1.4557,
"step": 98300
},
{
"epoch": 31.80349062702004,
"grad_norm": 2.067758321762085,
"learning_rate": 0.001,
"loss": 1.4437,
"step": 98400
},
{
"epoch": 31.835811247575954,
"grad_norm": 1.5721486806869507,
"learning_rate": 0.001,
"loss": 1.4621,
"step": 98500
},
{
"epoch": 31.86813186813187,
"grad_norm": 1.4480345249176025,
"learning_rate": 0.001,
"loss": 1.4543,
"step": 98600
},
{
"epoch": 31.900452488687783,
"grad_norm": 1.422499179840088,
"learning_rate": 0.001,
"loss": 1.463,
"step": 98700
},
{
"epoch": 31.932773109243698,
"grad_norm": 1.4816770553588867,
"learning_rate": 0.001,
"loss": 1.4736,
"step": 98800
},
{
"epoch": 31.965093729799612,
"grad_norm": 1.3213480710983276,
"learning_rate": 0.001,
"loss": 1.4433,
"step": 98900
},
{
"epoch": 31.997414350355527,
"grad_norm": 1.6971964836120605,
"learning_rate": 0.001,
"loss": 1.437,
"step": 99000
},
{
"epoch": 32.02973497091144,
"grad_norm": 1.3023521900177002,
"learning_rate": 0.001,
"loss": 1.384,
"step": 99100
},
{
"epoch": 32.062055591467356,
"grad_norm": 1.6234173774719238,
"learning_rate": 0.001,
"loss": 1.3549,
"step": 99200
},
{
"epoch": 32.09437621202327,
"grad_norm": 1.768607497215271,
"learning_rate": 0.001,
"loss": 1.3397,
"step": 99300
},
{
"epoch": 32.126696832579185,
"grad_norm": 1.3650957345962524,
"learning_rate": 0.001,
"loss": 1.3694,
"step": 99400
},
{
"epoch": 32.1590174531351,
"grad_norm": 1.4948350191116333,
"learning_rate": 0.001,
"loss": 1.3792,
"step": 99500
},
{
"epoch": 32.191338073691014,
"grad_norm": 1.908570647239685,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 99600
},
{
"epoch": 32.22365869424693,
"grad_norm": 1.3277240991592407,
"learning_rate": 0.001,
"loss": 1.3784,
"step": 99700
},
{
"epoch": 32.25597931480284,
"grad_norm": 1.5453720092773438,
"learning_rate": 0.001,
"loss": 1.382,
"step": 99800
},
{
"epoch": 32.28829993535876,
"grad_norm": 1.4696482419967651,
"learning_rate": 0.001,
"loss": 1.3842,
"step": 99900
},
{
"epoch": 32.32062055591467,
"grad_norm": 1.1971416473388672,
"learning_rate": 0.001,
"loss": 1.4028,
"step": 100000
},
{
"epoch": 32.35294117647059,
"grad_norm": 1.68018639087677,
"learning_rate": 0.001,
"loss": 1.4056,
"step": 100100
},
{
"epoch": 32.3852617970265,
"grad_norm": 1.378050446510315,
"learning_rate": 0.001,
"loss": 1.4084,
"step": 100200
},
{
"epoch": 32.417582417582416,
"grad_norm": 1.5301188230514526,
"learning_rate": 0.001,
"loss": 1.4073,
"step": 100300
},
{
"epoch": 32.44990303813833,
"grad_norm": 1.2459090948104858,
"learning_rate": 0.001,
"loss": 1.4053,
"step": 100400
},
{
"epoch": 32.482223658694245,
"grad_norm": 1.490999460220337,
"learning_rate": 0.001,
"loss": 1.3908,
"step": 100500
},
{
"epoch": 32.51454427925016,
"grad_norm": 1.8385999202728271,
"learning_rate": 0.001,
"loss": 1.3967,
"step": 100600
},
{
"epoch": 32.546864899806074,
"grad_norm": 1.4233742952346802,
"learning_rate": 0.001,
"loss": 1.4154,
"step": 100700
},
{
"epoch": 32.57918552036199,
"grad_norm": 1.403134822845459,
"learning_rate": 0.001,
"loss": 1.4066,
"step": 100800
},
{
"epoch": 32.6115061409179,
"grad_norm": 1.558266043663025,
"learning_rate": 0.001,
"loss": 1.4068,
"step": 100900
},
{
"epoch": 32.64382676147382,
"grad_norm": 1.7633984088897705,
"learning_rate": 0.001,
"loss": 1.429,
"step": 101000
},
{
"epoch": 32.67614738202973,
"grad_norm": 1.5486977100372314,
"learning_rate": 0.001,
"loss": 1.4175,
"step": 101100
},
{
"epoch": 32.70846800258565,
"grad_norm": 1.4046351909637451,
"learning_rate": 0.001,
"loss": 1.4407,
"step": 101200
},
{
"epoch": 32.74078862314156,
"grad_norm": 1.6478217840194702,
"learning_rate": 0.001,
"loss": 1.4231,
"step": 101300
},
{
"epoch": 32.773109243697476,
"grad_norm": 1.5102832317352295,
"learning_rate": 0.001,
"loss": 1.4342,
"step": 101400
},
{
"epoch": 32.80542986425339,
"grad_norm": 1.3012944459915161,
"learning_rate": 0.001,
"loss": 1.4379,
"step": 101500
},
{
"epoch": 32.837750484809305,
"grad_norm": 1.6932172775268555,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 101600
},
{
"epoch": 32.87007110536522,
"grad_norm": 1.916825771331787,
"learning_rate": 0.001,
"loss": 1.4309,
"step": 101700
},
{
"epoch": 32.902391725921134,
"grad_norm": 1.3492487668991089,
"learning_rate": 0.001,
"loss": 1.4423,
"step": 101800
},
{
"epoch": 32.93471234647705,
"grad_norm": 1.241912603378296,
"learning_rate": 0.001,
"loss": 1.4403,
"step": 101900
},
{
"epoch": 32.967032967032964,
"grad_norm": 1.467849612236023,
"learning_rate": 0.001,
"loss": 1.4475,
"step": 102000
},
{
"epoch": 32.999353587588885,
"grad_norm": 1.6124597787857056,
"learning_rate": 0.001,
"loss": 1.42,
"step": 102100
},
{
"epoch": 33.0316742081448,
"grad_norm": 1.371381163597107,
"learning_rate": 0.001,
"loss": 1.3293,
"step": 102200
},
{
"epoch": 33.063994828700714,
"grad_norm": 1.6305900812149048,
"learning_rate": 0.001,
"loss": 1.3276,
"step": 102300
},
{
"epoch": 33.09631544925663,
"grad_norm": 1.3269321918487549,
"learning_rate": 0.001,
"loss": 1.3424,
"step": 102400
},
{
"epoch": 33.12863606981254,
"grad_norm": 1.3419619798660278,
"learning_rate": 0.001,
"loss": 1.3428,
"step": 102500
},
{
"epoch": 33.16095669036846,
"grad_norm": 1.6693239212036133,
"learning_rate": 0.001,
"loss": 1.3468,
"step": 102600
},
{
"epoch": 33.19327731092437,
"grad_norm": 1.695776104927063,
"learning_rate": 0.001,
"loss": 1.3426,
"step": 102700
},
{
"epoch": 33.22559793148029,
"grad_norm": 1.5298768281936646,
"learning_rate": 0.001,
"loss": 1.3613,
"step": 102800
},
{
"epoch": 33.2579185520362,
"grad_norm": 1.1615654230117798,
"learning_rate": 0.001,
"loss": 1.3602,
"step": 102900
},
{
"epoch": 33.290239172592116,
"grad_norm": 1.2159184217453003,
"learning_rate": 0.001,
"loss": 1.3699,
"step": 103000
},
{
"epoch": 33.32255979314803,
"grad_norm": 1.51858389377594,
"learning_rate": 0.001,
"loss": 1.3637,
"step": 103100
},
{
"epoch": 33.354880413703945,
"grad_norm": 1.4186036586761475,
"learning_rate": 0.001,
"loss": 1.3848,
"step": 103200
},
{
"epoch": 33.38720103425986,
"grad_norm": 1.075792908668518,
"learning_rate": 0.001,
"loss": 1.386,
"step": 103300
},
{
"epoch": 33.419521654815775,
"grad_norm": 1.2305415868759155,
"learning_rate": 0.001,
"loss": 1.3859,
"step": 103400
},
{
"epoch": 33.45184227537169,
"grad_norm": 1.5435240268707275,
"learning_rate": 0.001,
"loss": 1.3855,
"step": 103500
},
{
"epoch": 33.484162895927604,
"grad_norm": 1.311586618423462,
"learning_rate": 0.001,
"loss": 1.3882,
"step": 103600
},
{
"epoch": 33.51648351648352,
"grad_norm": 1.5243502855300903,
"learning_rate": 0.001,
"loss": 1.3927,
"step": 103700
},
{
"epoch": 33.54880413703943,
"grad_norm": 1.2692375183105469,
"learning_rate": 0.001,
"loss": 1.4102,
"step": 103800
},
{
"epoch": 33.58112475759535,
"grad_norm": 1.051710844039917,
"learning_rate": 0.001,
"loss": 1.3751,
"step": 103900
},
{
"epoch": 33.61344537815126,
"grad_norm": 1.5681949853897095,
"learning_rate": 0.001,
"loss": 1.4064,
"step": 104000
},
{
"epoch": 33.645765998707176,
"grad_norm": 1.0312007665634155,
"learning_rate": 0.001,
"loss": 1.3998,
"step": 104100
},
{
"epoch": 33.67808661926309,
"grad_norm": 1.3208967447280884,
"learning_rate": 0.001,
"loss": 1.4146,
"step": 104200
},
{
"epoch": 33.710407239819006,
"grad_norm": 1.1144686937332153,
"learning_rate": 0.001,
"loss": 1.4013,
"step": 104300
},
{
"epoch": 33.74272786037492,
"grad_norm": 1.0032135248184204,
"learning_rate": 0.001,
"loss": 1.4139,
"step": 104400
},
{
"epoch": 33.775048480930835,
"grad_norm": 1.4895492792129517,
"learning_rate": 0.001,
"loss": 1.431,
"step": 104500
},
{
"epoch": 33.80736910148675,
"grad_norm": 1.7431833744049072,
"learning_rate": 0.001,
"loss": 1.4265,
"step": 104600
},
{
"epoch": 33.839689722042664,
"grad_norm": 1.2346980571746826,
"learning_rate": 0.001,
"loss": 1.4348,
"step": 104700
},
{
"epoch": 33.87201034259858,
"grad_norm": 1.3240312337875366,
"learning_rate": 0.001,
"loss": 1.423,
"step": 104800
},
{
"epoch": 33.90433096315449,
"grad_norm": 1.233581304550171,
"learning_rate": 0.001,
"loss": 1.4241,
"step": 104900
},
{
"epoch": 33.93665158371041,
"grad_norm": 1.156237006187439,
"learning_rate": 0.001,
"loss": 1.436,
"step": 105000
},
{
"epoch": 33.96897220426632,
"grad_norm": 1.8369982242584229,
"learning_rate": 0.001,
"loss": 1.467,
"step": 105100
},
{
"epoch": 34.00129282482224,
"grad_norm": 1.6478368043899536,
"learning_rate": 0.001,
"loss": 1.4149,
"step": 105200
},
{
"epoch": 34.03361344537815,
"grad_norm": 1.313193917274475,
"learning_rate": 0.001,
"loss": 1.3248,
"step": 105300
},
{
"epoch": 34.065934065934066,
"grad_norm": 1.7729835510253906,
"learning_rate": 0.001,
"loss": 1.3279,
"step": 105400
},
{
"epoch": 34.09825468648998,
"grad_norm": 1.3086352348327637,
"learning_rate": 0.001,
"loss": 1.3106,
"step": 105500
},
{
"epoch": 34.130575307045895,
"grad_norm": 1.9100786447525024,
"learning_rate": 0.001,
"loss": 1.341,
"step": 105600
},
{
"epoch": 34.16289592760181,
"grad_norm": 1.6001800298690796,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 105700
},
{
"epoch": 34.195216548157724,
"grad_norm": 2.048259973526001,
"learning_rate": 0.001,
"loss": 1.342,
"step": 105800
},
{
"epoch": 34.22753716871364,
"grad_norm": 1.946439266204834,
"learning_rate": 0.001,
"loss": 1.3327,
"step": 105900
},
{
"epoch": 34.25985778926955,
"grad_norm": 2.4914627075195312,
"learning_rate": 0.001,
"loss": 1.3502,
"step": 106000
},
{
"epoch": 34.29217840982547,
"grad_norm": 1.5728304386138916,
"learning_rate": 0.001,
"loss": 1.3537,
"step": 106100
},
{
"epoch": 34.32449903038138,
"grad_norm": 1.6155190467834473,
"learning_rate": 0.001,
"loss": 1.3541,
"step": 106200
},
{
"epoch": 34.3568196509373,
"grad_norm": 1.7084872722625732,
"learning_rate": 0.001,
"loss": 1.3779,
"step": 106300
},
{
"epoch": 34.38914027149321,
"grad_norm": 2.218944787979126,
"learning_rate": 0.001,
"loss": 1.3609,
"step": 106400
},
{
"epoch": 34.421460892049126,
"grad_norm": 1.687677025794983,
"learning_rate": 0.001,
"loss": 1.3636,
"step": 106500
},
{
"epoch": 34.45378151260504,
"grad_norm": 1.9398376941680908,
"learning_rate": 0.001,
"loss": 1.367,
"step": 106600
},
{
"epoch": 34.486102133160955,
"grad_norm": 1.5889030694961548,
"learning_rate": 0.001,
"loss": 1.3579,
"step": 106700
},
{
"epoch": 34.51842275371687,
"grad_norm": 1.5578453540802002,
"learning_rate": 0.001,
"loss": 1.3691,
"step": 106800
},
{
"epoch": 34.550743374272784,
"grad_norm": 1.7951077222824097,
"learning_rate": 0.001,
"loss": 1.3921,
"step": 106900
},
{
"epoch": 34.5830639948287,
"grad_norm": 1.4601072072982788,
"learning_rate": 0.001,
"loss": 1.3783,
"step": 107000
},
{
"epoch": 34.61538461538461,
"grad_norm": 1.6969603300094604,
"learning_rate": 0.001,
"loss": 1.3871,
"step": 107100
},
{
"epoch": 34.64770523594053,
"grad_norm": 2.027162551879883,
"learning_rate": 0.001,
"loss": 1.3973,
"step": 107200
},
{
"epoch": 34.68002585649644,
"grad_norm": 1.6259560585021973,
"learning_rate": 0.001,
"loss": 1.3925,
"step": 107300
},
{
"epoch": 34.71234647705236,
"grad_norm": 2.365182876586914,
"learning_rate": 0.001,
"loss": 1.3948,
"step": 107400
},
{
"epoch": 34.74466709760827,
"grad_norm": 1.8796085119247437,
"learning_rate": 0.001,
"loss": 1.3853,
"step": 107500
},
{
"epoch": 34.776987718164186,
"grad_norm": 2.121328115463257,
"learning_rate": 0.001,
"loss": 1.3946,
"step": 107600
},
{
"epoch": 34.8093083387201,
"grad_norm": 2.0668091773986816,
"learning_rate": 0.001,
"loss": 1.4044,
"step": 107700
},
{
"epoch": 34.841628959276015,
"grad_norm": 1.8705992698669434,
"learning_rate": 0.001,
"loss": 1.4063,
"step": 107800
},
{
"epoch": 34.87394957983193,
"grad_norm": 1.508862018585205,
"learning_rate": 0.001,
"loss": 1.4206,
"step": 107900
},
{
"epoch": 34.906270200387844,
"grad_norm": 1.69413161277771,
"learning_rate": 0.001,
"loss": 1.4,
"step": 108000
},
{
"epoch": 34.93859082094376,
"grad_norm": 1.6866521835327148,
"learning_rate": 0.001,
"loss": 1.4208,
"step": 108100
},
{
"epoch": 34.97091144149967,
"grad_norm": 1.5360355377197266,
"learning_rate": 0.001,
"loss": 1.4136,
"step": 108200
},
{
"epoch": 35.003232062055595,
"grad_norm": 1.6422771215438843,
"learning_rate": 0.001,
"loss": 1.4272,
"step": 108300
},
{
"epoch": 35.03555268261151,
"grad_norm": 1.7707010507583618,
"learning_rate": 0.001,
"loss": 1.301,
"step": 108400
},
{
"epoch": 35.067873303167424,
"grad_norm": 1.2861260175704956,
"learning_rate": 0.001,
"loss": 1.2999,
"step": 108500
},
{
"epoch": 35.10019392372334,
"grad_norm": 1.4097256660461426,
"learning_rate": 0.001,
"loss": 1.2971,
"step": 108600
},
{
"epoch": 35.13251454427925,
"grad_norm": 1.6693590879440308,
"learning_rate": 0.001,
"loss": 1.3326,
"step": 108700
},
{
"epoch": 35.16483516483517,
"grad_norm": 1.5808466672897339,
"learning_rate": 0.001,
"loss": 1.3262,
"step": 108800
},
{
"epoch": 35.19715578539108,
"grad_norm": 2.1594834327697754,
"learning_rate": 0.001,
"loss": 1.3371,
"step": 108900
},
{
"epoch": 35.229476405947,
"grad_norm": 1.4698948860168457,
"learning_rate": 0.001,
"loss": 1.3155,
"step": 109000
},
{
"epoch": 35.26179702650291,
"grad_norm": 1.303154706954956,
"learning_rate": 0.001,
"loss": 1.3431,
"step": 109100
},
{
"epoch": 35.294117647058826,
"grad_norm": 2.034785032272339,
"learning_rate": 0.001,
"loss": 1.3363,
"step": 109200
},
{
"epoch": 35.32643826761474,
"grad_norm": 1.606525182723999,
"learning_rate": 0.001,
"loss": 1.3313,
"step": 109300
},
{
"epoch": 35.358758888170655,
"grad_norm": 1.4785006046295166,
"learning_rate": 0.001,
"loss": 1.3454,
"step": 109400
},
{
"epoch": 35.39107950872657,
"grad_norm": 1.8039274215698242,
"learning_rate": 0.001,
"loss": 1.3362,
"step": 109500
},
{
"epoch": 35.423400129282484,
"grad_norm": 1.7395920753479004,
"learning_rate": 0.001,
"loss": 1.3266,
"step": 109600
},
{
"epoch": 35.4557207498384,
"grad_norm": 1.8243016004562378,
"learning_rate": 0.001,
"loss": 1.3594,
"step": 109700
},
{
"epoch": 35.48804137039431,
"grad_norm": 1.5380924940109253,
"learning_rate": 0.001,
"loss": 1.3546,
"step": 109800
},
{
"epoch": 35.52036199095023,
"grad_norm": 1.3779228925704956,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 109900
},
{
"epoch": 35.55268261150614,
"grad_norm": 2.3389909267425537,
"learning_rate": 0.001,
"loss": 1.3681,
"step": 110000
},
{
"epoch": 35.58500323206206,
"grad_norm": 1.4241420030593872,
"learning_rate": 0.001,
"loss": 1.3503,
"step": 110100
},
{
"epoch": 35.61732385261797,
"grad_norm": 1.442950963973999,
"learning_rate": 0.001,
"loss": 1.3631,
"step": 110200
},
{
"epoch": 35.649644473173886,
"grad_norm": 1.332844614982605,
"learning_rate": 0.001,
"loss": 1.3938,
"step": 110300
},
{
"epoch": 35.6819650937298,
"grad_norm": 1.8216750621795654,
"learning_rate": 0.001,
"loss": 1.3864,
"step": 110400
},
{
"epoch": 35.714285714285715,
"grad_norm": 1.7127039432525635,
"learning_rate": 0.001,
"loss": 1.3654,
"step": 110500
},
{
"epoch": 35.74660633484163,
"grad_norm": 2.0166015625,
"learning_rate": 0.001,
"loss": 1.3565,
"step": 110600
},
{
"epoch": 35.778926955397544,
"grad_norm": 1.2647751569747925,
"learning_rate": 0.001,
"loss": 1.3747,
"step": 110700
},
{
"epoch": 35.81124757595346,
"grad_norm": 1.3230773210525513,
"learning_rate": 0.001,
"loss": 1.3771,
"step": 110800
},
{
"epoch": 35.84356819650937,
"grad_norm": 1.4883989095687866,
"learning_rate": 0.001,
"loss": 1.42,
"step": 110900
},
{
"epoch": 35.87588881706529,
"grad_norm": 1.489632248878479,
"learning_rate": 0.001,
"loss": 1.37,
"step": 111000
},
{
"epoch": 35.9082094376212,
"grad_norm": 1.4182137250900269,
"learning_rate": 0.001,
"loss": 1.3863,
"step": 111100
},
{
"epoch": 35.94053005817712,
"grad_norm": 2.037407159805298,
"learning_rate": 0.001,
"loss": 1.3966,
"step": 111200
},
{
"epoch": 35.97285067873303,
"grad_norm": 1.4528826475143433,
"learning_rate": 0.001,
"loss": 1.4004,
"step": 111300
},
{
"epoch": 36.005171299288946,
"grad_norm": 1.3675060272216797,
"learning_rate": 0.001,
"loss": 1.4169,
"step": 111400
},
{
"epoch": 36.03749191984486,
"grad_norm": 1.653662919998169,
"learning_rate": 0.001,
"loss": 1.2959,
"step": 111500
},
{
"epoch": 36.069812540400775,
"grad_norm": 1.1706920862197876,
"learning_rate": 0.001,
"loss": 1.2975,
"step": 111600
},
{
"epoch": 36.10213316095669,
"grad_norm": 1.3043246269226074,
"learning_rate": 0.001,
"loss": 1.2886,
"step": 111700
},
{
"epoch": 36.134453781512605,
"grad_norm": 1.1849963665008545,
"learning_rate": 0.001,
"loss": 1.2927,
"step": 111800
},
{
"epoch": 36.16677440206852,
"grad_norm": 1.2800124883651733,
"learning_rate": 0.001,
"loss": 1.3027,
"step": 111900
},
{
"epoch": 36.199095022624434,
"grad_norm": 1.2872220277786255,
"learning_rate": 0.001,
"loss": 1.3163,
"step": 112000
},
{
"epoch": 36.23141564318035,
"grad_norm": 2.119028329849243,
"learning_rate": 0.001,
"loss": 1.3386,
"step": 112100
},
{
"epoch": 36.26373626373626,
"grad_norm": 1.3781286478042603,
"learning_rate": 0.001,
"loss": 1.3095,
"step": 112200
},
{
"epoch": 36.29605688429218,
"grad_norm": 1.3859822750091553,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 112300
},
{
"epoch": 36.32837750484809,
"grad_norm": 1.6260432004928589,
"learning_rate": 0.001,
"loss": 1.3324,
"step": 112400
},
{
"epoch": 36.36069812540401,
"grad_norm": 1.4716598987579346,
"learning_rate": 0.001,
"loss": 1.3418,
"step": 112500
},
{
"epoch": 36.39301874595992,
"grad_norm": 1.2116726636886597,
"learning_rate": 0.001,
"loss": 1.3201,
"step": 112600
},
{
"epoch": 36.425339366515836,
"grad_norm": 1.6337072849273682,
"learning_rate": 0.001,
"loss": 1.3316,
"step": 112700
},
{
"epoch": 36.45765998707175,
"grad_norm": 1.3227875232696533,
"learning_rate": 0.001,
"loss": 1.3599,
"step": 112800
},
{
"epoch": 36.489980607627665,
"grad_norm": 1.3499410152435303,
"learning_rate": 0.001,
"loss": 1.3502,
"step": 112900
},
{
"epoch": 36.52230122818358,
"grad_norm": 1.3329325914382935,
"learning_rate": 0.001,
"loss": 1.3339,
"step": 113000
},
{
"epoch": 36.554621848739494,
"grad_norm": 1.4539114236831665,
"learning_rate": 0.001,
"loss": 1.3449,
"step": 113100
},
{
"epoch": 36.58694246929541,
"grad_norm": 2.205564022064209,
"learning_rate": 0.001,
"loss": 1.352,
"step": 113200
},
{
"epoch": 36.61926308985132,
"grad_norm": 1.6035467386245728,
"learning_rate": 0.001,
"loss": 1.3914,
"step": 113300
},
{
"epoch": 36.65158371040724,
"grad_norm": 1.2898305654525757,
"learning_rate": 0.001,
"loss": 1.3443,
"step": 113400
},
{
"epoch": 36.68390433096315,
"grad_norm": 1.606456995010376,
"learning_rate": 0.001,
"loss": 1.3539,
"step": 113500
},
{
"epoch": 36.71622495151907,
"grad_norm": 1.2838432788848877,
"learning_rate": 0.001,
"loss": 1.3457,
"step": 113600
},
{
"epoch": 36.74854557207498,
"grad_norm": 1.1542515754699707,
"learning_rate": 0.001,
"loss": 1.3706,
"step": 113700
},
{
"epoch": 36.780866192630896,
"grad_norm": 1.388629674911499,
"learning_rate": 0.001,
"loss": 1.3513,
"step": 113800
},
{
"epoch": 36.81318681318681,
"grad_norm": 1.447067141532898,
"learning_rate": 0.001,
"loss": 1.3755,
"step": 113900
},
{
"epoch": 36.845507433742725,
"grad_norm": 1.297431468963623,
"learning_rate": 0.001,
"loss": 1.365,
"step": 114000
},
{
"epoch": 36.87782805429864,
"grad_norm": 1.7740482091903687,
"learning_rate": 0.001,
"loss": 1.3714,
"step": 114100
},
{
"epoch": 36.910148674854554,
"grad_norm": 1.3775901794433594,
"learning_rate": 0.001,
"loss": 1.3733,
"step": 114200
},
{
"epoch": 36.94246929541047,
"grad_norm": 1.302278995513916,
"learning_rate": 0.001,
"loss": 1.37,
"step": 114300
},
{
"epoch": 36.97478991596638,
"grad_norm": 1.3867436647415161,
"learning_rate": 0.001,
"loss": 1.3881,
"step": 114400
},
{
"epoch": 37.007110536522305,
"grad_norm": 1.459952473640442,
"learning_rate": 0.001,
"loss": 1.3532,
"step": 114500
},
{
"epoch": 37.03943115707822,
"grad_norm": 1.1632601022720337,
"learning_rate": 0.001,
"loss": 1.2744,
"step": 114600
},
{
"epoch": 37.071751777634134,
"grad_norm": 1.4252634048461914,
"learning_rate": 0.001,
"loss": 1.2723,
"step": 114700
},
{
"epoch": 37.10407239819005,
"grad_norm": 1.4162706136703491,
"learning_rate": 0.001,
"loss": 1.2911,
"step": 114800
},
{
"epoch": 37.13639301874596,
"grad_norm": 1.4186283349990845,
"learning_rate": 0.001,
"loss": 1.273,
"step": 114900
},
{
"epoch": 37.16871363930188,
"grad_norm": 1.3726344108581543,
"learning_rate": 0.001,
"loss": 1.2772,
"step": 115000
},
{
"epoch": 37.20103425985779,
"grad_norm": 1.6328552961349487,
"learning_rate": 0.001,
"loss": 1.2846,
"step": 115100
},
{
"epoch": 37.23335488041371,
"grad_norm": 1.3136317729949951,
"learning_rate": 0.001,
"loss": 1.3262,
"step": 115200
},
{
"epoch": 37.26567550096962,
"grad_norm": 1.3501302003860474,
"learning_rate": 0.001,
"loss": 1.3047,
"step": 115300
},
{
"epoch": 37.297996121525536,
"grad_norm": 1.55678391456604,
"learning_rate": 0.001,
"loss": 1.3003,
"step": 115400
},
{
"epoch": 37.33031674208145,
"grad_norm": 1.5524173974990845,
"learning_rate": 0.001,
"loss": 1.3146,
"step": 115500
},
{
"epoch": 37.362637362637365,
"grad_norm": 1.2920715808868408,
"learning_rate": 0.001,
"loss": 1.3089,
"step": 115600
},
{
"epoch": 37.39495798319328,
"grad_norm": 1.1357648372650146,
"learning_rate": 0.001,
"loss": 1.3199,
"step": 115700
},
{
"epoch": 37.427278603749194,
"grad_norm": 1.5290732383728027,
"learning_rate": 0.001,
"loss": 1.3319,
"step": 115800
},
{
"epoch": 37.45959922430511,
"grad_norm": 1.258813500404358,
"learning_rate": 0.001,
"loss": 1.3225,
"step": 115900
},
{
"epoch": 37.49191984486102,
"grad_norm": 1.464493751525879,
"learning_rate": 0.001,
"loss": 1.3389,
"step": 116000
},
{
"epoch": 37.52424046541694,
"grad_norm": 1.4164890050888062,
"learning_rate": 0.001,
"loss": 1.3241,
"step": 116100
},
{
"epoch": 37.55656108597285,
"grad_norm": 1.4241828918457031,
"learning_rate": 0.001,
"loss": 1.3247,
"step": 116200
},
{
"epoch": 37.58888170652877,
"grad_norm": 1.3933970928192139,
"learning_rate": 0.001,
"loss": 1.3393,
"step": 116300
},
{
"epoch": 37.62120232708468,
"grad_norm": 1.3898669481277466,
"learning_rate": 0.001,
"loss": 1.3299,
"step": 116400
},
{
"epoch": 37.653522947640596,
"grad_norm": 1.510009527206421,
"learning_rate": 0.001,
"loss": 1.3366,
"step": 116500
},
{
"epoch": 37.68584356819651,
"grad_norm": 1.4854003190994263,
"learning_rate": 0.001,
"loss": 1.3391,
"step": 116600
},
{
"epoch": 37.718164188752425,
"grad_norm": 1.470606803894043,
"learning_rate": 0.001,
"loss": 1.3546,
"step": 116700
},
{
"epoch": 37.75048480930834,
"grad_norm": 1.4459855556488037,
"learning_rate": 0.001,
"loss": 1.3478,
"step": 116800
},
{
"epoch": 37.782805429864254,
"grad_norm": 1.8643691539764404,
"learning_rate": 0.001,
"loss": 1.3612,
"step": 116900
},
{
"epoch": 37.81512605042017,
"grad_norm": 1.3574726581573486,
"learning_rate": 0.001,
"loss": 1.338,
"step": 117000
},
{
"epoch": 37.84744667097608,
"grad_norm": 1.159055233001709,
"learning_rate": 0.001,
"loss": 1.3554,
"step": 117100
},
{
"epoch": 37.879767291532,
"grad_norm": 1.482819676399231,
"learning_rate": 0.001,
"loss": 1.3272,
"step": 117200
},
{
"epoch": 37.91208791208791,
"grad_norm": 1.246833086013794,
"learning_rate": 0.001,
"loss": 1.3687,
"step": 117300
},
{
"epoch": 37.94440853264383,
"grad_norm": 1.2404323816299438,
"learning_rate": 0.001,
"loss": 1.3753,
"step": 117400
},
{
"epoch": 37.97672915319974,
"grad_norm": 1.2670542001724243,
"learning_rate": 0.001,
"loss": 1.3756,
"step": 117500
},
{
"epoch": 38.009049773755656,
"grad_norm": 1.3727078437805176,
"learning_rate": 0.001,
"loss": 1.3417,
"step": 117600
},
{
"epoch": 38.04137039431157,
"grad_norm": 1.7219202518463135,
"learning_rate": 0.001,
"loss": 1.2648,
"step": 117700
},
{
"epoch": 38.073691014867485,
"grad_norm": 1.3733896017074585,
"learning_rate": 0.001,
"loss": 1.255,
"step": 117800
},
{
"epoch": 38.1060116354234,
"grad_norm": 1.310388445854187,
"learning_rate": 0.001,
"loss": 1.2708,
"step": 117900
},
{
"epoch": 38.138332255979314,
"grad_norm": 1.933166742324829,
"learning_rate": 0.001,
"loss": 1.2782,
"step": 118000
},
{
"epoch": 38.17065287653523,
"grad_norm": 1.2118735313415527,
"learning_rate": 0.001,
"loss": 1.2703,
"step": 118100
},
{
"epoch": 38.20297349709114,
"grad_norm": 1.679825782775879,
"learning_rate": 0.001,
"loss": 1.293,
"step": 118200
},
{
"epoch": 38.23529411764706,
"grad_norm": 1.510223150253296,
"learning_rate": 0.001,
"loss": 1.2627,
"step": 118300
},
{
"epoch": 38.26761473820297,
"grad_norm": 1.2792211771011353,
"learning_rate": 0.001,
"loss": 1.2951,
"step": 118400
},
{
"epoch": 38.29993535875889,
"grad_norm": 1.5114414691925049,
"learning_rate": 0.001,
"loss": 1.2875,
"step": 118500
},
{
"epoch": 38.3322559793148,
"grad_norm": 2.0232787132263184,
"learning_rate": 0.001,
"loss": 1.2998,
"step": 118600
},
{
"epoch": 38.364576599870716,
"grad_norm": 1.484636664390564,
"learning_rate": 0.001,
"loss": 1.2964,
"step": 118700
},
{
"epoch": 38.39689722042663,
"grad_norm": 1.48430597782135,
"learning_rate": 0.001,
"loss": 1.3011,
"step": 118800
},
{
"epoch": 38.429217840982545,
"grad_norm": 1.327134370803833,
"learning_rate": 0.001,
"loss": 1.2844,
"step": 118900
},
{
"epoch": 38.46153846153846,
"grad_norm": 1.5213978290557861,
"learning_rate": 0.001,
"loss": 1.3078,
"step": 119000
},
{
"epoch": 38.493859082094374,
"grad_norm": 1.2812081575393677,
"learning_rate": 0.001,
"loss": 1.3009,
"step": 119100
},
{
"epoch": 38.52617970265029,
"grad_norm": 1.4381201267242432,
"learning_rate": 0.001,
"loss": 1.318,
"step": 119200
},
{
"epoch": 38.558500323206204,
"grad_norm": 1.292160987854004,
"learning_rate": 0.001,
"loss": 1.3231,
"step": 119300
},
{
"epoch": 38.59082094376212,
"grad_norm": 6.911149978637695,
"learning_rate": 0.001,
"loss": 1.3112,
"step": 119400
},
{
"epoch": 38.62314156431803,
"grad_norm": 1.2957707643508911,
"learning_rate": 0.001,
"loss": 1.3244,
"step": 119500
},
{
"epoch": 38.65546218487395,
"grad_norm": 1.4078917503356934,
"learning_rate": 0.001,
"loss": 1.3198,
"step": 119600
},
{
"epoch": 38.68778280542986,
"grad_norm": 1.5980867147445679,
"learning_rate": 0.001,
"loss": 1.3277,
"step": 119700
},
{
"epoch": 38.720103425985776,
"grad_norm": 1.592467188835144,
"learning_rate": 0.001,
"loss": 1.3444,
"step": 119800
},
{
"epoch": 38.75242404654169,
"grad_norm": 1.4102036952972412,
"learning_rate": 0.001,
"loss": 1.3247,
"step": 119900
},
{
"epoch": 38.784744667097605,
"grad_norm": 1.3911519050598145,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 120000
},
{
"epoch": 38.81706528765352,
"grad_norm": 1.5132681131362915,
"learning_rate": 0.001,
"loss": 1.3244,
"step": 120100
},
{
"epoch": 38.849385908209435,
"grad_norm": 1.2358170747756958,
"learning_rate": 0.001,
"loss": 1.3629,
"step": 120200
},
{
"epoch": 38.88170652876535,
"grad_norm": 1.7172753810882568,
"learning_rate": 0.001,
"loss": 1.3467,
"step": 120300
},
{
"epoch": 38.914027149321264,
"grad_norm": 1.4436832666397095,
"learning_rate": 0.001,
"loss": 1.3567,
"step": 120400
},
{
"epoch": 38.94634776987718,
"grad_norm": 1.5369603633880615,
"learning_rate": 0.001,
"loss": 1.3385,
"step": 120500
},
{
"epoch": 38.97866839043309,
"grad_norm": 1.888579249382019,
"learning_rate": 0.001,
"loss": 1.3517,
"step": 120600
},
{
"epoch": 39.010989010989015,
"grad_norm": 1.46559476852417,
"learning_rate": 0.001,
"loss": 1.2974,
"step": 120700
},
{
"epoch": 39.04330963154493,
"grad_norm": 1.5635486841201782,
"learning_rate": 0.001,
"loss": 1.2502,
"step": 120800
},
{
"epoch": 39.075630252100844,
"grad_norm": 1.7329033613204956,
"learning_rate": 0.001,
"loss": 1.2601,
"step": 120900
},
{
"epoch": 39.10795087265676,
"grad_norm": 1.5442296266555786,
"learning_rate": 0.001,
"loss": 1.2569,
"step": 121000
},
{
"epoch": 39.14027149321267,
"grad_norm": 6.576416492462158,
"learning_rate": 0.001,
"loss": 1.2498,
"step": 121100
},
{
"epoch": 39.17259211376859,
"grad_norm": 1.594713568687439,
"learning_rate": 0.001,
"loss": 1.2479,
"step": 121200
},
{
"epoch": 39.2049127343245,
"grad_norm": 1.8197976350784302,
"learning_rate": 0.001,
"loss": 1.2542,
"step": 121300
},
{
"epoch": 39.237233354880416,
"grad_norm": 1.6412878036499023,
"learning_rate": 0.001,
"loss": 1.2893,
"step": 121400
},
{
"epoch": 39.26955397543633,
"grad_norm": 1.4169423580169678,
"learning_rate": 0.001,
"loss": 1.2744,
"step": 121500
},
{
"epoch": 39.301874595992246,
"grad_norm": 1.4842840433120728,
"learning_rate": 0.001,
"loss": 1.2591,
"step": 121600
},
{
"epoch": 39.33419521654816,
"grad_norm": 1.398146629333496,
"learning_rate": 0.001,
"loss": 1.2875,
"step": 121700
},
{
"epoch": 39.366515837104075,
"grad_norm": 1.357816219329834,
"learning_rate": 0.001,
"loss": 1.2726,
"step": 121800
},
{
"epoch": 39.39883645765999,
"grad_norm": 1.549759030342102,
"learning_rate": 0.001,
"loss": 1.2822,
"step": 121900
},
{
"epoch": 39.431157078215904,
"grad_norm": 2.409834146499634,
"learning_rate": 0.001,
"loss": 1.2891,
"step": 122000
},
{
"epoch": 39.46347769877182,
"grad_norm": 1.6706969738006592,
"learning_rate": 0.001,
"loss": 1.2912,
"step": 122100
},
{
"epoch": 39.49579831932773,
"grad_norm": 1.5066723823547363,
"learning_rate": 0.001,
"loss": 1.3089,
"step": 122200
},
{
"epoch": 39.52811893988365,
"grad_norm": 1.2610282897949219,
"learning_rate": 0.001,
"loss": 1.2997,
"step": 122300
},
{
"epoch": 39.56043956043956,
"grad_norm": 1.5646350383758545,
"learning_rate": 0.001,
"loss": 1.2968,
"step": 122400
},
{
"epoch": 39.59276018099548,
"grad_norm": 1.5079572200775146,
"learning_rate": 0.001,
"loss": 1.319,
"step": 122500
},
{
"epoch": 39.62508080155139,
"grad_norm": 1.5733426809310913,
"learning_rate": 0.001,
"loss": 1.3032,
"step": 122600
},
{
"epoch": 39.657401422107306,
"grad_norm": 2.1421279907226562,
"learning_rate": 0.001,
"loss": 1.3238,
"step": 122700
},
{
"epoch": 39.68972204266322,
"grad_norm": 1.4061602354049683,
"learning_rate": 0.001,
"loss": 1.3252,
"step": 122800
},
{
"epoch": 39.722042663219135,
"grad_norm": 2.070896625518799,
"learning_rate": 0.001,
"loss": 1.3183,
"step": 122900
},
{
"epoch": 39.75436328377505,
"grad_norm": 1.3934524059295654,
"learning_rate": 0.001,
"loss": 1.3248,
"step": 123000
},
{
"epoch": 39.786683904330964,
"grad_norm": 1.6643691062927246,
"learning_rate": 0.001,
"loss": 1.3273,
"step": 123100
},
{
"epoch": 39.81900452488688,
"grad_norm": 1.5145334005355835,
"learning_rate": 0.001,
"loss": 1.3264,
"step": 123200
},
{
"epoch": 39.85132514544279,
"grad_norm": 1.5545204877853394,
"learning_rate": 0.001,
"loss": 1.3218,
"step": 123300
},
{
"epoch": 39.88364576599871,
"grad_norm": 1.9395331144332886,
"learning_rate": 0.001,
"loss": 1.3351,
"step": 123400
},
{
"epoch": 39.91596638655462,
"grad_norm": 1.5554918050765991,
"learning_rate": 0.001,
"loss": 1.3308,
"step": 123500
},
{
"epoch": 39.94828700711054,
"grad_norm": 1.4148260354995728,
"learning_rate": 0.001,
"loss": 1.3201,
"step": 123600
},
{
"epoch": 39.98060762766645,
"grad_norm": 1.5533839464187622,
"learning_rate": 0.001,
"loss": 1.336,
"step": 123700
},
{
"epoch": 40.012928248222366,
"grad_norm": 1.5636522769927979,
"learning_rate": 0.001,
"loss": 1.2853,
"step": 123800
},
{
"epoch": 40.04524886877828,
"grad_norm": 1.8232518434524536,
"learning_rate": 0.001,
"loss": 1.2589,
"step": 123900
},
{
"epoch": 40.077569489334195,
"grad_norm": 1.736384391784668,
"learning_rate": 0.001,
"loss": 1.2304,
"step": 124000
},
{
"epoch": 40.10989010989011,
"grad_norm": 1.4018938541412354,
"learning_rate": 0.001,
"loss": 1.2321,
"step": 124100
},
{
"epoch": 40.142210730446024,
"grad_norm": 1.7340322732925415,
"learning_rate": 0.001,
"loss": 1.2276,
"step": 124200
},
{
"epoch": 40.17453135100194,
"grad_norm": 1.7465698719024658,
"learning_rate": 0.001,
"loss": 1.2472,
"step": 124300
},
{
"epoch": 40.20685197155785,
"grad_norm": 1.368248701095581,
"learning_rate": 0.001,
"loss": 1.2797,
"step": 124400
},
{
"epoch": 40.23917259211377,
"grad_norm": 1.4630776643753052,
"learning_rate": 0.001,
"loss": 1.2547,
"step": 124500
},
{
"epoch": 40.27149321266968,
"grad_norm": 1.3100026845932007,
"learning_rate": 0.001,
"loss": 1.2584,
"step": 124600
},
{
"epoch": 40.3038138332256,
"grad_norm": 1.499247431755066,
"learning_rate": 0.001,
"loss": 1.2662,
"step": 124700
},
{
"epoch": 40.33613445378151,
"grad_norm": 1.5203977823257446,
"learning_rate": 0.001,
"loss": 1.2632,
"step": 124800
},
{
"epoch": 40.368455074337426,
"grad_norm": 1.2384952306747437,
"learning_rate": 0.001,
"loss": 1.2653,
"step": 124900
},
{
"epoch": 40.40077569489334,
"grad_norm": 1.4984188079833984,
"learning_rate": 0.001,
"loss": 1.2775,
"step": 125000
},
{
"epoch": 40.433096315449255,
"grad_norm": 2.0445687770843506,
"learning_rate": 0.001,
"loss": 1.2666,
"step": 125100
},
{
"epoch": 40.46541693600517,
"grad_norm": 1.394576072692871,
"learning_rate": 0.001,
"loss": 1.2984,
"step": 125200
},
{
"epoch": 40.497737556561084,
"grad_norm": 1.9576348066329956,
"learning_rate": 0.001,
"loss": 1.2802,
"step": 125300
},
{
"epoch": 40.530058177117,
"grad_norm": 1.972380518913269,
"learning_rate": 0.001,
"loss": 1.295,
"step": 125400
},
{
"epoch": 40.56237879767291,
"grad_norm": 1.725782871246338,
"learning_rate": 0.001,
"loss": 1.2914,
"step": 125500
},
{
"epoch": 40.59469941822883,
"grad_norm": 1.329160451889038,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 125600
},
{
"epoch": 40.62702003878474,
"grad_norm": 2.017256259918213,
"learning_rate": 0.001,
"loss": 1.2944,
"step": 125700
},
{
"epoch": 40.65934065934066,
"grad_norm": 1.6846826076507568,
"learning_rate": 0.001,
"loss": 1.2814,
"step": 125800
},
{
"epoch": 40.69166127989657,
"grad_norm": 1.4076398611068726,
"learning_rate": 0.001,
"loss": 1.2891,
"step": 125900
},
{
"epoch": 40.723981900452486,
"grad_norm": 1.5176540613174438,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 126000
},
{
"epoch": 40.7563025210084,
"grad_norm": 1.530797004699707,
"learning_rate": 0.001,
"loss": 1.2988,
"step": 126100
},
{
"epoch": 40.788623141564315,
"grad_norm": 1.7626606225967407,
"learning_rate": 0.001,
"loss": 1.3186,
"step": 126200
},
{
"epoch": 40.82094376212023,
"grad_norm": 1.7462037801742554,
"learning_rate": 0.001,
"loss": 1.3148,
"step": 126300
},
{
"epoch": 40.853264382676144,
"grad_norm": 1.7329397201538086,
"learning_rate": 0.001,
"loss": 1.3184,
"step": 126400
},
{
"epoch": 40.88558500323206,
"grad_norm": 1.6765464544296265,
"learning_rate": 0.001,
"loss": 1.3073,
"step": 126500
},
{
"epoch": 40.91790562378797,
"grad_norm": 1.4348853826522827,
"learning_rate": 0.001,
"loss": 1.3172,
"step": 126600
},
{
"epoch": 40.95022624434389,
"grad_norm": 1.515199065208435,
"learning_rate": 0.001,
"loss": 1.3298,
"step": 126700
},
{
"epoch": 40.9825468648998,
"grad_norm": 1.3540042638778687,
"learning_rate": 0.001,
"loss": 1.3278,
"step": 126800
},
{
"epoch": 41.014867485455724,
"grad_norm": 1.2277718782424927,
"learning_rate": 0.001,
"loss": 1.2313,
"step": 126900
},
{
"epoch": 41.04718810601164,
"grad_norm": 1.4838180541992188,
"learning_rate": 0.001,
"loss": 1.2345,
"step": 127000
},
{
"epoch": 41.07950872656755,
"grad_norm": 1.5526901483535767,
"learning_rate": 0.001,
"loss": 1.2396,
"step": 127100
},
{
"epoch": 41.11182934712347,
"grad_norm": 1.3968197107315063,
"learning_rate": 0.001,
"loss": 1.244,
"step": 127200
},
{
"epoch": 41.14414996767938,
"grad_norm": 1.746369481086731,
"learning_rate": 0.001,
"loss": 1.2296,
"step": 127300
},
{
"epoch": 41.1764705882353,
"grad_norm": 1.4499101638793945,
"learning_rate": 0.001,
"loss": 1.2091,
"step": 127400
},
{
"epoch": 41.20879120879121,
"grad_norm": 1.3192867040634155,
"learning_rate": 0.001,
"loss": 1.2435,
"step": 127500
},
{
"epoch": 41.241111829347126,
"grad_norm": 2.4370741844177246,
"learning_rate": 0.001,
"loss": 1.2728,
"step": 127600
},
{
"epoch": 41.27343244990304,
"grad_norm": 1.7852426767349243,
"learning_rate": 0.001,
"loss": 1.2712,
"step": 127700
},
{
"epoch": 41.305753070458955,
"grad_norm": 1.3352060317993164,
"learning_rate": 0.001,
"loss": 1.2434,
"step": 127800
},
{
"epoch": 41.33807369101487,
"grad_norm": 1.5020313262939453,
"learning_rate": 0.001,
"loss": 1.2524,
"step": 127900
},
{
"epoch": 41.370394311570784,
"grad_norm": 2.2881219387054443,
"learning_rate": 0.001,
"loss": 1.2798,
"step": 128000
},
{
"epoch": 41.4027149321267,
"grad_norm": 1.1049096584320068,
"learning_rate": 0.001,
"loss": 1.2553,
"step": 128100
},
{
"epoch": 41.43503555268261,
"grad_norm": 1.5790807008743286,
"learning_rate": 0.001,
"loss": 1.264,
"step": 128200
},
{
"epoch": 41.46735617323853,
"grad_norm": 1.2970410585403442,
"learning_rate": 0.001,
"loss": 1.2725,
"step": 128300
},
{
"epoch": 41.49967679379444,
"grad_norm": 1.338108777999878,
"learning_rate": 0.001,
"loss": 1.272,
"step": 128400
},
{
"epoch": 41.53199741435036,
"grad_norm": 1.2910419702529907,
"learning_rate": 0.001,
"loss": 1.2712,
"step": 128500
},
{
"epoch": 41.56431803490627,
"grad_norm": 1.386048436164856,
"learning_rate": 0.001,
"loss": 1.2547,
"step": 128600
},
{
"epoch": 41.596638655462186,
"grad_norm": 1.3139433860778809,
"learning_rate": 0.001,
"loss": 1.2594,
"step": 128700
},
{
"epoch": 41.6289592760181,
"grad_norm": 1.7138657569885254,
"learning_rate": 0.001,
"loss": 1.2662,
"step": 128800
},
{
"epoch": 41.661279896574015,
"grad_norm": 1.7849534749984741,
"learning_rate": 0.001,
"loss": 1.2807,
"step": 128900
},
{
"epoch": 41.69360051712993,
"grad_norm": 1.5666148662567139,
"learning_rate": 0.001,
"loss": 1.2924,
"step": 129000
},
{
"epoch": 41.725921137685845,
"grad_norm": 1.6829689741134644,
"learning_rate": 0.001,
"loss": 1.2825,
"step": 129100
},
{
"epoch": 41.75824175824176,
"grad_norm": 2.1652581691741943,
"learning_rate": 0.001,
"loss": 1.3067,
"step": 129200
},
{
"epoch": 41.790562378797674,
"grad_norm": 1.6427013874053955,
"learning_rate": 0.001,
"loss": 1.267,
"step": 129300
},
{
"epoch": 41.82288299935359,
"grad_norm": 1.4795788526535034,
"learning_rate": 0.001,
"loss": 1.2871,
"step": 129400
},
{
"epoch": 41.8552036199095,
"grad_norm": 1.4031800031661987,
"learning_rate": 0.001,
"loss": 1.2955,
"step": 129500
},
{
"epoch": 41.88752424046542,
"grad_norm": 1.7825078964233398,
"learning_rate": 0.001,
"loss": 1.2859,
"step": 129600
},
{
"epoch": 41.91984486102133,
"grad_norm": 1.5589793920516968,
"learning_rate": 0.001,
"loss": 1.3138,
"step": 129700
},
{
"epoch": 41.95216548157725,
"grad_norm": 1.6905754804611206,
"learning_rate": 0.001,
"loss": 1.2902,
"step": 129800
},
{
"epoch": 41.98448610213316,
"grad_norm": 1.573258399963379,
"learning_rate": 0.001,
"loss": 1.3129,
"step": 129900
},
{
"epoch": 42.016806722689076,
"grad_norm": 2.0176637172698975,
"learning_rate": 0.001,
"loss": 1.2269,
"step": 130000
},
{
"epoch": 42.04912734324499,
"grad_norm": 2.4781367778778076,
"learning_rate": 0.001,
"loss": 1.1985,
"step": 130100
},
{
"epoch": 42.081447963800905,
"grad_norm": 2.272233009338379,
"learning_rate": 0.001,
"loss": 1.2131,
"step": 130200
},
{
"epoch": 42.11376858435682,
"grad_norm": 2.1599671840667725,
"learning_rate": 0.001,
"loss": 1.2109,
"step": 130300
},
{
"epoch": 42.146089204912734,
"grad_norm": 1.9499235153198242,
"learning_rate": 0.001,
"loss": 1.2144,
"step": 130400
},
{
"epoch": 42.17840982546865,
"grad_norm": 2.5726876258850098,
"learning_rate": 0.001,
"loss": 1.2383,
"step": 130500
},
{
"epoch": 42.21073044602456,
"grad_norm": 3.015073776245117,
"learning_rate": 0.001,
"loss": 1.2187,
"step": 130600
},
{
"epoch": 42.24305106658048,
"grad_norm": 1.6541657447814941,
"learning_rate": 0.001,
"loss": 1.2317,
"step": 130700
},
{
"epoch": 42.27537168713639,
"grad_norm": 2.5568466186523438,
"learning_rate": 0.001,
"loss": 1.2327,
"step": 130800
},
{
"epoch": 42.30769230769231,
"grad_norm": 2.0873799324035645,
"learning_rate": 0.001,
"loss": 1.2395,
"step": 130900
},
{
"epoch": 42.34001292824822,
"grad_norm": 2.5677318572998047,
"learning_rate": 0.001,
"loss": 1.2289,
"step": 131000
},
{
"epoch": 42.372333548804136,
"grad_norm": 2.3683345317840576,
"learning_rate": 0.001,
"loss": 1.257,
"step": 131100
},
{
"epoch": 42.40465416936005,
"grad_norm": 2.197979688644409,
"learning_rate": 0.001,
"loss": 1.2731,
"step": 131200
},
{
"epoch": 42.436974789915965,
"grad_norm": 1.855226993560791,
"learning_rate": 0.001,
"loss": 1.2417,
"step": 131300
},
{
"epoch": 42.46929541047188,
"grad_norm": 2.141434907913208,
"learning_rate": 0.001,
"loss": 1.2596,
"step": 131400
},
{
"epoch": 42.501616031027794,
"grad_norm": 1.755460500717163,
"learning_rate": 0.001,
"loss": 1.2524,
"step": 131500
},
{
"epoch": 42.53393665158371,
"grad_norm": 2.5488452911376953,
"learning_rate": 0.001,
"loss": 1.2532,
"step": 131600
},
{
"epoch": 42.56625727213962,
"grad_norm": 1.5715889930725098,
"learning_rate": 0.001,
"loss": 1.2617,
"step": 131700
},
{
"epoch": 42.59857789269554,
"grad_norm": 2.407876968383789,
"learning_rate": 0.001,
"loss": 1.2961,
"step": 131800
},
{
"epoch": 42.63089851325145,
"grad_norm": 1.963181734085083,
"learning_rate": 0.001,
"loss": 1.2588,
"step": 131900
},
{
"epoch": 42.66321913380737,
"grad_norm": 1.7608472108840942,
"learning_rate": 0.001,
"loss": 1.2694,
"step": 132000
},
{
"epoch": 42.69553975436328,
"grad_norm": 2.268115758895874,
"learning_rate": 0.001,
"loss": 1.2631,
"step": 132100
},
{
"epoch": 42.727860374919196,
"grad_norm": 2.737407684326172,
"learning_rate": 0.001,
"loss": 1.2707,
"step": 132200
},
{
"epoch": 42.76018099547511,
"grad_norm": 2.061769723892212,
"learning_rate": 0.001,
"loss": 1.2621,
"step": 132300
},
{
"epoch": 42.792501616031025,
"grad_norm": 2.5539047718048096,
"learning_rate": 0.001,
"loss": 1.2816,
"step": 132400
},
{
"epoch": 42.82482223658694,
"grad_norm": 2.357755661010742,
"learning_rate": 0.001,
"loss": 1.2758,
"step": 132500
},
{
"epoch": 42.857142857142854,
"grad_norm": 2.4219696521759033,
"learning_rate": 0.001,
"loss": 1.291,
"step": 132600
},
{
"epoch": 42.88946347769877,
"grad_norm": 2.248538017272949,
"learning_rate": 0.001,
"loss": 1.2678,
"step": 132700
},
{
"epoch": 42.92178409825468,
"grad_norm": 2.223904848098755,
"learning_rate": 0.001,
"loss": 1.3182,
"step": 132800
},
{
"epoch": 42.9541047188106,
"grad_norm": 2.2266910076141357,
"learning_rate": 0.001,
"loss": 1.3059,
"step": 132900
},
{
"epoch": 42.98642533936652,
"grad_norm": 2.4159045219421387,
"learning_rate": 0.001,
"loss": 1.2882,
"step": 133000
},
{
"epoch": 43.018745959922434,
"grad_norm": 1.578348159790039,
"learning_rate": 0.001,
"loss": 1.2544,
"step": 133100
},
{
"epoch": 43.05106658047835,
"grad_norm": 1.3469822406768799,
"learning_rate": 0.001,
"loss": 1.1898,
"step": 133200
},
{
"epoch": 43.08338720103426,
"grad_norm": 1.6561214923858643,
"learning_rate": 0.001,
"loss": 1.1783,
"step": 133300
},
{
"epoch": 43.11570782159018,
"grad_norm": 1.9601678848266602,
"learning_rate": 0.001,
"loss": 1.2017,
"step": 133400
},
{
"epoch": 43.14802844214609,
"grad_norm": 1.7265827655792236,
"learning_rate": 0.001,
"loss": 1.2033,
"step": 133500
},
{
"epoch": 43.18034906270201,
"grad_norm": 1.684269905090332,
"learning_rate": 0.001,
"loss": 1.2206,
"step": 133600
},
{
"epoch": 43.21266968325792,
"grad_norm": 1.8783414363861084,
"learning_rate": 0.001,
"loss": 1.2415,
"step": 133700
},
{
"epoch": 43.244990303813836,
"grad_norm": 1.572043776512146,
"learning_rate": 0.001,
"loss": 1.208,
"step": 133800
},
{
"epoch": 43.27731092436975,
"grad_norm": 1.7568895816802979,
"learning_rate": 0.001,
"loss": 1.2148,
"step": 133900
},
{
"epoch": 43.309631544925665,
"grad_norm": 1.3380953073501587,
"learning_rate": 0.001,
"loss": 1.2408,
"step": 134000
},
{
"epoch": 43.34195216548158,
"grad_norm": 1.442492127418518,
"learning_rate": 0.001,
"loss": 1.239,
"step": 134100
},
{
"epoch": 43.374272786037494,
"grad_norm": 1.6913920640945435,
"learning_rate": 0.001,
"loss": 1.2408,
"step": 134200
},
{
"epoch": 43.40659340659341,
"grad_norm": 1.5097508430480957,
"learning_rate": 0.001,
"loss": 1.2494,
"step": 134300
},
{
"epoch": 43.43891402714932,
"grad_norm": 1.304591417312622,
"learning_rate": 0.001,
"loss": 1.2255,
"step": 134400
},
{
"epoch": 43.47123464770524,
"grad_norm": 1.5187216997146606,
"learning_rate": 0.001,
"loss": 1.2574,
"step": 134500
},
{
"epoch": 43.50355526826115,
"grad_norm": 1.67637300491333,
"learning_rate": 0.001,
"loss": 1.2373,
"step": 134600
},
{
"epoch": 43.53587588881707,
"grad_norm": 1.3660882711410522,
"learning_rate": 0.001,
"loss": 1.2314,
"step": 134700
},
{
"epoch": 43.56819650937298,
"grad_norm": 1.8594450950622559,
"learning_rate": 0.001,
"loss": 1.2504,
"step": 134800
},
{
"epoch": 43.600517129928896,
"grad_norm": 1.4114196300506592,
"learning_rate": 0.001,
"loss": 1.2192,
"step": 134900
},
{
"epoch": 43.63283775048481,
"grad_norm": 1.8799229860305786,
"learning_rate": 0.001,
"loss": 1.2706,
"step": 135000
},
{
"epoch": 43.665158371040725,
"grad_norm": 1.6473530530929565,
"learning_rate": 0.001,
"loss": 1.2412,
"step": 135100
},
{
"epoch": 43.69747899159664,
"grad_norm": 1.7104618549346924,
"learning_rate": 0.001,
"loss": 1.2671,
"step": 135200
},
{
"epoch": 43.729799612152554,
"grad_norm": 1.5365654230117798,
"learning_rate": 0.001,
"loss": 1.2637,
"step": 135300
},
{
"epoch": 43.76212023270847,
"grad_norm": 1.5581703186035156,
"learning_rate": 0.001,
"loss": 1.2647,
"step": 135400
},
{
"epoch": 43.79444085326438,
"grad_norm": 1.283259630203247,
"learning_rate": 0.001,
"loss": 1.2827,
"step": 135500
},
{
"epoch": 43.8267614738203,
"grad_norm": 1.4676380157470703,
"learning_rate": 0.001,
"loss": 1.2488,
"step": 135600
},
{
"epoch": 43.85908209437621,
"grad_norm": 1.6356333494186401,
"learning_rate": 0.001,
"loss": 1.2684,
"step": 135700
},
{
"epoch": 43.89140271493213,
"grad_norm": 1.6816112995147705,
"learning_rate": 0.001,
"loss": 1.2755,
"step": 135800
},
{
"epoch": 43.92372333548804,
"grad_norm": 1.900416374206543,
"learning_rate": 0.001,
"loss": 1.2773,
"step": 135900
},
{
"epoch": 43.956043956043956,
"grad_norm": 1.8260838985443115,
"learning_rate": 0.001,
"loss": 1.2654,
"step": 136000
},
{
"epoch": 43.98836457659987,
"grad_norm": 1.3427294492721558,
"learning_rate": 0.001,
"loss": 1.2813,
"step": 136100
},
{
"epoch": 44.020685197155785,
"grad_norm": 1.6175090074539185,
"learning_rate": 0.001,
"loss": 1.2353,
"step": 136200
},
{
"epoch": 44.0530058177117,
"grad_norm": 1.451216220855713,
"learning_rate": 0.001,
"loss": 1.1861,
"step": 136300
},
{
"epoch": 44.085326438267614,
"grad_norm": 1.4623968601226807,
"learning_rate": 0.001,
"loss": 1.1686,
"step": 136400
},
{
"epoch": 44.11764705882353,
"grad_norm": 1.295665979385376,
"learning_rate": 0.001,
"loss": 1.1901,
"step": 136500
},
{
"epoch": 44.14996767937944,
"grad_norm": 1.691976547241211,
"learning_rate": 0.001,
"loss": 1.1879,
"step": 136600
},
{
"epoch": 44.18228829993536,
"grad_norm": 1.5531518459320068,
"learning_rate": 0.001,
"loss": 1.2036,
"step": 136700
},
{
"epoch": 44.21460892049127,
"grad_norm": 1.4985134601593018,
"learning_rate": 0.001,
"loss": 1.1843,
"step": 136800
},
{
"epoch": 44.24692954104719,
"grad_norm": 1.9060845375061035,
"learning_rate": 0.001,
"loss": 1.195,
"step": 136900
},
{
"epoch": 44.2792501616031,
"grad_norm": 1.326461911201477,
"learning_rate": 0.001,
"loss": 1.2133,
"step": 137000
},
{
"epoch": 44.311570782159016,
"grad_norm": 1.4161741733551025,
"learning_rate": 0.001,
"loss": 1.1989,
"step": 137100
},
{
"epoch": 44.34389140271493,
"grad_norm": 1.6229037046432495,
"learning_rate": 0.001,
"loss": 1.2134,
"step": 137200
},
{
"epoch": 44.376212023270845,
"grad_norm": 1.394170880317688,
"learning_rate": 0.001,
"loss": 1.2012,
"step": 137300
},
{
"epoch": 44.40853264382676,
"grad_norm": 1.5358482599258423,
"learning_rate": 0.001,
"loss": 1.2169,
"step": 137400
},
{
"epoch": 44.440853264382675,
"grad_norm": 1.7619261741638184,
"learning_rate": 0.001,
"loss": 1.2092,
"step": 137500
},
{
"epoch": 44.47317388493859,
"grad_norm": 1.7660161256790161,
"learning_rate": 0.001,
"loss": 1.231,
"step": 137600
},
{
"epoch": 44.505494505494504,
"grad_norm": 1.3661121129989624,
"learning_rate": 0.001,
"loss": 1.2354,
"step": 137700
},
{
"epoch": 44.53781512605042,
"grad_norm": 1.4903825521469116,
"learning_rate": 0.001,
"loss": 1.2407,
"step": 137800
},
{
"epoch": 44.57013574660633,
"grad_norm": 1.6031243801116943,
"learning_rate": 0.001,
"loss": 1.2485,
"step": 137900
},
{
"epoch": 44.60245636716225,
"grad_norm": 1.6402915716171265,
"learning_rate": 0.001,
"loss": 1.2353,
"step": 138000
},
{
"epoch": 44.63477698771816,
"grad_norm": 1.4186668395996094,
"learning_rate": 0.001,
"loss": 1.2482,
"step": 138100
},
{
"epoch": 44.66709760827408,
"grad_norm": 1.4842973947525024,
"learning_rate": 0.001,
"loss": 1.2324,
"step": 138200
},
{
"epoch": 44.69941822882999,
"grad_norm": 1.4359252452850342,
"learning_rate": 0.001,
"loss": 1.2624,
"step": 138300
},
{
"epoch": 44.731738849385906,
"grad_norm": 1.7322560548782349,
"learning_rate": 0.001,
"loss": 1.2446,
"step": 138400
},
{
"epoch": 44.76405946994182,
"grad_norm": 1.3641490936279297,
"learning_rate": 0.001,
"loss": 1.2614,
"step": 138500
},
{
"epoch": 44.796380090497735,
"grad_norm": 1.7803534269332886,
"learning_rate": 0.001,
"loss": 1.259,
"step": 138600
},
{
"epoch": 44.82870071105365,
"grad_norm": 1.528218388557434,
"learning_rate": 0.001,
"loss": 1.2661,
"step": 138700
},
{
"epoch": 44.861021331609564,
"grad_norm": 1.7628278732299805,
"learning_rate": 0.001,
"loss": 1.2614,
"step": 138800
},
{
"epoch": 44.89334195216548,
"grad_norm": 1.6271640062332153,
"learning_rate": 0.001,
"loss": 1.265,
"step": 138900
},
{
"epoch": 44.92566257272139,
"grad_norm": 2.0720269680023193,
"learning_rate": 0.001,
"loss": 1.2693,
"step": 139000
},
{
"epoch": 44.95798319327731,
"grad_norm": 1.4295275211334229,
"learning_rate": 0.001,
"loss": 1.2845,
"step": 139100
},
{
"epoch": 44.99030381383322,
"grad_norm": 1.7425401210784912,
"learning_rate": 0.001,
"loss": 1.258,
"step": 139200
},
{
"epoch": 45.022624434389144,
"grad_norm": 1.3533971309661865,
"learning_rate": 0.001,
"loss": 1.2069,
"step": 139300
},
{
"epoch": 45.05494505494506,
"grad_norm": 1.7541122436523438,
"learning_rate": 0.001,
"loss": 1.1611,
"step": 139400
},
{
"epoch": 45.08726567550097,
"grad_norm": 1.2720140218734741,
"learning_rate": 0.001,
"loss": 1.1736,
"step": 139500
},
{
"epoch": 45.11958629605689,
"grad_norm": 1.4547255039215088,
"learning_rate": 0.001,
"loss": 1.176,
"step": 139600
},
{
"epoch": 45.1519069166128,
"grad_norm": 1.6576576232910156,
"learning_rate": 0.001,
"loss": 1.1481,
"step": 139700
},
{
"epoch": 45.18422753716872,
"grad_norm": 1.6915053129196167,
"learning_rate": 0.001,
"loss": 1.1698,
"step": 139800
},
{
"epoch": 45.21654815772463,
"grad_norm": 1.5308271646499634,
"learning_rate": 0.001,
"loss": 1.1879,
"step": 139900
},
{
"epoch": 45.248868778280546,
"grad_norm": 1.6623047590255737,
"learning_rate": 0.001,
"loss": 1.1991,
"step": 140000
},
{
"epoch": 45.28118939883646,
"grad_norm": 1.9195517301559448,
"learning_rate": 0.001,
"loss": 1.1843,
"step": 140100
},
{
"epoch": 45.313510019392375,
"grad_norm": 1.644578456878662,
"learning_rate": 0.001,
"loss": 1.1975,
"step": 140200
},
{
"epoch": 45.34583063994829,
"grad_norm": 1.4461241960525513,
"learning_rate": 0.001,
"loss": 1.1956,
"step": 140300
},
{
"epoch": 45.378151260504204,
"grad_norm": 1.663171648979187,
"learning_rate": 0.001,
"loss": 1.2051,
"step": 140400
},
{
"epoch": 45.41047188106012,
"grad_norm": 1.6137675046920776,
"learning_rate": 0.001,
"loss": 1.2108,
"step": 140500
},
{
"epoch": 45.44279250161603,
"grad_norm": 1.6578792333602905,
"learning_rate": 0.001,
"loss": 1.2131,
"step": 140600
},
{
"epoch": 45.47511312217195,
"grad_norm": 1.2520636320114136,
"learning_rate": 0.001,
"loss": 1.208,
"step": 140700
},
{
"epoch": 45.50743374272786,
"grad_norm": 1.5555574893951416,
"learning_rate": 0.001,
"loss": 1.2404,
"step": 140800
},
{
"epoch": 45.53975436328378,
"grad_norm": 1.76326322555542,
"learning_rate": 0.001,
"loss": 1.2328,
"step": 140900
},
{
"epoch": 45.57207498383969,
"grad_norm": 1.5170855522155762,
"learning_rate": 0.001,
"loss": 1.2143,
"step": 141000
},
{
"epoch": 45.604395604395606,
"grad_norm": 1.3871099948883057,
"learning_rate": 0.001,
"loss": 1.2245,
"step": 141100
},
{
"epoch": 45.63671622495152,
"grad_norm": 1.281063437461853,
"learning_rate": 0.001,
"loss": 1.2288,
"step": 141200
},
{
"epoch": 45.669036845507435,
"grad_norm": 1.7094895839691162,
"learning_rate": 0.001,
"loss": 1.2412,
"step": 141300
},
{
"epoch": 45.70135746606335,
"grad_norm": 1.6494312286376953,
"learning_rate": 0.001,
"loss": 1.2116,
"step": 141400
},
{
"epoch": 45.733678086619264,
"grad_norm": 1.505626916885376,
"learning_rate": 0.001,
"loss": 1.2361,
"step": 141500
},
{
"epoch": 45.76599870717518,
"grad_norm": 1.4134547710418701,
"learning_rate": 0.001,
"loss": 1.2627,
"step": 141600
},
{
"epoch": 45.79831932773109,
"grad_norm": 1.3101989030838013,
"learning_rate": 0.001,
"loss": 1.2472,
"step": 141700
},
{
"epoch": 45.83063994828701,
"grad_norm": 1.5769306421279907,
"learning_rate": 0.001,
"loss": 1.2321,
"step": 141800
},
{
"epoch": 45.86296056884292,
"grad_norm": 1.8324518203735352,
"learning_rate": 0.001,
"loss": 1.2352,
"step": 141900
},
{
"epoch": 45.89528118939884,
"grad_norm": 1.7726576328277588,
"learning_rate": 0.001,
"loss": 1.2554,
"step": 142000
},
{
"epoch": 45.92760180995475,
"grad_norm": 1.480567455291748,
"learning_rate": 0.001,
"loss": 1.1472,
"step": 142100
},
{
"epoch": 45.959922430510666,
"grad_norm": 1.8656824827194214,
"learning_rate": 0.001,
"loss": 1.153,
"step": 142200
},
{
"epoch": 45.99224305106658,
"grad_norm": 1.5704985857009888,
"learning_rate": 0.001,
"loss": 1.1408,
"step": 142300
},
{
"epoch": 46.024563671622495,
"grad_norm": 1.6737463474273682,
"learning_rate": 0.001,
"loss": 1.1702,
"step": 142400
},
{
"epoch": 46.05688429217841,
"grad_norm": 1.4223302602767944,
"learning_rate": 0.001,
"loss": 1.1648,
"step": 142500
},
{
"epoch": 46.089204912734324,
"grad_norm": 1.4680297374725342,
"learning_rate": 0.001,
"loss": 1.1531,
"step": 142600
},
{
"epoch": 46.12152553329024,
"grad_norm": 1.2421369552612305,
"learning_rate": 0.001,
"loss": 1.152,
"step": 142700
},
{
"epoch": 46.15384615384615,
"grad_norm": 1.6976029872894287,
"learning_rate": 0.001,
"loss": 1.1664,
"step": 142800
},
{
"epoch": 46.18616677440207,
"grad_norm": 1.5253973007202148,
"learning_rate": 0.001,
"loss": 1.1621,
"step": 142900
},
{
"epoch": 46.21848739495798,
"grad_norm": 1.3489149808883667,
"learning_rate": 0.001,
"loss": 1.1926,
"step": 143000
},
{
"epoch": 46.2508080155139,
"grad_norm": 1.8092660903930664,
"learning_rate": 0.001,
"loss": 1.1874,
"step": 143100
},
{
"epoch": 46.28312863606981,
"grad_norm": 2.419577121734619,
"learning_rate": 0.001,
"loss": 1.1816,
"step": 143200
},
{
"epoch": 46.315449256625726,
"grad_norm": 1.3183753490447998,
"learning_rate": 0.001,
"loss": 1.1854,
"step": 143300
},
{
"epoch": 46.34776987718164,
"grad_norm": 1.8555301427841187,
"learning_rate": 0.001,
"loss": 1.2091,
"step": 143400
},
{
"epoch": 46.380090497737555,
"grad_norm": 1.7718279361724854,
"learning_rate": 0.001,
"loss": 1.1856,
"step": 143500
},
{
"epoch": 46.41241111829347,
"grad_norm": 1.52428138256073,
"learning_rate": 0.001,
"loss": 1.1845,
"step": 143600
},
{
"epoch": 46.444731738849384,
"grad_norm": 1.6636964082717896,
"learning_rate": 0.001,
"loss": 1.1983,
"step": 143700
},
{
"epoch": 46.4770523594053,
"grad_norm": 1.3807815313339233,
"learning_rate": 0.001,
"loss": 1.2193,
"step": 143800
},
{
"epoch": 46.50937297996121,
"grad_norm": 1.3139708042144775,
"learning_rate": 0.001,
"loss": 1.205,
"step": 143900
},
{
"epoch": 46.54169360051713,
"grad_norm": 1.2375787496566772,
"learning_rate": 0.001,
"loss": 1.1898,
"step": 144000
},
{
"epoch": 46.57401422107304,
"grad_norm": 1.7440154552459717,
"learning_rate": 0.001,
"loss": 1.1944,
"step": 144100
},
{
"epoch": 46.60633484162896,
"grad_norm": 1.7930964231491089,
"learning_rate": 0.001,
"loss": 1.2046,
"step": 144200
},
{
"epoch": 46.63865546218487,
"grad_norm": 1.7479157447814941,
"learning_rate": 0.001,
"loss": 1.2002,
"step": 144300
},
{
"epoch": 46.670976082740786,
"grad_norm": 1.5789364576339722,
"learning_rate": 0.001,
"loss": 1.2219,
"step": 144400
},
{
"epoch": 46.7032967032967,
"grad_norm": 1.5130776166915894,
"learning_rate": 0.001,
"loss": 1.2434,
"step": 144500
},
{
"epoch": 46.735617323852615,
"grad_norm": 2.0466105937957764,
"learning_rate": 0.001,
"loss": 1.2327,
"step": 144600
},
{
"epoch": 46.76793794440853,
"grad_norm": 1.5826852321624756,
"learning_rate": 0.001,
"loss": 1.2467,
"step": 144700
},
{
"epoch": 46.800258564964444,
"grad_norm": 1.5535780191421509,
"learning_rate": 0.001,
"loss": 1.2317,
"step": 144800
},
{
"epoch": 46.83257918552036,
"grad_norm": 1.7607159614562988,
"learning_rate": 0.001,
"loss": 1.2328,
"step": 144900
},
{
"epoch": 46.864899806076274,
"grad_norm": 1.4072030782699585,
"learning_rate": 0.001,
"loss": 1.2437,
"step": 145000
},
{
"epoch": 46.89722042663219,
"grad_norm": 1.9758942127227783,
"learning_rate": 0.001,
"loss": 1.2378,
"step": 145100
},
{
"epoch": 46.9295410471881,
"grad_norm": 1.6869163513183594,
"learning_rate": 0.001,
"loss": 1.2454,
"step": 145200
},
{
"epoch": 46.96186166774402,
"grad_norm": 1.5331294536590576,
"learning_rate": 0.001,
"loss": 1.2447,
"step": 145300
},
{
"epoch": 46.99418228829994,
"grad_norm": 1.357105016708374,
"learning_rate": 0.001,
"loss": 1.2275,
"step": 145400
},
{
"epoch": 47.02650290885585,
"grad_norm": 1.6140611171722412,
"learning_rate": 0.001,
"loss": 1.1603,
"step": 145500
},
{
"epoch": 47.05882352941177,
"grad_norm": 2.081415891647339,
"learning_rate": 0.001,
"loss": 1.1452,
"step": 145600
},
{
"epoch": 47.09114414996768,
"grad_norm": 2.0446853637695312,
"learning_rate": 0.001,
"loss": 1.1433,
"step": 145700
},
{
"epoch": 47.1234647705236,
"grad_norm": 1.3056678771972656,
"learning_rate": 0.001,
"loss": 1.1422,
"step": 145800
},
{
"epoch": 47.15578539107951,
"grad_norm": 1.6712660789489746,
"learning_rate": 0.001,
"loss": 1.1523,
"step": 145900
},
{
"epoch": 47.188106011635426,
"grad_norm": 1.5658693313598633,
"learning_rate": 0.001,
"loss": 1.1758,
"step": 146000
},
{
"epoch": 47.22042663219134,
"grad_norm": 1.599502682685852,
"learning_rate": 0.001,
"loss": 1.1551,
"step": 146100
},
{
"epoch": 47.252747252747255,
"grad_norm": 1.624306559562683,
"learning_rate": 0.001,
"loss": 1.1577,
"step": 146200
},
{
"epoch": 47.28506787330317,
"grad_norm": 1.6202889680862427,
"learning_rate": 0.001,
"loss": 1.1607,
"step": 146300
},
{
"epoch": 47.317388493859085,
"grad_norm": 1.5435707569122314,
"learning_rate": 0.001,
"loss": 1.1746,
"step": 146400
},
{
"epoch": 47.349709114415,
"grad_norm": 1.4827113151550293,
"learning_rate": 0.001,
"loss": 1.1569,
"step": 146500
},
{
"epoch": 47.382029734970914,
"grad_norm": 1.2994977235794067,
"learning_rate": 0.001,
"loss": 1.1738,
"step": 146600
},
{
"epoch": 47.41435035552683,
"grad_norm": 1.5433175563812256,
"learning_rate": 0.001,
"loss": 1.1945,
"step": 146700
},
{
"epoch": 47.44667097608274,
"grad_norm": 1.61589777469635,
"learning_rate": 0.001,
"loss": 1.1804,
"step": 146800
},
{
"epoch": 47.47899159663866,
"grad_norm": 2.19016695022583,
"learning_rate": 0.001,
"loss": 1.1817,
"step": 146900
},
{
"epoch": 47.51131221719457,
"grad_norm": 1.3823586702346802,
"learning_rate": 0.001,
"loss": 1.2015,
"step": 147000
},
{
"epoch": 47.543632837750486,
"grad_norm": 1.8267431259155273,
"learning_rate": 0.001,
"loss": 1.1772,
"step": 147100
},
{
"epoch": 47.5759534583064,
"grad_norm": 1.4151540994644165,
"learning_rate": 0.001,
"loss": 1.1982,
"step": 147200
},
{
"epoch": 47.608274078862316,
"grad_norm": 1.7475345134735107,
"learning_rate": 0.001,
"loss": 1.1941,
"step": 147300
},
{
"epoch": 47.64059469941823,
"grad_norm": 1.7757151126861572,
"learning_rate": 0.001,
"loss": 1.2097,
"step": 147400
},
{
"epoch": 47.672915319974145,
"grad_norm": 1.5026477575302124,
"learning_rate": 0.001,
"loss": 1.2292,
"step": 147500
},
{
"epoch": 47.70523594053006,
"grad_norm": 1.9475871324539185,
"learning_rate": 0.001,
"loss": 1.231,
"step": 147600
},
{
"epoch": 47.737556561085974,
"grad_norm": 1.2796332836151123,
"learning_rate": 0.001,
"loss": 1.2198,
"step": 147700
},
{
"epoch": 47.76987718164189,
"grad_norm": 1.7256336212158203,
"learning_rate": 0.001,
"loss": 1.2134,
"step": 147800
},
{
"epoch": 47.8021978021978,
"grad_norm": 1.6882165670394897,
"learning_rate": 0.001,
"loss": 1.2361,
"step": 147900
},
{
"epoch": 47.83451842275372,
"grad_norm": 1.5431451797485352,
"learning_rate": 0.001,
"loss": 1.2174,
"step": 148000
},
{
"epoch": 47.86683904330963,
"grad_norm": 1.7564618587493896,
"learning_rate": 0.001,
"loss": 1.2215,
"step": 148100
},
{
"epoch": 47.89915966386555,
"grad_norm": 1.5725071430206299,
"learning_rate": 0.001,
"loss": 1.2261,
"step": 148200
},
{
"epoch": 47.93148028442146,
"grad_norm": 1.5619136095046997,
"learning_rate": 0.001,
"loss": 1.2253,
"step": 148300
},
{
"epoch": 47.963800904977376,
"grad_norm": 1.8927035331726074,
"learning_rate": 0.001,
"loss": 1.2225,
"step": 148400
},
{
"epoch": 47.99612152553329,
"grad_norm": 1.8232885599136353,
"learning_rate": 0.001,
"loss": 1.2266,
"step": 148500
},
{
"epoch": 48.028442146089205,
"grad_norm": 1.9754383563995361,
"learning_rate": 0.001,
"loss": 1.1366,
"step": 148600
},
{
"epoch": 48.06076276664512,
"grad_norm": 1.661736249923706,
"learning_rate": 0.001,
"loss": 1.1406,
"step": 148700
},
{
"epoch": 48.093083387201034,
"grad_norm": 1.6042405366897583,
"learning_rate": 0.001,
"loss": 1.1258,
"step": 148800
},
{
"epoch": 48.12540400775695,
"grad_norm": 1.820178508758545,
"learning_rate": 0.001,
"loss": 1.1516,
"step": 148900
},
{
"epoch": 48.15772462831286,
"grad_norm": 1.3801450729370117,
"learning_rate": 0.001,
"loss": 1.1606,
"step": 149000
},
{
"epoch": 48.19004524886878,
"grad_norm": 1.8100758790969849,
"learning_rate": 0.001,
"loss": 1.1324,
"step": 149100
},
{
"epoch": 48.22236586942469,
"grad_norm": 1.4243206977844238,
"learning_rate": 0.001,
"loss": 1.1639,
"step": 149200
},
{
"epoch": 48.25468648998061,
"grad_norm": 1.5422807931900024,
"learning_rate": 0.001,
"loss": 1.1675,
"step": 149300
},
{
"epoch": 48.28700711053652,
"grad_norm": 1.4906775951385498,
"learning_rate": 0.001,
"loss": 1.1498,
"step": 149400
},
{
"epoch": 48.319327731092436,
"grad_norm": 1.2655715942382812,
"learning_rate": 0.001,
"loss": 1.1568,
"step": 149500
},
{
"epoch": 48.35164835164835,
"grad_norm": 2.0237317085266113,
"learning_rate": 0.001,
"loss": 1.1634,
"step": 149600
},
{
"epoch": 48.383968972204265,
"grad_norm": 1.5441726446151733,
"learning_rate": 0.001,
"loss": 1.1673,
"step": 149700
},
{
"epoch": 48.41628959276018,
"grad_norm": 1.5430012941360474,
"learning_rate": 0.001,
"loss": 1.1761,
"step": 149800
},
{
"epoch": 48.448610213316094,
"grad_norm": 1.3982590436935425,
"learning_rate": 0.001,
"loss": 1.1713,
"step": 149900
},
{
"epoch": 48.48093083387201,
"grad_norm": 2.1957812309265137,
"learning_rate": 0.001,
"loss": 1.1768,
"step": 150000
},
{
"epoch": 48.51325145442792,
"grad_norm": 1.6471282243728638,
"learning_rate": 0.001,
"loss": 1.1628,
"step": 150100
},
{
"epoch": 48.54557207498384,
"grad_norm": 1.5349053144454956,
"learning_rate": 0.001,
"loss": 1.1756,
"step": 150200
},
{
"epoch": 48.57789269553975,
"grad_norm": 1.5751968622207642,
"learning_rate": 0.001,
"loss": 1.1746,
"step": 150300
},
{
"epoch": 48.61021331609567,
"grad_norm": 1.6446707248687744,
"learning_rate": 0.001,
"loss": 1.1901,
"step": 150400
},
{
"epoch": 48.64253393665158,
"grad_norm": 1.5992244482040405,
"learning_rate": 0.001,
"loss": 1.2006,
"step": 150500
},
{
"epoch": 48.674854557207496,
"grad_norm": 1.4150946140289307,
"learning_rate": 0.001,
"loss": 1.1895,
"step": 150600
},
{
"epoch": 48.70717517776341,
"grad_norm": 2.0123841762542725,
"learning_rate": 0.001,
"loss": 1.2026,
"step": 150700
},
{
"epoch": 48.739495798319325,
"grad_norm": 1.5513368844985962,
"learning_rate": 0.001,
"loss": 1.2164,
"step": 150800
},
{
"epoch": 48.77181641887524,
"grad_norm": 1.7306681871414185,
"learning_rate": 0.001,
"loss": 1.2243,
"step": 150900
},
{
"epoch": 48.804137039431154,
"grad_norm": 1.7530173063278198,
"learning_rate": 0.001,
"loss": 1.1997,
"step": 151000
},
{
"epoch": 48.83645765998707,
"grad_norm": 1.6544743776321411,
"learning_rate": 0.001,
"loss": 1.2087,
"step": 151100
},
{
"epoch": 48.86877828054298,
"grad_norm": 1.404657006263733,
"learning_rate": 0.001,
"loss": 1.2177,
"step": 151200
},
{
"epoch": 48.9010989010989,
"grad_norm": 1.3040798902511597,
"learning_rate": 0.001,
"loss": 1.2182,
"step": 151300
},
{
"epoch": 48.93341952165481,
"grad_norm": 1.7273650169372559,
"learning_rate": 0.001,
"loss": 1.2262,
"step": 151400
},
{
"epoch": 48.96574014221073,
"grad_norm": 1.721778154373169,
"learning_rate": 0.001,
"loss": 1.2277,
"step": 151500
},
{
"epoch": 48.99806076276664,
"grad_norm": 2.63419246673584,
"learning_rate": 0.001,
"loss": 1.1779,
"step": 151600
},
{
"epoch": 49.03038138332256,
"grad_norm": 1.5806034803390503,
"learning_rate": 0.001,
"loss": 1.1105,
"step": 151700
},
{
"epoch": 49.06270200387848,
"grad_norm": 1.3670389652252197,
"learning_rate": 0.001,
"loss": 1.123,
"step": 151800
},
{
"epoch": 49.09502262443439,
"grad_norm": 2.055310010910034,
"learning_rate": 0.001,
"loss": 1.1219,
"step": 151900
},
{
"epoch": 49.12734324499031,
"grad_norm": 1.7734863758087158,
"learning_rate": 0.001,
"loss": 1.1289,
"step": 152000
},
{
"epoch": 49.15966386554622,
"grad_norm": 1.5310850143432617,
"learning_rate": 0.001,
"loss": 1.1494,
"step": 152100
},
{
"epoch": 49.191984486102136,
"grad_norm": 1.7340278625488281,
"learning_rate": 0.001,
"loss": 1.1401,
"step": 152200
},
{
"epoch": 49.22430510665805,
"grad_norm": 1.797639012336731,
"learning_rate": 0.001,
"loss": 1.1599,
"step": 152300
},
{
"epoch": 49.256625727213965,
"grad_norm": 1.9868476390838623,
"learning_rate": 0.001,
"loss": 1.1398,
"step": 152400
},
{
"epoch": 49.28894634776988,
"grad_norm": 1.5633907318115234,
"learning_rate": 0.001,
"loss": 1.1491,
"step": 152500
},
{
"epoch": 49.321266968325794,
"grad_norm": 1.5145635604858398,
"learning_rate": 0.001,
"loss": 1.1577,
"step": 152600
},
{
"epoch": 49.35358758888171,
"grad_norm": 1.5028250217437744,
"learning_rate": 0.001,
"loss": 1.1317,
"step": 152700
},
{
"epoch": 49.38590820943762,
"grad_norm": 1.316331148147583,
"learning_rate": 0.001,
"loss": 1.1653,
"step": 152800
},
{
"epoch": 49.41822882999354,
"grad_norm": 1.6597394943237305,
"learning_rate": 0.001,
"loss": 1.171,
"step": 152900
},
{
"epoch": 49.45054945054945,
"grad_norm": 1.6509952545166016,
"learning_rate": 0.001,
"loss": 1.1547,
"step": 153000
},
{
"epoch": 49.48287007110537,
"grad_norm": 1.3771039247512817,
"learning_rate": 0.001,
"loss": 1.1643,
"step": 153100
},
{
"epoch": 49.51519069166128,
"grad_norm": 1.25186026096344,
"learning_rate": 0.001,
"loss": 1.1755,
"step": 153200
},
{
"epoch": 49.547511312217196,
"grad_norm": 1.5495226383209229,
"learning_rate": 0.001,
"loss": 1.176,
"step": 153300
},
{
"epoch": 49.57983193277311,
"grad_norm": 1.9336113929748535,
"learning_rate": 0.001,
"loss": 1.168,
"step": 153400
},
{
"epoch": 49.612152553329025,
"grad_norm": 1.4076240062713623,
"learning_rate": 0.001,
"loss": 1.1858,
"step": 153500
},
{
"epoch": 49.64447317388494,
"grad_norm": 1.9218026399612427,
"learning_rate": 0.001,
"loss": 1.185,
"step": 153600
},
{
"epoch": 49.676793794440854,
"grad_norm": 1.3178162574768066,
"learning_rate": 0.001,
"loss": 1.1826,
"step": 153700
},
{
"epoch": 49.70911441499677,
"grad_norm": 1.7671688795089722,
"learning_rate": 0.001,
"loss": 1.2037,
"step": 153800
},
{
"epoch": 49.74143503555268,
"grad_norm": 1.5962954759597778,
"learning_rate": 0.001,
"loss": 1.1803,
"step": 153900
},
{
"epoch": 49.7737556561086,
"grad_norm": 1.9002434015274048,
"learning_rate": 0.001,
"loss": 1.1916,
"step": 154000
},
{
"epoch": 49.80607627666451,
"grad_norm": 1.387270450592041,
"learning_rate": 0.001,
"loss": 1.1982,
"step": 154100
},
{
"epoch": 49.83839689722043,
"grad_norm": 1.8704843521118164,
"learning_rate": 0.001,
"loss": 1.1858,
"step": 154200
},
{
"epoch": 49.87071751777634,
"grad_norm": 1.5286774635314941,
"learning_rate": 0.001,
"loss": 1.19,
"step": 154300
},
{
"epoch": 49.903038138332256,
"grad_norm": 1.6860578060150146,
"learning_rate": 0.001,
"loss": 1.2088,
"step": 154400
},
{
"epoch": 49.93535875888817,
"grad_norm": 1.7764240503311157,
"learning_rate": 0.001,
"loss": 1.214,
"step": 154500
},
{
"epoch": 49.967679379444085,
"grad_norm": 1.661325454711914,
"learning_rate": 0.001,
"loss": 1.2037,
"step": 154600
},
{
"epoch": 50.0,
"grad_norm": 1.262018084526062,
"learning_rate": 0.001,
"loss": 1.1552,
"step": 154700
},
{
"epoch": 50.032320620555915,
"grad_norm": 0.9459092617034912,
"learning_rate": 0.001,
"loss": 1.1055,
"step": 154800
},
{
"epoch": 50.06464124111183,
"grad_norm": 1.3187966346740723,
"learning_rate": 0.001,
"loss": 1.1186,
"step": 154900
},
{
"epoch": 50.096961861667744,
"grad_norm": 0.8036020398139954,
"learning_rate": 0.001,
"loss": 1.1067,
"step": 155000
},
{
"epoch": 50.12928248222366,
"grad_norm": 0.9802110195159912,
"learning_rate": 0.001,
"loss": 1.1072,
"step": 155100
},
{
"epoch": 50.16160310277957,
"grad_norm": 0.7400931715965271,
"learning_rate": 0.001,
"loss": 1.1048,
"step": 155200
},
{
"epoch": 50.19392372333549,
"grad_norm": 0.8337299823760986,
"learning_rate": 0.001,
"loss": 1.1185,
"step": 155300
},
{
"epoch": 50.2262443438914,
"grad_norm": 0.857559859752655,
"learning_rate": 0.001,
"loss": 1.1487,
"step": 155400
},
{
"epoch": 50.25856496444732,
"grad_norm": 0.9556663632392883,
"learning_rate": 0.001,
"loss": 1.1425,
"step": 155500
},
{
"epoch": 50.29088558500323,
"grad_norm": 1.762825846672058,
"learning_rate": 0.001,
"loss": 1.1336,
"step": 155600
},
{
"epoch": 50.323206205559146,
"grad_norm": 0.46932679414749146,
"learning_rate": 0.001,
"loss": 1.1445,
"step": 155700
},
{
"epoch": 50.35552682611506,
"grad_norm": 0.46966707706451416,
"learning_rate": 0.001,
"loss": 1.1263,
"step": 155800
},
{
"epoch": 50.387847446670975,
"grad_norm": 0.8840547204017639,
"learning_rate": 0.001,
"loss": 1.1478,
"step": 155900
},
{
"epoch": 50.42016806722689,
"grad_norm": 0.2904863655567169,
"learning_rate": 0.001,
"loss": 1.1441,
"step": 156000
},
{
"epoch": 50.452488687782804,
"grad_norm": 0.5406709909439087,
"learning_rate": 0.001,
"loss": 1.1493,
"step": 156100
},
{
"epoch": 50.48480930833872,
"grad_norm": 0.8404807448387146,
"learning_rate": 0.001,
"loss": 1.1593,
"step": 156200
},
{
"epoch": 50.51712992889463,
"grad_norm": 0.6578356623649597,
"learning_rate": 0.001,
"loss": 1.1599,
"step": 156300
},
{
"epoch": 50.54945054945055,
"grad_norm": 1.5182676315307617,
"learning_rate": 0.001,
"loss": 1.1606,
"step": 156400
},
{
"epoch": 50.58177117000646,
"grad_norm": 0.5784684419631958,
"learning_rate": 0.001,
"loss": 1.1615,
"step": 156500
},
{
"epoch": 50.61409179056238,
"grad_norm": 0.5010991096496582,
"learning_rate": 0.001,
"loss": 1.182,
"step": 156600
},
{
"epoch": 50.64641241111829,
"grad_norm": 1.0248159170150757,
"learning_rate": 0.001,
"loss": 1.157,
"step": 156700
},
{
"epoch": 50.678733031674206,
"grad_norm": 0.4912484884262085,
"learning_rate": 0.001,
"loss": 1.1554,
"step": 156800
},
{
"epoch": 50.71105365223012,
"grad_norm": 1.007466197013855,
"learning_rate": 0.001,
"loss": 1.2062,
"step": 156900
},
{
"epoch": 50.743374272786035,
"grad_norm": 0.6946946978569031,
"learning_rate": 0.001,
"loss": 1.1865,
"step": 157000
},
{
"epoch": 50.77569489334195,
"grad_norm": 1.3229173421859741,
"learning_rate": 0.001,
"loss": 1.2058,
"step": 157100
},
{
"epoch": 50.808015513897864,
"grad_norm": 0.8402013778686523,
"learning_rate": 0.001,
"loss": 1.1784,
"step": 157200
},
{
"epoch": 50.84033613445378,
"grad_norm": 0.7068992257118225,
"learning_rate": 0.001,
"loss": 1.1889,
"step": 157300
},
{
"epoch": 50.87265675500969,
"grad_norm": 1.3328818082809448,
"learning_rate": 0.001,
"loss": 1.1871,
"step": 157400
},
{
"epoch": 50.90497737556561,
"grad_norm": 0.7360666990280151,
"learning_rate": 0.001,
"loss": 1.1767,
"step": 157500
},
{
"epoch": 50.93729799612152,
"grad_norm": 0.4994688928127289,
"learning_rate": 0.001,
"loss": 1.1876,
"step": 157600
},
{
"epoch": 50.96961861667744,
"grad_norm": 0.7623894214630127,
"learning_rate": 0.001,
"loss": 1.2043,
"step": 157700
},
{
"epoch": 51.00193923723336,
"grad_norm": 1.4419955015182495,
"learning_rate": 0.001,
"loss": 1.2096,
"step": 157800
},
{
"epoch": 51.03425985778927,
"grad_norm": 1.7683582305908203,
"learning_rate": 0.001,
"loss": 1.084,
"step": 157900
},
{
"epoch": 51.06658047834519,
"grad_norm": 1.7946703433990479,
"learning_rate": 0.001,
"loss": 1.09,
"step": 158000
},
{
"epoch": 51.0989010989011,
"grad_norm": 1.7804067134857178,
"learning_rate": 0.001,
"loss": 1.1085,
"step": 158100
},
{
"epoch": 51.13122171945702,
"grad_norm": 2.1206374168395996,
"learning_rate": 0.001,
"loss": 1.0934,
"step": 158200
},
{
"epoch": 51.16354234001293,
"grad_norm": 1.6893900632858276,
"learning_rate": 0.001,
"loss": 1.1113,
"step": 158300
},
{
"epoch": 51.195862960568846,
"grad_norm": 1.67825186252594,
"learning_rate": 0.001,
"loss": 1.1348,
"step": 158400
},
{
"epoch": 51.22818358112476,
"grad_norm": 1.5735834836959839,
"learning_rate": 0.001,
"loss": 1.1227,
"step": 158500
},
{
"epoch": 51.260504201680675,
"grad_norm": 1.5846762657165527,
"learning_rate": 0.001,
"loss": 1.1272,
"step": 158600
},
{
"epoch": 51.29282482223659,
"grad_norm": 1.7856253385543823,
"learning_rate": 0.001,
"loss": 1.1188,
"step": 158700
},
{
"epoch": 51.325145442792504,
"grad_norm": 1.580972671508789,
"learning_rate": 0.001,
"loss": 1.1211,
"step": 158800
},
{
"epoch": 51.35746606334842,
"grad_norm": 1.6010007858276367,
"learning_rate": 0.001,
"loss": 1.1172,
"step": 158900
},
{
"epoch": 51.38978668390433,
"grad_norm": 1.87574303150177,
"learning_rate": 0.001,
"loss": 1.1303,
"step": 159000
},
{
"epoch": 51.42210730446025,
"grad_norm": 1.657053828239441,
"learning_rate": 0.001,
"loss": 1.1479,
"step": 159100
},
{
"epoch": 51.45442792501616,
"grad_norm": 1.626969814300537,
"learning_rate": 0.001,
"loss": 1.1429,
"step": 159200
},
{
"epoch": 51.48674854557208,
"grad_norm": 1.7134428024291992,
"learning_rate": 0.001,
"loss": 1.1462,
"step": 159300
},
{
"epoch": 51.51906916612799,
"grad_norm": 1.6238813400268555,
"learning_rate": 0.001,
"loss": 1.1381,
"step": 159400
},
{
"epoch": 51.551389786683906,
"grad_norm": 1.773478388786316,
"learning_rate": 0.001,
"loss": 1.1606,
"step": 159500
},
{
"epoch": 51.58371040723982,
"grad_norm": 1.8276687860488892,
"learning_rate": 0.001,
"loss": 1.1531,
"step": 159600
},
{
"epoch": 51.616031027795735,
"grad_norm": 1.6656960248947144,
"learning_rate": 0.001,
"loss": 1.167,
"step": 159700
},
{
"epoch": 51.64835164835165,
"grad_norm": 1.8340784311294556,
"learning_rate": 0.001,
"loss": 1.1476,
"step": 159800
},
{
"epoch": 51.680672268907564,
"grad_norm": 2.128314256668091,
"learning_rate": 0.001,
"loss": 1.1672,
"step": 159900
},
{
"epoch": 51.71299288946348,
"grad_norm": 1.9760433435440063,
"learning_rate": 0.001,
"loss": 1.1584,
"step": 160000
},
{
"epoch": 51.74531351001939,
"grad_norm": 2.1876275539398193,
"learning_rate": 0.001,
"loss": 1.1566,
"step": 160100
},
{
"epoch": 51.77763413057531,
"grad_norm": 1.6389830112457275,
"learning_rate": 0.001,
"loss": 1.1819,
"step": 160200
},
{
"epoch": 51.80995475113122,
"grad_norm": 1.499732255935669,
"learning_rate": 0.001,
"loss": 1.1732,
"step": 160300
},
{
"epoch": 51.84227537168714,
"grad_norm": 4.384315013885498,
"learning_rate": 0.001,
"loss": 1.1634,
"step": 160400
},
{
"epoch": 51.87459599224305,
"grad_norm": 1.833261489868164,
"learning_rate": 0.001,
"loss": 1.1801,
"step": 160500
},
{
"epoch": 51.906916612798966,
"grad_norm": 2.3746957778930664,
"learning_rate": 0.001,
"loss": 1.1738,
"step": 160600
},
{
"epoch": 51.93923723335488,
"grad_norm": 2.942558526992798,
"learning_rate": 0.001,
"loss": 1.1719,
"step": 160700
},
{
"epoch": 51.971557853910795,
"grad_norm": 1.7185624837875366,
"learning_rate": 0.001,
"loss": 1.2128,
"step": 160800
},
{
"epoch": 52.00387847446671,
"grad_norm": 1.84402334690094,
"learning_rate": 0.001,
"loss": 1.2243,
"step": 160900
},
{
"epoch": 52.036199095022624,
"grad_norm": 1.704215407371521,
"learning_rate": 0.001,
"loss": 1.0773,
"step": 161000
},
{
"epoch": 52.06851971557854,
"grad_norm": 1.5075286626815796,
"learning_rate": 0.001,
"loss": 1.0701,
"step": 161100
},
{
"epoch": 52.10084033613445,
"grad_norm": 1.6493642330169678,
"learning_rate": 0.001,
"loss": 1.1066,
"step": 161200
},
{
"epoch": 52.13316095669037,
"grad_norm": 1.6777901649475098,
"learning_rate": 0.001,
"loss": 1.1036,
"step": 161300
},
{
"epoch": 52.16548157724628,
"grad_norm": 1.7006093263626099,
"learning_rate": 0.001,
"loss": 1.092,
"step": 161400
},
{
"epoch": 52.1978021978022,
"grad_norm": 2.084979295730591,
"learning_rate": 0.001,
"loss": 1.1008,
"step": 161500
},
{
"epoch": 52.23012281835811,
"grad_norm": 1.5530115365982056,
"learning_rate": 0.001,
"loss": 1.1272,
"step": 161600
},
{
"epoch": 52.262443438914026,
"grad_norm": 1.7012377977371216,
"learning_rate": 0.001,
"loss": 1.1185,
"step": 161700
},
{
"epoch": 52.29476405946994,
"grad_norm": 1.9061259031295776,
"learning_rate": 0.001,
"loss": 1.1186,
"step": 161800
},
{
"epoch": 52.327084680025855,
"grad_norm": 1.660539984703064,
"learning_rate": 0.001,
"loss": 1.1255,
"step": 161900
},
{
"epoch": 52.35940530058177,
"grad_norm": 1.4890267848968506,
"learning_rate": 0.001,
"loss": 1.1091,
"step": 162000
},
{
"epoch": 52.391725921137684,
"grad_norm": 1.545951008796692,
"learning_rate": 0.001,
"loss": 1.1293,
"step": 162100
},
{
"epoch": 52.4240465416936,
"grad_norm": 1.6420233249664307,
"learning_rate": 0.001,
"loss": 1.1157,
"step": 162200
},
{
"epoch": 52.456367162249514,
"grad_norm": 1.496561050415039,
"learning_rate": 0.001,
"loss": 1.1222,
"step": 162300
},
{
"epoch": 52.48868778280543,
"grad_norm": 1.982618808746338,
"learning_rate": 0.001,
"loss": 1.1287,
"step": 162400
},
{
"epoch": 52.52100840336134,
"grad_norm": 1.955478549003601,
"learning_rate": 0.001,
"loss": 1.131,
"step": 162500
},
{
"epoch": 52.55332902391726,
"grad_norm": 1.3720135688781738,
"learning_rate": 0.001,
"loss": 1.1394,
"step": 162600
},
{
"epoch": 52.58564964447317,
"grad_norm": 1.5025126934051514,
"learning_rate": 0.001,
"loss": 1.1357,
"step": 162700
},
{
"epoch": 52.617970265029086,
"grad_norm": 1.3566020727157593,
"learning_rate": 0.001,
"loss": 1.1557,
"step": 162800
},
{
"epoch": 52.650290885585,
"grad_norm": 1.7265077829360962,
"learning_rate": 0.001,
"loss": 1.1595,
"step": 162900
},
{
"epoch": 52.682611506140915,
"grad_norm": 2.398094654083252,
"learning_rate": 0.001,
"loss": 1.1625,
"step": 163000
},
{
"epoch": 52.71493212669683,
"grad_norm": 1.7664134502410889,
"learning_rate": 0.001,
"loss": 1.1529,
"step": 163100
},
{
"epoch": 52.747252747252745,
"grad_norm": 1.6189780235290527,
"learning_rate": 0.001,
"loss": 1.1595,
"step": 163200
},
{
"epoch": 52.77957336780866,
"grad_norm": 1.4175572395324707,
"learning_rate": 0.001,
"loss": 1.1622,
"step": 163300
},
{
"epoch": 52.811893988364574,
"grad_norm": 1.6346251964569092,
"learning_rate": 0.001,
"loss": 1.1434,
"step": 163400
},
{
"epoch": 52.84421460892049,
"grad_norm": 1.599571704864502,
"learning_rate": 0.001,
"loss": 1.1716,
"step": 163500
},
{
"epoch": 52.8765352294764,
"grad_norm": 1.4147549867630005,
"learning_rate": 0.001,
"loss": 1.1765,
"step": 163600
},
{
"epoch": 52.90885585003232,
"grad_norm": 2.264082431793213,
"learning_rate": 0.001,
"loss": 1.1803,
"step": 163700
},
{
"epoch": 52.94117647058823,
"grad_norm": 1.427462100982666,
"learning_rate": 0.001,
"loss": 1.1808,
"step": 163800
},
{
"epoch": 52.97349709114415,
"grad_norm": 1.7261021137237549,
"learning_rate": 0.001,
"loss": 1.1801,
"step": 163900
},
{
"epoch": 53.00581771170007,
"grad_norm": 1.4574629068374634,
"learning_rate": 0.001,
"loss": 1.1513,
"step": 164000
},
{
"epoch": 53.03813833225598,
"grad_norm": 1.341528058052063,
"learning_rate": 0.001,
"loss": 1.0847,
"step": 164100
},
{
"epoch": 53.0704589528119,
"grad_norm": 1.2390453815460205,
"learning_rate": 0.001,
"loss": 1.0535,
"step": 164200
},
{
"epoch": 53.10277957336781,
"grad_norm": 1.554686427116394,
"learning_rate": 0.001,
"loss": 1.0839,
"step": 164300
},
{
"epoch": 53.135100193923726,
"grad_norm": 1.6500283479690552,
"learning_rate": 0.001,
"loss": 1.0895,
"step": 164400
},
{
"epoch": 53.16742081447964,
"grad_norm": 1.352728247642517,
"learning_rate": 0.001,
"loss": 1.1027,
"step": 164500
},
{
"epoch": 53.199741435035556,
"grad_norm": 1.5491443872451782,
"learning_rate": 0.001,
"loss": 1.0904,
"step": 164600
},
{
"epoch": 53.23206205559147,
"grad_norm": 1.5412120819091797,
"learning_rate": 0.001,
"loss": 1.096,
"step": 164700
},
{
"epoch": 53.264382676147385,
"grad_norm": 1.3460454940795898,
"learning_rate": 0.001,
"loss": 1.1034,
"step": 164800
},
{
"epoch": 53.2967032967033,
"grad_norm": 1.4456019401550293,
"learning_rate": 0.001,
"loss": 1.1198,
"step": 164900
},
{
"epoch": 53.329023917259214,
"grad_norm": 1.5322580337524414,
"learning_rate": 0.001,
"loss": 1.13,
"step": 165000
},
{
"epoch": 53.36134453781513,
"grad_norm": 1.5252654552459717,
"learning_rate": 0.001,
"loss": 1.1037,
"step": 165100
},
{
"epoch": 53.39366515837104,
"grad_norm": 1.5594348907470703,
"learning_rate": 0.001,
"loss": 1.1234,
"step": 165200
},
{
"epoch": 53.42598577892696,
"grad_norm": 1.6623215675354004,
"learning_rate": 0.001,
"loss": 1.1315,
"step": 165300
},
{
"epoch": 53.45830639948287,
"grad_norm": 1.8282194137573242,
"learning_rate": 0.001,
"loss": 1.1151,
"step": 165400
},
{
"epoch": 53.49062702003879,
"grad_norm": 1.407645344734192,
"learning_rate": 0.001,
"loss": 1.12,
"step": 165500
},
{
"epoch": 53.5229476405947,
"grad_norm": 1.8748985528945923,
"learning_rate": 0.001,
"loss": 1.1289,
"step": 165600
},
{
"epoch": 53.555268261150616,
"grad_norm": 1.391897201538086,
"learning_rate": 0.001,
"loss": 1.1274,
"step": 165700
},
{
"epoch": 53.58758888170653,
"grad_norm": 1.3943778276443481,
"learning_rate": 0.001,
"loss": 1.1019,
"step": 165800
},
{
"epoch": 53.619909502262445,
"grad_norm": 1.5813536643981934,
"learning_rate": 0.001,
"loss": 1.134,
"step": 165900
},
{
"epoch": 53.65223012281836,
"grad_norm": 2.096243143081665,
"learning_rate": 0.001,
"loss": 1.1326,
"step": 166000
},
{
"epoch": 53.684550743374274,
"grad_norm": 1.64535653591156,
"learning_rate": 0.001,
"loss": 1.1487,
"step": 166100
},
{
"epoch": 53.71687136393019,
"grad_norm": 1.6284700632095337,
"learning_rate": 0.001,
"loss": 1.1465,
"step": 166200
},
{
"epoch": 53.7491919844861,
"grad_norm": 1.598093867301941,
"learning_rate": 0.001,
"loss": 1.1414,
"step": 166300
},
{
"epoch": 53.78151260504202,
"grad_norm": 1.3568588495254517,
"learning_rate": 0.001,
"loss": 1.1602,
"step": 166400
},
{
"epoch": 53.81383322559793,
"grad_norm": 1.6616982221603394,
"learning_rate": 0.001,
"loss": 1.1318,
"step": 166500
},
{
"epoch": 53.84615384615385,
"grad_norm": 1.5817769765853882,
"learning_rate": 0.001,
"loss": 1.1405,
"step": 166600
},
{
"epoch": 53.87847446670976,
"grad_norm": 1.7250040769577026,
"learning_rate": 0.001,
"loss": 1.1601,
"step": 166700
},
{
"epoch": 53.910795087265676,
"grad_norm": 1.6254827976226807,
"learning_rate": 0.001,
"loss": 1.1653,
"step": 166800
},
{
"epoch": 53.94311570782159,
"grad_norm": 1.3446640968322754,
"learning_rate": 0.001,
"loss": 1.1598,
"step": 166900
},
{
"epoch": 53.975436328377505,
"grad_norm": 1.6438404321670532,
"learning_rate": 0.001,
"loss": 1.1655,
"step": 167000
},
{
"epoch": 54.00775694893342,
"grad_norm": 1.505845069885254,
"learning_rate": 0.001,
"loss": 1.1783,
"step": 167100
},
{
"epoch": 54.040077569489334,
"grad_norm": 1.787855863571167,
"learning_rate": 0.001,
"loss": 1.062,
"step": 167200
},
{
"epoch": 54.07239819004525,
"grad_norm": 1.8118529319763184,
"learning_rate": 0.001,
"loss": 1.0605,
"step": 167300
},
{
"epoch": 54.10471881060116,
"grad_norm": 2.0226893424987793,
"learning_rate": 0.001,
"loss": 1.0748,
"step": 167400
},
{
"epoch": 54.13703943115708,
"grad_norm": 1.669325590133667,
"learning_rate": 0.001,
"loss": 1.063,
"step": 167500
},
{
"epoch": 54.16936005171299,
"grad_norm": 1.4700037240982056,
"learning_rate": 0.001,
"loss": 1.073,
"step": 167600
},
{
"epoch": 54.20168067226891,
"grad_norm": 1.7410122156143188,
"learning_rate": 0.001,
"loss": 1.0722,
"step": 167700
},
{
"epoch": 54.23400129282482,
"grad_norm": 1.518440842628479,
"learning_rate": 0.001,
"loss": 1.0926,
"step": 167800
},
{
"epoch": 54.266321913380736,
"grad_norm": 1.6282093524932861,
"learning_rate": 0.001,
"loss": 1.0978,
"step": 167900
},
{
"epoch": 54.29864253393665,
"grad_norm": 1.4921144247055054,
"learning_rate": 0.001,
"loss": 1.0918,
"step": 168000
},
{
"epoch": 54.330963154492565,
"grad_norm": 1.4011250734329224,
"learning_rate": 0.001,
"loss": 1.102,
"step": 168100
},
{
"epoch": 54.36328377504848,
"grad_norm": 1.6125963926315308,
"learning_rate": 0.001,
"loss": 1.0986,
"step": 168200
},
{
"epoch": 54.395604395604394,
"grad_norm": 2.5859177112579346,
"learning_rate": 0.001,
"loss": 1.1138,
"step": 168300
},
{
"epoch": 54.42792501616031,
"grad_norm": 1.9405006170272827,
"learning_rate": 0.001,
"loss": 1.1107,
"step": 168400
},
{
"epoch": 54.46024563671622,
"grad_norm": 1.813139796257019,
"learning_rate": 0.001,
"loss": 1.102,
"step": 168500
},
{
"epoch": 54.49256625727214,
"grad_norm": 1.5077065229415894,
"learning_rate": 0.001,
"loss": 1.1122,
"step": 168600
},
{
"epoch": 54.52488687782805,
"grad_norm": 1.3443224430084229,
"learning_rate": 0.001,
"loss": 1.1176,
"step": 168700
},
{
"epoch": 54.55720749838397,
"grad_norm": 1.6900509595870972,
"learning_rate": 0.001,
"loss": 1.1348,
"step": 168800
},
{
"epoch": 54.58952811893988,
"grad_norm": 1.4742830991744995,
"learning_rate": 0.001,
"loss": 1.1336,
"step": 168900
},
{
"epoch": 54.621848739495796,
"grad_norm": 1.719673752784729,
"learning_rate": 0.001,
"loss": 1.1237,
"step": 169000
},
{
"epoch": 54.65416936005171,
"grad_norm": 1.4414795637130737,
"learning_rate": 0.001,
"loss": 1.1374,
"step": 169100
},
{
"epoch": 54.686489980607625,
"grad_norm": 1.6505208015441895,
"learning_rate": 0.001,
"loss": 1.1369,
"step": 169200
},
{
"epoch": 54.71881060116354,
"grad_norm": 1.6553674936294556,
"learning_rate": 0.001,
"loss": 1.111,
"step": 169300
},
{
"epoch": 54.751131221719454,
"grad_norm": 1.5446336269378662,
"learning_rate": 0.001,
"loss": 1.1443,
"step": 169400
},
{
"epoch": 54.78345184227537,
"grad_norm": 1.5632305145263672,
"learning_rate": 0.001,
"loss": 1.1349,
"step": 169500
},
{
"epoch": 54.81577246283128,
"grad_norm": 1.4174524545669556,
"learning_rate": 0.001,
"loss": 1.1394,
"step": 169600
},
{
"epoch": 54.8480930833872,
"grad_norm": 1.5752019882202148,
"learning_rate": 0.001,
"loss": 1.1356,
"step": 169700
},
{
"epoch": 54.88041370394311,
"grad_norm": 3.077443838119507,
"learning_rate": 0.001,
"loss": 1.1346,
"step": 169800
},
{
"epoch": 54.91273432449903,
"grad_norm": 1.3338394165039062,
"learning_rate": 0.001,
"loss": 1.1594,
"step": 169900
},
{
"epoch": 54.94505494505494,
"grad_norm": 1.565075397491455,
"learning_rate": 0.001,
"loss": 1.1667,
"step": 170000
},
{
"epoch": 54.977375565610856,
"grad_norm": 1.5464483499526978,
"learning_rate": 0.001,
"loss": 1.1592,
"step": 170100
},
{
"epoch": 55.00969618616678,
"grad_norm": 1.780253529548645,
"learning_rate": 0.001,
"loss": 1.1241,
"step": 170200
},
{
"epoch": 55.04201680672269,
"grad_norm": 1.3855479955673218,
"learning_rate": 0.001,
"loss": 1.0466,
"step": 170300
},
{
"epoch": 55.07433742727861,
"grad_norm": 1.5501574277877808,
"learning_rate": 0.001,
"loss": 1.0559,
"step": 170400
},
{
"epoch": 55.10665804783452,
"grad_norm": 1.5003669261932373,
"learning_rate": 0.001,
"loss": 1.0591,
"step": 170500
},
{
"epoch": 55.138978668390436,
"grad_norm": 1.529256820678711,
"learning_rate": 0.001,
"loss": 1.0588,
"step": 170600
},
{
"epoch": 55.17129928894635,
"grad_norm": 1.6695269346237183,
"learning_rate": 0.001,
"loss": 1.0455,
"step": 170700
},
{
"epoch": 55.203619909502265,
"grad_norm": 1.1769028902053833,
"learning_rate": 0.001,
"loss": 1.0813,
"step": 170800
},
{
"epoch": 55.23594053005818,
"grad_norm": 1.33823823928833,
"learning_rate": 0.001,
"loss": 1.0969,
"step": 170900
},
{
"epoch": 55.268261150614094,
"grad_norm": 1.3728437423706055,
"learning_rate": 0.001,
"loss": 1.0731,
"step": 171000
},
{
"epoch": 55.30058177117001,
"grad_norm": 1.431822419166565,
"learning_rate": 0.001,
"loss": 1.0707,
"step": 171100
},
{
"epoch": 55.33290239172592,
"grad_norm": 1.5498169660568237,
"learning_rate": 0.001,
"loss": 1.0905,
"step": 171200
},
{
"epoch": 55.36522301228184,
"grad_norm": 1.598392367362976,
"learning_rate": 0.001,
"loss": 1.0835,
"step": 171300
},
{
"epoch": 55.39754363283775,
"grad_norm": 1.8495707511901855,
"learning_rate": 0.001,
"loss": 1.1046,
"step": 171400
},
{
"epoch": 55.42986425339367,
"grad_norm": 1.452956199645996,
"learning_rate": 0.001,
"loss": 1.1048,
"step": 171500
},
{
"epoch": 55.46218487394958,
"grad_norm": 1.5427874326705933,
"learning_rate": 0.001,
"loss": 1.0985,
"step": 171600
},
{
"epoch": 55.494505494505496,
"grad_norm": 1.5331876277923584,
"learning_rate": 0.001,
"loss": 1.0998,
"step": 171700
},
{
"epoch": 55.52682611506141,
"grad_norm": 1.4414092302322388,
"learning_rate": 0.001,
"loss": 1.0987,
"step": 171800
},
{
"epoch": 55.559146735617325,
"grad_norm": 1.469769835472107,
"learning_rate": 0.001,
"loss": 1.1062,
"step": 171900
},
{
"epoch": 55.59146735617324,
"grad_norm": 1.6705925464630127,
"learning_rate": 0.001,
"loss": 1.121,
"step": 172000
},
{
"epoch": 55.623787976729155,
"grad_norm": 2.0680267810821533,
"learning_rate": 0.001,
"loss": 1.1212,
"step": 172100
},
{
"epoch": 55.65610859728507,
"grad_norm": 1.1828604936599731,
"learning_rate": 0.001,
"loss": 1.1022,
"step": 172200
},
{
"epoch": 55.688429217840984,
"grad_norm": 2.1064138412475586,
"learning_rate": 0.001,
"loss": 1.1253,
"step": 172300
},
{
"epoch": 55.7207498383969,
"grad_norm": 1.25071382522583,
"learning_rate": 0.001,
"loss": 1.1336,
"step": 172400
},
{
"epoch": 55.75307045895281,
"grad_norm": 1.4712039232254028,
"learning_rate": 0.001,
"loss": 1.1207,
"step": 172500
},
{
"epoch": 55.78539107950873,
"grad_norm": 1.976820468902588,
"learning_rate": 0.001,
"loss": 1.127,
"step": 172600
},
{
"epoch": 55.81771170006464,
"grad_norm": 1.6924251317977905,
"learning_rate": 0.001,
"loss": 1.1265,
"step": 172700
},
{
"epoch": 55.85003232062056,
"grad_norm": 1.5243760347366333,
"learning_rate": 0.001,
"loss": 1.1382,
"step": 172800
},
{
"epoch": 55.88235294117647,
"grad_norm": 1.5966765880584717,
"learning_rate": 0.001,
"loss": 1.1386,
"step": 172900
},
{
"epoch": 55.914673561732386,
"grad_norm": 1.5467320680618286,
"learning_rate": 0.001,
"loss": 1.1288,
"step": 173000
},
{
"epoch": 55.9469941822883,
"grad_norm": 1.509267807006836,
"learning_rate": 0.001,
"loss": 1.157,
"step": 173100
},
{
"epoch": 55.979314802844215,
"grad_norm": 1.3964108228683472,
"learning_rate": 0.001,
"loss": 1.1531,
"step": 173200
},
{
"epoch": 56.01163542340013,
"grad_norm": 1.5852636098861694,
"learning_rate": 0.001,
"loss": 1.097,
"step": 173300
},
{
"epoch": 56.043956043956044,
"grad_norm": 1.3627901077270508,
"learning_rate": 0.001,
"loss": 1.0334,
"step": 173400
},
{
"epoch": 56.07627666451196,
"grad_norm": 1.454304575920105,
"learning_rate": 0.001,
"loss": 1.0504,
"step": 173500
},
{
"epoch": 56.10859728506787,
"grad_norm": 1.7767127752304077,
"learning_rate": 0.001,
"loss": 1.0563,
"step": 173600
},
{
"epoch": 56.14091790562379,
"grad_norm": 1.4236656427383423,
"learning_rate": 0.001,
"loss": 1.0453,
"step": 173700
},
{
"epoch": 56.1732385261797,
"grad_norm": 1.743483066558838,
"learning_rate": 0.001,
"loss": 1.0616,
"step": 173800
},
{
"epoch": 56.20555914673562,
"grad_norm": 1.520132064819336,
"learning_rate": 0.001,
"loss": 1.0532,
"step": 173900
},
{
"epoch": 56.23787976729153,
"grad_norm": 2.2348384857177734,
"learning_rate": 0.001,
"loss": 1.0477,
"step": 174000
},
{
"epoch": 56.270200387847446,
"grad_norm": 2.293739080429077,
"learning_rate": 0.001,
"loss": 1.0707,
"step": 174100
},
{
"epoch": 56.30252100840336,
"grad_norm": 1.6483992338180542,
"learning_rate": 0.001,
"loss": 1.0814,
"step": 174200
},
{
"epoch": 56.334841628959275,
"grad_norm": 1.7153490781784058,
"learning_rate": 0.001,
"loss": 1.0742,
"step": 174300
},
{
"epoch": 56.36716224951519,
"grad_norm": 1.7525187730789185,
"learning_rate": 0.001,
"loss": 1.0837,
"step": 174400
},
{
"epoch": 56.399482870071104,
"grad_norm": 2.0571489334106445,
"learning_rate": 0.001,
"loss": 1.0844,
"step": 174500
},
{
"epoch": 56.43180349062702,
"grad_norm": 1.923480749130249,
"learning_rate": 0.001,
"loss": 1.0849,
"step": 174600
},
{
"epoch": 56.46412411118293,
"grad_norm": 1.9075169563293457,
"learning_rate": 0.001,
"loss": 1.0761,
"step": 174700
},
{
"epoch": 56.49644473173885,
"grad_norm": 1.8042218685150146,
"learning_rate": 0.001,
"loss": 1.0983,
"step": 174800
},
{
"epoch": 56.52876535229476,
"grad_norm": 1.6176084280014038,
"learning_rate": 0.001,
"loss": 1.0871,
"step": 174900
},
{
"epoch": 56.56108597285068,
"grad_norm": 2.0587964057922363,
"learning_rate": 0.001,
"loss": 1.1045,
"step": 175000
},
{
"epoch": 56.59340659340659,
"grad_norm": 1.8761169910430908,
"learning_rate": 0.001,
"loss": 1.1241,
"step": 175100
},
{
"epoch": 56.625727213962506,
"grad_norm": 2.422200918197632,
"learning_rate": 0.001,
"loss": 1.1027,
"step": 175200
},
{
"epoch": 56.65804783451842,
"grad_norm": 1.628890037536621,
"learning_rate": 0.001,
"loss": 1.1181,
"step": 175300
},
{
"epoch": 56.690368455074335,
"grad_norm": 1.7099559307098389,
"learning_rate": 0.001,
"loss": 1.0967,
"step": 175400
},
{
"epoch": 56.72268907563025,
"grad_norm": 1.6226879358291626,
"learning_rate": 0.001,
"loss": 1.1141,
"step": 175500
},
{
"epoch": 56.755009696186164,
"grad_norm": 1.4959603548049927,
"learning_rate": 0.001,
"loss": 1.1177,
"step": 175600
},
{
"epoch": 56.78733031674208,
"grad_norm": 1.6804662942886353,
"learning_rate": 0.001,
"loss": 1.1291,
"step": 175700
},
{
"epoch": 56.81965093729799,
"grad_norm": 1.7406642436981201,
"learning_rate": 0.001,
"loss": 1.1339,
"step": 175800
},
{
"epoch": 56.85197155785391,
"grad_norm": 1.5311131477355957,
"learning_rate": 0.001,
"loss": 1.1195,
"step": 175900
},
{
"epoch": 56.88429217840982,
"grad_norm": 1.861016869544983,
"learning_rate": 0.001,
"loss": 1.1357,
"step": 176000
},
{
"epoch": 56.91661279896574,
"grad_norm": 1.7938547134399414,
"learning_rate": 0.001,
"loss": 1.1281,
"step": 176100
},
{
"epoch": 56.94893341952165,
"grad_norm": 1.6675164699554443,
"learning_rate": 0.001,
"loss": 1.1311,
"step": 176200
},
{
"epoch": 56.981254040077566,
"grad_norm": 1.8921256065368652,
"learning_rate": 0.001,
"loss": 1.132,
"step": 176300
},
{
"epoch": 57.01357466063349,
"grad_norm": 1.6421865224838257,
"learning_rate": 0.001,
"loss": 1.0711,
"step": 176400
},
{
"epoch": 57.0458952811894,
"grad_norm": 1.6362299919128418,
"learning_rate": 0.001,
"loss": 1.0239,
"step": 176500
},
{
"epoch": 57.07821590174532,
"grad_norm": 1.5909955501556396,
"learning_rate": 0.001,
"loss": 1.0368,
"step": 176600
},
{
"epoch": 57.11053652230123,
"grad_norm": 1.624410629272461,
"learning_rate": 0.001,
"loss": 1.0494,
"step": 176700
},
{
"epoch": 57.142857142857146,
"grad_norm": 1.610818862915039,
"learning_rate": 0.001,
"loss": 1.0437,
"step": 176800
},
{
"epoch": 57.17517776341306,
"grad_norm": 1.4707742929458618,
"learning_rate": 0.001,
"loss": 1.0494,
"step": 176900
},
{
"epoch": 57.207498383968975,
"grad_norm": 1.3964228630065918,
"learning_rate": 0.001,
"loss": 1.0635,
"step": 177000
},
{
"epoch": 57.23981900452489,
"grad_norm": 1.5611023902893066,
"learning_rate": 0.001,
"loss": 1.0516,
"step": 177100
},
{
"epoch": 57.272139625080804,
"grad_norm": 1.4848252534866333,
"learning_rate": 0.001,
"loss": 1.0631,
"step": 177200
},
{
"epoch": 57.30446024563672,
"grad_norm": 1.2606589794158936,
"learning_rate": 0.001,
"loss": 1.0566,
"step": 177300
},
{
"epoch": 57.33678086619263,
"grad_norm": 1.6222645044326782,
"learning_rate": 0.001,
"loss": 1.0771,
"step": 177400
},
{
"epoch": 57.36910148674855,
"grad_norm": 2.2820799350738525,
"learning_rate": 0.001,
"loss": 1.0735,
"step": 177500
},
{
"epoch": 57.40142210730446,
"grad_norm": 1.7914892435073853,
"learning_rate": 0.001,
"loss": 1.076,
"step": 177600
},
{
"epoch": 57.43374272786038,
"grad_norm": 1.5209670066833496,
"learning_rate": 0.001,
"loss": 1.0958,
"step": 177700
},
{
"epoch": 57.46606334841629,
"grad_norm": 2.0354158878326416,
"learning_rate": 0.001,
"loss": 1.0731,
"step": 177800
},
{
"epoch": 57.498383968972206,
"grad_norm": 1.5465794801712036,
"learning_rate": 0.001,
"loss": 1.082,
"step": 177900
},
{
"epoch": 57.53070458952812,
"grad_norm": 1.4770228862762451,
"learning_rate": 0.001,
"loss": 1.0939,
"step": 178000
},
{
"epoch": 57.563025210084035,
"grad_norm": 1.5884677171707153,
"learning_rate": 0.001,
"loss": 1.0881,
"step": 178100
},
{
"epoch": 57.59534583063995,
"grad_norm": 1.3796273469924927,
"learning_rate": 0.001,
"loss": 1.0723,
"step": 178200
},
{
"epoch": 57.627666451195864,
"grad_norm": 1.7390011548995972,
"learning_rate": 0.001,
"loss": 1.0946,
"step": 178300
},
{
"epoch": 57.65998707175178,
"grad_norm": 1.8054834604263306,
"learning_rate": 0.001,
"loss": 1.0912,
"step": 178400
},
{
"epoch": 57.69230769230769,
"grad_norm": 1.5965912342071533,
"learning_rate": 0.001,
"loss": 1.1132,
"step": 178500
},
{
"epoch": 57.72462831286361,
"grad_norm": 1.4859302043914795,
"learning_rate": 0.001,
"loss": 1.0969,
"step": 178600
},
{
"epoch": 57.75694893341952,
"grad_norm": 1.8760346174240112,
"learning_rate": 0.001,
"loss": 1.1142,
"step": 178700
},
{
"epoch": 57.78926955397544,
"grad_norm": 1.9528679847717285,
"learning_rate": 0.001,
"loss": 1.0951,
"step": 178800
},
{
"epoch": 57.82159017453135,
"grad_norm": 1.9293407201766968,
"learning_rate": 0.001,
"loss": 1.1017,
"step": 178900
},
{
"epoch": 57.853910795087266,
"grad_norm": 1.3426704406738281,
"learning_rate": 0.001,
"loss": 1.105,
"step": 179000
},
{
"epoch": 57.88623141564318,
"grad_norm": 1.448326587677002,
"learning_rate": 0.001,
"loss": 1.1075,
"step": 179100
},
{
"epoch": 57.918552036199095,
"grad_norm": 1.7974193096160889,
"learning_rate": 0.001,
"loss": 1.1323,
"step": 179200
},
{
"epoch": 57.95087265675501,
"grad_norm": 1.621390461921692,
"learning_rate": 0.001,
"loss": 1.1397,
"step": 179300
},
{
"epoch": 57.983193277310924,
"grad_norm": 1.6346193552017212,
"learning_rate": 0.001,
"loss": 1.1445,
"step": 179400
},
{
"epoch": 58.01551389786684,
"grad_norm": 1.8142673969268799,
"learning_rate": 0.001,
"loss": 1.0451,
"step": 179500
},
{
"epoch": 58.04783451842275,
"grad_norm": 1.33307945728302,
"learning_rate": 0.001,
"loss": 1.0349,
"step": 179600
},
{
"epoch": 58.08015513897867,
"grad_norm": 1.182541012763977,
"learning_rate": 0.001,
"loss": 1.0273,
"step": 179700
},
{
"epoch": 58.11247575953458,
"grad_norm": 1.4469527006149292,
"learning_rate": 0.001,
"loss": 1.0151,
"step": 179800
},
{
"epoch": 58.1447963800905,
"grad_norm": 1.3350833654403687,
"learning_rate": 0.001,
"loss": 1.0442,
"step": 179900
},
{
"epoch": 58.17711700064641,
"grad_norm": 1.4207303524017334,
"learning_rate": 0.001,
"loss": 1.0211,
"step": 180000
},
{
"epoch": 58.209437621202326,
"grad_norm": 1.633563756942749,
"learning_rate": 0.001,
"loss": 1.0392,
"step": 180100
},
{
"epoch": 58.24175824175824,
"grad_norm": 1.3849244117736816,
"learning_rate": 0.001,
"loss": 1.0623,
"step": 180200
},
{
"epoch": 58.274078862314155,
"grad_norm": 1.8374276161193848,
"learning_rate": 0.001,
"loss": 1.0388,
"step": 180300
},
{
"epoch": 58.30639948287007,
"grad_norm": 1.3447884321212769,
"learning_rate": 0.001,
"loss": 1.0626,
"step": 180400
},
{
"epoch": 58.338720103425985,
"grad_norm": 1.5665903091430664,
"learning_rate": 0.001,
"loss": 1.0754,
"step": 180500
},
{
"epoch": 58.3710407239819,
"grad_norm": 1.3040165901184082,
"learning_rate": 0.001,
"loss": 1.0746,
"step": 180600
},
{
"epoch": 58.403361344537814,
"grad_norm": 1.3881635665893555,
"learning_rate": 0.001,
"loss": 1.0525,
"step": 180700
},
{
"epoch": 58.43568196509373,
"grad_norm": 1.1611268520355225,
"learning_rate": 0.001,
"loss": 1.0717,
"step": 180800
},
{
"epoch": 58.46800258564964,
"grad_norm": 1.2324254512786865,
"learning_rate": 0.001,
"loss": 1.0696,
"step": 180900
},
{
"epoch": 58.50032320620556,
"grad_norm": 1.5421253442764282,
"learning_rate": 0.001,
"loss": 1.0742,
"step": 181000
},
{
"epoch": 58.53264382676147,
"grad_norm": 0.9455136656761169,
"learning_rate": 0.001,
"loss": 1.0723,
"step": 181100
},
{
"epoch": 58.56496444731739,
"grad_norm": 1.49685800075531,
"learning_rate": 0.001,
"loss": 1.0705,
"step": 181200
},
{
"epoch": 58.5972850678733,
"grad_norm": 3.2262043952941895,
"learning_rate": 0.001,
"loss": 1.0927,
"step": 181300
},
{
"epoch": 58.629605688429216,
"grad_norm": 1.138930320739746,
"learning_rate": 0.001,
"loss": 1.0768,
"step": 181400
},
{
"epoch": 58.66192630898513,
"grad_norm": 1.6434274911880493,
"learning_rate": 0.001,
"loss": 1.086,
"step": 181500
},
{
"epoch": 58.694246929541045,
"grad_norm": 1.4245071411132812,
"learning_rate": 0.001,
"loss": 1.0806,
"step": 181600
},
{
"epoch": 58.72656755009696,
"grad_norm": 1.1249699592590332,
"learning_rate": 0.001,
"loss": 1.1123,
"step": 181700
},
{
"epoch": 58.758888170652874,
"grad_norm": 1.3846380710601807,
"learning_rate": 0.001,
"loss": 1.1084,
"step": 181800
},
{
"epoch": 58.79120879120879,
"grad_norm": 1.39168119430542,
"learning_rate": 0.001,
"loss": 1.1256,
"step": 181900
},
{
"epoch": 58.8235294117647,
"grad_norm": 1.3924428224563599,
"learning_rate": 0.001,
"loss": 1.1124,
"step": 182000
},
{
"epoch": 58.85585003232062,
"grad_norm": 1.7155895233154297,
"learning_rate": 0.001,
"loss": 0.9987,
"step": 182100
},
{
"epoch": 58.88817065287653,
"grad_norm": 1.6909483671188354,
"learning_rate": 0.001,
"loss": 1.008,
"step": 182200
},
{
"epoch": 58.92049127343245,
"grad_norm": 1.2806386947631836,
"learning_rate": 0.001,
"loss": 1.0156,
"step": 182300
},
{
"epoch": 58.95281189398836,
"grad_norm": 1.5416556596755981,
"learning_rate": 0.001,
"loss": 1.0119,
"step": 182400
},
{
"epoch": 58.985132514544276,
"grad_norm": 1.9270284175872803,
"learning_rate": 0.001,
"loss": 0.9978,
"step": 182500
},
{
"epoch": 59.0174531351002,
"grad_norm": 1.8223607540130615,
"learning_rate": 0.001,
"loss": 1.0246,
"step": 182600
},
{
"epoch": 59.04977375565611,
"grad_norm": 2.0636425018310547,
"learning_rate": 0.001,
"loss": 1.0184,
"step": 182700
},
{
"epoch": 59.08209437621203,
"grad_norm": 1.6684495210647583,
"learning_rate": 0.001,
"loss": 1.008,
"step": 182800
},
{
"epoch": 59.11441499676794,
"grad_norm": 2.0706138610839844,
"learning_rate": 0.001,
"loss": 1.0314,
"step": 182900
},
{
"epoch": 59.146735617323856,
"grad_norm": 2.2243518829345703,
"learning_rate": 0.001,
"loss": 1.0092,
"step": 183000
},
{
"epoch": 59.17905623787977,
"grad_norm": 1.8938789367675781,
"learning_rate": 0.001,
"loss": 1.0275,
"step": 183100
},
{
"epoch": 59.211376858435685,
"grad_norm": 1.7461127042770386,
"learning_rate": 0.001,
"loss": 1.0462,
"step": 183200
},
{
"epoch": 59.2436974789916,
"grad_norm": 2.4900665283203125,
"learning_rate": 0.001,
"loss": 1.0474,
"step": 183300
},
{
"epoch": 59.276018099547514,
"grad_norm": 2.173794984817505,
"learning_rate": 0.001,
"loss": 1.0452,
"step": 183400
},
{
"epoch": 59.30833872010343,
"grad_norm": 2.100980043411255,
"learning_rate": 0.001,
"loss": 1.0621,
"step": 183500
},
{
"epoch": 59.34065934065934,
"grad_norm": 2.124846935272217,
"learning_rate": 0.001,
"loss": 1.077,
"step": 183600
},
{
"epoch": 59.37297996121526,
"grad_norm": 2.6659579277038574,
"learning_rate": 0.001,
"loss": 1.06,
"step": 183700
},
{
"epoch": 59.40530058177117,
"grad_norm": 2.3515841960906982,
"learning_rate": 0.001,
"loss": 1.0685,
"step": 183800
},
{
"epoch": 59.43762120232709,
"grad_norm": 2.4156997203826904,
"learning_rate": 0.001,
"loss": 1.061,
"step": 183900
},
{
"epoch": 59.469941822883,
"grad_norm": 1.7681801319122314,
"learning_rate": 0.001,
"loss": 1.068,
"step": 184000
},
{
"epoch": 59.502262443438916,
"grad_norm": 1.7692102193832397,
"learning_rate": 0.001,
"loss": 1.0839,
"step": 184100
},
{
"epoch": 59.53458306399483,
"grad_norm": 1.82084059715271,
"learning_rate": 0.001,
"loss": 1.0723,
"step": 184200
},
{
"epoch": 59.566903684550745,
"grad_norm": 2.9844210147857666,
"learning_rate": 0.001,
"loss": 1.0736,
"step": 184300
},
{
"epoch": 59.59922430510666,
"grad_norm": 1.6782689094543457,
"learning_rate": 0.001,
"loss": 1.079,
"step": 184400
},
{
"epoch": 59.631544925662574,
"grad_norm": 2.3039188385009766,
"learning_rate": 0.001,
"loss": 1.077,
"step": 184500
},
{
"epoch": 59.66386554621849,
"grad_norm": 1.5084682703018188,
"learning_rate": 0.001,
"loss": 1.0895,
"step": 184600
},
{
"epoch": 59.6961861667744,
"grad_norm": 2.922206163406372,
"learning_rate": 0.001,
"loss": 1.0795,
"step": 184700
},
{
"epoch": 59.72850678733032,
"grad_norm": 1.4033228158950806,
"learning_rate": 0.001,
"loss": 1.0939,
"step": 184800
},
{
"epoch": 59.76082740788623,
"grad_norm": 1.722129464149475,
"learning_rate": 0.001,
"loss": 1.0868,
"step": 184900
},
{
"epoch": 59.79314802844215,
"grad_norm": 1.9006403684616089,
"learning_rate": 0.001,
"loss": 1.0985,
"step": 185000
},
{
"epoch": 59.82546864899806,
"grad_norm": 2.96142315864563,
"learning_rate": 0.001,
"loss": 1.1029,
"step": 185100
},
{
"epoch": 59.857789269553976,
"grad_norm": 1.7573572397232056,
"learning_rate": 0.001,
"loss": 1.101,
"step": 185200
},
{
"epoch": 59.89010989010989,
"grad_norm": 1.8144474029541016,
"learning_rate": 0.001,
"loss": 1.0882,
"step": 185300
},
{
"epoch": 59.922430510665805,
"grad_norm": 1.9638746976852417,
"learning_rate": 0.001,
"loss": 1.0941,
"step": 185400
},
{
"epoch": 59.95475113122172,
"grad_norm": 1.8024924993515015,
"learning_rate": 0.001,
"loss": 1.1102,
"step": 185500
},
{
"epoch": 59.987071751777634,
"grad_norm": 1.905829668045044,
"learning_rate": 0.001,
"loss": 1.0942,
"step": 185600
},
{
"epoch": 60.01939237233355,
"grad_norm": 1.6812539100646973,
"learning_rate": 0.001,
"loss": 1.0748,
"step": 185700
},
{
"epoch": 60.05171299288946,
"grad_norm": 1.7039813995361328,
"learning_rate": 0.001,
"loss": 1.0093,
"step": 185800
},
{
"epoch": 60.08403361344538,
"grad_norm": 1.8435310125350952,
"learning_rate": 0.001,
"loss": 1.0105,
"step": 185900
},
{
"epoch": 60.11635423400129,
"grad_norm": 2.0761561393737793,
"learning_rate": 0.001,
"loss": 1.0328,
"step": 186000
},
{
"epoch": 60.14867485455721,
"grad_norm": 4.478841304779053,
"learning_rate": 0.001,
"loss": 1.03,
"step": 186100
},
{
"epoch": 60.18099547511312,
"grad_norm": 1.5484015941619873,
"learning_rate": 0.001,
"loss": 1.0442,
"step": 186200
},
{
"epoch": 60.213316095669036,
"grad_norm": 1.8544293642044067,
"learning_rate": 0.001,
"loss": 1.0356,
"step": 186300
},
{
"epoch": 60.24563671622495,
"grad_norm": 1.7366797924041748,
"learning_rate": 0.001,
"loss": 1.0181,
"step": 186400
},
{
"epoch": 60.277957336780865,
"grad_norm": 1.8898407220840454,
"learning_rate": 0.001,
"loss": 1.0362,
"step": 186500
},
{
"epoch": 60.31027795733678,
"grad_norm": 1.422088623046875,
"learning_rate": 0.001,
"loss": 1.0307,
"step": 186600
},
{
"epoch": 60.342598577892694,
"grad_norm": 2.02425217628479,
"learning_rate": 0.001,
"loss": 1.0591,
"step": 186700
},
{
"epoch": 60.37491919844861,
"grad_norm": 1.5161786079406738,
"learning_rate": 0.001,
"loss": 1.03,
"step": 186800
},
{
"epoch": 60.40723981900452,
"grad_norm": 1.4360978603363037,
"learning_rate": 0.001,
"loss": 1.0376,
"step": 186900
},
{
"epoch": 60.43956043956044,
"grad_norm": 1.6159266233444214,
"learning_rate": 0.001,
"loss": 1.0403,
"step": 187000
},
{
"epoch": 60.47188106011635,
"grad_norm": 1.6211012601852417,
"learning_rate": 0.001,
"loss": 1.0488,
"step": 187100
},
{
"epoch": 60.50420168067227,
"grad_norm": 1.6320511102676392,
"learning_rate": 0.001,
"loss": 1.0632,
"step": 187200
},
{
"epoch": 60.53652230122818,
"grad_norm": 1.4999951124191284,
"learning_rate": 0.001,
"loss": 1.0759,
"step": 187300
},
{
"epoch": 60.568842921784096,
"grad_norm": 1.581023931503296,
"learning_rate": 0.001,
"loss": 1.0658,
"step": 187400
},
{
"epoch": 60.60116354234001,
"grad_norm": 1.7817587852478027,
"learning_rate": 0.001,
"loss": 1.054,
"step": 187500
},
{
"epoch": 60.633484162895925,
"grad_norm": 1.7636590003967285,
"learning_rate": 0.001,
"loss": 1.0684,
"step": 187600
},
{
"epoch": 60.66580478345184,
"grad_norm": 1.7448927164077759,
"learning_rate": 0.001,
"loss": 1.0625,
"step": 187700
},
{
"epoch": 60.698125404007754,
"grad_norm": 2.117954969406128,
"learning_rate": 0.001,
"loss": 1.0824,
"step": 187800
},
{
"epoch": 60.73044602456367,
"grad_norm": 1.5040390491485596,
"learning_rate": 0.001,
"loss": 1.0652,
"step": 187900
},
{
"epoch": 60.762766645119584,
"grad_norm": 1.752811074256897,
"learning_rate": 0.001,
"loss": 1.0901,
"step": 188000
},
{
"epoch": 60.7950872656755,
"grad_norm": 1.800222396850586,
"learning_rate": 0.001,
"loss": 1.0856,
"step": 188100
},
{
"epoch": 60.82740788623141,
"grad_norm": 1.8694863319396973,
"learning_rate": 0.001,
"loss": 1.0953,
"step": 188200
},
{
"epoch": 60.85972850678733,
"grad_norm": 1.5264275074005127,
"learning_rate": 0.001,
"loss": 1.0727,
"step": 188300
},
{
"epoch": 60.89204912734324,
"grad_norm": 1.5535075664520264,
"learning_rate": 0.001,
"loss": 1.0954,
"step": 188400
},
{
"epoch": 60.924369747899156,
"grad_norm": 1.4875106811523438,
"learning_rate": 0.001,
"loss": 1.0942,
"step": 188500
},
{
"epoch": 60.95669036845507,
"grad_norm": 1.7024391889572144,
"learning_rate": 0.001,
"loss": 1.0822,
"step": 188600
},
{
"epoch": 60.98901098901099,
"grad_norm": 1.6542145013809204,
"learning_rate": 0.001,
"loss": 1.0996,
"step": 188700
},
{
"epoch": 61.02133160956691,
"grad_norm": 1.7969130277633667,
"learning_rate": 0.001,
"loss": 1.0485,
"step": 188800
},
{
"epoch": 61.05365223012282,
"grad_norm": 1.9887769222259521,
"learning_rate": 0.001,
"loss": 0.9878,
"step": 188900
},
{
"epoch": 61.085972850678736,
"grad_norm": 1.2860554456710815,
"learning_rate": 0.001,
"loss": 1.0076,
"step": 189000
},
{
"epoch": 61.11829347123465,
"grad_norm": 2.320146083831787,
"learning_rate": 0.001,
"loss": 1.0179,
"step": 189100
},
{
"epoch": 61.150614091790565,
"grad_norm": 1.7772401571273804,
"learning_rate": 0.001,
"loss": 1.0105,
"step": 189200
},
{
"epoch": 61.18293471234648,
"grad_norm": 1.7308861017227173,
"learning_rate": 0.001,
"loss": 1.0254,
"step": 189300
},
{
"epoch": 61.215255332902395,
"grad_norm": 1.4736839532852173,
"learning_rate": 0.001,
"loss": 1.0321,
"step": 189400
},
{
"epoch": 61.24757595345831,
"grad_norm": 1.709177851676941,
"learning_rate": 0.001,
"loss": 1.0102,
"step": 189500
},
{
"epoch": 61.279896574014224,
"grad_norm": 1.6960136890411377,
"learning_rate": 0.001,
"loss": 1.017,
"step": 189600
},
{
"epoch": 61.31221719457014,
"grad_norm": 1.7163701057434082,
"learning_rate": 0.001,
"loss": 1.0547,
"step": 189700
},
{
"epoch": 61.34453781512605,
"grad_norm": 1.6530094146728516,
"learning_rate": 0.001,
"loss": 1.0263,
"step": 189800
},
{
"epoch": 61.37685843568197,
"grad_norm": 2.045729398727417,
"learning_rate": 0.001,
"loss": 1.0378,
"step": 189900
},
{
"epoch": 61.40917905623788,
"grad_norm": 1.541930913925171,
"learning_rate": 0.001,
"loss": 1.0491,
"step": 190000
},
{
"epoch": 61.441499676793796,
"grad_norm": 1.3723949193954468,
"learning_rate": 0.001,
"loss": 1.0508,
"step": 190100
},
{
"epoch": 61.47382029734971,
"grad_norm": 1.6087392568588257,
"learning_rate": 0.001,
"loss": 1.0278,
"step": 190200
},
{
"epoch": 61.506140917905626,
"grad_norm": 1.7446894645690918,
"learning_rate": 0.001,
"loss": 1.0405,
"step": 190300
},
{
"epoch": 61.53846153846154,
"grad_norm": 1.5976572036743164,
"learning_rate": 0.001,
"loss": 1.062,
"step": 190400
},
{
"epoch": 61.570782159017455,
"grad_norm": 1.484781265258789,
"learning_rate": 0.001,
"loss": 1.0507,
"step": 190500
},
{
"epoch": 61.60310277957337,
"grad_norm": 1.9391828775405884,
"learning_rate": 0.001,
"loss": 1.049,
"step": 190600
},
{
"epoch": 61.635423400129284,
"grad_norm": 1.4382134675979614,
"learning_rate": 0.001,
"loss": 1.0471,
"step": 190700
},
{
"epoch": 61.6677440206852,
"grad_norm": 1.4284454584121704,
"learning_rate": 0.001,
"loss": 1.0551,
"step": 190800
},
{
"epoch": 61.70006464124111,
"grad_norm": 1.9543735980987549,
"learning_rate": 0.001,
"loss": 1.064,
"step": 190900
},
{
"epoch": 61.73238526179703,
"grad_norm": 1.5626345872879028,
"learning_rate": 0.001,
"loss": 1.0552,
"step": 191000
},
{
"epoch": 61.76470588235294,
"grad_norm": 1.6643787622451782,
"learning_rate": 0.001,
"loss": 1.0775,
"step": 191100
},
{
"epoch": 61.79702650290886,
"grad_norm": 1.4800466299057007,
"learning_rate": 0.001,
"loss": 1.07,
"step": 191200
},
{
"epoch": 61.82934712346477,
"grad_norm": 1.5903040170669556,
"learning_rate": 0.001,
"loss": 1.0834,
"step": 191300
},
{
"epoch": 61.861667744020686,
"grad_norm": 1.492759108543396,
"learning_rate": 0.001,
"loss": 1.0834,
"step": 191400
},
{
"epoch": 61.8939883645766,
"grad_norm": 1.6342939138412476,
"learning_rate": 0.001,
"loss": 1.0651,
"step": 191500
},
{
"epoch": 61.926308985132515,
"grad_norm": 1.6058002710342407,
"learning_rate": 0.001,
"loss": 1.0673,
"step": 191600
},
{
"epoch": 61.95862960568843,
"grad_norm": 1.3571763038635254,
"learning_rate": 0.001,
"loss": 1.0744,
"step": 191700
},
{
"epoch": 61.990950226244344,
"grad_norm": 1.5555781126022339,
"learning_rate": 0.001,
"loss": 1.0971,
"step": 191800
},
{
"epoch": 62.02327084680026,
"grad_norm": 1.648055911064148,
"learning_rate": 0.001,
"loss": 1.0267,
"step": 191900
},
{
"epoch": 62.05559146735617,
"grad_norm": 1.420931339263916,
"learning_rate": 0.001,
"loss": 0.9998,
"step": 192000
},
{
"epoch": 62.08791208791209,
"grad_norm": 1.436596393585205,
"learning_rate": 0.001,
"loss": 0.9868,
"step": 192100
},
{
"epoch": 62.120232708468,
"grad_norm": 1.863795518875122,
"learning_rate": 0.001,
"loss": 1.0002,
"step": 192200
},
{
"epoch": 62.15255332902392,
"grad_norm": 2.2537336349487305,
"learning_rate": 0.001,
"loss": 1.0086,
"step": 192300
},
{
"epoch": 62.18487394957983,
"grad_norm": 1.4402319192886353,
"learning_rate": 0.001,
"loss": 1.0017,
"step": 192400
},
{
"epoch": 62.217194570135746,
"grad_norm": 1.756102442741394,
"learning_rate": 0.001,
"loss": 1.0109,
"step": 192500
},
{
"epoch": 62.24951519069166,
"grad_norm": 1.5982434749603271,
"learning_rate": 0.001,
"loss": 0.9953,
"step": 192600
},
{
"epoch": 62.281835811247575,
"grad_norm": 1.6266897916793823,
"learning_rate": 0.001,
"loss": 1.0191,
"step": 192700
},
{
"epoch": 62.31415643180349,
"grad_norm": 1.8158379793167114,
"learning_rate": 0.001,
"loss": 1.0156,
"step": 192800
},
{
"epoch": 62.346477052359404,
"grad_norm": 1.6840484142303467,
"learning_rate": 0.001,
"loss": 1.0194,
"step": 192900
},
{
"epoch": 62.37879767291532,
"grad_norm": 1.6763032674789429,
"learning_rate": 0.001,
"loss": 1.0243,
"step": 193000
},
{
"epoch": 62.41111829347123,
"grad_norm": 2.340099334716797,
"learning_rate": 0.001,
"loss": 1.0222,
"step": 193100
},
{
"epoch": 62.44343891402715,
"grad_norm": 1.2376492023468018,
"learning_rate": 0.001,
"loss": 1.0288,
"step": 193200
},
{
"epoch": 62.47575953458306,
"grad_norm": 1.5578148365020752,
"learning_rate": 0.001,
"loss": 1.0442,
"step": 193300
},
{
"epoch": 62.50808015513898,
"grad_norm": 1.7190568447113037,
"learning_rate": 0.001,
"loss": 1.0338,
"step": 193400
},
{
"epoch": 62.54040077569489,
"grad_norm": 1.876619577407837,
"learning_rate": 0.001,
"loss": 1.0451,
"step": 193500
},
{
"epoch": 62.572721396250806,
"grad_norm": 1.5763674974441528,
"learning_rate": 0.001,
"loss": 1.0518,
"step": 193600
},
{
"epoch": 62.60504201680672,
"grad_norm": 1.3858126401901245,
"learning_rate": 0.001,
"loss": 1.0555,
"step": 193700
},
{
"epoch": 62.637362637362635,
"grad_norm": 1.5947778224945068,
"learning_rate": 0.001,
"loss": 1.0683,
"step": 193800
},
{
"epoch": 62.66968325791855,
"grad_norm": 1.6285808086395264,
"learning_rate": 0.001,
"loss": 1.0541,
"step": 193900
},
{
"epoch": 62.702003878474464,
"grad_norm": 1.5361659526824951,
"learning_rate": 0.001,
"loss": 1.0422,
"step": 194000
},
{
"epoch": 62.73432449903038,
"grad_norm": 1.534832239151001,
"learning_rate": 0.001,
"loss": 1.0602,
"step": 194100
},
{
"epoch": 62.76664511958629,
"grad_norm": 1.6349917650222778,
"learning_rate": 0.001,
"loss": 1.0666,
"step": 194200
},
{
"epoch": 62.79896574014221,
"grad_norm": 2.202749729156494,
"learning_rate": 0.001,
"loss": 1.0618,
"step": 194300
},
{
"epoch": 62.83128636069812,
"grad_norm": 1.2954192161560059,
"learning_rate": 0.001,
"loss": 1.0573,
"step": 194400
},
{
"epoch": 62.86360698125404,
"grad_norm": 1.7332080602645874,
"learning_rate": 0.001,
"loss": 1.0763,
"step": 194500
},
{
"epoch": 62.89592760180995,
"grad_norm": 1.8932256698608398,
"learning_rate": 0.001,
"loss": 1.0794,
"step": 194600
},
{
"epoch": 62.928248222365866,
"grad_norm": 1.3601816892623901,
"learning_rate": 0.001,
"loss": 1.0828,
"step": 194700
},
{
"epoch": 62.96056884292178,
"grad_norm": 1.5028172731399536,
"learning_rate": 0.001,
"loss": 1.0879,
"step": 194800
},
{
"epoch": 62.992889463477695,
"grad_norm": 1.466861367225647,
"learning_rate": 0.001,
"loss": 1.0798,
"step": 194900
},
{
"epoch": 63.02521008403362,
"grad_norm": 1.4932770729064941,
"learning_rate": 0.001,
"loss": 1.0139,
"step": 195000
},
{
"epoch": 63.05753070458953,
"grad_norm": 1.7395883798599243,
"learning_rate": 0.001,
"loss": 0.9744,
"step": 195100
},
{
"epoch": 63.089851325145446,
"grad_norm": 1.5448013544082642,
"learning_rate": 0.001,
"loss": 0.9972,
"step": 195200
},
{
"epoch": 63.12217194570136,
"grad_norm": 1.6690627336502075,
"learning_rate": 0.001,
"loss": 0.9984,
"step": 195300
},
{
"epoch": 63.154492566257275,
"grad_norm": 1.3744710683822632,
"learning_rate": 0.001,
"loss": 0.989,
"step": 195400
},
{
"epoch": 63.18681318681319,
"grad_norm": 1.4976576566696167,
"learning_rate": 0.001,
"loss": 1.0029,
"step": 195500
},
{
"epoch": 63.219133807369104,
"grad_norm": 1.7103939056396484,
"learning_rate": 0.001,
"loss": 1.0041,
"step": 195600
},
{
"epoch": 63.25145442792502,
"grad_norm": 1.535509467124939,
"learning_rate": 0.001,
"loss": 1.0122,
"step": 195700
},
{
"epoch": 63.28377504848093,
"grad_norm": 1.9888761043548584,
"learning_rate": 0.001,
"loss": 1.0109,
"step": 195800
},
{
"epoch": 63.31609566903685,
"grad_norm": 1.6202621459960938,
"learning_rate": 0.001,
"loss": 1.0244,
"step": 195900
},
{
"epoch": 63.34841628959276,
"grad_norm": 1.4541372060775757,
"learning_rate": 0.001,
"loss": 1.0023,
"step": 196000
},
{
"epoch": 63.38073691014868,
"grad_norm": 1.5561103820800781,
"learning_rate": 0.001,
"loss": 1.0055,
"step": 196100
},
{
"epoch": 63.41305753070459,
"grad_norm": 1.6799334287643433,
"learning_rate": 0.001,
"loss": 1.0261,
"step": 196200
},
{
"epoch": 63.445378151260506,
"grad_norm": 1.6529778242111206,
"learning_rate": 0.001,
"loss": 1.0308,
"step": 196300
},
{
"epoch": 63.47769877181642,
"grad_norm": 1.3067463636398315,
"learning_rate": 0.001,
"loss": 1.018,
"step": 196400
},
{
"epoch": 63.510019392372335,
"grad_norm": 1.8982113599777222,
"learning_rate": 0.001,
"loss": 1.0239,
"step": 196500
},
{
"epoch": 63.54234001292825,
"grad_norm": 1.3777034282684326,
"learning_rate": 0.001,
"loss": 1.0361,
"step": 196600
},
{
"epoch": 63.574660633484164,
"grad_norm": 1.6532855033874512,
"learning_rate": 0.001,
"loss": 1.0291,
"step": 196700
},
{
"epoch": 63.60698125404008,
"grad_norm": 1.9859448671340942,
"learning_rate": 0.001,
"loss": 1.0361,
"step": 196800
},
{
"epoch": 63.63930187459599,
"grad_norm": 1.6853806972503662,
"learning_rate": 0.001,
"loss": 1.0609,
"step": 196900
},
{
"epoch": 63.67162249515191,
"grad_norm": 1.7648768424987793,
"learning_rate": 0.001,
"loss": 1.0455,
"step": 197000
},
{
"epoch": 63.70394311570782,
"grad_norm": 1.4905048608779907,
"learning_rate": 0.001,
"loss": 1.0397,
"step": 197100
},
{
"epoch": 63.73626373626374,
"grad_norm": 1.4467679262161255,
"learning_rate": 0.001,
"loss": 1.0418,
"step": 197200
},
{
"epoch": 63.76858435681965,
"grad_norm": 1.4468274116516113,
"learning_rate": 0.001,
"loss": 1.069,
"step": 197300
},
{
"epoch": 63.800904977375566,
"grad_norm": 1.6732083559036255,
"learning_rate": 0.001,
"loss": 1.0555,
"step": 197400
},
{
"epoch": 63.83322559793148,
"grad_norm": 1.6796603202819824,
"learning_rate": 0.001,
"loss": 1.0448,
"step": 197500
},
{
"epoch": 63.865546218487395,
"grad_norm": 1.566683053970337,
"learning_rate": 0.001,
"loss": 1.073,
"step": 197600
},
{
"epoch": 63.89786683904331,
"grad_norm": 1.954232931137085,
"learning_rate": 0.001,
"loss": 1.0571,
"step": 197700
},
{
"epoch": 63.930187459599225,
"grad_norm": 1.7381691932678223,
"learning_rate": 0.001,
"loss": 1.0625,
"step": 197800
},
{
"epoch": 63.96250808015514,
"grad_norm": 1.313048005104065,
"learning_rate": 0.001,
"loss": 1.0704,
"step": 197900
},
{
"epoch": 63.994828700711054,
"grad_norm": 1.5494202375411987,
"learning_rate": 0.001,
"loss": 1.0526,
"step": 198000
},
{
"epoch": 64.02714932126698,
"grad_norm": 1.574831247329712,
"learning_rate": 0.001,
"loss": 1.0078,
"step": 198100
},
{
"epoch": 64.05946994182288,
"grad_norm": 1.8430677652359009,
"learning_rate": 0.001,
"loss": 0.9663,
"step": 198200
},
{
"epoch": 64.0917905623788,
"grad_norm": 1.352602243423462,
"learning_rate": 0.001,
"loss": 0.9762,
"step": 198300
},
{
"epoch": 64.12411118293471,
"grad_norm": 17.603384017944336,
"learning_rate": 0.001,
"loss": 0.9659,
"step": 198400
},
{
"epoch": 64.15643180349063,
"grad_norm": 1.5747212171554565,
"learning_rate": 0.001,
"loss": 0.9795,
"step": 198500
},
{
"epoch": 64.18875242404654,
"grad_norm": 1.5215107202529907,
"learning_rate": 0.001,
"loss": 0.9927,
"step": 198600
},
{
"epoch": 64.22107304460246,
"grad_norm": 1.9847549200057983,
"learning_rate": 0.001,
"loss": 0.9911,
"step": 198700
},
{
"epoch": 64.25339366515837,
"grad_norm": 1.9193583726882935,
"learning_rate": 0.001,
"loss": 0.9902,
"step": 198800
},
{
"epoch": 64.28571428571429,
"grad_norm": 2.0221829414367676,
"learning_rate": 0.001,
"loss": 1.0085,
"step": 198900
},
{
"epoch": 64.3180349062702,
"grad_norm": 1.7887248992919922,
"learning_rate": 0.001,
"loss": 1.0029,
"step": 199000
},
{
"epoch": 64.35035552682612,
"grad_norm": 1.7705152034759521,
"learning_rate": 0.001,
"loss": 1.0082,
"step": 199100
},
{
"epoch": 64.38267614738203,
"grad_norm": 1.9016677141189575,
"learning_rate": 0.001,
"loss": 1.0053,
"step": 199200
},
{
"epoch": 64.41499676793795,
"grad_norm": 1.694682240486145,
"learning_rate": 0.001,
"loss": 0.9946,
"step": 199300
},
{
"epoch": 64.44731738849386,
"grad_norm": 1.935779333114624,
"learning_rate": 0.001,
"loss": 1.0154,
"step": 199400
},
{
"epoch": 64.47963800904978,
"grad_norm": 1.625313639640808,
"learning_rate": 0.001,
"loss": 1.0126,
"step": 199500
},
{
"epoch": 64.51195862960569,
"grad_norm": 1.7697724103927612,
"learning_rate": 0.001,
"loss": 1.0334,
"step": 199600
},
{
"epoch": 64.54427925016161,
"grad_norm": 2.2513742446899414,
"learning_rate": 0.001,
"loss": 1.0362,
"step": 199700
},
{
"epoch": 64.57659987071752,
"grad_norm": 1.630536437034607,
"learning_rate": 0.001,
"loss": 1.0259,
"step": 199800
},
{
"epoch": 64.60892049127344,
"grad_norm": 1.9813055992126465,
"learning_rate": 0.001,
"loss": 1.0455,
"step": 199900
},
{
"epoch": 64.64124111182934,
"grad_norm": 1.6457490921020508,
"learning_rate": 0.001,
"loss": 1.0369,
"step": 200000
},
{
"epoch": 64.67356173238527,
"grad_norm": 1.7242528200149536,
"learning_rate": 0.001,
"loss": 1.028,
"step": 200100
},
{
"epoch": 64.70588235294117,
"grad_norm": 1.8536429405212402,
"learning_rate": 0.001,
"loss": 1.052,
"step": 200200
},
{
"epoch": 64.7382029734971,
"grad_norm": 2.031980037689209,
"learning_rate": 0.001,
"loss": 1.0352,
"step": 200300
},
{
"epoch": 64.770523594053,
"grad_norm": 2.1543500423431396,
"learning_rate": 0.001,
"loss": 1.0238,
"step": 200400
},
{
"epoch": 64.80284421460892,
"grad_norm": 1.4823232889175415,
"learning_rate": 0.001,
"loss": 1.0573,
"step": 200500
},
{
"epoch": 64.83516483516483,
"grad_norm": 1.7782244682312012,
"learning_rate": 0.001,
"loss": 1.0476,
"step": 200600
},
{
"epoch": 64.86748545572075,
"grad_norm": 2.004187822341919,
"learning_rate": 0.001,
"loss": 1.0579,
"step": 200700
},
{
"epoch": 64.89980607627666,
"grad_norm": 1.5076520442962646,
"learning_rate": 0.001,
"loss": 1.055,
"step": 200800
},
{
"epoch": 64.93212669683258,
"grad_norm": 2.1601130962371826,
"learning_rate": 0.001,
"loss": 1.0416,
"step": 200900
},
{
"epoch": 64.96444731738849,
"grad_norm": 1.9222103357315063,
"learning_rate": 0.001,
"loss": 1.0593,
"step": 201000
},
{
"epoch": 64.99676793794441,
"grad_norm": 2.148522138595581,
"learning_rate": 0.001,
"loss": 1.0256,
"step": 201100
},
{
"epoch": 65.02908855850032,
"grad_norm": 1.4371371269226074,
"learning_rate": 0.001,
"loss": 0.9674,
"step": 201200
},
{
"epoch": 65.06140917905624,
"grad_norm": 1.54662024974823,
"learning_rate": 0.001,
"loss": 0.9374,
"step": 201300
},
{
"epoch": 65.09372979961215,
"grad_norm": 1.464662790298462,
"learning_rate": 0.001,
"loss": 0.9625,
"step": 201400
},
{
"epoch": 65.12605042016807,
"grad_norm": 1.7494497299194336,
"learning_rate": 0.001,
"loss": 0.9913,
"step": 201500
},
{
"epoch": 65.15837104072398,
"grad_norm": 1.547864317893982,
"learning_rate": 0.001,
"loss": 0.9913,
"step": 201600
},
{
"epoch": 65.1906916612799,
"grad_norm": 1.4280509948730469,
"learning_rate": 0.001,
"loss": 0.9887,
"step": 201700
},
{
"epoch": 65.2230122818358,
"grad_norm": 1.948360800743103,
"learning_rate": 0.001,
"loss": 0.9843,
"step": 201800
},
{
"epoch": 65.25533290239173,
"grad_norm": 1.9437856674194336,
"learning_rate": 0.001,
"loss": 0.9906,
"step": 201900
},
{
"epoch": 65.28765352294764,
"grad_norm": 2.2211198806762695,
"learning_rate": 0.001,
"loss": 0.9781,
"step": 202000
},
{
"epoch": 65.31997414350356,
"grad_norm": 1.444524884223938,
"learning_rate": 0.001,
"loss": 1.0022,
"step": 202100
},
{
"epoch": 65.35229476405947,
"grad_norm": 1.7573517560958862,
"learning_rate": 0.001,
"loss": 1.0147,
"step": 202200
},
{
"epoch": 65.38461538461539,
"grad_norm": 1.4979794025421143,
"learning_rate": 0.001,
"loss": 0.9888,
"step": 202300
},
{
"epoch": 65.4169360051713,
"grad_norm": 1.5916309356689453,
"learning_rate": 0.001,
"loss": 0.982,
"step": 202400
},
{
"epoch": 65.44925662572722,
"grad_norm": 1.95499849319458,
"learning_rate": 0.001,
"loss": 1.0023,
"step": 202500
},
{
"epoch": 65.48157724628312,
"grad_norm": 1.6247013807296753,
"learning_rate": 0.001,
"loss": 0.9973,
"step": 202600
},
{
"epoch": 65.51389786683905,
"grad_norm": 1.7667204141616821,
"learning_rate": 0.001,
"loss": 1.0227,
"step": 202700
},
{
"epoch": 65.54621848739495,
"grad_norm": 1.7284674644470215,
"learning_rate": 0.001,
"loss": 1.0197,
"step": 202800
},
{
"epoch": 65.57853910795087,
"grad_norm": 1.888560175895691,
"learning_rate": 0.001,
"loss": 1.028,
"step": 202900
},
{
"epoch": 65.61085972850678,
"grad_norm": 1.4070155620574951,
"learning_rate": 0.001,
"loss": 1.039,
"step": 203000
},
{
"epoch": 65.6431803490627,
"grad_norm": 2.212310552597046,
"learning_rate": 0.001,
"loss": 1.0261,
"step": 203100
},
{
"epoch": 65.67550096961861,
"grad_norm": 2.192976474761963,
"learning_rate": 0.001,
"loss": 1.0353,
"step": 203200
},
{
"epoch": 65.70782159017453,
"grad_norm": 1.7333697080612183,
"learning_rate": 0.001,
"loss": 1.0293,
"step": 203300
},
{
"epoch": 65.74014221073044,
"grad_norm": 1.5100023746490479,
"learning_rate": 0.001,
"loss": 1.0205,
"step": 203400
},
{
"epoch": 65.77246283128636,
"grad_norm": 2.0148110389709473,
"learning_rate": 0.001,
"loss": 1.027,
"step": 203500
},
{
"epoch": 65.80478345184227,
"grad_norm": 1.6597840785980225,
"learning_rate": 0.001,
"loss": 1.0248,
"step": 203600
},
{
"epoch": 65.83710407239819,
"grad_norm": 1.6659585237503052,
"learning_rate": 0.001,
"loss": 1.0414,
"step": 203700
},
{
"epoch": 65.8694246929541,
"grad_norm": 1.5771119594573975,
"learning_rate": 0.001,
"loss": 1.0326,
"step": 203800
},
{
"epoch": 65.90174531351002,
"grad_norm": 1.8948572874069214,
"learning_rate": 0.001,
"loss": 1.0511,
"step": 203900
},
{
"epoch": 65.93406593406593,
"grad_norm": 1.6274173259735107,
"learning_rate": 0.001,
"loss": 1.0555,
"step": 204000
},
{
"epoch": 65.96638655462185,
"grad_norm": 1.5236989259719849,
"learning_rate": 0.001,
"loss": 1.0338,
"step": 204100
},
{
"epoch": 65.99870717517777,
"grad_norm": 1.388321042060852,
"learning_rate": 0.001,
"loss": 1.018,
"step": 204200
},
{
"epoch": 66.03102779573368,
"grad_norm": 1.1837842464447021,
"learning_rate": 0.001,
"loss": 0.9499,
"step": 204300
},
{
"epoch": 66.0633484162896,
"grad_norm": 1.6892098188400269,
"learning_rate": 0.001,
"loss": 0.9493,
"step": 204400
},
{
"epoch": 66.0956690368455,
"grad_norm": 1.2390347719192505,
"learning_rate": 0.001,
"loss": 0.9732,
"step": 204500
},
{
"epoch": 66.12798965740143,
"grad_norm": 1.2077933549880981,
"learning_rate": 0.001,
"loss": 0.9703,
"step": 204600
},
{
"epoch": 66.16031027795734,
"grad_norm": 1.7999850511550903,
"learning_rate": 0.001,
"loss": 0.968,
"step": 204700
},
{
"epoch": 66.19263089851326,
"grad_norm": 1.8721860647201538,
"learning_rate": 0.001,
"loss": 0.9882,
"step": 204800
},
{
"epoch": 66.22495151906917,
"grad_norm": 1.4129645824432373,
"learning_rate": 0.001,
"loss": 0.9709,
"step": 204900
},
{
"epoch": 66.25727213962509,
"grad_norm": 1.5184195041656494,
"learning_rate": 0.001,
"loss": 0.9852,
"step": 205000
},
{
"epoch": 66.289592760181,
"grad_norm": 1.9621903896331787,
"learning_rate": 0.001,
"loss": 0.971,
"step": 205100
},
{
"epoch": 66.32191338073692,
"grad_norm": 1.714373230934143,
"learning_rate": 0.001,
"loss": 0.9942,
"step": 205200
},
{
"epoch": 66.35423400129282,
"grad_norm": 1.475232720375061,
"learning_rate": 0.001,
"loss": 0.987,
"step": 205300
},
{
"epoch": 66.38655462184875,
"grad_norm": 1.7224111557006836,
"learning_rate": 0.001,
"loss": 0.9769,
"step": 205400
},
{
"epoch": 66.41887524240465,
"grad_norm": 1.6576015949249268,
"learning_rate": 0.001,
"loss": 0.998,
"step": 205500
},
{
"epoch": 66.45119586296057,
"grad_norm": 1.739248275756836,
"learning_rate": 0.001,
"loss": 1.0116,
"step": 205600
},
{
"epoch": 66.48351648351648,
"grad_norm": 1.4287973642349243,
"learning_rate": 0.001,
"loss": 0.9823,
"step": 205700
},
{
"epoch": 66.5158371040724,
"grad_norm": 2.0759823322296143,
"learning_rate": 0.001,
"loss": 0.9991,
"step": 205800
},
{
"epoch": 66.54815772462831,
"grad_norm": 1.8355493545532227,
"learning_rate": 0.001,
"loss": 1.0104,
"step": 205900
},
{
"epoch": 66.58047834518423,
"grad_norm": 2.031424045562744,
"learning_rate": 0.001,
"loss": 1.0043,
"step": 206000
},
{
"epoch": 66.61279896574014,
"grad_norm": 1.8214043378829956,
"learning_rate": 0.001,
"loss": 1.0088,
"step": 206100
},
{
"epoch": 66.64511958629606,
"grad_norm": 1.7729460000991821,
"learning_rate": 0.001,
"loss": 1.0072,
"step": 206200
},
{
"epoch": 66.67744020685197,
"grad_norm": 1.6359024047851562,
"learning_rate": 0.001,
"loss": 1.0232,
"step": 206300
},
{
"epoch": 66.70976082740789,
"grad_norm": 1.8432180881500244,
"learning_rate": 0.001,
"loss": 1.0086,
"step": 206400
},
{
"epoch": 66.7420814479638,
"grad_norm": 1.3775078058242798,
"learning_rate": 0.001,
"loss": 1.0241,
"step": 206500
},
{
"epoch": 66.77440206851972,
"grad_norm": 1.574852705001831,
"learning_rate": 0.001,
"loss": 1.0194,
"step": 206600
},
{
"epoch": 66.80672268907563,
"grad_norm": 1.415818691253662,
"learning_rate": 0.001,
"loss": 1.0053,
"step": 206700
},
{
"epoch": 66.83904330963155,
"grad_norm": 1.3461685180664062,
"learning_rate": 0.001,
"loss": 1.0263,
"step": 206800
},
{
"epoch": 66.87136393018746,
"grad_norm": 1.530918836593628,
"learning_rate": 0.001,
"loss": 1.0466,
"step": 206900
},
{
"epoch": 66.90368455074338,
"grad_norm": 1.2895759344100952,
"learning_rate": 0.001,
"loss": 1.0355,
"step": 207000
},
{
"epoch": 66.93600517129929,
"grad_norm": 1.4702177047729492,
"learning_rate": 0.001,
"loss": 1.0289,
"step": 207100
},
{
"epoch": 66.96832579185521,
"grad_norm": 1.5006325244903564,
"learning_rate": 0.001,
"loss": 1.0585,
"step": 207200
},
{
"epoch": 67.00064641241111,
"grad_norm": 2.1274211406707764,
"learning_rate": 0.001,
"loss": 1.016,
"step": 207300
},
{
"epoch": 67.03296703296704,
"grad_norm": 1.4935500621795654,
"learning_rate": 0.001,
"loss": 0.9389,
"step": 207400
},
{
"epoch": 67.06528765352294,
"grad_norm": 1.7395673990249634,
"learning_rate": 0.001,
"loss": 0.9416,
"step": 207500
},
{
"epoch": 67.09760827407887,
"grad_norm": 2.2219367027282715,
"learning_rate": 0.001,
"loss": 0.9603,
"step": 207600
},
{
"epoch": 67.12992889463477,
"grad_norm": 2.3106393814086914,
"learning_rate": 0.001,
"loss": 0.9557,
"step": 207700
},
{
"epoch": 67.1622495151907,
"grad_norm": 2.637923002243042,
"learning_rate": 0.001,
"loss": 0.9448,
"step": 207800
},
{
"epoch": 67.1945701357466,
"grad_norm": 1.9894449710845947,
"learning_rate": 0.001,
"loss": 0.9693,
"step": 207900
},
{
"epoch": 67.22689075630252,
"grad_norm": 2.0248193740844727,
"learning_rate": 0.001,
"loss": 0.9842,
"step": 208000
},
{
"epoch": 67.25921137685843,
"grad_norm": 2.42965030670166,
"learning_rate": 0.001,
"loss": 0.954,
"step": 208100
},
{
"epoch": 67.29153199741435,
"grad_norm": 2.4134888648986816,
"learning_rate": 0.001,
"loss": 0.9746,
"step": 208200
},
{
"epoch": 67.32385261797026,
"grad_norm": 2.5889575481414795,
"learning_rate": 0.001,
"loss": 0.9787,
"step": 208300
},
{
"epoch": 67.35617323852618,
"grad_norm": 2.4583373069763184,
"learning_rate": 0.001,
"loss": 0.977,
"step": 208400
},
{
"epoch": 67.38849385908209,
"grad_norm": 1.9895371198654175,
"learning_rate": 0.001,
"loss": 0.981,
"step": 208500
},
{
"epoch": 67.42081447963801,
"grad_norm": 2.2331972122192383,
"learning_rate": 0.001,
"loss": 1.005,
"step": 208600
},
{
"epoch": 67.45313510019392,
"grad_norm": 1.9639379978179932,
"learning_rate": 0.001,
"loss": 0.9807,
"step": 208700
},
{
"epoch": 67.48545572074984,
"grad_norm": 1.7588427066802979,
"learning_rate": 0.001,
"loss": 1.001,
"step": 208800
},
{
"epoch": 67.51777634130575,
"grad_norm": 2.256831169128418,
"learning_rate": 0.001,
"loss": 1.0069,
"step": 208900
},
{
"epoch": 67.55009696186167,
"grad_norm": 1.9242079257965088,
"learning_rate": 0.001,
"loss": 0.9981,
"step": 209000
},
{
"epoch": 67.58241758241758,
"grad_norm": 2.6344449520111084,
"learning_rate": 0.001,
"loss": 1.0119,
"step": 209100
},
{
"epoch": 67.6147382029735,
"grad_norm": 2.7711076736450195,
"learning_rate": 0.001,
"loss": 1.0018,
"step": 209200
},
{
"epoch": 67.6470588235294,
"grad_norm": 1.9800853729248047,
"learning_rate": 0.001,
"loss": 0.9975,
"step": 209300
},
{
"epoch": 67.67937944408533,
"grad_norm": 1.8756169080734253,
"learning_rate": 0.001,
"loss": 1.0123,
"step": 209400
},
{
"epoch": 67.71170006464124,
"grad_norm": 3.0076963901519775,
"learning_rate": 0.001,
"loss": 1.0041,
"step": 209500
},
{
"epoch": 67.74402068519716,
"grad_norm": 1.6423379182815552,
"learning_rate": 0.001,
"loss": 1.0102,
"step": 209600
},
{
"epoch": 67.77634130575306,
"grad_norm": 2.183377504348755,
"learning_rate": 0.001,
"loss": 1.025,
"step": 209700
},
{
"epoch": 67.80866192630899,
"grad_norm": 3.2337539196014404,
"learning_rate": 0.001,
"loss": 1.0051,
"step": 209800
},
{
"epoch": 67.8409825468649,
"grad_norm": 1.9558117389678955,
"learning_rate": 0.001,
"loss": 1.0139,
"step": 209900
},
{
"epoch": 67.87330316742081,
"grad_norm": 2.0811750888824463,
"learning_rate": 0.001,
"loss": 1.043,
"step": 210000
},
{
"epoch": 67.90562378797672,
"grad_norm": 2.108379602432251,
"learning_rate": 0.001,
"loss": 1.0259,
"step": 210100
},
{
"epoch": 67.93794440853264,
"grad_norm": 2.1077399253845215,
"learning_rate": 0.001,
"loss": 1.0235,
"step": 210200
},
{
"epoch": 67.97026502908855,
"grad_norm": 1.871891975402832,
"learning_rate": 0.001,
"loss": 1.0415,
"step": 210300
},
{
"epoch": 68.00258564964447,
"grad_norm": 1.6436115503311157,
"learning_rate": 0.001,
"loss": 1.0467,
"step": 210400
},
{
"epoch": 68.0349062702004,
"grad_norm": 1.6314891576766968,
"learning_rate": 0.001,
"loss": 0.9279,
"step": 210500
},
{
"epoch": 68.0672268907563,
"grad_norm": 1.5454355478286743,
"learning_rate": 0.001,
"loss": 0.9457,
"step": 210600
},
{
"epoch": 68.09954751131222,
"grad_norm": 1.8596247434616089,
"learning_rate": 0.001,
"loss": 0.9512,
"step": 210700
},
{
"epoch": 68.13186813186813,
"grad_norm": 1.685611367225647,
"learning_rate": 0.001,
"loss": 0.9767,
"step": 210800
},
{
"epoch": 68.16418875242405,
"grad_norm": 1.754593849182129,
"learning_rate": 0.001,
"loss": 0.9568,
"step": 210900
},
{
"epoch": 68.19650937297996,
"grad_norm": 1.898842453956604,
"learning_rate": 0.001,
"loss": 0.9628,
"step": 211000
},
{
"epoch": 68.22882999353588,
"grad_norm": 2.4915871620178223,
"learning_rate": 0.001,
"loss": 0.9551,
"step": 211100
},
{
"epoch": 68.26115061409179,
"grad_norm": 1.8018248081207275,
"learning_rate": 0.001,
"loss": 0.9756,
"step": 211200
},
{
"epoch": 68.29347123464771,
"grad_norm": 1.7845089435577393,
"learning_rate": 0.001,
"loss": 0.9681,
"step": 211300
},
{
"epoch": 68.32579185520362,
"grad_norm": 1.6771483421325684,
"learning_rate": 0.001,
"loss": 0.9567,
"step": 211400
},
{
"epoch": 68.35811247575954,
"grad_norm": 1.7854185104370117,
"learning_rate": 0.001,
"loss": 0.9724,
"step": 211500
},
{
"epoch": 68.39043309631545,
"grad_norm": 3.660853147506714,
"learning_rate": 0.001,
"loss": 0.981,
"step": 211600
},
{
"epoch": 68.42275371687137,
"grad_norm": 1.621956706047058,
"learning_rate": 0.001,
"loss": 0.9833,
"step": 211700
},
{
"epoch": 68.45507433742728,
"grad_norm": 1.5680972337722778,
"learning_rate": 0.001,
"loss": 0.9766,
"step": 211800
},
{
"epoch": 68.4873949579832,
"grad_norm": 1.4635026454925537,
"learning_rate": 0.001,
"loss": 0.986,
"step": 211900
},
{
"epoch": 68.5197155785391,
"grad_norm": 1.9271141290664673,
"learning_rate": 0.001,
"loss": 0.9871,
"step": 212000
},
{
"epoch": 68.55203619909503,
"grad_norm": 1.8947614431381226,
"learning_rate": 0.001,
"loss": 0.9893,
"step": 212100
},
{
"epoch": 68.58435681965094,
"grad_norm": 1.6011998653411865,
"learning_rate": 0.001,
"loss": 0.9805,
"step": 212200
},
{
"epoch": 68.61667744020686,
"grad_norm": 1.5570119619369507,
"learning_rate": 0.001,
"loss": 1.0029,
"step": 212300
},
{
"epoch": 68.64899806076276,
"grad_norm": 1.8010640144348145,
"learning_rate": 0.001,
"loss": 0.9896,
"step": 212400
},
{
"epoch": 68.68131868131869,
"grad_norm": 1.5905638933181763,
"learning_rate": 0.001,
"loss": 0.9957,
"step": 212500
},
{
"epoch": 68.7136393018746,
"grad_norm": 1.521748423576355,
"learning_rate": 0.001,
"loss": 0.9887,
"step": 212600
},
{
"epoch": 68.74595992243052,
"grad_norm": 2.2083487510681152,
"learning_rate": 0.001,
"loss": 1.002,
"step": 212700
},
{
"epoch": 68.77828054298642,
"grad_norm": 1.520521879196167,
"learning_rate": 0.001,
"loss": 0.9973,
"step": 212800
},
{
"epoch": 68.81060116354234,
"grad_norm": 2.090383291244507,
"learning_rate": 0.001,
"loss": 1.0091,
"step": 212900
},
{
"epoch": 68.84292178409825,
"grad_norm": 1.6483055353164673,
"learning_rate": 0.001,
"loss": 1.0292,
"step": 213000
},
{
"epoch": 68.87524240465417,
"grad_norm": 1.4829121828079224,
"learning_rate": 0.001,
"loss": 1.03,
"step": 213100
},
{
"epoch": 68.90756302521008,
"grad_norm": 1.6526368856430054,
"learning_rate": 0.001,
"loss": 1.0217,
"step": 213200
},
{
"epoch": 68.939883645766,
"grad_norm": 1.554659128189087,
"learning_rate": 0.001,
"loss": 1.0293,
"step": 213300
},
{
"epoch": 68.97220426632191,
"grad_norm": 1.735242486000061,
"learning_rate": 0.001,
"loss": 1.0124,
"step": 213400
},
{
"epoch": 69.00452488687783,
"grad_norm": 1.4963144063949585,
"learning_rate": 0.001,
"loss": 1.0175,
"step": 213500
},
{
"epoch": 69.03684550743374,
"grad_norm": 1.603499412536621,
"learning_rate": 0.001,
"loss": 0.9341,
"step": 213600
},
{
"epoch": 69.06916612798966,
"grad_norm": 1.6510496139526367,
"learning_rate": 0.001,
"loss": 0.9354,
"step": 213700
},
{
"epoch": 69.10148674854557,
"grad_norm": 1.66983962059021,
"learning_rate": 0.001,
"loss": 0.9496,
"step": 213800
},
{
"epoch": 69.13380736910149,
"grad_norm": 1.7199987173080444,
"learning_rate": 0.001,
"loss": 0.9496,
"step": 213900
},
{
"epoch": 69.1661279896574,
"grad_norm": 1.773880124092102,
"learning_rate": 0.001,
"loss": 0.949,
"step": 214000
},
{
"epoch": 69.19844861021332,
"grad_norm": 1.6391410827636719,
"learning_rate": 0.001,
"loss": 0.9458,
"step": 214100
},
{
"epoch": 69.23076923076923,
"grad_norm": 1.5789082050323486,
"learning_rate": 0.001,
"loss": 0.9601,
"step": 214200
},
{
"epoch": 69.26308985132515,
"grad_norm": 1.9064000844955444,
"learning_rate": 0.001,
"loss": 0.9513,
"step": 214300
},
{
"epoch": 69.29541047188106,
"grad_norm": 1.6485440731048584,
"learning_rate": 0.001,
"loss": 0.9489,
"step": 214400
},
{
"epoch": 69.32773109243698,
"grad_norm": 1.4657033681869507,
"learning_rate": 0.001,
"loss": 0.9752,
"step": 214500
},
{
"epoch": 69.36005171299288,
"grad_norm": 1.8752721548080444,
"learning_rate": 0.001,
"loss": 0.9484,
"step": 214600
},
{
"epoch": 69.3923723335488,
"grad_norm": 1.4221304655075073,
"learning_rate": 0.001,
"loss": 0.9518,
"step": 214700
},
{
"epoch": 69.42469295410471,
"grad_norm": 1.6716110706329346,
"learning_rate": 0.001,
"loss": 0.9571,
"step": 214800
},
{
"epoch": 69.45701357466064,
"grad_norm": 1.6328706741333008,
"learning_rate": 0.001,
"loss": 0.9617,
"step": 214900
},
{
"epoch": 69.48933419521654,
"grad_norm": 1.467498779296875,
"learning_rate": 0.001,
"loss": 0.993,
"step": 215000
},
{
"epoch": 69.52165481577246,
"grad_norm": 1.6579245328903198,
"learning_rate": 0.001,
"loss": 0.9775,
"step": 215100
},
{
"epoch": 69.55397543632837,
"grad_norm": 1.9347474575042725,
"learning_rate": 0.001,
"loss": 0.9802,
"step": 215200
},
{
"epoch": 69.5862960568843,
"grad_norm": 1.8056132793426514,
"learning_rate": 0.001,
"loss": 0.9917,
"step": 215300
},
{
"epoch": 69.6186166774402,
"grad_norm": 1.5170385837554932,
"learning_rate": 0.001,
"loss": 0.989,
"step": 215400
},
{
"epoch": 69.65093729799612,
"grad_norm": 1.9028490781784058,
"learning_rate": 0.001,
"loss": 1.0019,
"step": 215500
},
{
"epoch": 69.68325791855203,
"grad_norm": 1.7090563774108887,
"learning_rate": 0.001,
"loss": 0.9913,
"step": 215600
},
{
"epoch": 69.71557853910795,
"grad_norm": 3.4204039573669434,
"learning_rate": 0.001,
"loss": 0.9862,
"step": 215700
},
{
"epoch": 69.74789915966386,
"grad_norm": 1.559909462928772,
"learning_rate": 0.001,
"loss": 0.9973,
"step": 215800
},
{
"epoch": 69.78021978021978,
"grad_norm": 1.948533058166504,
"learning_rate": 0.001,
"loss": 0.9941,
"step": 215900
},
{
"epoch": 69.81254040077569,
"grad_norm": 2.0749168395996094,
"learning_rate": 0.001,
"loss": 0.9959,
"step": 216000
},
{
"epoch": 69.84486102133161,
"grad_norm": 1.929993748664856,
"learning_rate": 0.001,
"loss": 0.9985,
"step": 216100
},
{
"epoch": 69.87718164188752,
"grad_norm": 1.607664942741394,
"learning_rate": 0.001,
"loss": 0.9967,
"step": 216200
},
{
"epoch": 69.90950226244344,
"grad_norm": 1.9539837837219238,
"learning_rate": 0.001,
"loss": 1.0014,
"step": 216300
},
{
"epoch": 69.94182288299935,
"grad_norm": 1.4872267246246338,
"learning_rate": 0.001,
"loss": 1.0314,
"step": 216400
},
{
"epoch": 69.97414350355527,
"grad_norm": 1.7044696807861328,
"learning_rate": 0.001,
"loss": 1.0059,
"step": 216500
},
{
"epoch": 70.00646412411119,
"grad_norm": 56.00265121459961,
"learning_rate": 0.001,
"loss": 1.0061,
"step": 216600
},
{
"epoch": 70.0387847446671,
"grad_norm": 1.7173796892166138,
"learning_rate": 0.001,
"loss": 0.9109,
"step": 216700
},
{
"epoch": 70.07110536522302,
"grad_norm": 1.9290728569030762,
"learning_rate": 0.001,
"loss": 0.9124,
"step": 216800
},
{
"epoch": 70.10342598577893,
"grad_norm": 1.516363263130188,
"learning_rate": 0.001,
"loss": 0.9359,
"step": 216900
},
{
"epoch": 70.13574660633485,
"grad_norm": 2.09360408782959,
"learning_rate": 0.001,
"loss": 0.9359,
"step": 217000
},
{
"epoch": 70.16806722689076,
"grad_norm": 1.570351004600525,
"learning_rate": 0.001,
"loss": 0.933,
"step": 217100
},
{
"epoch": 70.20038784744668,
"grad_norm": 1.773545742034912,
"learning_rate": 0.001,
"loss": 0.9225,
"step": 217200
},
{
"epoch": 70.23270846800258,
"grad_norm": 1.9614516496658325,
"learning_rate": 0.001,
"loss": 0.9409,
"step": 217300
},
{
"epoch": 70.2650290885585,
"grad_norm": 1.9093842506408691,
"learning_rate": 0.001,
"loss": 0.945,
"step": 217400
},
{
"epoch": 70.29734970911441,
"grad_norm": 1.4649842977523804,
"learning_rate": 0.001,
"loss": 0.9564,
"step": 217500
},
{
"epoch": 70.32967032967034,
"grad_norm": 1.7459298372268677,
"learning_rate": 0.001,
"loss": 0.9691,
"step": 217600
},
{
"epoch": 70.36199095022624,
"grad_norm": 1.5805401802062988,
"learning_rate": 0.001,
"loss": 0.9633,
"step": 217700
},
{
"epoch": 70.39431157078216,
"grad_norm": 1.8536136150360107,
"learning_rate": 0.001,
"loss": 0.9546,
"step": 217800
},
{
"epoch": 70.42663219133807,
"grad_norm": 1.3968721628189087,
"learning_rate": 0.001,
"loss": 0.9538,
"step": 217900
},
{
"epoch": 70.458952811894,
"grad_norm": 2.076166868209839,
"learning_rate": 0.001,
"loss": 0.9602,
"step": 218000
},
{
"epoch": 70.4912734324499,
"grad_norm": 1.5355348587036133,
"learning_rate": 0.001,
"loss": 0.9816,
"step": 218100
},
{
"epoch": 70.52359405300582,
"grad_norm": 1.6751102209091187,
"learning_rate": 0.001,
"loss": 0.9691,
"step": 218200
},
{
"epoch": 70.55591467356173,
"grad_norm": 1.6039971113204956,
"learning_rate": 0.001,
"loss": 0.9945,
"step": 218300
},
{
"epoch": 70.58823529411765,
"grad_norm": 1.4298136234283447,
"learning_rate": 0.001,
"loss": 0.998,
"step": 218400
},
{
"epoch": 70.62055591467356,
"grad_norm": 1.6534899473190308,
"learning_rate": 0.001,
"loss": 0.9642,
"step": 218500
},
{
"epoch": 70.65287653522948,
"grad_norm": 1.645822286605835,
"learning_rate": 0.001,
"loss": 0.9765,
"step": 218600
},
{
"epoch": 70.68519715578539,
"grad_norm": 1.5585241317749023,
"learning_rate": 0.001,
"loss": 0.9683,
"step": 218700
},
{
"epoch": 70.71751777634131,
"grad_norm": 2.097666025161743,
"learning_rate": 0.001,
"loss": 0.9765,
"step": 218800
},
{
"epoch": 70.74983839689722,
"grad_norm": 1.694786787033081,
"learning_rate": 0.001,
"loss": 0.9764,
"step": 218900
},
{
"epoch": 70.78215901745314,
"grad_norm": 1.9305164813995361,
"learning_rate": 0.001,
"loss": 0.9903,
"step": 219000
},
{
"epoch": 70.81447963800905,
"grad_norm": 1.6934314966201782,
"learning_rate": 0.001,
"loss": 0.9976,
"step": 219100
},
{
"epoch": 70.84680025856497,
"grad_norm": 1.8787205219268799,
"learning_rate": 0.001,
"loss": 1.0127,
"step": 219200
},
{
"epoch": 70.87912087912088,
"grad_norm": 1.8684394359588623,
"learning_rate": 0.001,
"loss": 0.9992,
"step": 219300
},
{
"epoch": 70.9114414996768,
"grad_norm": 1.5081862211227417,
"learning_rate": 0.001,
"loss": 1.005,
"step": 219400
},
{
"epoch": 70.9437621202327,
"grad_norm": 1.4214324951171875,
"learning_rate": 0.001,
"loss": 0.9998,
"step": 219500
},
{
"epoch": 70.97608274078863,
"grad_norm": 1.6040948629379272,
"learning_rate": 0.001,
"loss": 0.9977,
"step": 219600
},
{
"epoch": 71.00840336134453,
"grad_norm": 1.9321045875549316,
"learning_rate": 0.001,
"loss": 1.0009,
"step": 219700
},
{
"epoch": 71.04072398190046,
"grad_norm": 1.6411948204040527,
"learning_rate": 0.001,
"loss": 0.9007,
"step": 219800
},
{
"epoch": 71.07304460245636,
"grad_norm": 1.5181108713150024,
"learning_rate": 0.001,
"loss": 0.9252,
"step": 219900
},
{
"epoch": 71.10536522301229,
"grad_norm": 1.6294838190078735,
"learning_rate": 0.001,
"loss": 0.9105,
"step": 220000
},
{
"epoch": 71.13768584356819,
"grad_norm": 1.670454502105713,
"learning_rate": 0.001,
"loss": 0.9215,
"step": 220100
},
{
"epoch": 71.17000646412411,
"grad_norm": 1.461074948310852,
"learning_rate": 0.001,
"loss": 0.9223,
"step": 220200
},
{
"epoch": 71.20232708468002,
"grad_norm": 1.9516828060150146,
"learning_rate": 0.001,
"loss": 0.9358,
"step": 220300
},
{
"epoch": 71.23464770523594,
"grad_norm": 1.5460422039031982,
"learning_rate": 0.001,
"loss": 0.9511,
"step": 220400
},
{
"epoch": 71.26696832579185,
"grad_norm": 1.5585349798202515,
"learning_rate": 0.001,
"loss": 0.9239,
"step": 220500
},
{
"epoch": 71.29928894634777,
"grad_norm": 1.466973900794983,
"learning_rate": 0.001,
"loss": 0.9475,
"step": 220600
},
{
"epoch": 71.33160956690368,
"grad_norm": 1.8937549591064453,
"learning_rate": 0.001,
"loss": 0.9497,
"step": 220700
},
{
"epoch": 71.3639301874596,
"grad_norm": 1.702267050743103,
"learning_rate": 0.001,
"loss": 0.9463,
"step": 220800
},
{
"epoch": 71.39625080801551,
"grad_norm": 2.2292089462280273,
"learning_rate": 0.001,
"loss": 0.9696,
"step": 220900
},
{
"epoch": 71.42857142857143,
"grad_norm": 1.7881437540054321,
"learning_rate": 0.001,
"loss": 0.9697,
"step": 221000
},
{
"epoch": 71.46089204912734,
"grad_norm": 1.5955432653427124,
"learning_rate": 0.001,
"loss": 0.9586,
"step": 221100
},
{
"epoch": 71.49321266968326,
"grad_norm": 1.5466030836105347,
"learning_rate": 0.001,
"loss": 0.9773,
"step": 221200
},
{
"epoch": 71.52553329023917,
"grad_norm": 1.5583151578903198,
"learning_rate": 0.001,
"loss": 0.9608,
"step": 221300
},
{
"epoch": 71.55785391079509,
"grad_norm": 1.478674292564392,
"learning_rate": 0.001,
"loss": 0.9787,
"step": 221400
},
{
"epoch": 71.590174531351,
"grad_norm": 1.3500618934631348,
"learning_rate": 0.001,
"loss": 0.9751,
"step": 221500
},
{
"epoch": 71.62249515190692,
"grad_norm": 1.8969987630844116,
"learning_rate": 0.001,
"loss": 0.9751,
"step": 221600
},
{
"epoch": 71.65481577246283,
"grad_norm": 1.7005226612091064,
"learning_rate": 0.001,
"loss": 0.9813,
"step": 221700
},
{
"epoch": 71.68713639301875,
"grad_norm": 1.568214774131775,
"learning_rate": 0.001,
"loss": 0.9743,
"step": 221800
},
{
"epoch": 71.71945701357465,
"grad_norm": 1.417033314704895,
"learning_rate": 0.001,
"loss": 0.9677,
"step": 221900
},
{
"epoch": 71.75177763413058,
"grad_norm": 1.7298557758331299,
"learning_rate": 0.001,
"loss": 0.9805,
"step": 222000
},
{
"epoch": 71.78409825468648,
"grad_norm": 1.6061108112335205,
"learning_rate": 0.001,
"loss": 0.9917,
"step": 222100
},
{
"epoch": 71.8164188752424,
"grad_norm": 1.9525036811828613,
"learning_rate": 0.001,
"loss": 0.9839,
"step": 222200
},
{
"epoch": 71.84873949579831,
"grad_norm": 1.677465796470642,
"learning_rate": 0.001,
"loss": 1.0039,
"step": 222300
},
{
"epoch": 71.88106011635423,
"grad_norm": 2.25289249420166,
"learning_rate": 0.001,
"loss": 0.9813,
"step": 222400
},
{
"epoch": 71.91338073691014,
"grad_norm": 1.4465690851211548,
"learning_rate": 0.001,
"loss": 0.9896,
"step": 222500
},
{
"epoch": 71.94570135746606,
"grad_norm": 1.8012648820877075,
"learning_rate": 0.001,
"loss": 0.9868,
"step": 222600
},
{
"epoch": 71.97802197802197,
"grad_norm": 1.6948093175888062,
"learning_rate": 0.001,
"loss": 1.0119,
"step": 222700
},
{
"epoch": 72.01034259857789,
"grad_norm": 1.7720965147018433,
"learning_rate": 0.001,
"loss": 0.9562,
"step": 222800
},
{
"epoch": 72.04266321913381,
"grad_norm": 1.5736439228057861,
"learning_rate": 0.001,
"loss": 0.8985,
"step": 222900
},
{
"epoch": 72.07498383968972,
"grad_norm": 1.4669991731643677,
"learning_rate": 0.001,
"loss": 0.9112,
"step": 223000
},
{
"epoch": 72.10730446024564,
"grad_norm": 1.6496646404266357,
"learning_rate": 0.001,
"loss": 0.9098,
"step": 223100
},
{
"epoch": 72.13962508080155,
"grad_norm": 1.6294455528259277,
"learning_rate": 0.001,
"loss": 0.9193,
"step": 223200
},
{
"epoch": 72.17194570135747,
"grad_norm": 1.664873480796814,
"learning_rate": 0.001,
"loss": 0.9216,
"step": 223300
},
{
"epoch": 72.20426632191338,
"grad_norm": 1.6307495832443237,
"learning_rate": 0.001,
"loss": 0.9311,
"step": 223400
},
{
"epoch": 72.2365869424693,
"grad_norm": 1.928640604019165,
"learning_rate": 0.001,
"loss": 0.9326,
"step": 223500
},
{
"epoch": 72.26890756302521,
"grad_norm": 1.5909250974655151,
"learning_rate": 0.001,
"loss": 0.9341,
"step": 223600
},
{
"epoch": 72.30122818358113,
"grad_norm": 1.8960169553756714,
"learning_rate": 0.001,
"loss": 0.9352,
"step": 223700
},
{
"epoch": 72.33354880413704,
"grad_norm": 1.693377137184143,
"learning_rate": 0.001,
"loss": 0.9254,
"step": 223800
},
{
"epoch": 72.36586942469296,
"grad_norm": 1.8388135433197021,
"learning_rate": 0.001,
"loss": 0.9422,
"step": 223900
},
{
"epoch": 72.39819004524887,
"grad_norm": 1.7362223863601685,
"learning_rate": 0.001,
"loss": 0.9542,
"step": 224000
},
{
"epoch": 72.43051066580479,
"grad_norm": 2.2575292587280273,
"learning_rate": 0.001,
"loss": 0.9544,
"step": 224100
},
{
"epoch": 72.4628312863607,
"grad_norm": 1.8790425062179565,
"learning_rate": 0.001,
"loss": 0.9703,
"step": 224200
},
{
"epoch": 72.49515190691662,
"grad_norm": 1.7574528455734253,
"learning_rate": 0.001,
"loss": 0.9609,
"step": 224300
},
{
"epoch": 72.52747252747253,
"grad_norm": 1.6900725364685059,
"learning_rate": 0.001,
"loss": 0.9607,
"step": 224400
},
{
"epoch": 72.55979314802845,
"grad_norm": 1.4746700525283813,
"learning_rate": 0.001,
"loss": 0.9459,
"step": 224500
},
{
"epoch": 72.59211376858435,
"grad_norm": 1.7143659591674805,
"learning_rate": 0.001,
"loss": 0.9697,
"step": 224600
},
{
"epoch": 72.62443438914028,
"grad_norm": 1.6546257734298706,
"learning_rate": 0.001,
"loss": 0.9662,
"step": 224700
},
{
"epoch": 72.65675500969618,
"grad_norm": 1.6653079986572266,
"learning_rate": 0.001,
"loss": 0.96,
"step": 224800
},
{
"epoch": 72.6890756302521,
"grad_norm": 6.541753768920898,
"learning_rate": 0.001,
"loss": 0.9632,
"step": 224900
},
{
"epoch": 72.72139625080801,
"grad_norm": 1.6017827987670898,
"learning_rate": 0.001,
"loss": 0.9813,
"step": 225000
},
{
"epoch": 72.75371687136393,
"grad_norm": 1.894031286239624,
"learning_rate": 0.001,
"loss": 0.9779,
"step": 225100
},
{
"epoch": 72.78603749191984,
"grad_norm": 2.0835087299346924,
"learning_rate": 0.001,
"loss": 0.9904,
"step": 225200
},
{
"epoch": 72.81835811247576,
"grad_norm": 1.5882728099822998,
"learning_rate": 0.001,
"loss": 0.9841,
"step": 225300
},
{
"epoch": 72.85067873303167,
"grad_norm": 1.8402472734451294,
"learning_rate": 0.001,
"loss": 0.9843,
"step": 225400
},
{
"epoch": 72.88299935358759,
"grad_norm": 22.45509910583496,
"learning_rate": 0.001,
"loss": 0.9746,
"step": 225500
},
{
"epoch": 72.9153199741435,
"grad_norm": 1.7956304550170898,
"learning_rate": 0.001,
"loss": 0.9798,
"step": 225600
},
{
"epoch": 72.94764059469942,
"grad_norm": 1.881117582321167,
"learning_rate": 0.001,
"loss": 0.9862,
"step": 225700
},
{
"epoch": 72.97996121525533,
"grad_norm": 1.8564492464065552,
"learning_rate": 0.001,
"loss": 0.9993,
"step": 225800
},
{
"epoch": 73.01228183581125,
"grad_norm": 1.7108012437820435,
"learning_rate": 0.001,
"loss": 0.9341,
"step": 225900
},
{
"epoch": 73.04460245636716,
"grad_norm": 3.303130865097046,
"learning_rate": 0.001,
"loss": 0.8865,
"step": 226000
},
{
"epoch": 73.07692307692308,
"grad_norm": 1.7715092897415161,
"learning_rate": 0.001,
"loss": 0.9084,
"step": 226100
},
{
"epoch": 73.10924369747899,
"grad_norm": 2.105649948120117,
"learning_rate": 0.001,
"loss": 0.9055,
"step": 226200
},
{
"epoch": 73.14156431803491,
"grad_norm": 1.626462697982788,
"learning_rate": 0.001,
"loss": 0.9072,
"step": 226300
},
{
"epoch": 73.17388493859082,
"grad_norm": 1.675002932548523,
"learning_rate": 0.001,
"loss": 0.9123,
"step": 226400
},
{
"epoch": 73.20620555914674,
"grad_norm": 1.7024825811386108,
"learning_rate": 0.001,
"loss": 0.9245,
"step": 226500
},
{
"epoch": 73.23852617970265,
"grad_norm": 1.7466365098953247,
"learning_rate": 0.001,
"loss": 0.9124,
"step": 226600
},
{
"epoch": 73.27084680025857,
"grad_norm": 2.140228748321533,
"learning_rate": 0.001,
"loss": 0.9284,
"step": 226700
},
{
"epoch": 73.30316742081448,
"grad_norm": 1.6669248342514038,
"learning_rate": 0.001,
"loss": 0.934,
"step": 226800
},
{
"epoch": 73.3354880413704,
"grad_norm": 1.5409226417541504,
"learning_rate": 0.001,
"loss": 0.93,
"step": 226900
},
{
"epoch": 73.3678086619263,
"grad_norm": 1.4296633005142212,
"learning_rate": 0.001,
"loss": 0.9363,
"step": 227000
},
{
"epoch": 73.40012928248223,
"grad_norm": 1.7514222860336304,
"learning_rate": 0.001,
"loss": 0.9472,
"step": 227100
},
{
"epoch": 73.43244990303813,
"grad_norm": 1.5675846338272095,
"learning_rate": 0.001,
"loss": 0.9367,
"step": 227200
},
{
"epoch": 73.46477052359405,
"grad_norm": 1.4617664813995361,
"learning_rate": 0.001,
"loss": 0.9389,
"step": 227300
},
{
"epoch": 73.49709114414996,
"grad_norm": 1.8842968940734863,
"learning_rate": 0.001,
"loss": 0.9441,
"step": 227400
},
{
"epoch": 73.52941176470588,
"grad_norm": 1.7531907558441162,
"learning_rate": 0.001,
"loss": 0.9572,
"step": 227500
},
{
"epoch": 73.56173238526179,
"grad_norm": 1.828428864479065,
"learning_rate": 0.001,
"loss": 0.9802,
"step": 227600
},
{
"epoch": 73.59405300581771,
"grad_norm": 5.751527786254883,
"learning_rate": 0.001,
"loss": 0.9551,
"step": 227700
},
{
"epoch": 73.62637362637362,
"grad_norm": 1.549415111541748,
"learning_rate": 0.001,
"loss": 0.9573,
"step": 227800
},
{
"epoch": 73.65869424692954,
"grad_norm": 2.144838333129883,
"learning_rate": 0.001,
"loss": 0.9626,
"step": 227900
},
{
"epoch": 73.69101486748545,
"grad_norm": 1.6326838731765747,
"learning_rate": 0.001,
"loss": 0.9568,
"step": 228000
},
{
"epoch": 73.72333548804137,
"grad_norm": 1.6560914516448975,
"learning_rate": 0.001,
"loss": 0.9749,
"step": 228100
},
{
"epoch": 73.75565610859728,
"grad_norm": 1.5580803155899048,
"learning_rate": 0.001,
"loss": 0.9803,
"step": 228200
},
{
"epoch": 73.7879767291532,
"grad_norm": 1.5272198915481567,
"learning_rate": 0.001,
"loss": 0.9738,
"step": 228300
},
{
"epoch": 73.82029734970911,
"grad_norm": 1.8833277225494385,
"learning_rate": 0.001,
"loss": 0.9753,
"step": 228400
},
{
"epoch": 73.85261797026503,
"grad_norm": 1.8136698007583618,
"learning_rate": 0.001,
"loss": 0.9737,
"step": 228500
},
{
"epoch": 73.88493859082094,
"grad_norm": 1.8745663166046143,
"learning_rate": 0.001,
"loss": 0.9822,
"step": 228600
},
{
"epoch": 73.91725921137686,
"grad_norm": 1.5352058410644531,
"learning_rate": 0.001,
"loss": 0.9901,
"step": 228700
},
{
"epoch": 73.94957983193277,
"grad_norm": 1.4566476345062256,
"learning_rate": 0.001,
"loss": 0.9663,
"step": 228800
},
{
"epoch": 73.98190045248869,
"grad_norm": 1.6003284454345703,
"learning_rate": 0.001,
"loss": 0.9907,
"step": 228900
},
{
"epoch": 74.01422107304461,
"grad_norm": 1.3612558841705322,
"learning_rate": 0.001,
"loss": 0.909,
"step": 229000
},
{
"epoch": 74.04654169360052,
"grad_norm": 1.8809579610824585,
"learning_rate": 0.001,
"loss": 0.9056,
"step": 229100
},
{
"epoch": 74.07886231415644,
"grad_norm": 1.5369434356689453,
"learning_rate": 0.001,
"loss": 0.895,
"step": 229200
},
{
"epoch": 74.11118293471235,
"grad_norm": 2.4605133533477783,
"learning_rate": 0.001,
"loss": 0.9107,
"step": 229300
},
{
"epoch": 74.14350355526827,
"grad_norm": 1.4680637121200562,
"learning_rate": 0.001,
"loss": 0.9069,
"step": 229400
},
{
"epoch": 74.17582417582418,
"grad_norm": 2.1316113471984863,
"learning_rate": 0.001,
"loss": 0.912,
"step": 229500
},
{
"epoch": 74.2081447963801,
"grad_norm": 2.0725295543670654,
"learning_rate": 0.001,
"loss": 0.9133,
"step": 229600
},
{
"epoch": 74.240465416936,
"grad_norm": 1.671245813369751,
"learning_rate": 0.001,
"loss": 0.9045,
"step": 229700
},
{
"epoch": 74.27278603749193,
"grad_norm": 1.4853520393371582,
"learning_rate": 0.001,
"loss": 0.9284,
"step": 229800
},
{
"epoch": 74.30510665804783,
"grad_norm": 2.123257875442505,
"learning_rate": 0.001,
"loss": 0.9156,
"step": 229900
},
{
"epoch": 74.33742727860376,
"grad_norm": 1.213315486907959,
"learning_rate": 0.001,
"loss": 0.9246,
"step": 230000
},
{
"epoch": 74.36974789915966,
"grad_norm": 1.6315170526504517,
"learning_rate": 0.001,
"loss": 0.9245,
"step": 230100
},
{
"epoch": 74.40206851971558,
"grad_norm": 1.4847023487091064,
"learning_rate": 0.001,
"loss": 0.9439,
"step": 230200
},
{
"epoch": 74.43438914027149,
"grad_norm": 2.0054616928100586,
"learning_rate": 0.001,
"loss": 0.944,
"step": 230300
},
{
"epoch": 74.46670976082741,
"grad_norm": 1.3979027271270752,
"learning_rate": 0.001,
"loss": 0.9408,
"step": 230400
},
{
"epoch": 74.49903038138332,
"grad_norm": 1.9914604425430298,
"learning_rate": 0.001,
"loss": 0.9585,
"step": 230500
},
{
"epoch": 74.53135100193924,
"grad_norm": 2.074437379837036,
"learning_rate": 0.001,
"loss": 0.9411,
"step": 230600
},
{
"epoch": 74.56367162249515,
"grad_norm": 1.5431259870529175,
"learning_rate": 0.001,
"loss": 0.9452,
"step": 230700
},
{
"epoch": 74.59599224305107,
"grad_norm": 1.7445917129516602,
"learning_rate": 0.001,
"loss": 0.9415,
"step": 230800
},
{
"epoch": 74.62831286360698,
"grad_norm": 1.5942051410675049,
"learning_rate": 0.001,
"loss": 0.9594,
"step": 230900
},
{
"epoch": 74.6606334841629,
"grad_norm": 1.7823764085769653,
"learning_rate": 0.001,
"loss": 0.9477,
"step": 231000
},
{
"epoch": 74.69295410471881,
"grad_norm": 1.778059959411621,
"learning_rate": 0.001,
"loss": 0.9554,
"step": 231100
},
{
"epoch": 74.72527472527473,
"grad_norm": 1.4539250135421753,
"learning_rate": 0.001,
"loss": 0.9525,
"step": 231200
},
{
"epoch": 74.75759534583064,
"grad_norm": 1.9402769804000854,
"learning_rate": 0.001,
"loss": 0.9544,
"step": 231300
},
{
"epoch": 74.78991596638656,
"grad_norm": 1.6188675165176392,
"learning_rate": 0.001,
"loss": 0.9697,
"step": 231400
},
{
"epoch": 74.82223658694247,
"grad_norm": 1.5256141424179077,
"learning_rate": 0.001,
"loss": 0.9508,
"step": 231500
},
{
"epoch": 74.85455720749839,
"grad_norm": 1.6139826774597168,
"learning_rate": 0.001,
"loss": 0.9579,
"step": 231600
},
{
"epoch": 74.8868778280543,
"grad_norm": 1.597104787826538,
"learning_rate": 0.001,
"loss": 0.9626,
"step": 231700
},
{
"epoch": 74.91919844861022,
"grad_norm": 1.529085636138916,
"learning_rate": 0.001,
"loss": 0.9808,
"step": 231800
},
{
"epoch": 74.95151906916612,
"grad_norm": 1.8991279602050781,
"learning_rate": 0.001,
"loss": 0.9758,
"step": 231900
},
{
"epoch": 74.98383968972205,
"grad_norm": 1.5026214122772217,
"learning_rate": 0.001,
"loss": 0.984,
"step": 232000
},
{
"epoch": 75.01616031027795,
"grad_norm": 1.071177363395691,
"learning_rate": 0.001,
"loss": 0.8859,
"step": 232100
},
{
"epoch": 75.04848093083388,
"grad_norm": 1.1679174900054932,
"learning_rate": 0.001,
"loss": 0.8884,
"step": 232200
},
{
"epoch": 75.08080155138978,
"grad_norm": 0.9388474225997925,
"learning_rate": 0.001,
"loss": 0.8708,
"step": 232300
},
{
"epoch": 75.1131221719457,
"grad_norm": 0.4192671775817871,
"learning_rate": 0.001,
"loss": 0.896,
"step": 232400
},
{
"epoch": 75.14544279250161,
"grad_norm": 0.6754704117774963,
"learning_rate": 0.001,
"loss": 0.9114,
"step": 232500
},
{
"epoch": 75.17776341305753,
"grad_norm": 0.7327497005462646,
"learning_rate": 0.001,
"loss": 0.8957,
"step": 232600
},
{
"epoch": 75.21008403361344,
"grad_norm": 1.0593340396881104,
"learning_rate": 0.001,
"loss": 0.9038,
"step": 232700
},
{
"epoch": 75.24240465416936,
"grad_norm": 0.8486055135726929,
"learning_rate": 0.001,
"loss": 0.9076,
"step": 232800
},
{
"epoch": 75.27472527472527,
"grad_norm": 0.3056071996688843,
"learning_rate": 0.001,
"loss": 0.9377,
"step": 232900
},
{
"epoch": 75.30704589528119,
"grad_norm": 0.7975426316261292,
"learning_rate": 0.001,
"loss": 0.9175,
"step": 233000
},
{
"epoch": 75.3393665158371,
"grad_norm": 1.1826505661010742,
"learning_rate": 0.001,
"loss": 0.9082,
"step": 233100
},
{
"epoch": 75.37168713639302,
"grad_norm": 0.6041250824928284,
"learning_rate": 0.001,
"loss": 0.9184,
"step": 233200
},
{
"epoch": 75.40400775694893,
"grad_norm": 0.9403012990951538,
"learning_rate": 0.001,
"loss": 0.9225,
"step": 233300
},
{
"epoch": 75.43632837750485,
"grad_norm": 0.2614690959453583,
"learning_rate": 0.001,
"loss": 0.9362,
"step": 233400
},
{
"epoch": 75.46864899806076,
"grad_norm": 0.6960493922233582,
"learning_rate": 0.001,
"loss": 0.9441,
"step": 233500
},
{
"epoch": 75.50096961861668,
"grad_norm": 0.4896499216556549,
"learning_rate": 0.001,
"loss": 0.9179,
"step": 233600
},
{
"epoch": 75.53329023917259,
"grad_norm": 0.5739967226982117,
"learning_rate": 0.001,
"loss": 0.9323,
"step": 233700
},
{
"epoch": 75.56561085972851,
"grad_norm": 0.287398099899292,
"learning_rate": 0.001,
"loss": 0.9521,
"step": 233800
},
{
"epoch": 75.59793148028442,
"grad_norm": 0.7290223240852356,
"learning_rate": 0.001,
"loss": 0.9349,
"step": 233900
},
{
"epoch": 75.63025210084034,
"grad_norm": 0.48519167304039,
"learning_rate": 0.001,
"loss": 0.9373,
"step": 234000
},
{
"epoch": 75.66257272139624,
"grad_norm": 0.6618010401725769,
"learning_rate": 0.001,
"loss": 0.9405,
"step": 234100
},
{
"epoch": 75.69489334195217,
"grad_norm": 1.3536745309829712,
"learning_rate": 0.001,
"loss": 0.945,
"step": 234200
},
{
"epoch": 75.72721396250807,
"grad_norm": 1.1508914232254028,
"learning_rate": 0.001,
"loss": 0.9543,
"step": 234300
},
{
"epoch": 75.759534583064,
"grad_norm": 0.33295896649360657,
"learning_rate": 0.001,
"loss": 0.9693,
"step": 234400
},
{
"epoch": 75.7918552036199,
"grad_norm": 0.6864619255065918,
"learning_rate": 0.001,
"loss": 0.9433,
"step": 234500
},
{
"epoch": 75.82417582417582,
"grad_norm": 0.7642561197280884,
"learning_rate": 0.001,
"loss": 0.9418,
"step": 234600
},
{
"epoch": 75.85649644473173,
"grad_norm": 1.3688499927520752,
"learning_rate": 0.001,
"loss": 0.9591,
"step": 234700
},
{
"epoch": 75.88881706528765,
"grad_norm": 0.7520803809165955,
"learning_rate": 0.001,
"loss": 0.9798,
"step": 234800
},
{
"epoch": 75.92113768584356,
"grad_norm": 0.8843240737915039,
"learning_rate": 0.001,
"loss": 0.9686,
"step": 234900
},
{
"epoch": 75.95345830639948,
"grad_norm": 0.34943240880966187,
"learning_rate": 0.001,
"loss": 0.9566,
"step": 235000
},
{
"epoch": 75.98577892695539,
"grad_norm": 0.8272552490234375,
"learning_rate": 0.001,
"loss": 0.9776,
"step": 235100
},
{
"epoch": 76.01809954751131,
"grad_norm": 1.922287940979004,
"learning_rate": 0.001,
"loss": 0.9338,
"step": 235200
},
{
"epoch": 76.05042016806723,
"grad_norm": 6.02496337890625,
"learning_rate": 0.001,
"loss": 0.878,
"step": 235300
},
{
"epoch": 76.08274078862314,
"grad_norm": 1.6479023694992065,
"learning_rate": 0.001,
"loss": 0.8833,
"step": 235400
},
{
"epoch": 76.11506140917906,
"grad_norm": 2.0861504077911377,
"learning_rate": 0.001,
"loss": 0.8836,
"step": 235500
},
{
"epoch": 76.14738202973497,
"grad_norm": 2.1388919353485107,
"learning_rate": 0.001,
"loss": 0.8864,
"step": 235600
},
{
"epoch": 76.17970265029089,
"grad_norm": 1.9114896059036255,
"learning_rate": 0.001,
"loss": 0.9015,
"step": 235700
},
{
"epoch": 76.2120232708468,
"grad_norm": 1.6289089918136597,
"learning_rate": 0.001,
"loss": 0.9024,
"step": 235800
},
{
"epoch": 76.24434389140272,
"grad_norm": 1.781545639038086,
"learning_rate": 0.001,
"loss": 0.9032,
"step": 235900
},
{
"epoch": 76.27666451195863,
"grad_norm": 1.6694800853729248,
"learning_rate": 0.001,
"loss": 0.9199,
"step": 236000
},
{
"epoch": 76.30898513251455,
"grad_norm": 1.621127724647522,
"learning_rate": 0.001,
"loss": 0.9102,
"step": 236100
},
{
"epoch": 76.34130575307046,
"grad_norm": 1.9022835493087769,
"learning_rate": 0.001,
"loss": 0.9136,
"step": 236200
},
{
"epoch": 76.37362637362638,
"grad_norm": 1.7457150220870972,
"learning_rate": 0.001,
"loss": 0.9216,
"step": 236300
},
{
"epoch": 76.40594699418229,
"grad_norm": 1.751098394393921,
"learning_rate": 0.001,
"loss": 0.9272,
"step": 236400
},
{
"epoch": 76.43826761473821,
"grad_norm": 2.129150390625,
"learning_rate": 0.001,
"loss": 0.92,
"step": 236500
},
{
"epoch": 76.47058823529412,
"grad_norm": 1.7513155937194824,
"learning_rate": 0.001,
"loss": 0.9246,
"step": 236600
},
{
"epoch": 76.50290885585004,
"grad_norm": 1.6849470138549805,
"learning_rate": 0.001,
"loss": 0.9175,
"step": 236700
},
{
"epoch": 76.53522947640595,
"grad_norm": 1.7795782089233398,
"learning_rate": 0.001,
"loss": 0.9305,
"step": 236800
},
{
"epoch": 76.56755009696187,
"grad_norm": 1.9670932292938232,
"learning_rate": 0.001,
"loss": 0.9366,
"step": 236900
},
{
"epoch": 76.59987071751777,
"grad_norm": 1.8679801225662231,
"learning_rate": 0.001,
"loss": 0.9519,
"step": 237000
},
{
"epoch": 76.6321913380737,
"grad_norm": 1.8227111101150513,
"learning_rate": 0.001,
"loss": 0.9325,
"step": 237100
},
{
"epoch": 76.6645119586296,
"grad_norm": 6.633680820465088,
"learning_rate": 0.001,
"loss": 0.9339,
"step": 237200
},
{
"epoch": 76.69683257918552,
"grad_norm": 1.7457256317138672,
"learning_rate": 0.001,
"loss": 0.9416,
"step": 237300
},
{
"epoch": 76.72915319974143,
"grad_norm": 2.4800775051116943,
"learning_rate": 0.001,
"loss": 0.9526,
"step": 237400
},
{
"epoch": 76.76147382029735,
"grad_norm": 1.730302333831787,
"learning_rate": 0.001,
"loss": 0.9285,
"step": 237500
},
{
"epoch": 76.79379444085326,
"grad_norm": 1.8279228210449219,
"learning_rate": 0.001,
"loss": 0.9723,
"step": 237600
},
{
"epoch": 76.82611506140918,
"grad_norm": 2.014887809753418,
"learning_rate": 0.001,
"loss": 0.9415,
"step": 237700
},
{
"epoch": 76.85843568196509,
"grad_norm": 2.5039334297180176,
"learning_rate": 0.001,
"loss": 0.9643,
"step": 237800
},
{
"epoch": 76.89075630252101,
"grad_norm": 2.5656185150146484,
"learning_rate": 0.001,
"loss": 0.9451,
"step": 237900
},
{
"epoch": 76.92307692307692,
"grad_norm": 1.8565692901611328,
"learning_rate": 0.001,
"loss": 0.941,
"step": 238000
},
{
"epoch": 76.95539754363284,
"grad_norm": 1.7059361934661865,
"learning_rate": 0.001,
"loss": 0.9776,
"step": 238100
},
{
"epoch": 76.98771816418875,
"grad_norm": 1.8570643663406372,
"learning_rate": 0.001,
"loss": 0.9439,
"step": 238200
},
{
"epoch": 77.02003878474467,
"grad_norm": 2.594362258911133,
"learning_rate": 0.001,
"loss": 0.9292,
"step": 238300
},
{
"epoch": 77.05235940530058,
"grad_norm": 1.7089303731918335,
"learning_rate": 0.001,
"loss": 0.8733,
"step": 238400
},
{
"epoch": 77.0846800258565,
"grad_norm": 2.044188976287842,
"learning_rate": 0.001,
"loss": 0.8921,
"step": 238500
},
{
"epoch": 77.11700064641241,
"grad_norm": 1.9743757247924805,
"learning_rate": 0.001,
"loss": 0.8858,
"step": 238600
},
{
"epoch": 77.14932126696833,
"grad_norm": 1.9437718391418457,
"learning_rate": 0.001,
"loss": 0.8908,
"step": 238700
},
{
"epoch": 77.18164188752424,
"grad_norm": 2.1930902004241943,
"learning_rate": 0.001,
"loss": 0.8885,
"step": 238800
},
{
"epoch": 77.21396250808016,
"grad_norm": 1.930568814277649,
"learning_rate": 0.001,
"loss": 0.8924,
"step": 238900
},
{
"epoch": 77.24628312863607,
"grad_norm": 1.827075719833374,
"learning_rate": 0.001,
"loss": 0.8958,
"step": 239000
},
{
"epoch": 77.27860374919199,
"grad_norm": 1.926182746887207,
"learning_rate": 0.001,
"loss": 0.9041,
"step": 239100
},
{
"epoch": 77.3109243697479,
"grad_norm": 22.807456970214844,
"learning_rate": 0.001,
"loss": 0.9079,
"step": 239200
},
{
"epoch": 77.34324499030382,
"grad_norm": 2.4273526668548584,
"learning_rate": 0.001,
"loss": 0.9035,
"step": 239300
},
{
"epoch": 77.37556561085972,
"grad_norm": 1.7550523281097412,
"learning_rate": 0.001,
"loss": 0.9272,
"step": 239400
},
{
"epoch": 77.40788623141565,
"grad_norm": 1.4827191829681396,
"learning_rate": 0.001,
"loss": 0.9148,
"step": 239500
},
{
"epoch": 77.44020685197155,
"grad_norm": 1.5911911725997925,
"learning_rate": 0.001,
"loss": 0.9168,
"step": 239600
},
{
"epoch": 77.47252747252747,
"grad_norm": 1.583900809288025,
"learning_rate": 0.001,
"loss": 0.9196,
"step": 239700
},
{
"epoch": 77.50484809308338,
"grad_norm": 2.1414811611175537,
"learning_rate": 0.001,
"loss": 0.9164,
"step": 239800
},
{
"epoch": 77.5371687136393,
"grad_norm": 1.7510583400726318,
"learning_rate": 0.001,
"loss": 0.9213,
"step": 239900
},
{
"epoch": 77.56948933419521,
"grad_norm": 1.7050837278366089,
"learning_rate": 0.001,
"loss": 0.9308,
"step": 240000
},
{
"epoch": 77.60180995475113,
"grad_norm": 2.887627363204956,
"learning_rate": 0.001,
"loss": 0.9263,
"step": 240100
},
{
"epoch": 77.63413057530704,
"grad_norm": 1.8627345561981201,
"learning_rate": 0.001,
"loss": 0.935,
"step": 240200
},
{
"epoch": 77.66645119586296,
"grad_norm": 1.4883463382720947,
"learning_rate": 0.001,
"loss": 0.9439,
"step": 240300
},
{
"epoch": 77.69877181641887,
"grad_norm": 2.6425867080688477,
"learning_rate": 0.001,
"loss": 0.9367,
"step": 240400
},
{
"epoch": 77.73109243697479,
"grad_norm": 1.971907615661621,
"learning_rate": 0.001,
"loss": 0.9371,
"step": 240500
},
{
"epoch": 77.7634130575307,
"grad_norm": 1.6637628078460693,
"learning_rate": 0.001,
"loss": 0.9277,
"step": 240600
},
{
"epoch": 77.79573367808662,
"grad_norm": 4.366883754730225,
"learning_rate": 0.001,
"loss": 0.9409,
"step": 240700
},
{
"epoch": 77.82805429864253,
"grad_norm": 1.700300931930542,
"learning_rate": 0.001,
"loss": 0.9582,
"step": 240800
},
{
"epoch": 77.86037491919845,
"grad_norm": 1.6653016805648804,
"learning_rate": 0.001,
"loss": 0.9337,
"step": 240900
},
{
"epoch": 77.89269553975436,
"grad_norm": 1.5066953897476196,
"learning_rate": 0.001,
"loss": 0.9384,
"step": 241000
},
{
"epoch": 77.92501616031028,
"grad_norm": 1.970182180404663,
"learning_rate": 0.001,
"loss": 0.9514,
"step": 241100
},
{
"epoch": 77.95733678086619,
"grad_norm": 1.863133192062378,
"learning_rate": 0.001,
"loss": 0.9502,
"step": 241200
},
{
"epoch": 77.98965740142211,
"grad_norm": 2.4191184043884277,
"learning_rate": 0.001,
"loss": 0.9507,
"step": 241300
},
{
"epoch": 78.02197802197803,
"grad_norm": 1.864750862121582,
"learning_rate": 0.001,
"loss": 0.9046,
"step": 241400
},
{
"epoch": 78.05429864253394,
"grad_norm": 1.5248106718063354,
"learning_rate": 0.001,
"loss": 0.8657,
"step": 241500
},
{
"epoch": 78.08661926308986,
"grad_norm": 1.8853729963302612,
"learning_rate": 0.001,
"loss": 0.8685,
"step": 241600
},
{
"epoch": 78.11893988364577,
"grad_norm": 1.8293319940567017,
"learning_rate": 0.001,
"loss": 0.871,
"step": 241700
},
{
"epoch": 78.15126050420169,
"grad_norm": 2.1890885829925537,
"learning_rate": 0.001,
"loss": 0.8768,
"step": 241800
},
{
"epoch": 78.1835811247576,
"grad_norm": 1.899473786354065,
"learning_rate": 0.001,
"loss": 0.8958,
"step": 241900
},
{
"epoch": 78.21590174531352,
"grad_norm": 1.888890027999878,
"learning_rate": 0.001,
"loss": 0.8698,
"step": 242000
},
{
"epoch": 78.24822236586942,
"grad_norm": 1.9390349388122559,
"learning_rate": 0.001,
"loss": 0.8938,
"step": 242100
},
{
"epoch": 78.28054298642535,
"grad_norm": 1.613270878791809,
"learning_rate": 0.001,
"loss": 0.8969,
"step": 242200
},
{
"epoch": 78.31286360698125,
"grad_norm": 1.510816216468811,
"learning_rate": 0.001,
"loss": 0.8928,
"step": 242300
},
{
"epoch": 78.34518422753717,
"grad_norm": 1.6668297052383423,
"learning_rate": 0.001,
"loss": 0.8946,
"step": 242400
},
{
"epoch": 78.37750484809308,
"grad_norm": 2.631298065185547,
"learning_rate": 0.001,
"loss": 0.9224,
"step": 242500
},
{
"epoch": 78.409825468649,
"grad_norm": 1.9295475482940674,
"learning_rate": 0.001,
"loss": 0.8946,
"step": 242600
},
{
"epoch": 78.44214608920491,
"grad_norm": 1.5661232471466064,
"learning_rate": 0.001,
"loss": 0.9005,
"step": 242700
},
{
"epoch": 78.47446670976083,
"grad_norm": 1.602188229560852,
"learning_rate": 0.001,
"loss": 0.9058,
"step": 242800
},
{
"epoch": 78.50678733031674,
"grad_norm": 2.1843175888061523,
"learning_rate": 0.001,
"loss": 0.9144,
"step": 242900
},
{
"epoch": 78.53910795087266,
"grad_norm": 1.603387475013733,
"learning_rate": 0.001,
"loss": 0.9284,
"step": 243000
},
{
"epoch": 78.57142857142857,
"grad_norm": 2.4746646881103516,
"learning_rate": 0.001,
"loss": 0.9248,
"step": 243100
},
{
"epoch": 78.60374919198449,
"grad_norm": 1.4542807340621948,
"learning_rate": 0.001,
"loss": 0.908,
"step": 243200
},
{
"epoch": 78.6360698125404,
"grad_norm": 2.2467586994171143,
"learning_rate": 0.001,
"loss": 0.9209,
"step": 243300
},
{
"epoch": 78.66839043309632,
"grad_norm": 1.670127034187317,
"learning_rate": 0.001,
"loss": 0.9106,
"step": 243400
},
{
"epoch": 78.70071105365223,
"grad_norm": 1.4943729639053345,
"learning_rate": 0.001,
"loss": 0.9295,
"step": 243500
},
{
"epoch": 78.73303167420815,
"grad_norm": 1.8003182411193848,
"learning_rate": 0.001,
"loss": 0.9323,
"step": 243600
},
{
"epoch": 78.76535229476406,
"grad_norm": 1.6677433252334595,
"learning_rate": 0.001,
"loss": 0.9376,
"step": 243700
},
{
"epoch": 78.79767291531998,
"grad_norm": 1.4286067485809326,
"learning_rate": 0.001,
"loss": 0.9452,
"step": 243800
},
{
"epoch": 78.82999353587589,
"grad_norm": 1.4021170139312744,
"learning_rate": 0.001,
"loss": 0.9488,
"step": 243900
},
{
"epoch": 78.86231415643181,
"grad_norm": 1.5756992101669312,
"learning_rate": 0.001,
"loss": 0.9512,
"step": 244000
},
{
"epoch": 78.89463477698771,
"grad_norm": 1.6208910942077637,
"learning_rate": 0.001,
"loss": 0.9351,
"step": 244100
},
{
"epoch": 78.92695539754364,
"grad_norm": 1.8639366626739502,
"learning_rate": 0.001,
"loss": 0.9422,
"step": 244200
},
{
"epoch": 78.95927601809954,
"grad_norm": 1.9582170248031616,
"learning_rate": 0.001,
"loss": 0.962,
"step": 244300
},
{
"epoch": 78.99159663865547,
"grad_norm": 1.5845123529434204,
"learning_rate": 0.001,
"loss": 0.9453,
"step": 244400
},
{
"epoch": 79.02391725921137,
"grad_norm": 1.7335338592529297,
"learning_rate": 0.001,
"loss": 0.8861,
"step": 244500
},
{
"epoch": 79.0562378797673,
"grad_norm": 1.3891186714172363,
"learning_rate": 0.001,
"loss": 0.8479,
"step": 244600
},
{
"epoch": 79.0885585003232,
"grad_norm": 2.224407434463501,
"learning_rate": 0.001,
"loss": 0.8521,
"step": 244700
},
{
"epoch": 79.12087912087912,
"grad_norm": 1.7175556421279907,
"learning_rate": 0.001,
"loss": 0.8663,
"step": 244800
},
{
"epoch": 79.15319974143503,
"grad_norm": 1.3971540927886963,
"learning_rate": 0.001,
"loss": 0.8854,
"step": 244900
},
{
"epoch": 79.18552036199095,
"grad_norm": 1.6051901578903198,
"learning_rate": 0.001,
"loss": 0.8806,
"step": 245000
},
{
"epoch": 79.21784098254686,
"grad_norm": 1.4144991636276245,
"learning_rate": 0.001,
"loss": 0.8854,
"step": 245100
},
{
"epoch": 79.25016160310278,
"grad_norm": 1.6836448907852173,
"learning_rate": 0.001,
"loss": 0.8852,
"step": 245200
},
{
"epoch": 79.28248222365869,
"grad_norm": 1.8142777681350708,
"learning_rate": 0.001,
"loss": 0.889,
"step": 245300
},
{
"epoch": 79.31480284421461,
"grad_norm": 1.4503638744354248,
"learning_rate": 0.001,
"loss": 0.8878,
"step": 245400
},
{
"epoch": 79.34712346477052,
"grad_norm": 1.810864806175232,
"learning_rate": 0.001,
"loss": 0.8944,
"step": 245500
},
{
"epoch": 79.37944408532644,
"grad_norm": 1.5955215692520142,
"learning_rate": 0.001,
"loss": 0.8924,
"step": 245600
},
{
"epoch": 79.41176470588235,
"grad_norm": 1.9111827611923218,
"learning_rate": 0.001,
"loss": 0.9071,
"step": 245700
},
{
"epoch": 79.44408532643827,
"grad_norm": 1.7505015134811401,
"learning_rate": 0.001,
"loss": 0.9116,
"step": 245800
},
{
"epoch": 79.47640594699418,
"grad_norm": 1.6506662368774414,
"learning_rate": 0.001,
"loss": 0.8994,
"step": 245900
},
{
"epoch": 79.5087265675501,
"grad_norm": 1.7561360597610474,
"learning_rate": 0.001,
"loss": 0.897,
"step": 246000
},
{
"epoch": 79.541047188106,
"grad_norm": 1.4854172468185425,
"learning_rate": 0.001,
"loss": 0.9047,
"step": 246100
},
{
"epoch": 79.57336780866193,
"grad_norm": 2.3282666206359863,
"learning_rate": 0.001,
"loss": 0.915,
"step": 246200
},
{
"epoch": 79.60568842921784,
"grad_norm": 1.461841344833374,
"learning_rate": 0.001,
"loss": 0.9143,
"step": 246300
},
{
"epoch": 79.63800904977376,
"grad_norm": 1.3623031377792358,
"learning_rate": 0.001,
"loss": 0.9193,
"step": 246400
},
{
"epoch": 79.67032967032966,
"grad_norm": 1.9313454627990723,
"learning_rate": 0.001,
"loss": 0.9184,
"step": 246500
},
{
"epoch": 79.70265029088559,
"grad_norm": 1.8738027811050415,
"learning_rate": 0.001,
"loss": 0.9146,
"step": 246600
},
{
"epoch": 79.7349709114415,
"grad_norm": 1.95305335521698,
"learning_rate": 0.001,
"loss": 0.9158,
"step": 246700
},
{
"epoch": 79.76729153199742,
"grad_norm": 1.6693203449249268,
"learning_rate": 0.001,
"loss": 0.9184,
"step": 246800
},
{
"epoch": 79.79961215255332,
"grad_norm": 1.764410376548767,
"learning_rate": 0.001,
"loss": 0.9463,
"step": 246900
},
{
"epoch": 79.83193277310924,
"grad_norm": 1.8955789804458618,
"learning_rate": 0.001,
"loss": 0.9434,
"step": 247000
},
{
"epoch": 79.86425339366515,
"grad_norm": 1.6161580085754395,
"learning_rate": 0.001,
"loss": 0.9376,
"step": 247100
},
{
"epoch": 79.89657401422107,
"grad_norm": 1.5458617210388184,
"learning_rate": 0.001,
"loss": 0.9327,
"step": 247200
},
{
"epoch": 79.92889463477698,
"grad_norm": 1.5969195365905762,
"learning_rate": 0.001,
"loss": 0.9487,
"step": 247300
},
{
"epoch": 79.9612152553329,
"grad_norm": 1.6383031606674194,
"learning_rate": 0.001,
"loss": 0.9546,
"step": 247400
},
{
"epoch": 79.99353587588882,
"grad_norm": 2.259331464767456,
"learning_rate": 0.001,
"loss": 0.9425,
"step": 247500
},
{
"epoch": 80.02585649644473,
"grad_norm": 2.12754225730896,
"learning_rate": 0.001,
"loss": 0.8788,
"step": 247600
},
{
"epoch": 80.05817711700065,
"grad_norm": 1.7799402475357056,
"learning_rate": 0.001,
"loss": 0.8519,
"step": 247700
},
{
"epoch": 80.09049773755656,
"grad_norm": 2.46228289604187,
"learning_rate": 0.001,
"loss": 0.8533,
"step": 247800
},
{
"epoch": 80.12281835811248,
"grad_norm": 1.5350501537322998,
"learning_rate": 0.001,
"loss": 0.861,
"step": 247900
},
{
"epoch": 80.15513897866839,
"grad_norm": 2.1066055297851562,
"learning_rate": 0.001,
"loss": 0.8711,
"step": 248000
},
{
"epoch": 80.18745959922431,
"grad_norm": 1.4657257795333862,
"learning_rate": 0.001,
"loss": 0.8665,
"step": 248100
},
{
"epoch": 80.21978021978022,
"grad_norm": 1.7249951362609863,
"learning_rate": 0.001,
"loss": 0.8784,
"step": 248200
},
{
"epoch": 80.25210084033614,
"grad_norm": 2.1803066730499268,
"learning_rate": 0.001,
"loss": 0.8675,
"step": 248300
},
{
"epoch": 80.28442146089205,
"grad_norm": 1.5328905582427979,
"learning_rate": 0.001,
"loss": 0.8967,
"step": 248400
},
{
"epoch": 80.31674208144797,
"grad_norm": 1.712321400642395,
"learning_rate": 0.001,
"loss": 0.8854,
"step": 248500
},
{
"epoch": 80.34906270200388,
"grad_norm": 1.821807861328125,
"learning_rate": 0.001,
"loss": 0.8878,
"step": 248600
},
{
"epoch": 80.3813833225598,
"grad_norm": 1.7244770526885986,
"learning_rate": 0.001,
"loss": 0.8968,
"step": 248700
},
{
"epoch": 80.4137039431157,
"grad_norm": 1.9649723768234253,
"learning_rate": 0.001,
"loss": 0.8979,
"step": 248800
},
{
"epoch": 80.44602456367163,
"grad_norm": 1.7593700885772705,
"learning_rate": 0.001,
"loss": 0.8921,
"step": 248900
},
{
"epoch": 80.47834518422754,
"grad_norm": 1.606793999671936,
"learning_rate": 0.001,
"loss": 0.9013,
"step": 249000
},
{
"epoch": 80.51066580478346,
"grad_norm": 1.2996211051940918,
"learning_rate": 0.001,
"loss": 0.8991,
"step": 249100
},
{
"epoch": 80.54298642533936,
"grad_norm": 1.8968507051467896,
"learning_rate": 0.001,
"loss": 0.8959,
"step": 249200
},
{
"epoch": 80.57530704589529,
"grad_norm": 1.6174746751785278,
"learning_rate": 0.001,
"loss": 0.9086,
"step": 249300
},
{
"epoch": 80.6076276664512,
"grad_norm": 1.8977900743484497,
"learning_rate": 0.001,
"loss": 0.916,
"step": 249400
},
{
"epoch": 80.63994828700712,
"grad_norm": 2.1842105388641357,
"learning_rate": 0.001,
"loss": 0.9061,
"step": 249500
},
{
"epoch": 80.67226890756302,
"grad_norm": 1.555891513824463,
"learning_rate": 0.001,
"loss": 0.9002,
"step": 249600
},
{
"epoch": 80.70458952811894,
"grad_norm": 1.8011150360107422,
"learning_rate": 0.001,
"loss": 0.9051,
"step": 249700
},
{
"epoch": 80.73691014867485,
"grad_norm": 2.098155975341797,
"learning_rate": 0.001,
"loss": 0.9128,
"step": 249800
},
{
"epoch": 80.76923076923077,
"grad_norm": 1.7902772426605225,
"learning_rate": 0.001,
"loss": 0.9092,
"step": 249900
},
{
"epoch": 80.80155138978668,
"grad_norm": 1.8167074918746948,
"learning_rate": 0.001,
"loss": 0.9232,
"step": 250000
},
{
"epoch": 80.8338720103426,
"grad_norm": 1.8946473598480225,
"learning_rate": 0.001,
"loss": 0.9256,
"step": 250100
},
{
"epoch": 80.86619263089851,
"grad_norm": 2.035849094390869,
"learning_rate": 0.001,
"loss": 0.9248,
"step": 250200
},
{
"epoch": 80.89851325145443,
"grad_norm": 2.2636032104492188,
"learning_rate": 0.001,
"loss": 0.9376,
"step": 250300
},
{
"epoch": 80.93083387201034,
"grad_norm": 2.28778338432312,
"learning_rate": 0.001,
"loss": 0.946,
"step": 250400
},
{
"epoch": 80.96315449256626,
"grad_norm": 2.307657480239868,
"learning_rate": 0.001,
"loss": 0.9468,
"step": 250500
},
{
"epoch": 80.99547511312217,
"grad_norm": 1.6172599792480469,
"learning_rate": 0.001,
"loss": 0.9319,
"step": 250600
},
{
"epoch": 81.02779573367809,
"grad_norm": 1.3265091180801392,
"learning_rate": 0.001,
"loss": 0.8342,
"step": 250700
},
{
"epoch": 81.060116354234,
"grad_norm": 1.4472734928131104,
"learning_rate": 0.001,
"loss": 0.8624,
"step": 250800
},
{
"epoch": 81.09243697478992,
"grad_norm": 1.6675355434417725,
"learning_rate": 0.001,
"loss": 0.8429,
"step": 250900
},
{
"epoch": 81.12475759534583,
"grad_norm": 1.8562703132629395,
"learning_rate": 0.001,
"loss": 0.8655,
"step": 251000
},
{
"epoch": 81.15707821590175,
"grad_norm": 2.034242630004883,
"learning_rate": 0.001,
"loss": 0.865,
"step": 251100
},
{
"epoch": 81.18939883645766,
"grad_norm": 1.5765374898910522,
"learning_rate": 0.001,
"loss": 0.8735,
"step": 251200
},
{
"epoch": 81.22171945701358,
"grad_norm": 1.7124650478363037,
"learning_rate": 0.001,
"loss": 0.8562,
"step": 251300
},
{
"epoch": 81.25404007756948,
"grad_norm": 1.5629198551177979,
"learning_rate": 0.001,
"loss": 0.8594,
"step": 251400
},
{
"epoch": 81.2863606981254,
"grad_norm": 1.4347710609436035,
"learning_rate": 0.001,
"loss": 0.8888,
"step": 251500
},
{
"epoch": 81.31868131868131,
"grad_norm": 1.937766671180725,
"learning_rate": 0.001,
"loss": 0.8832,
"step": 251600
},
{
"epoch": 81.35100193923724,
"grad_norm": 1.9682731628417969,
"learning_rate": 0.001,
"loss": 0.8715,
"step": 251700
},
{
"epoch": 81.38332255979314,
"grad_norm": 1.760391116142273,
"learning_rate": 0.001,
"loss": 0.8781,
"step": 251800
},
{
"epoch": 81.41564318034906,
"grad_norm": 1.8551701307296753,
"learning_rate": 0.001,
"loss": 0.8749,
"step": 251900
},
{
"epoch": 81.44796380090497,
"grad_norm": 1.7820264101028442,
"learning_rate": 0.001,
"loss": 0.8981,
"step": 252000
},
{
"epoch": 81.4802844214609,
"grad_norm": 1.556565284729004,
"learning_rate": 0.001,
"loss": 0.9005,
"step": 252100
},
{
"epoch": 81.5126050420168,
"grad_norm": 4.37351131439209,
"learning_rate": 0.001,
"loss": 0.8902,
"step": 252200
},
{
"epoch": 81.54492566257272,
"grad_norm": 1.8729733228683472,
"learning_rate": 0.001,
"loss": 0.8959,
"step": 252300
},
{
"epoch": 81.57724628312863,
"grad_norm": 1.4958561658859253,
"learning_rate": 0.001,
"loss": 0.9072,
"step": 252400
},
{
"epoch": 81.60956690368455,
"grad_norm": 1.8346575498580933,
"learning_rate": 0.001,
"loss": 0.9189,
"step": 252500
},
{
"epoch": 81.64188752424046,
"grad_norm": 1.9965070486068726,
"learning_rate": 0.001,
"loss": 0.8933,
"step": 252600
},
{
"epoch": 81.67420814479638,
"grad_norm": 1.7740451097488403,
"learning_rate": 0.001,
"loss": 0.9031,
"step": 252700
},
{
"epoch": 81.70652876535229,
"grad_norm": 1.8927515745162964,
"learning_rate": 0.001,
"loss": 0.9042,
"step": 252800
},
{
"epoch": 81.73884938590821,
"grad_norm": 1.7266933917999268,
"learning_rate": 0.001,
"loss": 0.9183,
"step": 252900
},
{
"epoch": 81.77117000646412,
"grad_norm": 1.5089515447616577,
"learning_rate": 0.001,
"loss": 0.9018,
"step": 253000
},
{
"epoch": 81.80349062702004,
"grad_norm": 17.170991897583008,
"learning_rate": 0.001,
"loss": 0.9344,
"step": 253100
},
{
"epoch": 81.83581124757595,
"grad_norm": 1.541385531425476,
"learning_rate": 0.001,
"loss": 0.9277,
"step": 253200
},
{
"epoch": 81.86813186813187,
"grad_norm": 1.8526593446731567,
"learning_rate": 0.001,
"loss": 0.9274,
"step": 253300
},
{
"epoch": 81.90045248868778,
"grad_norm": 1.4917140007019043,
"learning_rate": 0.001,
"loss": 0.914,
"step": 253400
},
{
"epoch": 81.9327731092437,
"grad_norm": 1.8501405715942383,
"learning_rate": 0.001,
"loss": 0.9397,
"step": 253500
},
{
"epoch": 81.9650937297996,
"grad_norm": 1.5764436721801758,
"learning_rate": 0.001,
"loss": 0.9387,
"step": 253600
},
{
"epoch": 81.99741435035553,
"grad_norm": 1.4233180284500122,
"learning_rate": 0.001,
"loss": 0.9267,
"step": 253700
},
{
"epoch": 82.02973497091145,
"grad_norm": 1.810560941696167,
"learning_rate": 0.001,
"loss": 0.8427,
"step": 253800
},
{
"epoch": 82.06205559146736,
"grad_norm": 1.6479710340499878,
"learning_rate": 0.001,
"loss": 0.8471,
"step": 253900
},
{
"epoch": 82.09437621202328,
"grad_norm": 1.6219289302825928,
"learning_rate": 0.001,
"loss": 0.8567,
"step": 254000
},
{
"epoch": 82.12669683257919,
"grad_norm": 1.7933605909347534,
"learning_rate": 0.001,
"loss": 0.859,
"step": 254100
},
{
"epoch": 82.1590174531351,
"grad_norm": 1.518423080444336,
"learning_rate": 0.001,
"loss": 0.8533,
"step": 254200
},
{
"epoch": 82.19133807369101,
"grad_norm": 1.4636738300323486,
"learning_rate": 0.001,
"loss": 0.8601,
"step": 254300
},
{
"epoch": 82.22365869424694,
"grad_norm": 1.5128989219665527,
"learning_rate": 0.001,
"loss": 0.8589,
"step": 254400
},
{
"epoch": 82.25597931480284,
"grad_norm": 1.463891863822937,
"learning_rate": 0.001,
"loss": 0.8611,
"step": 254500
},
{
"epoch": 82.28829993535876,
"grad_norm": 1.5017369985580444,
"learning_rate": 0.001,
"loss": 0.8683,
"step": 254600
},
{
"epoch": 82.32062055591467,
"grad_norm": 1.7159160375595093,
"learning_rate": 0.001,
"loss": 0.8748,
"step": 254700
},
{
"epoch": 82.3529411764706,
"grad_norm": 1.799309253692627,
"learning_rate": 0.001,
"loss": 0.8784,
"step": 254800
},
{
"epoch": 82.3852617970265,
"grad_norm": 1.742081880569458,
"learning_rate": 0.001,
"loss": 0.8949,
"step": 254900
},
{
"epoch": 82.41758241758242,
"grad_norm": 1.7784702777862549,
"learning_rate": 0.001,
"loss": 0.8735,
"step": 255000
},
{
"epoch": 82.44990303813833,
"grad_norm": 1.4071807861328125,
"learning_rate": 0.001,
"loss": 0.8704,
"step": 255100
},
{
"epoch": 82.48222365869425,
"grad_norm": 1.567540168762207,
"learning_rate": 0.001,
"loss": 0.877,
"step": 255200
},
{
"epoch": 82.51454427925016,
"grad_norm": 1.4727294445037842,
"learning_rate": 0.001,
"loss": 0.9026,
"step": 255300
},
{
"epoch": 82.54686489980608,
"grad_norm": 2.0840132236480713,
"learning_rate": 0.001,
"loss": 0.8898,
"step": 255400
},
{
"epoch": 82.57918552036199,
"grad_norm": 1.4543981552124023,
"learning_rate": 0.001,
"loss": 0.8851,
"step": 255500
},
{
"epoch": 82.61150614091791,
"grad_norm": 2.6201300621032715,
"learning_rate": 0.001,
"loss": 0.9079,
"step": 255600
},
{
"epoch": 82.64382676147382,
"grad_norm": 1.6429693698883057,
"learning_rate": 0.001,
"loss": 0.8945,
"step": 255700
},
{
"epoch": 82.67614738202974,
"grad_norm": 1.4842705726623535,
"learning_rate": 0.001,
"loss": 0.8983,
"step": 255800
},
{
"epoch": 82.70846800258565,
"grad_norm": 2.0955522060394287,
"learning_rate": 0.001,
"loss": 0.9124,
"step": 255900
},
{
"epoch": 82.74078862314157,
"grad_norm": 1.486534833908081,
"learning_rate": 0.001,
"loss": 0.9026,
"step": 256000
},
{
"epoch": 82.77310924369748,
"grad_norm": 1.7523531913757324,
"learning_rate": 0.001,
"loss": 0.9076,
"step": 256100
},
{
"epoch": 82.8054298642534,
"grad_norm": 1.7044719457626343,
"learning_rate": 0.001,
"loss": 0.9097,
"step": 256200
},
{
"epoch": 82.8377504848093,
"grad_norm": 1.5170854330062866,
"learning_rate": 0.001,
"loss": 0.9027,
"step": 256300
},
{
"epoch": 82.87007110536523,
"grad_norm": 1.520762324333191,
"learning_rate": 0.001,
"loss": 0.9069,
"step": 256400
},
{
"epoch": 82.90239172592113,
"grad_norm": 1.821850299835205,
"learning_rate": 0.001,
"loss": 0.925,
"step": 256500
},
{
"epoch": 82.93471234647706,
"grad_norm": 1.8839287757873535,
"learning_rate": 0.001,
"loss": 0.9241,
"step": 256600
},
{
"epoch": 82.96703296703296,
"grad_norm": 2.05893611907959,
"learning_rate": 0.001,
"loss": 0.9253,
"step": 256700
},
{
"epoch": 82.99935358758889,
"grad_norm": 1.280112862586975,
"learning_rate": 0.001,
"loss": 0.8939,
"step": 256800
},
{
"epoch": 83.03167420814479,
"grad_norm": 1.2705212831497192,
"learning_rate": 0.001,
"loss": 0.8337,
"step": 256900
},
{
"epoch": 83.06399482870071,
"grad_norm": 1.231841802597046,
"learning_rate": 0.001,
"loss": 0.8347,
"step": 257000
},
{
"epoch": 83.09631544925662,
"grad_norm": 1.311620831489563,
"learning_rate": 0.001,
"loss": 0.8409,
"step": 257100
},
{
"epoch": 83.12863606981254,
"grad_norm": 1.7882788181304932,
"learning_rate": 0.001,
"loss": 0.8443,
"step": 257200
},
{
"epoch": 83.16095669036845,
"grad_norm": 3.822011709213257,
"learning_rate": 0.001,
"loss": 0.8478,
"step": 257300
},
{
"epoch": 83.19327731092437,
"grad_norm": 1.924403190612793,
"learning_rate": 0.001,
"loss": 0.8459,
"step": 257400
},
{
"epoch": 83.22559793148028,
"grad_norm": 1.4485812187194824,
"learning_rate": 0.001,
"loss": 0.854,
"step": 257500
},
{
"epoch": 83.2579185520362,
"grad_norm": 1.1443918943405151,
"learning_rate": 0.001,
"loss": 0.8654,
"step": 257600
},
{
"epoch": 83.29023917259211,
"grad_norm": 1.3180463314056396,
"learning_rate": 0.001,
"loss": 0.8738,
"step": 257700
},
{
"epoch": 83.32255979314803,
"grad_norm": 1.2314362525939941,
"learning_rate": 0.001,
"loss": 0.8578,
"step": 257800
},
{
"epoch": 83.35488041370394,
"grad_norm": 1.101746678352356,
"learning_rate": 0.001,
"loss": 0.8723,
"step": 257900
},
{
"epoch": 83.38720103425986,
"grad_norm": 0.9859794974327087,
"learning_rate": 0.001,
"loss": 0.8622,
"step": 258000
},
{
"epoch": 83.41952165481577,
"grad_norm": 3.018252372741699,
"learning_rate": 0.001,
"loss": 0.8729,
"step": 258100
},
{
"epoch": 83.45184227537169,
"grad_norm": 1.3902442455291748,
"learning_rate": 0.001,
"loss": 0.8788,
"step": 258200
},
{
"epoch": 83.4841628959276,
"grad_norm": 1.4548698663711548,
"learning_rate": 0.001,
"loss": 0.8834,
"step": 258300
},
{
"epoch": 83.51648351648352,
"grad_norm": 1.4873329401016235,
"learning_rate": 0.001,
"loss": 0.8907,
"step": 258400
},
{
"epoch": 83.54880413703943,
"grad_norm": 1.177139163017273,
"learning_rate": 0.001,
"loss": 0.8801,
"step": 258500
},
{
"epoch": 83.58112475759535,
"grad_norm": 1.9104993343353271,
"learning_rate": 0.001,
"loss": 0.8869,
"step": 258600
},
{
"epoch": 83.61344537815125,
"grad_norm": 1.7681374549865723,
"learning_rate": 0.001,
"loss": 0.8976,
"step": 258700
},
{
"epoch": 83.64576599870718,
"grad_norm": 0.9657064080238342,
"learning_rate": 0.001,
"loss": 0.8977,
"step": 258800
},
{
"epoch": 83.67808661926308,
"grad_norm": 1.5095404386520386,
"learning_rate": 0.001,
"loss": 0.8891,
"step": 258900
},
{
"epoch": 83.710407239819,
"grad_norm": 0.9872686862945557,
"learning_rate": 0.001,
"loss": 0.8972,
"step": 259000
},
{
"epoch": 83.74272786037491,
"grad_norm": 1.5525736808776855,
"learning_rate": 0.001,
"loss": 0.9007,
"step": 259100
},
{
"epoch": 83.77504848093083,
"grad_norm": 1.1791164875030518,
"learning_rate": 0.001,
"loss": 0.904,
"step": 259200
},
{
"epoch": 83.80736910148674,
"grad_norm": 1.8377259969711304,
"learning_rate": 0.001,
"loss": 0.9127,
"step": 259300
},
{
"epoch": 83.83968972204266,
"grad_norm": 1.5384559631347656,
"learning_rate": 0.001,
"loss": 0.9098,
"step": 259400
},
{
"epoch": 83.87201034259857,
"grad_norm": 1.4044259786605835,
"learning_rate": 0.001,
"loss": 0.9075,
"step": 259500
},
{
"epoch": 83.9043309631545,
"grad_norm": 1.0175000429153442,
"learning_rate": 0.001,
"loss": 0.9107,
"step": 259600
},
{
"epoch": 83.9366515837104,
"grad_norm": 1.0417195558547974,
"learning_rate": 0.001,
"loss": 0.9235,
"step": 259700
},
{
"epoch": 83.96897220426632,
"grad_norm": 1.4625414609909058,
"learning_rate": 0.001,
"loss": 0.9279,
"step": 259800
},
{
"epoch": 84.00129282482224,
"grad_norm": 158.49737548828125,
"learning_rate": 0.001,
"loss": 0.9166,
"step": 259900
},
{
"epoch": 84.03361344537815,
"grad_norm": 3.5888888835906982,
"learning_rate": 0.001,
"loss": 0.825,
"step": 260000
},
{
"epoch": 84.06593406593407,
"grad_norm": 2.126845121383667,
"learning_rate": 0.001,
"loss": 0.8455,
"step": 260100
},
{
"epoch": 84.09825468648998,
"grad_norm": 2.2806663513183594,
"learning_rate": 0.001,
"loss": 0.8185,
"step": 260200
},
{
"epoch": 84.1305753070459,
"grad_norm": 1.8550175428390503,
"learning_rate": 0.001,
"loss": 0.8512,
"step": 260300
},
{
"epoch": 84.16289592760181,
"grad_norm": 2.1901886463165283,
"learning_rate": 0.001,
"loss": 0.8463,
"step": 260400
},
{
"epoch": 84.19521654815773,
"grad_norm": 2.431595802307129,
"learning_rate": 0.001,
"loss": 0.8452,
"step": 260500
},
{
"epoch": 84.22753716871364,
"grad_norm": 2.0532970428466797,
"learning_rate": 0.001,
"loss": 0.8521,
"step": 260600
},
{
"epoch": 84.25985778926956,
"grad_norm": 2.080876350402832,
"learning_rate": 0.001,
"loss": 0.8423,
"step": 260700
},
{
"epoch": 84.29217840982547,
"grad_norm": 2.1651861667633057,
"learning_rate": 0.001,
"loss": 0.8475,
"step": 260800
},
{
"epoch": 84.32449903038139,
"grad_norm": 2.6767241954803467,
"learning_rate": 0.001,
"loss": 0.8522,
"step": 260900
},
{
"epoch": 84.3568196509373,
"grad_norm": 2.2361319065093994,
"learning_rate": 0.001,
"loss": 0.8845,
"step": 261000
},
{
"epoch": 84.38914027149322,
"grad_norm": 2.157210111618042,
"learning_rate": 0.001,
"loss": 0.8596,
"step": 261100
},
{
"epoch": 84.42146089204913,
"grad_norm": 1.8781487941741943,
"learning_rate": 0.001,
"loss": 0.8816,
"step": 261200
},
{
"epoch": 84.45378151260505,
"grad_norm": 1.8453071117401123,
"learning_rate": 0.001,
"loss": 0.8741,
"step": 261300
},
{
"epoch": 84.48610213316095,
"grad_norm": 1.809593915939331,
"learning_rate": 0.001,
"loss": 0.8694,
"step": 261400
},
{
"epoch": 84.51842275371688,
"grad_norm": 2.3979721069335938,
"learning_rate": 0.001,
"loss": 0.8779,
"step": 261500
},
{
"epoch": 84.55074337427278,
"grad_norm": 2.704803228378296,
"learning_rate": 0.001,
"loss": 0.8825,
"step": 261600
},
{
"epoch": 84.5830639948287,
"grad_norm": 1.9731519222259521,
"learning_rate": 0.001,
"loss": 0.8836,
"step": 261700
},
{
"epoch": 84.61538461538461,
"grad_norm": 1.9622254371643066,
"learning_rate": 0.001,
"loss": 0.8799,
"step": 261800
},
{
"epoch": 84.64770523594053,
"grad_norm": 1.9820390939712524,
"learning_rate": 0.001,
"loss": 0.8798,
"step": 261900
},
{
"epoch": 84.68002585649644,
"grad_norm": 3.088493585586548,
"learning_rate": 0.001,
"loss": 0.8821,
"step": 262000
},
{
"epoch": 84.71234647705236,
"grad_norm": 2.2478713989257812,
"learning_rate": 0.001,
"loss": 0.9065,
"step": 262100
},
{
"epoch": 84.74466709760827,
"grad_norm": 1.8590142726898193,
"learning_rate": 0.001,
"loss": 0.8902,
"step": 262200
},
{
"epoch": 84.7769877181642,
"grad_norm": 1.954840064048767,
"learning_rate": 0.001,
"loss": 0.9051,
"step": 262300
},
{
"epoch": 84.8093083387201,
"grad_norm": 1.901276707649231,
"learning_rate": 0.001,
"loss": 0.9101,
"step": 262400
},
{
"epoch": 84.84162895927602,
"grad_norm": 2.1338517665863037,
"learning_rate": 0.001,
"loss": 0.8986,
"step": 262500
},
{
"epoch": 84.87394957983193,
"grad_norm": 9.55400276184082,
"learning_rate": 0.001,
"loss": 0.9106,
"step": 262600
},
{
"epoch": 84.90627020038785,
"grad_norm": 2.2221624851226807,
"learning_rate": 0.001,
"loss": 0.9136,
"step": 262700
},
{
"epoch": 84.93859082094376,
"grad_norm": 1.898995041847229,
"learning_rate": 0.001,
"loss": 0.916,
"step": 262800
},
{
"epoch": 84.97091144149968,
"grad_norm": 1.8953394889831543,
"learning_rate": 0.001,
"loss": 0.9177,
"step": 262900
},
{
"epoch": 85.00323206205559,
"grad_norm": 1.9160492420196533,
"learning_rate": 0.001,
"loss": 0.9063,
"step": 263000
},
{
"epoch": 85.03555268261151,
"grad_norm": 2.0379796028137207,
"learning_rate": 0.001,
"loss": 0.8174,
"step": 263100
},
{
"epoch": 85.06787330316742,
"grad_norm": 1.6912503242492676,
"learning_rate": 0.001,
"loss": 0.8196,
"step": 263200
},
{
"epoch": 85.10019392372334,
"grad_norm": 1.959588885307312,
"learning_rate": 0.001,
"loss": 0.8337,
"step": 263300
},
{
"epoch": 85.13251454427925,
"grad_norm": 2.460237503051758,
"learning_rate": 0.001,
"loss": 0.8392,
"step": 263400
},
{
"epoch": 85.16483516483517,
"grad_norm": 1.890858769416809,
"learning_rate": 0.001,
"loss": 0.8418,
"step": 263500
},
{
"epoch": 85.19715578539108,
"grad_norm": 52.027462005615234,
"learning_rate": 0.001,
"loss": 0.8351,
"step": 263600
},
{
"epoch": 85.229476405947,
"grad_norm": 2.1597611904144287,
"learning_rate": 0.001,
"loss": 0.8313,
"step": 263700
},
{
"epoch": 85.2617970265029,
"grad_norm": 1.7659682035446167,
"learning_rate": 0.001,
"loss": 0.8469,
"step": 263800
},
{
"epoch": 85.29411764705883,
"grad_norm": 2.0225253105163574,
"learning_rate": 0.001,
"loss": 0.8391,
"step": 263900
},
{
"epoch": 85.32643826761473,
"grad_norm": 1.9262751340866089,
"learning_rate": 0.001,
"loss": 0.8529,
"step": 264000
},
{
"epoch": 85.35875888817066,
"grad_norm": 2.231473684310913,
"learning_rate": 0.001,
"loss": 0.8587,
"step": 264100
},
{
"epoch": 85.39107950872656,
"grad_norm": 1.8628264665603638,
"learning_rate": 0.001,
"loss": 0.8519,
"step": 264200
},
{
"epoch": 85.42340012928248,
"grad_norm": 2.2745962142944336,
"learning_rate": 0.001,
"loss": 0.8694,
"step": 264300
},
{
"epoch": 85.45572074983839,
"grad_norm": 1.8171344995498657,
"learning_rate": 0.001,
"loss": 0.8683,
"step": 264400
},
{
"epoch": 85.48804137039431,
"grad_norm": 1.960572600364685,
"learning_rate": 0.001,
"loss": 0.8763,
"step": 264500
},
{
"epoch": 85.52036199095022,
"grad_norm": 1.8863012790679932,
"learning_rate": 0.001,
"loss": 0.8631,
"step": 264600
},
{
"epoch": 85.55268261150614,
"grad_norm": 2.6001973152160645,
"learning_rate": 0.001,
"loss": 0.8896,
"step": 264700
},
{
"epoch": 85.58500323206205,
"grad_norm": 1.8021491765975952,
"learning_rate": 0.001,
"loss": 0.8748,
"step": 264800
},
{
"epoch": 85.61732385261797,
"grad_norm": 2.0595309734344482,
"learning_rate": 0.001,
"loss": 0.8827,
"step": 264900
},
{
"epoch": 85.64964447317388,
"grad_norm": 2.0802087783813477,
"learning_rate": 0.001,
"loss": 0.8979,
"step": 265000
},
{
"epoch": 85.6819650937298,
"grad_norm": 1.499593734741211,
"learning_rate": 0.001,
"loss": 0.8839,
"step": 265100
},
{
"epoch": 85.71428571428571,
"grad_norm": 1.6684249639511108,
"learning_rate": 0.001,
"loss": 0.8892,
"step": 265200
},
{
"epoch": 85.74660633484163,
"grad_norm": 3.872833251953125,
"learning_rate": 0.001,
"loss": 0.8859,
"step": 265300
},
{
"epoch": 85.77892695539754,
"grad_norm": 2.0526092052459717,
"learning_rate": 0.001,
"loss": 0.8777,
"step": 265400
},
{
"epoch": 85.81124757595346,
"grad_norm": 1.7454049587249756,
"learning_rate": 0.001,
"loss": 0.9088,
"step": 265500
},
{
"epoch": 85.84356819650937,
"grad_norm": 1.6886650323867798,
"learning_rate": 0.001,
"loss": 0.8788,
"step": 265600
},
{
"epoch": 85.87588881706529,
"grad_norm": 1.5133458375930786,
"learning_rate": 0.001,
"loss": 0.9042,
"step": 265700
},
{
"epoch": 85.9082094376212,
"grad_norm": 2.204150676727295,
"learning_rate": 0.001,
"loss": 0.9147,
"step": 265800
},
{
"epoch": 85.94053005817712,
"grad_norm": 2.4142262935638428,
"learning_rate": 0.001,
"loss": 0.9162,
"step": 265900
},
{
"epoch": 85.97285067873302,
"grad_norm": 1.7503329515457153,
"learning_rate": 0.001,
"loss": 0.9047,
"step": 266000
},
{
"epoch": 86.00517129928895,
"grad_norm": 1.7632008790969849,
"learning_rate": 0.001,
"loss": 0.9059,
"step": 266100
},
{
"epoch": 86.03749191984487,
"grad_norm": 1.7511223554611206,
"learning_rate": 0.001,
"loss": 0.8161,
"step": 266200
},
{
"epoch": 86.06981254040078,
"grad_norm": 1.6529717445373535,
"learning_rate": 0.001,
"loss": 0.8258,
"step": 266300
},
{
"epoch": 86.1021331609567,
"grad_norm": 1.7957078218460083,
"learning_rate": 0.001,
"loss": 0.8162,
"step": 266400
},
{
"epoch": 86.1344537815126,
"grad_norm": 1.553620457649231,
"learning_rate": 0.001,
"loss": 0.834,
"step": 266500
},
{
"epoch": 86.16677440206853,
"grad_norm": 1.6290562152862549,
"learning_rate": 0.001,
"loss": 0.8362,
"step": 266600
},
{
"epoch": 86.19909502262443,
"grad_norm": 1.8925175666809082,
"learning_rate": 0.001,
"loss": 0.8177,
"step": 266700
},
{
"epoch": 86.23141564318036,
"grad_norm": 1.861567497253418,
"learning_rate": 0.001,
"loss": 0.8392,
"step": 266800
},
{
"epoch": 86.26373626373626,
"grad_norm": 1.780373215675354,
"learning_rate": 0.001,
"loss": 0.8536,
"step": 266900
},
{
"epoch": 86.29605688429218,
"grad_norm": 2.946547508239746,
"learning_rate": 0.001,
"loss": 0.8629,
"step": 267000
},
{
"epoch": 86.32837750484809,
"grad_norm": 1.7340573072433472,
"learning_rate": 0.001,
"loss": 0.8415,
"step": 267100
},
{
"epoch": 86.36069812540401,
"grad_norm": 1.4430204629898071,
"learning_rate": 0.001,
"loss": 0.847,
"step": 267200
},
{
"epoch": 86.39301874595992,
"grad_norm": 1.9690868854522705,
"learning_rate": 0.001,
"loss": 0.8584,
"step": 267300
},
{
"epoch": 86.42533936651584,
"grad_norm": 2.0398447513580322,
"learning_rate": 0.001,
"loss": 0.86,
"step": 267400
},
{
"epoch": 86.45765998707175,
"grad_norm": 4.0924577713012695,
"learning_rate": 0.001,
"loss": 0.8571,
"step": 267500
},
{
"epoch": 86.48998060762767,
"grad_norm": 1.7488676309585571,
"learning_rate": 0.001,
"loss": 0.8722,
"step": 267600
},
{
"epoch": 86.52230122818358,
"grad_norm": 1.8430218696594238,
"learning_rate": 0.001,
"loss": 0.8669,
"step": 267700
},
{
"epoch": 86.5546218487395,
"grad_norm": 1.694116473197937,
"learning_rate": 0.001,
"loss": 0.876,
"step": 267800
},
{
"epoch": 86.58694246929541,
"grad_norm": 1.5010638236999512,
"learning_rate": 0.001,
"loss": 0.8711,
"step": 267900
},
{
"epoch": 86.61926308985133,
"grad_norm": 17.831743240356445,
"learning_rate": 0.001,
"loss": 0.8698,
"step": 268000
},
{
"epoch": 86.65158371040724,
"grad_norm": 1.4711484909057617,
"learning_rate": 0.001,
"loss": 0.8819,
"step": 268100
},
{
"epoch": 86.68390433096316,
"grad_norm": 1.886446237564087,
"learning_rate": 0.001,
"loss": 0.8723,
"step": 268200
},
{
"epoch": 86.71622495151907,
"grad_norm": 1.6571316719055176,
"learning_rate": 0.001,
"loss": 0.882,
"step": 268300
},
{
"epoch": 86.74854557207499,
"grad_norm": 1.8453465700149536,
"learning_rate": 0.001,
"loss": 0.8937,
"step": 268400
},
{
"epoch": 86.7808661926309,
"grad_norm": 1.5359055995941162,
"learning_rate": 0.001,
"loss": 0.882,
"step": 268500
},
{
"epoch": 86.81318681318682,
"grad_norm": 1.6801995038986206,
"learning_rate": 0.001,
"loss": 0.8985,
"step": 268600
},
{
"epoch": 86.84550743374272,
"grad_norm": 1.4833015203475952,
"learning_rate": 0.001,
"loss": 0.8788,
"step": 268700
},
{
"epoch": 86.87782805429865,
"grad_norm": 1.8023815155029297,
"learning_rate": 0.001,
"loss": 0.8774,
"step": 268800
},
{
"epoch": 86.91014867485455,
"grad_norm": 1.7117823362350464,
"learning_rate": 0.001,
"loss": 0.8884,
"step": 268900
},
{
"epoch": 86.94246929541048,
"grad_norm": 1.8823192119598389,
"learning_rate": 0.001,
"loss": 0.8795,
"step": 269000
},
{
"epoch": 86.97478991596638,
"grad_norm": 5.488122463226318,
"learning_rate": 0.001,
"loss": 0.9028,
"step": 269100
},
{
"epoch": 87.0071105365223,
"grad_norm": 1.49545156955719,
"learning_rate": 0.001,
"loss": 0.8898,
"step": 269200
},
{
"epoch": 87.03943115707821,
"grad_norm": 1.3911610841751099,
"learning_rate": 0.001,
"loss": 0.8081,
"step": 269300
},
{
"epoch": 87.07175177763413,
"grad_norm": 1.6940593719482422,
"learning_rate": 0.001,
"loss": 0.8151,
"step": 269400
},
{
"epoch": 87.10407239819004,
"grad_norm": 1.702312707901001,
"learning_rate": 0.001,
"loss": 0.8274,
"step": 269500
},
{
"epoch": 87.13639301874596,
"grad_norm": 1.5827884674072266,
"learning_rate": 0.001,
"loss": 0.8283,
"step": 269600
},
{
"epoch": 87.16871363930187,
"grad_norm": 1.4666353464126587,
"learning_rate": 0.001,
"loss": 0.8203,
"step": 269700
},
{
"epoch": 87.20103425985779,
"grad_norm": 1.3696997165679932,
"learning_rate": 0.001,
"loss": 0.8289,
"step": 269800
},
{
"epoch": 87.2333548804137,
"grad_norm": 5.3828959465026855,
"learning_rate": 0.001,
"loss": 0.8233,
"step": 269900
},
{
"epoch": 87.26567550096962,
"grad_norm": 2.304546594619751,
"learning_rate": 0.001,
"loss": 0.8399,
"step": 270000
},
{
"epoch": 87.29799612152553,
"grad_norm": 1.6255569458007812,
"learning_rate": 0.001,
"loss": 0.8468,
"step": 270100
},
{
"epoch": 87.33031674208145,
"grad_norm": 1.8003710508346558,
"learning_rate": 0.001,
"loss": 0.8363,
"step": 270200
},
{
"epoch": 87.36263736263736,
"grad_norm": 1.5921064615249634,
"learning_rate": 0.001,
"loss": 0.8417,
"step": 270300
},
{
"epoch": 87.39495798319328,
"grad_norm": 1.5983816385269165,
"learning_rate": 0.001,
"loss": 0.838,
"step": 270400
},
{
"epoch": 87.42727860374919,
"grad_norm": 2.061232089996338,
"learning_rate": 0.001,
"loss": 0.8513,
"step": 270500
},
{
"epoch": 87.45959922430511,
"grad_norm": 1.5790170431137085,
"learning_rate": 0.001,
"loss": 0.8585,
"step": 270600
},
{
"epoch": 87.49191984486102,
"grad_norm": 1.874551773071289,
"learning_rate": 0.001,
"loss": 0.8671,
"step": 270700
},
{
"epoch": 87.52424046541694,
"grad_norm": 3.1544809341430664,
"learning_rate": 0.001,
"loss": 0.847,
"step": 270800
},
{
"epoch": 87.55656108597285,
"grad_norm": 1.533233880996704,
"learning_rate": 0.001,
"loss": 0.8618,
"step": 270900
},
{
"epoch": 87.58888170652877,
"grad_norm": 1.7700929641723633,
"learning_rate": 0.001,
"loss": 0.8678,
"step": 271000
},
{
"epoch": 87.62120232708467,
"grad_norm": 1.546187162399292,
"learning_rate": 0.001,
"loss": 0.8752,
"step": 271100
},
{
"epoch": 87.6535229476406,
"grad_norm": 1.6159831285476685,
"learning_rate": 0.001,
"loss": 0.8523,
"step": 271200
},
{
"epoch": 87.6858435681965,
"grad_norm": 1.5596297979354858,
"learning_rate": 0.001,
"loss": 0.8704,
"step": 271300
},
{
"epoch": 87.71816418875243,
"grad_norm": 1.7366268634796143,
"learning_rate": 0.001,
"loss": 0.8717,
"step": 271400
},
{
"epoch": 87.75048480930833,
"grad_norm": 5.429916858673096,
"learning_rate": 0.001,
"loss": 0.8716,
"step": 271500
},
{
"epoch": 87.78280542986425,
"grad_norm": 1.5415160655975342,
"learning_rate": 0.001,
"loss": 0.8722,
"step": 271600
},
{
"epoch": 87.81512605042016,
"grad_norm": 1.9593197107315063,
"learning_rate": 0.001,
"loss": 0.8802,
"step": 271700
},
{
"epoch": 87.84744667097608,
"grad_norm": 2.1193692684173584,
"learning_rate": 0.001,
"loss": 0.8842,
"step": 271800
},
{
"epoch": 87.87976729153199,
"grad_norm": 2.191317558288574,
"learning_rate": 0.001,
"loss": 0.8958,
"step": 271900
},
{
"epoch": 87.91208791208791,
"grad_norm": 2.4855754375457764,
"learning_rate": 0.001,
"loss": 0.899,
"step": 272000
},
{
"epoch": 87.94440853264382,
"grad_norm": 1.505005955696106,
"learning_rate": 0.001,
"loss": 0.8938,
"step": 272100
},
{
"epoch": 87.97672915319974,
"grad_norm": 1.9065461158752441,
"learning_rate": 0.001,
"loss": 0.907,
"step": 272200
},
{
"epoch": 88.00904977375566,
"grad_norm": 1.7264128923416138,
"learning_rate": 0.001,
"loss": 0.8799,
"step": 272300
},
{
"epoch": 88.04137039431157,
"grad_norm": 1.7042555809020996,
"learning_rate": 0.001,
"loss": 0.8046,
"step": 272400
},
{
"epoch": 88.07369101486749,
"grad_norm": 2.3185431957244873,
"learning_rate": 0.001,
"loss": 0.8033,
"step": 272500
},
{
"epoch": 88.1060116354234,
"grad_norm": 1.6043986082077026,
"learning_rate": 0.001,
"loss": 0.8068,
"step": 272600
},
{
"epoch": 88.13833225597932,
"grad_norm": 1.5428094863891602,
"learning_rate": 0.001,
"loss": 0.8138,
"step": 272700
},
{
"epoch": 88.17065287653523,
"grad_norm": 1.9031853675842285,
"learning_rate": 0.001,
"loss": 0.811,
"step": 272800
},
{
"epoch": 88.20297349709115,
"grad_norm": 2.3232343196868896,
"learning_rate": 0.001,
"loss": 0.8193,
"step": 272900
},
{
"epoch": 88.23529411764706,
"grad_norm": 1.4863861799240112,
"learning_rate": 0.001,
"loss": 0.8203,
"step": 273000
},
{
"epoch": 88.26761473820298,
"grad_norm": 1.7252644300460815,
"learning_rate": 0.001,
"loss": 0.836,
"step": 273100
},
{
"epoch": 88.29993535875889,
"grad_norm": 1.682450771331787,
"learning_rate": 0.001,
"loss": 0.8439,
"step": 273200
},
{
"epoch": 88.33225597931481,
"grad_norm": 1.6543453931808472,
"learning_rate": 0.001,
"loss": 0.8483,
"step": 273300
},
{
"epoch": 88.36457659987072,
"grad_norm": 1.7153080701828003,
"learning_rate": 0.001,
"loss": 0.8417,
"step": 273400
},
{
"epoch": 88.39689722042664,
"grad_norm": 2.134333372116089,
"learning_rate": 0.001,
"loss": 0.8499,
"step": 273500
},
{
"epoch": 88.42921784098255,
"grad_norm": 2.4157066345214844,
"learning_rate": 0.001,
"loss": 0.8418,
"step": 273600
},
{
"epoch": 88.46153846153847,
"grad_norm": 8.76519775390625,
"learning_rate": 0.001,
"loss": 0.8498,
"step": 273700
},
{
"epoch": 88.49385908209437,
"grad_norm": 1.6663837432861328,
"learning_rate": 0.001,
"loss": 0.854,
"step": 273800
},
{
"epoch": 88.5261797026503,
"grad_norm": 1.6248563528060913,
"learning_rate": 0.001,
"loss": 0.858,
"step": 273900
},
{
"epoch": 88.5585003232062,
"grad_norm": 1.6257447004318237,
"learning_rate": 0.001,
"loss": 0.865,
"step": 274000
},
{
"epoch": 88.59082094376213,
"grad_norm": 2.0399746894836426,
"learning_rate": 0.001,
"loss": 0.868,
"step": 274100
},
{
"epoch": 88.62314156431803,
"grad_norm": 1.530640721321106,
"learning_rate": 0.001,
"loss": 0.8678,
"step": 274200
},
{
"epoch": 88.65546218487395,
"grad_norm": 1.7557321786880493,
"learning_rate": 0.001,
"loss": 0.8672,
"step": 274300
},
{
"epoch": 88.68778280542986,
"grad_norm": 1.8893425464630127,
"learning_rate": 0.001,
"loss": 0.8677,
"step": 274400
},
{
"epoch": 88.72010342598578,
"grad_norm": 1.8820282220840454,
"learning_rate": 0.001,
"loss": 0.8846,
"step": 274500
},
{
"epoch": 88.75242404654169,
"grad_norm": 1.5749956369400024,
"learning_rate": 0.001,
"loss": 0.8603,
"step": 274600
},
{
"epoch": 88.78474466709761,
"grad_norm": 1.4271475076675415,
"learning_rate": 0.001,
"loss": 0.8767,
"step": 274700
},
{
"epoch": 88.81706528765352,
"grad_norm": 1.6466025114059448,
"learning_rate": 0.001,
"loss": 0.8762,
"step": 274800
},
{
"epoch": 88.84938590820944,
"grad_norm": 1.7966150045394897,
"learning_rate": 0.001,
"loss": 0.8847,
"step": 274900
},
{
"epoch": 88.88170652876535,
"grad_norm": 1.834599256515503,
"learning_rate": 0.001,
"loss": 0.8873,
"step": 275000
},
{
"epoch": 88.91402714932127,
"grad_norm": 1.5994890928268433,
"learning_rate": 0.001,
"loss": 0.8755,
"step": 275100
},
{
"epoch": 88.94634776987718,
"grad_norm": 6.51814079284668,
"learning_rate": 0.001,
"loss": 0.8676,
"step": 275200
},
{
"epoch": 88.9786683904331,
"grad_norm": 1.8895727396011353,
"learning_rate": 0.001,
"loss": 0.884,
"step": 275300
},
{
"epoch": 89.01098901098901,
"grad_norm": 2.0515339374542236,
"learning_rate": 0.001,
"loss": 0.8533,
"step": 275400
},
{
"epoch": 89.04330963154493,
"grad_norm": 2.124110698699951,
"learning_rate": 0.001,
"loss": 0.8042,
"step": 275500
},
{
"epoch": 89.07563025210084,
"grad_norm": 1.8615392446517944,
"learning_rate": 0.001,
"loss": 0.7895,
"step": 275600
},
{
"epoch": 89.10795087265676,
"grad_norm": 1.9530836343765259,
"learning_rate": 0.001,
"loss": 0.8206,
"step": 275700
},
{
"epoch": 89.14027149321267,
"grad_norm": 2.4263992309570312,
"learning_rate": 0.001,
"loss": 0.8153,
"step": 275800
},
{
"epoch": 89.17259211376859,
"grad_norm": 1.5851964950561523,
"learning_rate": 0.001,
"loss": 0.8218,
"step": 275900
},
{
"epoch": 89.2049127343245,
"grad_norm": 2.3962132930755615,
"learning_rate": 0.001,
"loss": 0.8249,
"step": 276000
},
{
"epoch": 89.23723335488042,
"grad_norm": 2.4741365909576416,
"learning_rate": 0.001,
"loss": 0.8316,
"step": 276100
},
{
"epoch": 89.26955397543632,
"grad_norm": 1.6986922025680542,
"learning_rate": 0.001,
"loss": 0.8327,
"step": 276200
},
{
"epoch": 89.30187459599225,
"grad_norm": 1.554884910583496,
"learning_rate": 0.001,
"loss": 0.8264,
"step": 276300
},
{
"epoch": 89.33419521654815,
"grad_norm": 7.18624210357666,
"learning_rate": 0.001,
"loss": 0.8393,
"step": 276400
},
{
"epoch": 89.36651583710407,
"grad_norm": 2.092738628387451,
"learning_rate": 0.001,
"loss": 0.8523,
"step": 276500
},
{
"epoch": 89.39883645765998,
"grad_norm": 2.584104299545288,
"learning_rate": 0.001,
"loss": 0.8507,
"step": 276600
},
{
"epoch": 89.4311570782159,
"grad_norm": 1.88053297996521,
"learning_rate": 0.001,
"loss": 0.8363,
"step": 276700
},
{
"epoch": 89.46347769877181,
"grad_norm": 1.7402923107147217,
"learning_rate": 0.001,
"loss": 0.8536,
"step": 276800
},
{
"epoch": 89.49579831932773,
"grad_norm": 1.7702052593231201,
"learning_rate": 0.001,
"loss": 0.8553,
"step": 276900
},
{
"epoch": 89.52811893988364,
"grad_norm": 1.5800539255142212,
"learning_rate": 0.001,
"loss": 0.8519,
"step": 277000
},
{
"epoch": 89.56043956043956,
"grad_norm": 1.7110040187835693,
"learning_rate": 0.001,
"loss": 0.8445,
"step": 277100
},
{
"epoch": 89.59276018099547,
"grad_norm": 1.5943825244903564,
"learning_rate": 0.001,
"loss": 0.8441,
"step": 277200
},
{
"epoch": 89.62508080155139,
"grad_norm": 1.785972237586975,
"learning_rate": 0.001,
"loss": 0.8622,
"step": 277300
},
{
"epoch": 89.6574014221073,
"grad_norm": 1.652089238166809,
"learning_rate": 0.001,
"loss": 0.8598,
"step": 277400
},
{
"epoch": 89.68972204266322,
"grad_norm": 1.664921522140503,
"learning_rate": 0.001,
"loss": 0.8516,
"step": 277500
},
{
"epoch": 89.72204266321913,
"grad_norm": 1.6153144836425781,
"learning_rate": 0.001,
"loss": 0.869,
"step": 277600
},
{
"epoch": 89.75436328377505,
"grad_norm": 2.071241617202759,
"learning_rate": 0.001,
"loss": 0.8643,
"step": 277700
},
{
"epoch": 89.78668390433096,
"grad_norm": 1.825007438659668,
"learning_rate": 0.001,
"loss": 0.8663,
"step": 277800
},
{
"epoch": 89.81900452488688,
"grad_norm": 2.0584936141967773,
"learning_rate": 0.001,
"loss": 0.8828,
"step": 277900
},
{
"epoch": 89.85132514544279,
"grad_norm": 2.5568366050720215,
"learning_rate": 0.001,
"loss": 0.8628,
"step": 278000
},
{
"epoch": 89.88364576599871,
"grad_norm": 1.7184977531433105,
"learning_rate": 0.001,
"loss": 0.8739,
"step": 278100
},
{
"epoch": 89.91596638655462,
"grad_norm": 1.8462786674499512,
"learning_rate": 0.001,
"loss": 0.8731,
"step": 278200
},
{
"epoch": 89.94828700711054,
"grad_norm": 3.5282583236694336,
"learning_rate": 0.001,
"loss": 0.8852,
"step": 278300
},
{
"epoch": 89.98060762766644,
"grad_norm": 1.5443145036697388,
"learning_rate": 0.001,
"loss": 0.8737,
"step": 278400
},
{
"epoch": 90.01292824822237,
"grad_norm": 13.76854133605957,
"learning_rate": 0.001,
"loss": 0.821,
"step": 278500
},
{
"epoch": 90.04524886877829,
"grad_norm": 1.5974069833755493,
"learning_rate": 0.001,
"loss": 0.7964,
"step": 278600
},
{
"epoch": 90.0775694893342,
"grad_norm": 1.4872169494628906,
"learning_rate": 0.001,
"loss": 0.8018,
"step": 278700
},
{
"epoch": 90.10989010989012,
"grad_norm": 2.4473085403442383,
"learning_rate": 0.001,
"loss": 0.7932,
"step": 278800
},
{
"epoch": 90.14221073044602,
"grad_norm": 1.7129226922988892,
"learning_rate": 0.001,
"loss": 0.8181,
"step": 278900
},
{
"epoch": 90.17453135100195,
"grad_norm": 1.6856908798217773,
"learning_rate": 0.001,
"loss": 0.8114,
"step": 279000
},
{
"epoch": 90.20685197155785,
"grad_norm": 1.5722812414169312,
"learning_rate": 0.001,
"loss": 0.8026,
"step": 279100
},
{
"epoch": 90.23917259211377,
"grad_norm": 1.3502541780471802,
"learning_rate": 0.001,
"loss": 0.8319,
"step": 279200
},
{
"epoch": 90.27149321266968,
"grad_norm": 1.4687424898147583,
"learning_rate": 0.001,
"loss": 0.8336,
"step": 279300
},
{
"epoch": 90.3038138332256,
"grad_norm": 1.8175430297851562,
"learning_rate": 0.001,
"loss": 0.8138,
"step": 279400
},
{
"epoch": 90.33613445378151,
"grad_norm": 1.4699106216430664,
"learning_rate": 0.001,
"loss": 0.8276,
"step": 279500
},
{
"epoch": 90.36845507433743,
"grad_norm": 2.0859620571136475,
"learning_rate": 0.001,
"loss": 0.8293,
"step": 279600
},
{
"epoch": 90.40077569489334,
"grad_norm": 1.5765058994293213,
"learning_rate": 0.001,
"loss": 0.8308,
"step": 279700
},
{
"epoch": 90.43309631544926,
"grad_norm": 1.4347844123840332,
"learning_rate": 0.001,
"loss": 0.825,
"step": 279800
},
{
"epoch": 90.46541693600517,
"grad_norm": 1.8731211423873901,
"learning_rate": 0.001,
"loss": 0.8459,
"step": 279900
},
{
"epoch": 90.49773755656109,
"grad_norm": 1.7345112562179565,
"learning_rate": 0.001,
"loss": 0.8469,
"step": 280000
},
{
"epoch": 90.530058177117,
"grad_norm": 1.9544506072998047,
"learning_rate": 0.001,
"loss": 0.8388,
"step": 280100
},
{
"epoch": 90.56237879767292,
"grad_norm": 1.8892264366149902,
"learning_rate": 0.001,
"loss": 0.8392,
"step": 280200
},
{
"epoch": 90.59469941822883,
"grad_norm": 2.320929765701294,
"learning_rate": 0.001,
"loss": 0.8473,
"step": 280300
},
{
"epoch": 90.62702003878475,
"grad_norm": 1.7178595066070557,
"learning_rate": 0.001,
"loss": 0.8469,
"step": 280400
},
{
"epoch": 90.65934065934066,
"grad_norm": 1.7863913774490356,
"learning_rate": 0.001,
"loss": 0.8319,
"step": 280500
},
{
"epoch": 90.69166127989658,
"grad_norm": 1.957006573677063,
"learning_rate": 0.001,
"loss": 0.8543,
"step": 280600
},
{
"epoch": 90.72398190045249,
"grad_norm": 1.944828748703003,
"learning_rate": 0.001,
"loss": 0.8531,
"step": 280700
},
{
"epoch": 90.75630252100841,
"grad_norm": 1.5873464345932007,
"learning_rate": 0.001,
"loss": 0.8541,
"step": 280800
},
{
"epoch": 90.78862314156432,
"grad_norm": 1.8036553859710693,
"learning_rate": 0.001,
"loss": 0.8562,
"step": 280900
},
{
"epoch": 90.82094376212024,
"grad_norm": 1.9891469478607178,
"learning_rate": 0.001,
"loss": 0.8705,
"step": 281000
},
{
"epoch": 90.85326438267614,
"grad_norm": 2.126394510269165,
"learning_rate": 0.001,
"loss": 0.8843,
"step": 281100
},
{
"epoch": 90.88558500323207,
"grad_norm": 2.3660521507263184,
"learning_rate": 0.001,
"loss": 0.883,
"step": 281200
},
{
"epoch": 90.91790562378797,
"grad_norm": 1.8034424781799316,
"learning_rate": 0.001,
"loss": 0.8748,
"step": 281300
},
{
"epoch": 90.9502262443439,
"grad_norm": 2.120250701904297,
"learning_rate": 0.001,
"loss": 0.8603,
"step": 281400
},
{
"epoch": 90.9825468648998,
"grad_norm": 1.402356743812561,
"learning_rate": 0.001,
"loss": 0.8768,
"step": 281500
},
{
"epoch": 91.01486748545572,
"grad_norm": 1.658185362815857,
"learning_rate": 0.001,
"loss": 0.8144,
"step": 281600
},
{
"epoch": 91.04718810601163,
"grad_norm": 1.5982550382614136,
"learning_rate": 0.001,
"loss": 0.7909,
"step": 281700
},
{
"epoch": 91.07950872656755,
"grad_norm": 1.4638272523880005,
"learning_rate": 0.001,
"loss": 0.7907,
"step": 281800
},
{
"epoch": 91.11182934712346,
"grad_norm": 1.9809621572494507,
"learning_rate": 0.001,
"loss": 0.8007,
"step": 281900
},
{
"epoch": 91.14414996767938,
"grad_norm": 1.5442944765090942,
"learning_rate": 0.001,
"loss": 0.8162,
"step": 282000
},
{
"epoch": 91.17647058823529,
"grad_norm": 1.343239665031433,
"learning_rate": 0.001,
"loss": 0.8132,
"step": 282100
},
{
"epoch": 91.20879120879121,
"grad_norm": 1.709086537361145,
"learning_rate": 0.001,
"loss": 0.8112,
"step": 282200
},
{
"epoch": 91.24111182934712,
"grad_norm": 1.8876216411590576,
"learning_rate": 0.001,
"loss": 0.8102,
"step": 282300
},
{
"epoch": 91.27343244990304,
"grad_norm": 1.4226528406143188,
"learning_rate": 0.001,
"loss": 0.8129,
"step": 282400
},
{
"epoch": 91.30575307045895,
"grad_norm": 1.2886971235275269,
"learning_rate": 0.001,
"loss": 0.8256,
"step": 282500
},
{
"epoch": 91.33807369101487,
"grad_norm": 1.4405816793441772,
"learning_rate": 0.001,
"loss": 0.8168,
"step": 282600
},
{
"epoch": 91.37039431157078,
"grad_norm": 1.7710850238800049,
"learning_rate": 0.001,
"loss": 0.8317,
"step": 282700
},
{
"epoch": 91.4027149321267,
"grad_norm": 1.5474108457565308,
"learning_rate": 0.001,
"loss": 0.8283,
"step": 282800
},
{
"epoch": 91.4350355526826,
"grad_norm": 1.4025057554244995,
"learning_rate": 0.001,
"loss": 0.8198,
"step": 282900
},
{
"epoch": 91.46735617323853,
"grad_norm": 1.5733214616775513,
"learning_rate": 0.001,
"loss": 0.8275,
"step": 283000
},
{
"epoch": 91.49967679379444,
"grad_norm": 1.640556812286377,
"learning_rate": 0.001,
"loss": 0.8366,
"step": 283100
},
{
"epoch": 91.53199741435036,
"grad_norm": 1.873395562171936,
"learning_rate": 0.001,
"loss": 0.8338,
"step": 283200
},
{
"epoch": 91.56431803490626,
"grad_norm": 1.2778507471084595,
"learning_rate": 0.001,
"loss": 0.8279,
"step": 283300
},
{
"epoch": 91.59663865546219,
"grad_norm": 1.236539363861084,
"learning_rate": 0.001,
"loss": 0.8453,
"step": 283400
},
{
"epoch": 91.6289592760181,
"grad_norm": 1.8194011449813843,
"learning_rate": 0.001,
"loss": 0.8474,
"step": 283500
},
{
"epoch": 91.66127989657402,
"grad_norm": 1.5453321933746338,
"learning_rate": 0.001,
"loss": 0.8523,
"step": 283600
},
{
"epoch": 91.69360051712992,
"grad_norm": 2.883462429046631,
"learning_rate": 0.001,
"loss": 0.8488,
"step": 283700
},
{
"epoch": 91.72592113768584,
"grad_norm": 1.4357653856277466,
"learning_rate": 0.001,
"loss": 0.8514,
"step": 283800
},
{
"epoch": 91.75824175824175,
"grad_norm": 1.421116590499878,
"learning_rate": 0.001,
"loss": 0.8597,
"step": 283900
},
{
"epoch": 91.79056237879767,
"grad_norm": 1.4790443181991577,
"learning_rate": 0.001,
"loss": 0.8558,
"step": 284000
},
{
"epoch": 91.82288299935358,
"grad_norm": 1.6023317575454712,
"learning_rate": 0.001,
"loss": 0.8622,
"step": 284100
},
{
"epoch": 91.8552036199095,
"grad_norm": 1.4575282335281372,
"learning_rate": 0.001,
"loss": 0.8699,
"step": 284200
},
{
"epoch": 91.88752424046541,
"grad_norm": 1.5747337341308594,
"learning_rate": 0.001,
"loss": 0.8589,
"step": 284300
},
{
"epoch": 91.91984486102133,
"grad_norm": 1.5780296325683594,
"learning_rate": 0.001,
"loss": 0.8587,
"step": 284400
},
{
"epoch": 91.95216548157724,
"grad_norm": 1.3909651041030884,
"learning_rate": 0.001,
"loss": 0.8712,
"step": 284500
},
{
"epoch": 91.98448610213316,
"grad_norm": 1.6297881603240967,
"learning_rate": 0.001,
"loss": 0.8557,
"step": 284600
},
{
"epoch": 92.01680672268908,
"grad_norm": 9.393735885620117,
"learning_rate": 0.001,
"loss": 0.7966,
"step": 284700
},
{
"epoch": 92.04912734324499,
"grad_norm": 2.9230968952178955,
"learning_rate": 0.001,
"loss": 0.7938,
"step": 284800
},
{
"epoch": 92.08144796380091,
"grad_norm": 3.2949323654174805,
"learning_rate": 0.001,
"loss": 0.7879,
"step": 284900
},
{
"epoch": 92.11376858435682,
"grad_norm": 2.239393949508667,
"learning_rate": 0.001,
"loss": 0.7979,
"step": 285000
},
{
"epoch": 92.14608920491274,
"grad_norm": 2.370790958404541,
"learning_rate": 0.001,
"loss": 0.8035,
"step": 285100
},
{
"epoch": 92.17840982546865,
"grad_norm": 2.7069642543792725,
"learning_rate": 0.001,
"loss": 0.8146,
"step": 285200
},
{
"epoch": 92.21073044602457,
"grad_norm": 2.8528716564178467,
"learning_rate": 0.001,
"loss": 0.8003,
"step": 285300
},
{
"epoch": 92.24305106658048,
"grad_norm": 2.916715383529663,
"learning_rate": 0.001,
"loss": 0.8079,
"step": 285400
},
{
"epoch": 92.2753716871364,
"grad_norm": 2.335475444793701,
"learning_rate": 0.001,
"loss": 0.8268,
"step": 285500
},
{
"epoch": 92.3076923076923,
"grad_norm": 2.3702704906463623,
"learning_rate": 0.001,
"loss": 0.8217,
"step": 285600
},
{
"epoch": 92.34001292824823,
"grad_norm": 2.541818380355835,
"learning_rate": 0.001,
"loss": 0.8306,
"step": 285700
},
{
"epoch": 92.37233354880414,
"grad_norm": 2.096249580383301,
"learning_rate": 0.001,
"loss": 0.814,
"step": 285800
},
{
"epoch": 92.40465416936006,
"grad_norm": 1.8846521377563477,
"learning_rate": 0.001,
"loss": 0.8156,
"step": 285900
},
{
"epoch": 92.43697478991596,
"grad_norm": 2.9999451637268066,
"learning_rate": 0.001,
"loss": 0.825,
"step": 286000
},
{
"epoch": 92.46929541047189,
"grad_norm": 1.9357293844223022,
"learning_rate": 0.001,
"loss": 0.8355,
"step": 286100
},
{
"epoch": 92.5016160310278,
"grad_norm": 2.528646469116211,
"learning_rate": 0.001,
"loss": 0.8257,
"step": 286200
},
{
"epoch": 92.53393665158372,
"grad_norm": 2.2945749759674072,
"learning_rate": 0.001,
"loss": 0.8417,
"step": 286300
},
{
"epoch": 92.56625727213962,
"grad_norm": 4.917908668518066,
"learning_rate": 0.001,
"loss": 0.8357,
"step": 286400
},
{
"epoch": 92.59857789269554,
"grad_norm": 1.9125192165374756,
"learning_rate": 0.001,
"loss": 0.8315,
"step": 286500
},
{
"epoch": 92.63089851325145,
"grad_norm": 2.123912811279297,
"learning_rate": 0.001,
"loss": 0.8342,
"step": 286600
},
{
"epoch": 92.66321913380737,
"grad_norm": 1.902299404144287,
"learning_rate": 0.001,
"loss": 0.8473,
"step": 286700
},
{
"epoch": 92.69553975436328,
"grad_norm": 2.312319278717041,
"learning_rate": 0.001,
"loss": 0.8454,
"step": 286800
},
{
"epoch": 92.7278603749192,
"grad_norm": 2.930506944656372,
"learning_rate": 0.001,
"loss": 0.8352,
"step": 286900
},
{
"epoch": 92.76018099547511,
"grad_norm": 2.298229932785034,
"learning_rate": 0.001,
"loss": 0.8647,
"step": 287000
},
{
"epoch": 92.79250161603103,
"grad_norm": 2.502454996109009,
"learning_rate": 0.001,
"loss": 0.8454,
"step": 287100
},
{
"epoch": 92.82482223658694,
"grad_norm": 2.3575501441955566,
"learning_rate": 0.001,
"loss": 0.8612,
"step": 287200
},
{
"epoch": 92.85714285714286,
"grad_norm": 2.6692607402801514,
"learning_rate": 0.001,
"loss": 0.8579,
"step": 287300
},
{
"epoch": 92.88946347769877,
"grad_norm": 2.066166400909424,
"learning_rate": 0.001,
"loss": 0.8441,
"step": 287400
},
{
"epoch": 92.92178409825469,
"grad_norm": 2.153433084487915,
"learning_rate": 0.001,
"loss": 0.8543,
"step": 287500
},
{
"epoch": 92.9541047188106,
"grad_norm": 2.5967466831207275,
"learning_rate": 0.001,
"loss": 0.8561,
"step": 287600
},
{
"epoch": 92.98642533936652,
"grad_norm": 2.4448750019073486,
"learning_rate": 0.001,
"loss": 0.863,
"step": 287700
},
{
"epoch": 93.01874595992243,
"grad_norm": 2.5534839630126953,
"learning_rate": 0.001,
"loss": 0.8194,
"step": 287800
},
{
"epoch": 93.05106658047835,
"grad_norm": 2.0508859157562256,
"learning_rate": 0.001,
"loss": 0.7717,
"step": 287900
},
{
"epoch": 93.08338720103426,
"grad_norm": 1.9136475324630737,
"learning_rate": 0.001,
"loss": 0.7865,
"step": 288000
},
{
"epoch": 93.11570782159018,
"grad_norm": 1.851262092590332,
"learning_rate": 0.001,
"loss": 0.7909,
"step": 288100
},
{
"epoch": 93.14802844214609,
"grad_norm": 1.6321786642074585,
"learning_rate": 0.001,
"loss": 0.7939,
"step": 288200
},
{
"epoch": 93.180349062702,
"grad_norm": 2.0846023559570312,
"learning_rate": 0.001,
"loss": 0.7887,
"step": 288300
},
{
"epoch": 93.21266968325791,
"grad_norm": 1.5281572341918945,
"learning_rate": 0.001,
"loss": 0.794,
"step": 288400
},
{
"epoch": 93.24499030381384,
"grad_norm": 1.4659923315048218,
"learning_rate": 0.001,
"loss": 0.7852,
"step": 288500
},
{
"epoch": 93.27731092436974,
"grad_norm": 1.5710750818252563,
"learning_rate": 0.001,
"loss": 0.7999,
"step": 288600
},
{
"epoch": 93.30963154492567,
"grad_norm": 1.6601274013519287,
"learning_rate": 0.001,
"loss": 0.8061,
"step": 288700
},
{
"epoch": 93.34195216548157,
"grad_norm": 2.21152663230896,
"learning_rate": 0.001,
"loss": 0.8111,
"step": 288800
},
{
"epoch": 93.3742727860375,
"grad_norm": 1.6584738492965698,
"learning_rate": 0.001,
"loss": 0.8111,
"step": 288900
},
{
"epoch": 93.4065934065934,
"grad_norm": 1.7776179313659668,
"learning_rate": 0.001,
"loss": 0.8083,
"step": 289000
},
{
"epoch": 93.43891402714932,
"grad_norm": 1.865175485610962,
"learning_rate": 0.001,
"loss": 0.8207,
"step": 289100
},
{
"epoch": 93.47123464770523,
"grad_norm": 2.4652998447418213,
"learning_rate": 0.001,
"loss": 0.8225,
"step": 289200
},
{
"epoch": 93.50355526826115,
"grad_norm": 1.636830449104309,
"learning_rate": 0.001,
"loss": 0.8235,
"step": 289300
},
{
"epoch": 93.53587588881706,
"grad_norm": 1.8473851680755615,
"learning_rate": 0.001,
"loss": 0.8259,
"step": 289400
},
{
"epoch": 93.56819650937298,
"grad_norm": 1.7582794427871704,
"learning_rate": 0.001,
"loss": 0.8427,
"step": 289500
},
{
"epoch": 93.60051712992889,
"grad_norm": 1.745152473449707,
"learning_rate": 0.001,
"loss": 0.8249,
"step": 289600
},
{
"epoch": 93.63283775048481,
"grad_norm": 1.648447036743164,
"learning_rate": 0.001,
"loss": 0.8273,
"step": 289700
},
{
"epoch": 93.66515837104072,
"grad_norm": 1.4837515354156494,
"learning_rate": 0.001,
"loss": 0.8356,
"step": 289800
},
{
"epoch": 93.69747899159664,
"grad_norm": 1.7128311395645142,
"learning_rate": 0.001,
"loss": 0.8498,
"step": 289900
},
{
"epoch": 93.72979961215255,
"grad_norm": 2.1904046535491943,
"learning_rate": 0.001,
"loss": 0.842,
"step": 290000
},
{
"epoch": 93.76212023270847,
"grad_norm": 1.8092793226242065,
"learning_rate": 0.001,
"loss": 0.8392,
"step": 290100
},
{
"epoch": 93.79444085326438,
"grad_norm": 1.474500060081482,
"learning_rate": 0.001,
"loss": 0.8514,
"step": 290200
},
{
"epoch": 93.8267614738203,
"grad_norm": 1.9503815174102783,
"learning_rate": 0.001,
"loss": 0.8679,
"step": 290300
},
{
"epoch": 93.8590820943762,
"grad_norm": 1.7040289640426636,
"learning_rate": 0.001,
"loss": 0.8419,
"step": 290400
},
{
"epoch": 93.89140271493213,
"grad_norm": 1.7533595561981201,
"learning_rate": 0.001,
"loss": 0.8657,
"step": 290500
},
{
"epoch": 93.92372333548803,
"grad_norm": 24.08082389831543,
"learning_rate": 0.001,
"loss": 0.8406,
"step": 290600
},
{
"epoch": 93.95604395604396,
"grad_norm": 1.9626394510269165,
"learning_rate": 0.001,
"loss": 0.8745,
"step": 290700
},
{
"epoch": 93.98836457659988,
"grad_norm": 2.037595272064209,
"learning_rate": 0.001,
"loss": 0.8533,
"step": 290800
},
{
"epoch": 94.02068519715579,
"grad_norm": 1.519704818725586,
"learning_rate": 0.001,
"loss": 0.8253,
"step": 290900
},
{
"epoch": 94.0530058177117,
"grad_norm": 1.9078086614608765,
"learning_rate": 0.001,
"loss": 0.7688,
"step": 291000
},
{
"epoch": 94.08532643826761,
"grad_norm": 1.8446297645568848,
"learning_rate": 0.001,
"loss": 0.785,
"step": 291100
},
{
"epoch": 94.11764705882354,
"grad_norm": 2.2081801891326904,
"learning_rate": 0.001,
"loss": 0.7876,
"step": 291200
},
{
"epoch": 94.14996767937944,
"grad_norm": 1.5179626941680908,
"learning_rate": 0.001,
"loss": 0.7979,
"step": 291300
},
{
"epoch": 94.18228829993537,
"grad_norm": 1.627453088760376,
"learning_rate": 0.001,
"loss": 0.7764,
"step": 291400
},
{
"epoch": 94.21460892049127,
"grad_norm": 1.617651343345642,
"learning_rate": 0.001,
"loss": 0.7927,
"step": 291500
},
{
"epoch": 94.2469295410472,
"grad_norm": 2.2514536380767822,
"learning_rate": 0.001,
"loss": 0.7932,
"step": 291600
},
{
"epoch": 94.2792501616031,
"grad_norm": 5.1066460609436035,
"learning_rate": 0.001,
"loss": 0.8039,
"step": 291700
},
{
"epoch": 94.31157078215902,
"grad_norm": 1.6531953811645508,
"learning_rate": 0.001,
"loss": 0.8012,
"step": 291800
},
{
"epoch": 94.34389140271493,
"grad_norm": 1.9666928052902222,
"learning_rate": 0.001,
"loss": 0.8011,
"step": 291900
},
{
"epoch": 94.37621202327085,
"grad_norm": 1.7861919403076172,
"learning_rate": 0.001,
"loss": 0.8004,
"step": 292000
},
{
"epoch": 94.40853264382676,
"grad_norm": 1.7683416604995728,
"learning_rate": 0.001,
"loss": 0.8082,
"step": 292100
},
{
"epoch": 94.44085326438268,
"grad_norm": 2.031372308731079,
"learning_rate": 0.001,
"loss": 0.8202,
"step": 292200
},
{
"epoch": 94.47317388493859,
"grad_norm": 1.7173404693603516,
"learning_rate": 0.001,
"loss": 0.8169,
"step": 292300
},
{
"epoch": 94.50549450549451,
"grad_norm": 1.615218162536621,
"learning_rate": 0.001,
"loss": 0.8257,
"step": 292400
},
{
"epoch": 94.53781512605042,
"grad_norm": 1.601548671722412,
"learning_rate": 0.001,
"loss": 0.8292,
"step": 292500
},
{
"epoch": 94.57013574660634,
"grad_norm": 5.030603408813477,
"learning_rate": 0.001,
"loss": 0.8219,
"step": 292600
},
{
"epoch": 94.60245636716225,
"grad_norm": 1.6010547876358032,
"learning_rate": 0.001,
"loss": 0.8289,
"step": 292700
},
{
"epoch": 94.63477698771817,
"grad_norm": 1.8350695371627808,
"learning_rate": 0.001,
"loss": 0.8296,
"step": 292800
},
{
"epoch": 94.66709760827408,
"grad_norm": 2.2363829612731934,
"learning_rate": 0.001,
"loss": 0.8278,
"step": 292900
},
{
"epoch": 94.69941822883,
"grad_norm": 1.8991843461990356,
"learning_rate": 0.001,
"loss": 0.8389,
"step": 293000
},
{
"epoch": 94.7317388493859,
"grad_norm": 3.0361104011535645,
"learning_rate": 0.001,
"loss": 0.8456,
"step": 293100
},
{
"epoch": 94.76405946994183,
"grad_norm": 1.8470691442489624,
"learning_rate": 0.001,
"loss": 0.8518,
"step": 293200
},
{
"epoch": 94.79638009049773,
"grad_norm": 1.6898738145828247,
"learning_rate": 0.001,
"loss": 0.8523,
"step": 293300
},
{
"epoch": 94.82870071105366,
"grad_norm": 1.8268592357635498,
"learning_rate": 0.001,
"loss": 0.8349,
"step": 293400
},
{
"epoch": 94.86102133160956,
"grad_norm": 1.5458370447158813,
"learning_rate": 0.001,
"loss": 0.8191,
"step": 293500
},
{
"epoch": 94.89334195216549,
"grad_norm": 1.5829081535339355,
"learning_rate": 0.001,
"loss": 0.8532,
"step": 293600
},
{
"epoch": 94.9256625727214,
"grad_norm": 2.295776844024658,
"learning_rate": 0.001,
"loss": 0.8528,
"step": 293700
},
{
"epoch": 94.95798319327731,
"grad_norm": 3.22947359085083,
"learning_rate": 0.001,
"loss": 0.8541,
"step": 293800
},
{
"epoch": 94.99030381383322,
"grad_norm": 2.0059330463409424,
"learning_rate": 0.001,
"loss": 0.858,
"step": 293900
},
{
"epoch": 95.02262443438914,
"grad_norm": 1.7929795980453491,
"learning_rate": 0.001,
"loss": 0.8092,
"step": 294000
},
{
"epoch": 95.05494505494505,
"grad_norm": 1.520918369293213,
"learning_rate": 0.001,
"loss": 0.7758,
"step": 294100
},
{
"epoch": 95.08726567550097,
"grad_norm": 1.6936134099960327,
"learning_rate": 0.001,
"loss": 0.7783,
"step": 294200
},
{
"epoch": 95.11958629605688,
"grad_norm": 1.9839485883712769,
"learning_rate": 0.001,
"loss": 0.7663,
"step": 294300
},
{
"epoch": 95.1519069166128,
"grad_norm": 1.593344807624817,
"learning_rate": 0.001,
"loss": 0.7913,
"step": 294400
},
{
"epoch": 95.18422753716871,
"grad_norm": 2.5860588550567627,
"learning_rate": 0.001,
"loss": 0.7833,
"step": 294500
},
{
"epoch": 95.21654815772463,
"grad_norm": 2.560615301132202,
"learning_rate": 0.001,
"loss": 0.7896,
"step": 294600
},
{
"epoch": 95.24886877828054,
"grad_norm": 2.0831151008605957,
"learning_rate": 0.001,
"loss": 0.7706,
"step": 294700
},
{
"epoch": 95.28118939883646,
"grad_norm": 1.564112663269043,
"learning_rate": 0.001,
"loss": 0.7903,
"step": 294800
},
{
"epoch": 95.31351001939237,
"grad_norm": 1.4641203880310059,
"learning_rate": 0.001,
"loss": 0.8028,
"step": 294900
},
{
"epoch": 95.34583063994829,
"grad_norm": 1.5678726434707642,
"learning_rate": 0.001,
"loss": 0.7956,
"step": 295000
},
{
"epoch": 95.3781512605042,
"grad_norm": 2.574636220932007,
"learning_rate": 0.001,
"loss": 0.8273,
"step": 295100
},
{
"epoch": 95.41047188106012,
"grad_norm": 2.0223448276519775,
"learning_rate": 0.001,
"loss": 0.7968,
"step": 295200
},
{
"epoch": 95.44279250161603,
"grad_norm": 1.6964439153671265,
"learning_rate": 0.001,
"loss": 0.8138,
"step": 295300
},
{
"epoch": 95.47511312217195,
"grad_norm": 1.6102707386016846,
"learning_rate": 0.001,
"loss": 0.8149,
"step": 295400
},
{
"epoch": 95.50743374272786,
"grad_norm": 1.607711672782898,
"learning_rate": 0.001,
"loss": 0.8176,
"step": 295500
},
{
"epoch": 95.53975436328378,
"grad_norm": 1.8629106283187866,
"learning_rate": 0.001,
"loss": 0.8331,
"step": 295600
},
{
"epoch": 95.57207498383968,
"grad_norm": 1.388291358947754,
"learning_rate": 0.001,
"loss": 0.8185,
"step": 295700
},
{
"epoch": 95.6043956043956,
"grad_norm": 2.2941222190856934,
"learning_rate": 0.001,
"loss": 0.8138,
"step": 295800
},
{
"epoch": 95.63671622495151,
"grad_norm": 2.3601090908050537,
"learning_rate": 0.001,
"loss": 0.8259,
"step": 295900
},
{
"epoch": 95.66903684550743,
"grad_norm": 1.7740150690078735,
"learning_rate": 0.001,
"loss": 0.812,
"step": 296000
},
{
"epoch": 95.70135746606334,
"grad_norm": 1.7329515218734741,
"learning_rate": 0.001,
"loss": 0.8219,
"step": 296100
},
{
"epoch": 95.73367808661926,
"grad_norm": 1.57270348072052,
"learning_rate": 0.001,
"loss": 0.8476,
"step": 296200
},
{
"epoch": 95.76599870717517,
"grad_norm": 1.5133745670318604,
"learning_rate": 0.001,
"loss": 0.8147,
"step": 296300
},
{
"epoch": 95.7983193277311,
"grad_norm": 1.8256583213806152,
"learning_rate": 0.001,
"loss": 0.8433,
"step": 296400
},
{
"epoch": 95.830639948287,
"grad_norm": 1.820281744003296,
"learning_rate": 0.001,
"loss": 0.8387,
"step": 296500
},
{
"epoch": 95.86296056884292,
"grad_norm": 1.661638617515564,
"learning_rate": 0.001,
"loss": 0.8386,
"step": 296600
},
{
"epoch": 95.89528118939883,
"grad_norm": 1.9161064624786377,
"learning_rate": 0.001,
"loss": 0.8376,
"step": 296700
},
{
"epoch": 95.92760180995475,
"grad_norm": 155.10792541503906,
"learning_rate": 0.001,
"loss": 0.8513,
"step": 296800
},
{
"epoch": 95.95992243051066,
"grad_norm": 1.6141908168792725,
"learning_rate": 0.001,
"loss": 0.8456,
"step": 296900
},
{
"epoch": 95.99224305106658,
"grad_norm": 1.7733116149902344,
"learning_rate": 0.001,
"loss": 0.8441,
"step": 297000
},
{
"epoch": 96.0245636716225,
"grad_norm": 1.8193919658660889,
"learning_rate": 0.001,
"loss": 0.8051,
"step": 297100
},
{
"epoch": 96.05688429217841,
"grad_norm": 1.5573382377624512,
"learning_rate": 0.001,
"loss": 0.7617,
"step": 297200
},
{
"epoch": 96.08920491273433,
"grad_norm": 1.5014259815216064,
"learning_rate": 0.001,
"loss": 0.7728,
"step": 297300
},
{
"epoch": 96.12152553329024,
"grad_norm": 1.7022926807403564,
"learning_rate": 0.001,
"loss": 0.7714,
"step": 297400
},
{
"epoch": 96.15384615384616,
"grad_norm": 1.7341800928115845,
"learning_rate": 0.001,
"loss": 0.7813,
"step": 297500
},
{
"epoch": 96.18616677440207,
"grad_norm": 1.7341227531433105,
"learning_rate": 0.001,
"loss": 0.7863,
"step": 297600
},
{
"epoch": 96.21848739495799,
"grad_norm": 1.5951087474822998,
"learning_rate": 0.001,
"loss": 0.7917,
"step": 297700
},
{
"epoch": 96.2508080155139,
"grad_norm": 1.7377108335494995,
"learning_rate": 0.001,
"loss": 0.787,
"step": 297800
},
{
"epoch": 96.28312863606982,
"grad_norm": 1.6977343559265137,
"learning_rate": 0.001,
"loss": 0.7687,
"step": 297900
},
{
"epoch": 96.31544925662573,
"grad_norm": 1.503747820854187,
"learning_rate": 0.001,
"loss": 0.7973,
"step": 298000
},
{
"epoch": 96.34776987718165,
"grad_norm": 1.4911261796951294,
"learning_rate": 0.001,
"loss": 0.8041,
"step": 298100
},
{
"epoch": 96.38009049773756,
"grad_norm": 1.8095695972442627,
"learning_rate": 0.001,
"loss": 0.8138,
"step": 298200
},
{
"epoch": 96.41241111829348,
"grad_norm": 1.5566781759262085,
"learning_rate": 0.001,
"loss": 0.7972,
"step": 298300
},
{
"epoch": 96.44473173884938,
"grad_norm": 1.5418801307678223,
"learning_rate": 0.001,
"loss": 0.8008,
"step": 298400
},
{
"epoch": 96.4770523594053,
"grad_norm": 1.9948352575302124,
"learning_rate": 0.001,
"loss": 0.8103,
"step": 298500
},
{
"epoch": 96.50937297996121,
"grad_norm": 2.468416213989258,
"learning_rate": 0.001,
"loss": 0.8254,
"step": 298600
},
{
"epoch": 96.54169360051714,
"grad_norm": 1.474953293800354,
"learning_rate": 0.001,
"loss": 0.8181,
"step": 298700
},
{
"epoch": 96.57401422107304,
"grad_norm": 1.533799409866333,
"learning_rate": 0.001,
"loss": 0.8145,
"step": 298800
},
{
"epoch": 96.60633484162896,
"grad_norm": 1.4074748754501343,
"learning_rate": 0.001,
"loss": 0.8215,
"step": 298900
},
{
"epoch": 96.63865546218487,
"grad_norm": 1.8802385330200195,
"learning_rate": 0.001,
"loss": 0.803,
"step": 299000
},
{
"epoch": 96.6709760827408,
"grad_norm": 1.875154972076416,
"learning_rate": 0.001,
"loss": 0.8164,
"step": 299100
},
{
"epoch": 96.7032967032967,
"grad_norm": 1.6148853302001953,
"learning_rate": 0.001,
"loss": 0.8259,
"step": 299200
},
{
"epoch": 96.73561732385262,
"grad_norm": 1.7623929977416992,
"learning_rate": 0.001,
"loss": 0.825,
"step": 299300
},
{
"epoch": 96.76793794440853,
"grad_norm": 1.9254313707351685,
"learning_rate": 0.001,
"loss": 0.8376,
"step": 299400
},
{
"epoch": 96.80025856496445,
"grad_norm": 1.6274827718734741,
"learning_rate": 0.001,
"loss": 0.8356,
"step": 299500
},
{
"epoch": 96.83257918552036,
"grad_norm": 1.6711455583572388,
"learning_rate": 0.001,
"loss": 0.8385,
"step": 299600
},
{
"epoch": 96.86489980607628,
"grad_norm": 2.029143810272217,
"learning_rate": 0.001,
"loss": 0.8333,
"step": 299700
},
{
"epoch": 96.89722042663219,
"grad_norm": 1.801000714302063,
"learning_rate": 0.001,
"loss": 0.8393,
"step": 299800
},
{
"epoch": 96.92954104718811,
"grad_norm": 1.6309932470321655,
"learning_rate": 0.001,
"loss": 0.8271,
"step": 299900
},
{
"epoch": 96.96186166774402,
"grad_norm": 1.8438305854797363,
"learning_rate": 0.001,
"loss": 0.8348,
"step": 300000
},
{
"epoch": 96.99418228829994,
"grad_norm": 1.7183477878570557,
"learning_rate": 0.001,
"loss": 0.8257,
"step": 300100
},
{
"epoch": 97.02650290885585,
"grad_norm": 1.5698041915893555,
"learning_rate": 0.001,
"loss": 0.7702,
"step": 300200
},
{
"epoch": 97.05882352941177,
"grad_norm": 1.6669820547103882,
"learning_rate": 0.001,
"loss": 0.7739,
"step": 300300
},
{
"epoch": 97.09114414996768,
"grad_norm": 1.6277503967285156,
"learning_rate": 0.001,
"loss": 0.7652,
"step": 300400
},
{
"epoch": 97.1234647705236,
"grad_norm": 1.3346500396728516,
"learning_rate": 0.001,
"loss": 0.7631,
"step": 300500
},
{
"epoch": 97.1557853910795,
"grad_norm": 1.852993130683899,
"learning_rate": 0.001,
"loss": 0.7823,
"step": 300600
},
{
"epoch": 97.18810601163543,
"grad_norm": 1.7382128238677979,
"learning_rate": 0.001,
"loss": 0.7705,
"step": 300700
},
{
"epoch": 97.22042663219133,
"grad_norm": 1.5095046758651733,
"learning_rate": 0.001,
"loss": 0.7798,
"step": 300800
},
{
"epoch": 97.25274725274726,
"grad_norm": 1.847023606300354,
"learning_rate": 0.001,
"loss": 0.7838,
"step": 300900
},
{
"epoch": 97.28506787330316,
"grad_norm": 1.7964493036270142,
"learning_rate": 0.001,
"loss": 0.7896,
"step": 301000
},
{
"epoch": 97.31738849385908,
"grad_norm": 2.9618778228759766,
"learning_rate": 0.001,
"loss": 0.7969,
"step": 301100
},
{
"epoch": 97.34970911441499,
"grad_norm": 1.6932001113891602,
"learning_rate": 0.001,
"loss": 0.795,
"step": 301200
},
{
"epoch": 97.38202973497091,
"grad_norm": 1.504896879196167,
"learning_rate": 0.001,
"loss": 0.794,
"step": 301300
},
{
"epoch": 97.41435035552682,
"grad_norm": 1.857369065284729,
"learning_rate": 0.001,
"loss": 0.7876,
"step": 301400
},
{
"epoch": 97.44667097608274,
"grad_norm": 1.5941156148910522,
"learning_rate": 0.001,
"loss": 0.8076,
"step": 301500
},
{
"epoch": 97.47899159663865,
"grad_norm": 2.0202865600585938,
"learning_rate": 0.001,
"loss": 0.8003,
"step": 301600
},
{
"epoch": 97.51131221719457,
"grad_norm": 1.77725088596344,
"learning_rate": 0.001,
"loss": 0.8186,
"step": 301700
},
{
"epoch": 97.54363283775048,
"grad_norm": 1.4381173849105835,
"learning_rate": 0.001,
"loss": 0.8083,
"step": 301800
},
{
"epoch": 97.5759534583064,
"grad_norm": 1.4481972455978394,
"learning_rate": 0.001,
"loss": 0.8015,
"step": 301900
},
{
"epoch": 97.60827407886231,
"grad_norm": 2.0220069885253906,
"learning_rate": 0.001,
"loss": 0.8175,
"step": 302000
},
{
"epoch": 97.64059469941823,
"grad_norm": 1.730670690536499,
"learning_rate": 0.001,
"loss": 0.8144,
"step": 302100
},
{
"epoch": 97.67291531997414,
"grad_norm": 1.6534379720687866,
"learning_rate": 0.001,
"loss": 0.8065,
"step": 302200
},
{
"epoch": 97.70523594053006,
"grad_norm": 1.6358345746994019,
"learning_rate": 0.001,
"loss": 0.8051,
"step": 302300
},
{
"epoch": 97.73755656108597,
"grad_norm": 1.6618595123291016,
"learning_rate": 0.001,
"loss": 0.8125,
"step": 302400
},
{
"epoch": 97.76987718164189,
"grad_norm": 1.5252058506011963,
"learning_rate": 0.001,
"loss": 0.8192,
"step": 302500
},
{
"epoch": 97.8021978021978,
"grad_norm": 1.7979437112808228,
"learning_rate": 0.001,
"loss": 0.8292,
"step": 302600
},
{
"epoch": 97.83451842275372,
"grad_norm": 1.6801445484161377,
"learning_rate": 0.001,
"loss": 0.8272,
"step": 302700
},
{
"epoch": 97.86683904330962,
"grad_norm": 1.5387403964996338,
"learning_rate": 0.001,
"loss": 0.8236,
"step": 302800
},
{
"epoch": 97.89915966386555,
"grad_norm": 1.9808905124664307,
"learning_rate": 0.001,
"loss": 0.8208,
"step": 302900
},
{
"epoch": 97.93148028442145,
"grad_norm": 1.9140608310699463,
"learning_rate": 0.001,
"loss": 0.8367,
"step": 303000
},
{
"epoch": 97.96380090497738,
"grad_norm": 2.240133762359619,
"learning_rate": 0.001,
"loss": 0.8299,
"step": 303100
},
{
"epoch": 97.99612152553328,
"grad_norm": 2.0139851570129395,
"learning_rate": 0.001,
"loss": 0.8189,
"step": 303200
},
{
"epoch": 98.0284421460892,
"grad_norm": 1.701872706413269,
"learning_rate": 0.001,
"loss": 0.7438,
"step": 303300
},
{
"epoch": 98.06076276664513,
"grad_norm": 1.5965427160263062,
"learning_rate": 0.001,
"loss": 0.7551,
"step": 303400
},
{
"epoch": 98.09308338720103,
"grad_norm": 1.8786427974700928,
"learning_rate": 0.001,
"loss": 0.766,
"step": 303500
},
{
"epoch": 98.12540400775696,
"grad_norm": 1.8245842456817627,
"learning_rate": 0.001,
"loss": 0.7673,
"step": 303600
},
{
"epoch": 98.15772462831286,
"grad_norm": 2.078568935394287,
"learning_rate": 0.001,
"loss": 0.7756,
"step": 303700
},
{
"epoch": 98.19004524886878,
"grad_norm": 1.7483006715774536,
"learning_rate": 0.001,
"loss": 0.7743,
"step": 303800
},
{
"epoch": 98.22236586942469,
"grad_norm": 1.639719843864441,
"learning_rate": 0.001,
"loss": 0.7776,
"step": 303900
},
{
"epoch": 98.25468648998061,
"grad_norm": 1.6719263792037964,
"learning_rate": 0.001,
"loss": 0.7783,
"step": 304000
},
{
"epoch": 98.28700711053652,
"grad_norm": 1.8221156597137451,
"learning_rate": 0.001,
"loss": 0.7849,
"step": 304100
},
{
"epoch": 98.31932773109244,
"grad_norm": 1.5369601249694824,
"learning_rate": 0.001,
"loss": 0.783,
"step": 304200
},
{
"epoch": 98.35164835164835,
"grad_norm": 1.405853033065796,
"learning_rate": 0.001,
"loss": 0.7821,
"step": 304300
},
{
"epoch": 98.38396897220427,
"grad_norm": 2.133615255355835,
"learning_rate": 0.001,
"loss": 0.7812,
"step": 304400
},
{
"epoch": 98.41628959276018,
"grad_norm": 1.5420911312103271,
"learning_rate": 0.001,
"loss": 0.7965,
"step": 304500
},
{
"epoch": 98.4486102133161,
"grad_norm": 1.6347417831420898,
"learning_rate": 0.001,
"loss": 0.7933,
"step": 304600
},
{
"epoch": 98.48093083387201,
"grad_norm": 1.4630227088928223,
"learning_rate": 0.001,
"loss": 0.7793,
"step": 304700
},
{
"epoch": 98.51325145442793,
"grad_norm": 1.563585638999939,
"learning_rate": 0.001,
"loss": 0.7818,
"step": 304800
},
{
"epoch": 98.54557207498384,
"grad_norm": 1.7685762643814087,
"learning_rate": 0.001,
"loss": 0.8194,
"step": 304900
},
{
"epoch": 98.57789269553976,
"grad_norm": 1.683583378791809,
"learning_rate": 0.001,
"loss": 0.8049,
"step": 305000
},
{
"epoch": 98.61021331609567,
"grad_norm": 1.280974268913269,
"learning_rate": 0.001,
"loss": 0.8025,
"step": 305100
},
{
"epoch": 98.64253393665159,
"grad_norm": 1.4059722423553467,
"learning_rate": 0.001,
"loss": 0.81,
"step": 305200
},
{
"epoch": 98.6748545572075,
"grad_norm": 1.575561761856079,
"learning_rate": 0.001,
"loss": 0.8082,
"step": 305300
},
{
"epoch": 98.70717517776342,
"grad_norm": 2.3904173374176025,
"learning_rate": 0.001,
"loss": 0.8047,
"step": 305400
},
{
"epoch": 98.73949579831933,
"grad_norm": 2.197852849960327,
"learning_rate": 0.001,
"loss": 0.8068,
"step": 305500
},
{
"epoch": 98.77181641887525,
"grad_norm": 1.8822154998779297,
"learning_rate": 0.001,
"loss": 0.8173,
"step": 305600
},
{
"epoch": 98.80413703943115,
"grad_norm": 1.5784273147583008,
"learning_rate": 0.001,
"loss": 0.8124,
"step": 305700
},
{
"epoch": 98.83645765998708,
"grad_norm": 1.572724461555481,
"learning_rate": 0.001,
"loss": 0.81,
"step": 305800
},
{
"epoch": 98.86877828054298,
"grad_norm": 1.875663161277771,
"learning_rate": 0.001,
"loss": 0.8151,
"step": 305900
},
{
"epoch": 98.9010989010989,
"grad_norm": 3.32124662399292,
"learning_rate": 0.001,
"loss": 0.809,
"step": 306000
},
{
"epoch": 98.93341952165481,
"grad_norm": 2.008251190185547,
"learning_rate": 0.001,
"loss": 0.825,
"step": 306100
},
{
"epoch": 98.96574014221073,
"grad_norm": 14.357972145080566,
"learning_rate": 0.001,
"loss": 0.8296,
"step": 306200
},
{
"epoch": 98.99806076276664,
"grad_norm": 1.7346726655960083,
"learning_rate": 0.001,
"loss": 0.8113,
"step": 306300
},
{
"epoch": 99.03038138332256,
"grad_norm": 1.5255569219589233,
"learning_rate": 0.001,
"loss": 0.753,
"step": 306400
},
{
"epoch": 99.06270200387847,
"grad_norm": 1.8542404174804688,
"learning_rate": 0.001,
"loss": 0.7537,
"step": 306500
},
{
"epoch": 99.09502262443439,
"grad_norm": 1.9389482736587524,
"learning_rate": 0.001,
"loss": 0.7604,
"step": 306600
},
{
"epoch": 99.1273432449903,
"grad_norm": 1.9633482694625854,
"learning_rate": 0.001,
"loss": 0.7669,
"step": 306700
},
{
"epoch": 99.15966386554622,
"grad_norm": 1.3970327377319336,
"learning_rate": 0.001,
"loss": 0.7502,
"step": 306800
},
{
"epoch": 99.19198448610213,
"grad_norm": 1.400421380996704,
"learning_rate": 0.001,
"loss": 0.7697,
"step": 306900
},
{
"epoch": 99.22430510665805,
"grad_norm": 2.0532190799713135,
"learning_rate": 0.001,
"loss": 0.7604,
"step": 307000
},
{
"epoch": 99.25662572721396,
"grad_norm": 1.6515733003616333,
"learning_rate": 0.001,
"loss": 0.7727,
"step": 307100
},
{
"epoch": 99.28894634776988,
"grad_norm": 1.4215999841690063,
"learning_rate": 0.001,
"loss": 0.7935,
"step": 307200
},
{
"epoch": 99.32126696832579,
"grad_norm": 1.2796798944473267,
"learning_rate": 0.001,
"loss": 0.783,
"step": 307300
},
{
"epoch": 99.35358758888171,
"grad_norm": 1.8550664186477661,
"learning_rate": 0.001,
"loss": 0.7806,
"step": 307400
},
{
"epoch": 99.38590820943762,
"grad_norm": 1.508309245109558,
"learning_rate": 0.001,
"loss": 0.7955,
"step": 307500
},
{
"epoch": 99.41822882999354,
"grad_norm": 5.713343620300293,
"learning_rate": 0.001,
"loss": 0.7914,
"step": 307600
},
{
"epoch": 99.45054945054945,
"grad_norm": 1.6815954446792603,
"learning_rate": 0.001,
"loss": 0.788,
"step": 307700
},
{
"epoch": 99.48287007110537,
"grad_norm": 1.4613316059112549,
"learning_rate": 0.001,
"loss": 0.7896,
"step": 307800
},
{
"epoch": 99.51519069166127,
"grad_norm": 1.8426566123962402,
"learning_rate": 0.001,
"loss": 0.8139,
"step": 307900
},
{
"epoch": 99.5475113122172,
"grad_norm": 1.4660000801086426,
"learning_rate": 0.001,
"loss": 0.795,
"step": 308000
},
{
"epoch": 99.5798319327731,
"grad_norm": 1.5318522453308105,
"learning_rate": 0.001,
"loss": 0.8095,
"step": 308100
},
{
"epoch": 99.61215255332903,
"grad_norm": 1.5603666305541992,
"learning_rate": 0.001,
"loss": 0.8079,
"step": 308200
},
{
"epoch": 99.64447317388493,
"grad_norm": 1.7400113344192505,
"learning_rate": 0.001,
"loss": 0.8057,
"step": 308300
},
{
"epoch": 99.67679379444085,
"grad_norm": 1.3645039796829224,
"learning_rate": 0.001,
"loss": 0.8052,
"step": 308400
},
{
"epoch": 99.70911441499676,
"grad_norm": 1.9990870952606201,
"learning_rate": 0.001,
"loss": 0.8038,
"step": 308500
},
{
"epoch": 99.74143503555268,
"grad_norm": 1.7578907012939453,
"learning_rate": 0.001,
"loss": 0.7997,
"step": 308600
},
{
"epoch": 99.77375565610859,
"grad_norm": 1.4323079586029053,
"learning_rate": 0.001,
"loss": 0.8035,
"step": 308700
},
{
"epoch": 99.80607627666451,
"grad_norm": 1.3833441734313965,
"learning_rate": 0.001,
"loss": 0.7939,
"step": 308800
},
{
"epoch": 99.83839689722042,
"grad_norm": 1.8678525686264038,
"learning_rate": 0.001,
"loss": 0.8105,
"step": 308900
},
{
"epoch": 99.87071751777634,
"grad_norm": 1.543296456336975,
"learning_rate": 0.001,
"loss": 0.8166,
"step": 309000
},
{
"epoch": 99.90303813833225,
"grad_norm": 1.2817319631576538,
"learning_rate": 0.001,
"loss": 0.8154,
"step": 309100
},
{
"epoch": 99.93535875888817,
"grad_norm": 1.460585594177246,
"learning_rate": 0.001,
"loss": 0.8102,
"step": 309200
},
{
"epoch": 99.96767937944408,
"grad_norm": 1.6103432178497314,
"learning_rate": 0.001,
"loss": 0.8096,
"step": 309300
},
{
"epoch": 100.0,
"grad_norm": 0.6271578073501587,
"learning_rate": 0.001,
"loss": 0.7634,
"step": 309400
},
{
"epoch": 100.0,
"step": 309400,
"total_flos": 3.297635183404155e+17,
"train_loss": 0.3764197817446728,
"train_runtime": 18030.0585,
"train_samples_per_second": 549.083,
"train_steps_per_second": 17.16
}
],
"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": 3.297635183404155e+17,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}