AusmitM's picture
Upload 31 files
c17f472 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": 336.53814697265625,
"learning_rate": 9.900000000000002e-06,
"loss": 4.9566,
"step": 100
},
{
"epoch": 0.06464124111182935,
"grad_norm": 167.75621032714844,
"learning_rate": 1.9900000000000003e-05,
"loss": 4.3463,
"step": 200
},
{
"epoch": 0.09696186166774402,
"grad_norm": 16.510164260864258,
"learning_rate": 2.9900000000000002e-05,
"loss": 4.1394,
"step": 300
},
{
"epoch": 0.1292824822236587,
"grad_norm": 6.610813617706299,
"learning_rate": 3.99e-05,
"loss": 4.0682,
"step": 400
},
{
"epoch": 0.16160310277957338,
"grad_norm": 2.8748035430908203,
"learning_rate": 4.99e-05,
"loss": 4.0366,
"step": 500
},
{
"epoch": 0.19392372333548805,
"grad_norm": 10.500384330749512,
"learning_rate": 5.9900000000000006e-05,
"loss": 3.9721,
"step": 600
},
{
"epoch": 0.22624434389140272,
"grad_norm": 3.2947850227355957,
"learning_rate": 6.99e-05,
"loss": 3.9745,
"step": 700
},
{
"epoch": 0.2585649644473174,
"grad_norm": 10.11046314239502,
"learning_rate": 7.99e-05,
"loss": 3.9394,
"step": 800
},
{
"epoch": 0.2908855850032321,
"grad_norm": 6.4788031578063965,
"learning_rate": 8.989999999999999e-05,
"loss": 3.9099,
"step": 900
},
{
"epoch": 0.32320620555914675,
"grad_norm": 4.389022350311279,
"learning_rate": 9.99e-05,
"loss": 3.8901,
"step": 1000
},
{
"epoch": 0.3555268261150614,
"grad_norm": 9.506305694580078,
"learning_rate": 0.0001099,
"loss": 3.8477,
"step": 1100
},
{
"epoch": 0.3878474466709761,
"grad_norm": 2.450131416320801,
"learning_rate": 0.00011990000000000001,
"loss": 3.8067,
"step": 1200
},
{
"epoch": 0.42016806722689076,
"grad_norm": 2.8842737674713135,
"learning_rate": 0.00012989999999999999,
"loss": 3.8179,
"step": 1300
},
{
"epoch": 0.45248868778280543,
"grad_norm": 2.5014400482177734,
"learning_rate": 0.0001399,
"loss": 3.794,
"step": 1400
},
{
"epoch": 0.4848093083387201,
"grad_norm": 2.4967105388641357,
"learning_rate": 0.0001499,
"loss": 3.7666,
"step": 1500
},
{
"epoch": 0.5171299288946348,
"grad_norm": 2.405489683151245,
"learning_rate": 0.00015989999999999998,
"loss": 3.7654,
"step": 1600
},
{
"epoch": 0.5494505494505495,
"grad_norm": 2.7155001163482666,
"learning_rate": 0.0001699,
"loss": 3.7365,
"step": 1700
},
{
"epoch": 0.5817711700064642,
"grad_norm": 2.614349365234375,
"learning_rate": 0.0001799,
"loss": 3.7264,
"step": 1800
},
{
"epoch": 0.6140917905623788,
"grad_norm": 2.1537041664123535,
"learning_rate": 0.0001899,
"loss": 3.7404,
"step": 1900
},
{
"epoch": 0.6464124111182935,
"grad_norm": 2.622662305831909,
"learning_rate": 0.0001999,
"loss": 3.6887,
"step": 2000
},
{
"epoch": 0.6787330316742082,
"grad_norm": 2.4469103813171387,
"learning_rate": 0.0002099,
"loss": 3.68,
"step": 2100
},
{
"epoch": 0.7110536522301228,
"grad_norm": 3.942584991455078,
"learning_rate": 0.0002199,
"loss": 3.6882,
"step": 2200
},
{
"epoch": 0.7433742727860375,
"grad_norm": 2.064706325531006,
"learning_rate": 0.0002299,
"loss": 3.6612,
"step": 2300
},
{
"epoch": 0.7756948933419522,
"grad_norm": 2.058544874191284,
"learning_rate": 0.0002399,
"loss": 3.6678,
"step": 2400
},
{
"epoch": 0.8080155138978669,
"grad_norm": 1.8396873474121094,
"learning_rate": 0.0002499,
"loss": 3.6486,
"step": 2500
},
{
"epoch": 0.8403361344537815,
"grad_norm": 2.7682371139526367,
"learning_rate": 0.00025990000000000003,
"loss": 3.6495,
"step": 2600
},
{
"epoch": 0.8726567550096962,
"grad_norm": 2.4739015102386475,
"learning_rate": 0.0002699,
"loss": 3.6604,
"step": 2700
},
{
"epoch": 0.9049773755656109,
"grad_norm": 3.238330125808716,
"learning_rate": 0.0002799,
"loss": 3.6534,
"step": 2800
},
{
"epoch": 0.9372979961215255,
"grad_norm": 2.4490232467651367,
"learning_rate": 0.0002899,
"loss": 3.6253,
"step": 2900
},
{
"epoch": 0.9696186166774402,
"grad_norm": 2.2283108234405518,
"learning_rate": 0.0002999,
"loss": 3.616,
"step": 3000
},
{
"epoch": 1.0019392372333549,
"grad_norm": 1.6477761268615723,
"learning_rate": 0.0003099,
"loss": 3.626,
"step": 3100
},
{
"epoch": 1.0342598577892697,
"grad_norm": 6.880061626434326,
"learning_rate": 0.0003199,
"loss": 3.5737,
"step": 3200
},
{
"epoch": 1.0665804783451842,
"grad_norm": 1.3400031328201294,
"learning_rate": 0.00032990000000000005,
"loss": 3.5692,
"step": 3300
},
{
"epoch": 1.098901098901099,
"grad_norm": 1.1662441492080688,
"learning_rate": 0.00033989999999999997,
"loss": 3.5527,
"step": 3400
},
{
"epoch": 1.1312217194570136,
"grad_norm": 1.404117465019226,
"learning_rate": 0.0003499,
"loss": 3.5722,
"step": 3500
},
{
"epoch": 1.1635423400129283,
"grad_norm": 1.5539703369140625,
"learning_rate": 0.0003599,
"loss": 3.5598,
"step": 3600
},
{
"epoch": 1.195862960568843,
"grad_norm": 1.4064580202102661,
"learning_rate": 0.0003699,
"loss": 3.5511,
"step": 3700
},
{
"epoch": 1.2281835811247577,
"grad_norm": 1.407173752784729,
"learning_rate": 0.0003799,
"loss": 3.5479,
"step": 3800
},
{
"epoch": 1.2605042016806722,
"grad_norm": 1.0030317306518555,
"learning_rate": 0.00038990000000000004,
"loss": 3.5413,
"step": 3900
},
{
"epoch": 1.292824822236587,
"grad_norm": 1.1430760622024536,
"learning_rate": 0.00039989999999999996,
"loss": 3.5226,
"step": 4000
},
{
"epoch": 1.3251454427925016,
"grad_norm": 1.2244195938110352,
"learning_rate": 0.0004099,
"loss": 3.5279,
"step": 4100
},
{
"epoch": 1.3574660633484164,
"grad_norm": 1.249206304550171,
"learning_rate": 0.0004199,
"loss": 3.5319,
"step": 4200
},
{
"epoch": 1.389786683904331,
"grad_norm": 1.2100669145584106,
"learning_rate": 0.0004299,
"loss": 3.5284,
"step": 4300
},
{
"epoch": 1.4221073044602457,
"grad_norm": 1.4405732154846191,
"learning_rate": 0.0004399,
"loss": 3.5119,
"step": 4400
},
{
"epoch": 1.4544279250161603,
"grad_norm": 3.148793935775757,
"learning_rate": 0.00044990000000000004,
"loss": 3.5069,
"step": 4500
},
{
"epoch": 1.486748545572075,
"grad_norm": 1.1930956840515137,
"learning_rate": 0.0004599,
"loss": 3.5064,
"step": 4600
},
{
"epoch": 1.5190691661279896,
"grad_norm": 1.3880915641784668,
"learning_rate": 0.0004699,
"loss": 3.5099,
"step": 4700
},
{
"epoch": 1.5513897866839044,
"grad_norm": 1.3251196146011353,
"learning_rate": 0.0004799,
"loss": 3.5005,
"step": 4800
},
{
"epoch": 1.5837104072398192,
"grad_norm": 0.9314677119255066,
"learning_rate": 0.0004899,
"loss": 3.4933,
"step": 4900
},
{
"epoch": 1.6160310277957337,
"grad_norm": 1.07969069480896,
"learning_rate": 0.0004999000000000001,
"loss": 3.47,
"step": 5000
},
{
"epoch": 1.6483516483516483,
"grad_norm": 1.0096684694290161,
"learning_rate": 0.0005099,
"loss": 3.4935,
"step": 5100
},
{
"epoch": 1.680672268907563,
"grad_norm": 0.885674238204956,
"learning_rate": 0.0005199,
"loss": 3.469,
"step": 5200
},
{
"epoch": 1.7129928894634778,
"grad_norm": 1.0472047328948975,
"learning_rate": 0.0005299,
"loss": 3.477,
"step": 5300
},
{
"epoch": 1.7453135100193924,
"grad_norm": 0.9959677457809448,
"learning_rate": 0.0005399000000000001,
"loss": 3.4711,
"step": 5400
},
{
"epoch": 1.777634130575307,
"grad_norm": 0.9637390375137329,
"learning_rate": 0.0005499000000000001,
"loss": 3.4786,
"step": 5500
},
{
"epoch": 1.8099547511312217,
"grad_norm": 0.9888423681259155,
"learning_rate": 0.0005599,
"loss": 3.4616,
"step": 5600
},
{
"epoch": 1.8422753716871365,
"grad_norm": 1.0575670003890991,
"learning_rate": 0.0005698999999999999,
"loss": 3.4529,
"step": 5700
},
{
"epoch": 1.874595992243051,
"grad_norm": 0.9052221179008484,
"learning_rate": 0.0005799,
"loss": 3.44,
"step": 5800
},
{
"epoch": 1.9069166127989656,
"grad_norm": 1.0884742736816406,
"learning_rate": 0.0005899,
"loss": 3.4564,
"step": 5900
},
{
"epoch": 1.9392372333548804,
"grad_norm": 1.5821954011917114,
"learning_rate": 0.0005999,
"loss": 3.4485,
"step": 6000
},
{
"epoch": 1.9715578539107952,
"grad_norm": 1.2174369096755981,
"learning_rate": 0.0006099,
"loss": 3.4562,
"step": 6100
},
{
"epoch": 2.0038784744667097,
"grad_norm": 1.0810948610305786,
"learning_rate": 0.0006199,
"loss": 3.4583,
"step": 6200
},
{
"epoch": 2.0361990950226243,
"grad_norm": 0.825448215007782,
"learning_rate": 0.0006299000000000001,
"loss": 3.357,
"step": 6300
},
{
"epoch": 2.0685197155785393,
"grad_norm": 1.1348732709884644,
"learning_rate": 0.0006399,
"loss": 3.375,
"step": 6400
},
{
"epoch": 2.100840336134454,
"grad_norm": 1.0036207437515259,
"learning_rate": 0.0006499,
"loss": 3.3689,
"step": 6500
},
{
"epoch": 2.1331609566903684,
"grad_norm": 1.0743062496185303,
"learning_rate": 0.0006599,
"loss": 3.3407,
"step": 6600
},
{
"epoch": 2.165481577246283,
"grad_norm": 1.4043335914611816,
"learning_rate": 0.0006699000000000001,
"loss": 3.3826,
"step": 6700
},
{
"epoch": 2.197802197802198,
"grad_norm": 0.9674347639083862,
"learning_rate": 0.0006799,
"loss": 3.3838,
"step": 6800
},
{
"epoch": 2.2301228183581125,
"grad_norm": 0.8897255063056946,
"learning_rate": 0.0006899,
"loss": 3.3658,
"step": 6900
},
{
"epoch": 2.262443438914027,
"grad_norm": 1.0808757543563843,
"learning_rate": 0.0006998999999999999,
"loss": 3.3755,
"step": 7000
},
{
"epoch": 2.2947640594699417,
"grad_norm": 1.1210312843322754,
"learning_rate": 0.0007099,
"loss": 3.3559,
"step": 7100
},
{
"epoch": 2.3270846800258567,
"grad_norm": 1.207525610923767,
"learning_rate": 0.0007199,
"loss": 3.3618,
"step": 7200
},
{
"epoch": 2.3594053005817712,
"grad_norm": 1.0012303590774536,
"learning_rate": 0.0007299,
"loss": 3.3726,
"step": 7300
},
{
"epoch": 2.391725921137686,
"grad_norm": 0.9295409917831421,
"learning_rate": 0.0007399,
"loss": 3.3754,
"step": 7400
},
{
"epoch": 2.4240465416936003,
"grad_norm": 1.305206537246704,
"learning_rate": 0.0007499000000000001,
"loss": 3.3728,
"step": 7500
},
{
"epoch": 2.4563671622495153,
"grad_norm": 0.9094693064689636,
"learning_rate": 0.0007599,
"loss": 3.3555,
"step": 7600
},
{
"epoch": 2.48868778280543,
"grad_norm": 1.1543060541152954,
"learning_rate": 0.0007699,
"loss": 3.3486,
"step": 7700
},
{
"epoch": 2.5210084033613445,
"grad_norm": 0.9114538431167603,
"learning_rate": 0.0007799,
"loss": 3.354,
"step": 7800
},
{
"epoch": 2.553329023917259,
"grad_norm": 0.8820834755897522,
"learning_rate": 0.0007899000000000001,
"loss": 3.3641,
"step": 7900
},
{
"epoch": 2.585649644473174,
"grad_norm": 0.9091460108757019,
"learning_rate": 0.0007999000000000001,
"loss": 3.3505,
"step": 8000
},
{
"epoch": 2.6179702650290886,
"grad_norm": 0.914205014705658,
"learning_rate": 0.0008099,
"loss": 3.3469,
"step": 8100
},
{
"epoch": 2.650290885585003,
"grad_norm": 0.9343860149383545,
"learning_rate": 0.0008198999999999999,
"loss": 3.3458,
"step": 8200
},
{
"epoch": 2.682611506140918,
"grad_norm": 0.8911165595054626,
"learning_rate": 0.0008299,
"loss": 3.3592,
"step": 8300
},
{
"epoch": 2.7149321266968327,
"grad_norm": 0.987874448299408,
"learning_rate": 0.0008399,
"loss": 3.3484,
"step": 8400
},
{
"epoch": 2.7472527472527473,
"grad_norm": 0.7139260172843933,
"learning_rate": 0.0008499,
"loss": 3.3369,
"step": 8500
},
{
"epoch": 2.779573367808662,
"grad_norm": 1.1197693347930908,
"learning_rate": 0.0008599,
"loss": 3.3217,
"step": 8600
},
{
"epoch": 2.8118939883645764,
"grad_norm": 0.9437581896781921,
"learning_rate": 0.0008699000000000001,
"loss": 3.3324,
"step": 8700
},
{
"epoch": 2.8442146089204914,
"grad_norm": 0.8579062819480896,
"learning_rate": 0.0008799000000000001,
"loss": 3.3395,
"step": 8800
},
{
"epoch": 2.876535229476406,
"grad_norm": 0.8377363085746765,
"learning_rate": 0.0008899,
"loss": 3.3239,
"step": 8900
},
{
"epoch": 2.9088558500323205,
"grad_norm": 1.500247597694397,
"learning_rate": 0.0008999,
"loss": 3.352,
"step": 9000
},
{
"epoch": 2.9411764705882355,
"grad_norm": 1.1210975646972656,
"learning_rate": 0.0009099,
"loss": 3.3573,
"step": 9100
},
{
"epoch": 2.97349709114415,
"grad_norm": 1.0669493675231934,
"learning_rate": 0.0009199000000000001,
"loss": 3.3113,
"step": 9200
},
{
"epoch": 3.0058177117000646,
"grad_norm": 0.8837397694587708,
"learning_rate": 0.0009299,
"loss": 3.3223,
"step": 9300
},
{
"epoch": 3.038138332255979,
"grad_norm": 1.0612775087356567,
"learning_rate": 0.0009399,
"loss": 3.2188,
"step": 9400
},
{
"epoch": 3.070458952811894,
"grad_norm": 0.9017513394355774,
"learning_rate": 0.0009498999999999999,
"loss": 3.25,
"step": 9500
},
{
"epoch": 3.1027795733678087,
"grad_norm": 0.8819141983985901,
"learning_rate": 0.0009599,
"loss": 3.247,
"step": 9600
},
{
"epoch": 3.1351001939237233,
"grad_norm": 0.8478001952171326,
"learning_rate": 0.0009699,
"loss": 3.2326,
"step": 9700
},
{
"epoch": 3.167420814479638,
"grad_norm": 1.109015703201294,
"learning_rate": 0.0009799,
"loss": 3.2437,
"step": 9800
},
{
"epoch": 3.199741435035553,
"grad_norm": 1.012062907218933,
"learning_rate": 0.0009899,
"loss": 3.2565,
"step": 9900
},
{
"epoch": 3.2320620555914674,
"grad_norm": 0.8651048541069031,
"learning_rate": 0.0009999,
"loss": 3.2521,
"step": 10000
},
{
"epoch": 3.264382676147382,
"grad_norm": 1.0444825887680054,
"learning_rate": 0.001,
"loss": 3.2492,
"step": 10100
},
{
"epoch": 3.2967032967032965,
"grad_norm": 1.1113461256027222,
"learning_rate": 0.001,
"loss": 3.2493,
"step": 10200
},
{
"epoch": 3.3290239172592115,
"grad_norm": 1.1284892559051514,
"learning_rate": 0.001,
"loss": 3.2354,
"step": 10300
},
{
"epoch": 3.361344537815126,
"grad_norm": 1.0857617855072021,
"learning_rate": 0.001,
"loss": 3.2582,
"step": 10400
},
{
"epoch": 3.3936651583710407,
"grad_norm": 1.2226123809814453,
"learning_rate": 0.001,
"loss": 3.2354,
"step": 10500
},
{
"epoch": 3.425985778926955,
"grad_norm": 1.0216729640960693,
"learning_rate": 0.001,
"loss": 3.233,
"step": 10600
},
{
"epoch": 3.45830639948287,
"grad_norm": 0.8395372629165649,
"learning_rate": 0.001,
"loss": 3.2255,
"step": 10700
},
{
"epoch": 3.490627020038785,
"grad_norm": 0.9143357872962952,
"learning_rate": 0.001,
"loss": 3.2337,
"step": 10800
},
{
"epoch": 3.5229476405946993,
"grad_norm": 0.8431721329689026,
"learning_rate": 0.001,
"loss": 3.2581,
"step": 10900
},
{
"epoch": 3.555268261150614,
"grad_norm": 0.8012754917144775,
"learning_rate": 0.001,
"loss": 3.2293,
"step": 11000
},
{
"epoch": 3.587588881706529,
"grad_norm": 1.1557896137237549,
"learning_rate": 0.001,
"loss": 3.2369,
"step": 11100
},
{
"epoch": 3.6199095022624435,
"grad_norm": 0.6755484938621521,
"learning_rate": 0.001,
"loss": 3.2297,
"step": 11200
},
{
"epoch": 3.652230122818358,
"grad_norm": 1.1369458436965942,
"learning_rate": 0.001,
"loss": 3.2337,
"step": 11300
},
{
"epoch": 3.684550743374273,
"grad_norm": 1.2520990371704102,
"learning_rate": 0.001,
"loss": 3.2183,
"step": 11400
},
{
"epoch": 3.7168713639301876,
"grad_norm": 1.36247980594635,
"learning_rate": 0.001,
"loss": 3.2403,
"step": 11500
},
{
"epoch": 3.749191984486102,
"grad_norm": 1.0469595193862915,
"learning_rate": 0.001,
"loss": 3.2342,
"step": 11600
},
{
"epoch": 3.7815126050420167,
"grad_norm": 0.9629628658294678,
"learning_rate": 0.001,
"loss": 3.241,
"step": 11700
},
{
"epoch": 3.8138332255979313,
"grad_norm": 1.02808678150177,
"learning_rate": 0.001,
"loss": 3.2199,
"step": 11800
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.8492690324783325,
"learning_rate": 0.001,
"loss": 3.2259,
"step": 11900
},
{
"epoch": 3.878474466709761,
"grad_norm": 0.9183644652366638,
"learning_rate": 0.001,
"loss": 3.2332,
"step": 12000
},
{
"epoch": 3.9107950872656754,
"grad_norm": 0.8981543779373169,
"learning_rate": 0.001,
"loss": 3.2172,
"step": 12100
},
{
"epoch": 3.9431157078215904,
"grad_norm": 0.7916227579116821,
"learning_rate": 0.001,
"loss": 3.2156,
"step": 12200
},
{
"epoch": 3.975436328377505,
"grad_norm": 1.0210102796554565,
"learning_rate": 0.001,
"loss": 3.218,
"step": 12300
},
{
"epoch": 4.0077569489334195,
"grad_norm": 0.8385574817657471,
"learning_rate": 0.001,
"loss": 3.1961,
"step": 12400
},
{
"epoch": 4.040077569489334,
"grad_norm": 0.8699992895126343,
"learning_rate": 0.001,
"loss": 3.0753,
"step": 12500
},
{
"epoch": 4.072398190045249,
"grad_norm": 0.9136468768119812,
"learning_rate": 0.001,
"loss": 3.0923,
"step": 12600
},
{
"epoch": 4.104718810601163,
"grad_norm": 0.9570497870445251,
"learning_rate": 0.001,
"loss": 3.1073,
"step": 12700
},
{
"epoch": 4.137039431157079,
"grad_norm": 1.0214887857437134,
"learning_rate": 0.001,
"loss": 3.1003,
"step": 12800
},
{
"epoch": 4.169360051712993,
"grad_norm": 0.8436883687973022,
"learning_rate": 0.001,
"loss": 3.1088,
"step": 12900
},
{
"epoch": 4.201680672268908,
"grad_norm": 0.9118756055831909,
"learning_rate": 0.001,
"loss": 3.1061,
"step": 13000
},
{
"epoch": 4.234001292824822,
"grad_norm": 0.9556608200073242,
"learning_rate": 0.001,
"loss": 3.1143,
"step": 13100
},
{
"epoch": 4.266321913380737,
"grad_norm": 1.2082922458648682,
"learning_rate": 0.001,
"loss": 3.1058,
"step": 13200
},
{
"epoch": 4.298642533936651,
"grad_norm": 0.7971954941749573,
"learning_rate": 0.001,
"loss": 3.0948,
"step": 13300
},
{
"epoch": 4.330963154492566,
"grad_norm": 0.8421384692192078,
"learning_rate": 0.001,
"loss": 3.1195,
"step": 13400
},
{
"epoch": 4.3632837750484805,
"grad_norm": 0.9221621155738831,
"learning_rate": 0.001,
"loss": 3.1236,
"step": 13500
},
{
"epoch": 4.395604395604396,
"grad_norm": 0.8832914233207703,
"learning_rate": 0.001,
"loss": 3.1135,
"step": 13600
},
{
"epoch": 4.4279250161603105,
"grad_norm": 0.8086226582527161,
"learning_rate": 0.001,
"loss": 3.0951,
"step": 13700
},
{
"epoch": 4.460245636716225,
"grad_norm": 1.0011699199676514,
"learning_rate": 0.001,
"loss": 3.1231,
"step": 13800
},
{
"epoch": 4.49256625727214,
"grad_norm": 0.8410008549690247,
"learning_rate": 0.001,
"loss": 3.1299,
"step": 13900
},
{
"epoch": 4.524886877828054,
"grad_norm": 0.986880362033844,
"learning_rate": 0.001,
"loss": 3.0891,
"step": 14000
},
{
"epoch": 4.557207498383969,
"grad_norm": 0.9187092781066895,
"learning_rate": 0.001,
"loss": 3.1157,
"step": 14100
},
{
"epoch": 4.589528118939883,
"grad_norm": 0.9055022597312927,
"learning_rate": 0.001,
"loss": 3.1278,
"step": 14200
},
{
"epoch": 4.621848739495798,
"grad_norm": 0.8347965478897095,
"learning_rate": 0.001,
"loss": 3.1235,
"step": 14300
},
{
"epoch": 4.654169360051713,
"grad_norm": 0.9672727584838867,
"learning_rate": 0.001,
"loss": 3.1257,
"step": 14400
},
{
"epoch": 4.686489980607628,
"grad_norm": 0.7583087682723999,
"learning_rate": 0.001,
"loss": 3.1226,
"step": 14500
},
{
"epoch": 4.7188106011635425,
"grad_norm": 0.8194339871406555,
"learning_rate": 0.001,
"loss": 3.0979,
"step": 14600
},
{
"epoch": 4.751131221719457,
"grad_norm": 0.9744320511817932,
"learning_rate": 0.001,
"loss": 3.1104,
"step": 14700
},
{
"epoch": 4.783451842275372,
"grad_norm": 1.0151008367538452,
"learning_rate": 0.001,
"loss": 3.0929,
"step": 14800
},
{
"epoch": 4.815772462831286,
"grad_norm": 1.2430899143218994,
"learning_rate": 0.001,
"loss": 3.0993,
"step": 14900
},
{
"epoch": 4.848093083387201,
"grad_norm": 0.9484360814094543,
"learning_rate": 0.001,
"loss": 3.1123,
"step": 15000
},
{
"epoch": 4.880413703943116,
"grad_norm": 0.8659531474113464,
"learning_rate": 0.001,
"loss": 3.0993,
"step": 15100
},
{
"epoch": 4.912734324499031,
"grad_norm": 0.9681306481361389,
"learning_rate": 0.001,
"loss": 3.1079,
"step": 15200
},
{
"epoch": 4.945054945054945,
"grad_norm": 0.931349515914917,
"learning_rate": 0.001,
"loss": 3.1038,
"step": 15300
},
{
"epoch": 4.97737556561086,
"grad_norm": 1.1522624492645264,
"learning_rate": 0.001,
"loss": 3.1147,
"step": 15400
},
{
"epoch": 5.009696186166774,
"grad_norm": 0.9395429491996765,
"learning_rate": 0.001,
"loss": 3.0746,
"step": 15500
},
{
"epoch": 5.042016806722689,
"grad_norm": 0.8363929390907288,
"learning_rate": 0.001,
"loss": 2.9545,
"step": 15600
},
{
"epoch": 5.0743374272786035,
"grad_norm": 0.8244017362594604,
"learning_rate": 0.001,
"loss": 2.9895,
"step": 15700
},
{
"epoch": 5.106658047834518,
"grad_norm": 1.0075528621673584,
"learning_rate": 0.001,
"loss": 2.9878,
"step": 15800
},
{
"epoch": 5.1389786683904335,
"grad_norm": 0.8172798156738281,
"learning_rate": 0.001,
"loss": 2.9957,
"step": 15900
},
{
"epoch": 5.171299288946348,
"grad_norm": 0.8578673005104065,
"learning_rate": 0.001,
"loss": 3.0058,
"step": 16000
},
{
"epoch": 5.203619909502263,
"grad_norm": 0.9416762590408325,
"learning_rate": 0.001,
"loss": 2.997,
"step": 16100
},
{
"epoch": 5.235940530058177,
"grad_norm": 0.9278176426887512,
"learning_rate": 0.001,
"loss": 2.9863,
"step": 16200
},
{
"epoch": 5.268261150614092,
"grad_norm": 0.7316229343414307,
"learning_rate": 0.001,
"loss": 3.0075,
"step": 16300
},
{
"epoch": 5.300581771170006,
"grad_norm": 1.0285266637802124,
"learning_rate": 0.001,
"loss": 3.0141,
"step": 16400
},
{
"epoch": 5.332902391725921,
"grad_norm": 1.1114362478256226,
"learning_rate": 0.001,
"loss": 3.0004,
"step": 16500
},
{
"epoch": 5.365223012281835,
"grad_norm": 1.0547707080841064,
"learning_rate": 0.001,
"loss": 3.0005,
"step": 16600
},
{
"epoch": 5.397543632837751,
"grad_norm": 1.1820578575134277,
"learning_rate": 0.001,
"loss": 2.9994,
"step": 16700
},
{
"epoch": 5.429864253393665,
"grad_norm": 0.7730932831764221,
"learning_rate": 0.001,
"loss": 3.0133,
"step": 16800
},
{
"epoch": 5.46218487394958,
"grad_norm": 0.971493124961853,
"learning_rate": 0.001,
"loss": 2.9977,
"step": 16900
},
{
"epoch": 5.4945054945054945,
"grad_norm": 0.9448038935661316,
"learning_rate": 0.001,
"loss": 3.0203,
"step": 17000
},
{
"epoch": 5.526826115061409,
"grad_norm": 1.1640589237213135,
"learning_rate": 0.001,
"loss": 2.999,
"step": 17100
},
{
"epoch": 5.559146735617324,
"grad_norm": 0.9492186903953552,
"learning_rate": 0.001,
"loss": 3.0076,
"step": 17200
},
{
"epoch": 5.591467356173238,
"grad_norm": 0.8014913201332092,
"learning_rate": 0.001,
"loss": 2.9883,
"step": 17300
},
{
"epoch": 5.623787976729153,
"grad_norm": 1.0996049642562866,
"learning_rate": 0.001,
"loss": 3.0055,
"step": 17400
},
{
"epoch": 5.656108597285068,
"grad_norm": 0.958639919757843,
"learning_rate": 0.001,
"loss": 3.0093,
"step": 17500
},
{
"epoch": 5.688429217840983,
"grad_norm": 0.9354169368743896,
"learning_rate": 0.001,
"loss": 3.031,
"step": 17600
},
{
"epoch": 5.720749838396897,
"grad_norm": 0.8590208292007446,
"learning_rate": 0.001,
"loss": 3.0183,
"step": 17700
},
{
"epoch": 5.753070458952812,
"grad_norm": 0.93686842918396,
"learning_rate": 0.001,
"loss": 3.0124,
"step": 17800
},
{
"epoch": 5.785391079508726,
"grad_norm": 0.9742192029953003,
"learning_rate": 0.001,
"loss": 2.9985,
"step": 17900
},
{
"epoch": 5.817711700064641,
"grad_norm": 1.2014163732528687,
"learning_rate": 0.001,
"loss": 3.0124,
"step": 18000
},
{
"epoch": 5.850032320620556,
"grad_norm": 1.0222500562667847,
"learning_rate": 0.001,
"loss": 3.0188,
"step": 18100
},
{
"epoch": 5.882352941176471,
"grad_norm": 0.8603215217590332,
"learning_rate": 0.001,
"loss": 3.02,
"step": 18200
},
{
"epoch": 5.914673561732386,
"grad_norm": 0.9950960874557495,
"learning_rate": 0.001,
"loss": 3.0372,
"step": 18300
},
{
"epoch": 5.9469941822883,
"grad_norm": 1.1442996263504028,
"learning_rate": 0.001,
"loss": 3.0098,
"step": 18400
},
{
"epoch": 5.979314802844215,
"grad_norm": 1.1008858680725098,
"learning_rate": 0.001,
"loss": 3.015,
"step": 18500
},
{
"epoch": 6.011635423400129,
"grad_norm": 0.8677753806114197,
"learning_rate": 0.001,
"loss": 2.9662,
"step": 18600
},
{
"epoch": 6.043956043956044,
"grad_norm": 0.9199807643890381,
"learning_rate": 0.001,
"loss": 2.8779,
"step": 18700
},
{
"epoch": 6.076276664511958,
"grad_norm": 1.2043673992156982,
"learning_rate": 0.001,
"loss": 2.8758,
"step": 18800
},
{
"epoch": 6.108597285067873,
"grad_norm": 1.0407578945159912,
"learning_rate": 0.001,
"loss": 2.8885,
"step": 18900
},
{
"epoch": 6.140917905623788,
"grad_norm": 0.9558809399604797,
"learning_rate": 0.001,
"loss": 2.8996,
"step": 19000
},
{
"epoch": 6.173238526179703,
"grad_norm": 0.96955406665802,
"learning_rate": 0.001,
"loss": 2.8962,
"step": 19100
},
{
"epoch": 6.2055591467356175,
"grad_norm": 0.8183215260505676,
"learning_rate": 0.001,
"loss": 2.9015,
"step": 19200
},
{
"epoch": 6.237879767291532,
"grad_norm": 0.902918815612793,
"learning_rate": 0.001,
"loss": 2.9103,
"step": 19300
},
{
"epoch": 6.270200387847447,
"grad_norm": 1.2891950607299805,
"learning_rate": 0.001,
"loss": 2.9301,
"step": 19400
},
{
"epoch": 6.302521008403361,
"grad_norm": 1.128111481666565,
"learning_rate": 0.001,
"loss": 2.901,
"step": 19500
},
{
"epoch": 6.334841628959276,
"grad_norm": 1.0123465061187744,
"learning_rate": 0.001,
"loss": 2.9011,
"step": 19600
},
{
"epoch": 6.36716224951519,
"grad_norm": 0.8102533221244812,
"learning_rate": 0.001,
"loss": 2.9126,
"step": 19700
},
{
"epoch": 6.399482870071106,
"grad_norm": 0.9812631607055664,
"learning_rate": 0.001,
"loss": 2.9284,
"step": 19800
},
{
"epoch": 6.43180349062702,
"grad_norm": 0.8977174758911133,
"learning_rate": 0.001,
"loss": 2.9122,
"step": 19900
},
{
"epoch": 6.464124111182935,
"grad_norm": 0.9242091774940491,
"learning_rate": 0.001,
"loss": 2.9026,
"step": 20000
},
{
"epoch": 6.496444731738849,
"grad_norm": 0.8112374544143677,
"learning_rate": 0.001,
"loss": 2.9236,
"step": 20100
},
{
"epoch": 6.528765352294764,
"grad_norm": 1.3237755298614502,
"learning_rate": 0.001,
"loss": 2.9227,
"step": 20200
},
{
"epoch": 6.5610859728506785,
"grad_norm": 1.166825294494629,
"learning_rate": 0.001,
"loss": 2.9198,
"step": 20300
},
{
"epoch": 6.593406593406593,
"grad_norm": 0.8015363812446594,
"learning_rate": 0.001,
"loss": 2.923,
"step": 20400
},
{
"epoch": 6.625727213962508,
"grad_norm": 1.1751559972763062,
"learning_rate": 0.001,
"loss": 2.9269,
"step": 20500
},
{
"epoch": 6.658047834518423,
"grad_norm": 0.9034355878829956,
"learning_rate": 0.001,
"loss": 2.9247,
"step": 20600
},
{
"epoch": 6.690368455074338,
"grad_norm": 1.0978208780288696,
"learning_rate": 0.001,
"loss": 2.9178,
"step": 20700
},
{
"epoch": 6.722689075630252,
"grad_norm": 0.6817880272865295,
"learning_rate": 0.001,
"loss": 2.9334,
"step": 20800
},
{
"epoch": 6.755009696186167,
"grad_norm": 1.0009515285491943,
"learning_rate": 0.001,
"loss": 2.9468,
"step": 20900
},
{
"epoch": 6.787330316742081,
"grad_norm": 0.9897229075431824,
"learning_rate": 0.001,
"loss": 2.9367,
"step": 21000
},
{
"epoch": 6.819650937297996,
"grad_norm": 0.9082470536231995,
"learning_rate": 0.001,
"loss": 2.9332,
"step": 21100
},
{
"epoch": 6.85197155785391,
"grad_norm": 0.7760358452796936,
"learning_rate": 0.001,
"loss": 2.955,
"step": 21200
},
{
"epoch": 6.884292178409826,
"grad_norm": 1.381992220878601,
"learning_rate": 0.001,
"loss": 2.9355,
"step": 21300
},
{
"epoch": 6.91661279896574,
"grad_norm": 0.9245606660842896,
"learning_rate": 0.001,
"loss": 2.9553,
"step": 21400
},
{
"epoch": 6.948933419521655,
"grad_norm": 1.1271631717681885,
"learning_rate": 0.001,
"loss": 2.94,
"step": 21500
},
{
"epoch": 6.98125404007757,
"grad_norm": 1.0409258604049683,
"learning_rate": 0.001,
"loss": 2.9205,
"step": 21600
},
{
"epoch": 7.013574660633484,
"grad_norm": 1.003588080406189,
"learning_rate": 0.001,
"loss": 2.8729,
"step": 21700
},
{
"epoch": 7.045895281189399,
"grad_norm": 0.9518575072288513,
"learning_rate": 0.001,
"loss": 2.8026,
"step": 21800
},
{
"epoch": 7.078215901745313,
"grad_norm": 1.1394156217575073,
"learning_rate": 0.001,
"loss": 2.7928,
"step": 21900
},
{
"epoch": 7.110536522301228,
"grad_norm": 1.1586087942123413,
"learning_rate": 0.001,
"loss": 2.8063,
"step": 22000
},
{
"epoch": 7.142857142857143,
"grad_norm": 0.8844541311264038,
"learning_rate": 0.001,
"loss": 2.8112,
"step": 22100
},
{
"epoch": 7.175177763413058,
"grad_norm": 1.3279544115066528,
"learning_rate": 0.001,
"loss": 2.814,
"step": 22200
},
{
"epoch": 7.207498383968972,
"grad_norm": 1.0511810779571533,
"learning_rate": 0.001,
"loss": 2.8295,
"step": 22300
},
{
"epoch": 7.239819004524887,
"grad_norm": 1.2267463207244873,
"learning_rate": 0.001,
"loss": 2.8351,
"step": 22400
},
{
"epoch": 7.2721396250808015,
"grad_norm": 1.1739734411239624,
"learning_rate": 0.001,
"loss": 2.8367,
"step": 22500
},
{
"epoch": 7.304460245636716,
"grad_norm": 0.8629540801048279,
"learning_rate": 0.001,
"loss": 2.8326,
"step": 22600
},
{
"epoch": 7.336780866192631,
"grad_norm": 1.048889398574829,
"learning_rate": 0.001,
"loss": 2.8328,
"step": 22700
},
{
"epoch": 7.369101486748546,
"grad_norm": 1.1405242681503296,
"learning_rate": 0.001,
"loss": 2.8529,
"step": 22800
},
{
"epoch": 7.401422107304461,
"grad_norm": 1.045923113822937,
"learning_rate": 0.001,
"loss": 2.8592,
"step": 22900
},
{
"epoch": 7.433742727860375,
"grad_norm": 0.9941197037696838,
"learning_rate": 0.001,
"loss": 2.8476,
"step": 23000
},
{
"epoch": 7.46606334841629,
"grad_norm": 1.1987438201904297,
"learning_rate": 0.001,
"loss": 2.8538,
"step": 23100
},
{
"epoch": 7.498383968972204,
"grad_norm": 1.1423817873001099,
"learning_rate": 0.001,
"loss": 2.8587,
"step": 23200
},
{
"epoch": 7.530704589528119,
"grad_norm": 1.3632299900054932,
"learning_rate": 0.001,
"loss": 2.8536,
"step": 23300
},
{
"epoch": 7.563025210084033,
"grad_norm": 1.2440208196640015,
"learning_rate": 0.001,
"loss": 2.8737,
"step": 23400
},
{
"epoch": 7.595345830639948,
"grad_norm": 0.7769545316696167,
"learning_rate": 0.001,
"loss": 2.8429,
"step": 23500
},
{
"epoch": 7.6276664511958625,
"grad_norm": 0.9705239534378052,
"learning_rate": 0.001,
"loss": 2.8554,
"step": 23600
},
{
"epoch": 7.659987071751778,
"grad_norm": 0.9184752702713013,
"learning_rate": 0.001,
"loss": 2.8536,
"step": 23700
},
{
"epoch": 7.6923076923076925,
"grad_norm": 1.2383073568344116,
"learning_rate": 0.001,
"loss": 2.8456,
"step": 23800
},
{
"epoch": 7.724628312863607,
"grad_norm": 1.2039756774902344,
"learning_rate": 0.001,
"loss": 2.8566,
"step": 23900
},
{
"epoch": 7.756948933419522,
"grad_norm": 1.2664653062820435,
"learning_rate": 0.001,
"loss": 2.8605,
"step": 24000
},
{
"epoch": 7.789269553975436,
"grad_norm": 1.2550541162490845,
"learning_rate": 0.001,
"loss": 2.8762,
"step": 24100
},
{
"epoch": 7.821590174531351,
"grad_norm": 0.9475702047348022,
"learning_rate": 0.001,
"loss": 2.8658,
"step": 24200
},
{
"epoch": 7.853910795087265,
"grad_norm": 1.1510746479034424,
"learning_rate": 0.001,
"loss": 2.8507,
"step": 24300
},
{
"epoch": 7.886231415643181,
"grad_norm": 1.0832042694091797,
"learning_rate": 0.001,
"loss": 2.8572,
"step": 24400
},
{
"epoch": 7.918552036199095,
"grad_norm": 1.154191493988037,
"learning_rate": 0.001,
"loss": 2.8605,
"step": 24500
},
{
"epoch": 7.95087265675501,
"grad_norm": 1.1333051919937134,
"learning_rate": 0.001,
"loss": 2.8948,
"step": 24600
},
{
"epoch": 7.983193277310924,
"grad_norm": 0.9192237257957458,
"learning_rate": 0.001,
"loss": 2.8726,
"step": 24700
},
{
"epoch": 8.015513897866839,
"grad_norm": 1.6592745780944824,
"learning_rate": 0.001,
"loss": 2.787,
"step": 24800
},
{
"epoch": 8.047834518422754,
"grad_norm": 1.409105896949768,
"learning_rate": 0.001,
"loss": 2.7394,
"step": 24900
},
{
"epoch": 8.080155138978668,
"grad_norm": 1.5304033756256104,
"learning_rate": 0.001,
"loss": 2.7462,
"step": 25000
},
{
"epoch": 8.112475759534583,
"grad_norm": 1.15884268283844,
"learning_rate": 0.001,
"loss": 2.7546,
"step": 25100
},
{
"epoch": 8.144796380090497,
"grad_norm": 1.3049625158309937,
"learning_rate": 0.001,
"loss": 2.7683,
"step": 25200
},
{
"epoch": 8.177117000646412,
"grad_norm": 1.4045416116714478,
"learning_rate": 0.001,
"loss": 2.7515,
"step": 25300
},
{
"epoch": 8.209437621202326,
"grad_norm": 1.4240517616271973,
"learning_rate": 0.001,
"loss": 2.757,
"step": 25400
},
{
"epoch": 8.241758241758241,
"grad_norm": 1.0942411422729492,
"learning_rate": 0.001,
"loss": 2.758,
"step": 25500
},
{
"epoch": 8.274078862314157,
"grad_norm": 1.3089206218719482,
"learning_rate": 0.001,
"loss": 2.7449,
"step": 25600
},
{
"epoch": 8.306399482870072,
"grad_norm": 1.1560167074203491,
"learning_rate": 0.001,
"loss": 2.757,
"step": 25700
},
{
"epoch": 8.338720103425986,
"grad_norm": 1.2630512714385986,
"learning_rate": 0.001,
"loss": 2.7637,
"step": 25800
},
{
"epoch": 8.371040723981901,
"grad_norm": 1.4474283456802368,
"learning_rate": 0.001,
"loss": 2.7716,
"step": 25900
},
{
"epoch": 8.403361344537815,
"grad_norm": 1.3015059232711792,
"learning_rate": 0.001,
"loss": 2.7403,
"step": 26000
},
{
"epoch": 8.43568196509373,
"grad_norm": 1.5766280889511108,
"learning_rate": 0.001,
"loss": 2.7634,
"step": 26100
},
{
"epoch": 8.468002585649645,
"grad_norm": 1.2736643552780151,
"learning_rate": 0.001,
"loss": 2.7885,
"step": 26200
},
{
"epoch": 8.50032320620556,
"grad_norm": 1.2255592346191406,
"learning_rate": 0.001,
"loss": 2.788,
"step": 26300
},
{
"epoch": 8.532643826761474,
"grad_norm": 1.0317450761795044,
"learning_rate": 0.001,
"loss": 2.7814,
"step": 26400
},
{
"epoch": 8.564964447317388,
"grad_norm": 1.1160918474197388,
"learning_rate": 0.001,
"loss": 2.7902,
"step": 26500
},
{
"epoch": 8.597285067873303,
"grad_norm": 1.5986204147338867,
"learning_rate": 0.001,
"loss": 2.791,
"step": 26600
},
{
"epoch": 8.629605688429217,
"grad_norm": 1.0425474643707275,
"learning_rate": 0.001,
"loss": 2.8,
"step": 26700
},
{
"epoch": 8.661926308985132,
"grad_norm": 1.5469958782196045,
"learning_rate": 0.001,
"loss": 2.7774,
"step": 26800
},
{
"epoch": 8.694246929541046,
"grad_norm": 1.0598360300064087,
"learning_rate": 0.001,
"loss": 2.7974,
"step": 26900
},
{
"epoch": 8.726567550096961,
"grad_norm": 1.1809691190719604,
"learning_rate": 0.001,
"loss": 2.8031,
"step": 27000
},
{
"epoch": 8.758888170652877,
"grad_norm": 1.5478875637054443,
"learning_rate": 0.001,
"loss": 2.7882,
"step": 27100
},
{
"epoch": 8.791208791208792,
"grad_norm": 1.3363964557647705,
"learning_rate": 0.001,
"loss": 2.7842,
"step": 27200
},
{
"epoch": 8.823529411764707,
"grad_norm": 1.1710869073867798,
"learning_rate": 0.001,
"loss": 2.794,
"step": 27300
},
{
"epoch": 8.855850032320621,
"grad_norm": 1.5282068252563477,
"learning_rate": 0.001,
"loss": 2.817,
"step": 27400
},
{
"epoch": 8.888170652876536,
"grad_norm": 1.2685834169387817,
"learning_rate": 0.001,
"loss": 2.8,
"step": 27500
},
{
"epoch": 8.92049127343245,
"grad_norm": 1.236366868019104,
"learning_rate": 0.001,
"loss": 2.7856,
"step": 27600
},
{
"epoch": 8.952811893988365,
"grad_norm": 1.238660454750061,
"learning_rate": 0.001,
"loss": 2.8202,
"step": 27700
},
{
"epoch": 8.98513251454428,
"grad_norm": 1.3011889457702637,
"learning_rate": 0.001,
"loss": 2.8083,
"step": 27800
},
{
"epoch": 9.017453135100194,
"grad_norm": 1.592627763748169,
"learning_rate": 0.001,
"loss": 2.7183,
"step": 27900
},
{
"epoch": 9.049773755656108,
"grad_norm": 1.7187007665634155,
"learning_rate": 0.001,
"loss": 2.6633,
"step": 28000
},
{
"epoch": 9.082094376212023,
"grad_norm": 1.1323732137680054,
"learning_rate": 0.001,
"loss": 2.6867,
"step": 28100
},
{
"epoch": 9.114414996767938,
"grad_norm": 1.3170040845870972,
"learning_rate": 0.001,
"loss": 2.679,
"step": 28200
},
{
"epoch": 9.146735617323852,
"grad_norm": 1.7077041864395142,
"learning_rate": 0.001,
"loss": 2.6861,
"step": 28300
},
{
"epoch": 9.179056237879767,
"grad_norm": 1.4596565961837769,
"learning_rate": 0.001,
"loss": 2.6885,
"step": 28400
},
{
"epoch": 9.211376858435681,
"grad_norm": 1.6344329118728638,
"learning_rate": 0.001,
"loss": 2.6774,
"step": 28500
},
{
"epoch": 9.243697478991596,
"grad_norm": 1.6928468942642212,
"learning_rate": 0.001,
"loss": 2.6804,
"step": 28600
},
{
"epoch": 9.276018099547512,
"grad_norm": 1.280487060546875,
"learning_rate": 0.001,
"loss": 2.6826,
"step": 28700
},
{
"epoch": 9.308338720103427,
"grad_norm": 1.200545072555542,
"learning_rate": 0.001,
"loss": 2.7226,
"step": 28800
},
{
"epoch": 9.340659340659341,
"grad_norm": 1.246785044670105,
"learning_rate": 0.001,
"loss": 2.7231,
"step": 28900
},
{
"epoch": 9.372979961215256,
"grad_norm": 1.3656576871871948,
"learning_rate": 0.001,
"loss": 2.7194,
"step": 29000
},
{
"epoch": 9.40530058177117,
"grad_norm": 1.3054988384246826,
"learning_rate": 0.001,
"loss": 2.7225,
"step": 29100
},
{
"epoch": 9.437621202327085,
"grad_norm": 1.3954949378967285,
"learning_rate": 0.001,
"loss": 2.719,
"step": 29200
},
{
"epoch": 9.469941822883,
"grad_norm": 0.9482543468475342,
"learning_rate": 0.001,
"loss": 2.728,
"step": 29300
},
{
"epoch": 9.502262443438914,
"grad_norm": 1.694023609161377,
"learning_rate": 0.001,
"loss": 2.7264,
"step": 29400
},
{
"epoch": 9.534583063994829,
"grad_norm": 1.1679247617721558,
"learning_rate": 0.001,
"loss": 2.7203,
"step": 29500
},
{
"epoch": 9.566903684550743,
"grad_norm": 1.1120076179504395,
"learning_rate": 0.001,
"loss": 2.706,
"step": 29600
},
{
"epoch": 9.599224305106658,
"grad_norm": 1.646690011024475,
"learning_rate": 0.001,
"loss": 2.7319,
"step": 29700
},
{
"epoch": 9.631544925662572,
"grad_norm": 1.382325530052185,
"learning_rate": 0.001,
"loss": 2.7578,
"step": 29800
},
{
"epoch": 9.663865546218487,
"grad_norm": 1.5380642414093018,
"learning_rate": 0.001,
"loss": 2.7056,
"step": 29900
},
{
"epoch": 9.696186166774401,
"grad_norm": 1.095766544342041,
"learning_rate": 0.001,
"loss": 2.7167,
"step": 30000
},
{
"epoch": 9.728506787330316,
"grad_norm": 1.8738212585449219,
"learning_rate": 0.001,
"loss": 2.6211,
"step": 30100
},
{
"epoch": 9.760827407886232,
"grad_norm": 1.6147414445877075,
"learning_rate": 0.001,
"loss": 2.6289,
"step": 30200
},
{
"epoch": 9.793148028442147,
"grad_norm": 1.3386214971542358,
"learning_rate": 0.001,
"loss": 2.6204,
"step": 30300
},
{
"epoch": 9.825468648998061,
"grad_norm": 1.342301845550537,
"learning_rate": 0.001,
"loss": 2.6572,
"step": 30400
},
{
"epoch": 9.857789269553976,
"grad_norm": 1.3404988050460815,
"learning_rate": 0.001,
"loss": 2.6799,
"step": 30500
},
{
"epoch": 9.89010989010989,
"grad_norm": 1.5837438106536865,
"learning_rate": 0.001,
"loss": 2.6697,
"step": 30600
},
{
"epoch": 9.922430510665805,
"grad_norm": 1.326537013053894,
"learning_rate": 0.001,
"loss": 2.6775,
"step": 30700
},
{
"epoch": 9.95475113122172,
"grad_norm": 1.5108411312103271,
"learning_rate": 0.001,
"loss": 2.654,
"step": 30800
},
{
"epoch": 9.987071751777634,
"grad_norm": 1.476011872291565,
"learning_rate": 0.001,
"loss": 2.6678,
"step": 30900
},
{
"epoch": 10.019392372333549,
"grad_norm": 1.0556116104125977,
"learning_rate": 0.001,
"loss": 2.6283,
"step": 31000
},
{
"epoch": 10.051712992889463,
"grad_norm": 1.3222134113311768,
"learning_rate": 0.001,
"loss": 2.6413,
"step": 31100
},
{
"epoch": 10.084033613445378,
"grad_norm": 1.0430011749267578,
"learning_rate": 0.001,
"loss": 2.6294,
"step": 31200
},
{
"epoch": 10.116354234001292,
"grad_norm": 0.773851215839386,
"learning_rate": 0.001,
"loss": 2.6359,
"step": 31300
},
{
"epoch": 10.148674854557207,
"grad_norm": 1.049518346786499,
"learning_rate": 0.001,
"loss": 2.6324,
"step": 31400
},
{
"epoch": 10.180995475113122,
"grad_norm": 1.0573190450668335,
"learning_rate": 0.001,
"loss": 2.6472,
"step": 31500
},
{
"epoch": 10.213316095669036,
"grad_norm": 1.0607960224151611,
"learning_rate": 0.001,
"loss": 2.6516,
"step": 31600
},
{
"epoch": 10.24563671622495,
"grad_norm": 1.3578628301620483,
"learning_rate": 0.001,
"loss": 2.6543,
"step": 31700
},
{
"epoch": 10.277957336780867,
"grad_norm": 1.242769718170166,
"learning_rate": 0.001,
"loss": 2.6501,
"step": 31800
},
{
"epoch": 10.310277957336782,
"grad_norm": 1.3352106809616089,
"learning_rate": 0.001,
"loss": 2.6632,
"step": 31900
},
{
"epoch": 10.342598577892696,
"grad_norm": 0.9553130865097046,
"learning_rate": 0.001,
"loss": 2.6334,
"step": 32000
},
{
"epoch": 10.37491919844861,
"grad_norm": 1.1314009428024292,
"learning_rate": 0.001,
"loss": 2.6529,
"step": 32100
},
{
"epoch": 10.407239819004525,
"grad_norm": 1.4260714054107666,
"learning_rate": 0.001,
"loss": 2.6845,
"step": 32200
},
{
"epoch": 10.43956043956044,
"grad_norm": 1.099668264389038,
"learning_rate": 0.001,
"loss": 2.6673,
"step": 32300
},
{
"epoch": 10.471881060116354,
"grad_norm": 1.0341742038726807,
"learning_rate": 0.001,
"loss": 2.6627,
"step": 32400
},
{
"epoch": 10.504201680672269,
"grad_norm": 0.9638963341712952,
"learning_rate": 0.001,
"loss": 2.6508,
"step": 32500
},
{
"epoch": 10.536522301228183,
"grad_norm": 0.9778535962104797,
"learning_rate": 0.001,
"loss": 2.666,
"step": 32600
},
{
"epoch": 10.568842921784098,
"grad_norm": 1.0867583751678467,
"learning_rate": 0.001,
"loss": 2.6959,
"step": 32700
},
{
"epoch": 10.601163542340013,
"grad_norm": 1.0800092220306396,
"learning_rate": 0.001,
"loss": 2.6805,
"step": 32800
},
{
"epoch": 10.633484162895927,
"grad_norm": 1.13056218624115,
"learning_rate": 0.001,
"loss": 2.6597,
"step": 32900
},
{
"epoch": 10.665804783451842,
"grad_norm": 0.9933321475982666,
"learning_rate": 0.001,
"loss": 2.6935,
"step": 33000
},
{
"epoch": 10.698125404007756,
"grad_norm": 0.7585985660552979,
"learning_rate": 0.001,
"loss": 2.6851,
"step": 33100
},
{
"epoch": 10.73044602456367,
"grad_norm": 0.8704131841659546,
"learning_rate": 0.001,
"loss": 2.6768,
"step": 33200
},
{
"epoch": 10.762766645119587,
"grad_norm": 1.0104848146438599,
"learning_rate": 0.001,
"loss": 2.6797,
"step": 33300
},
{
"epoch": 10.795087265675502,
"grad_norm": 0.8003226518630981,
"learning_rate": 0.001,
"loss": 2.6903,
"step": 33400
},
{
"epoch": 10.827407886231416,
"grad_norm": 1.0259637832641602,
"learning_rate": 0.001,
"loss": 2.6888,
"step": 33500
},
{
"epoch": 10.85972850678733,
"grad_norm": 1.269639015197754,
"learning_rate": 0.001,
"loss": 2.6946,
"step": 33600
},
{
"epoch": 10.892049127343245,
"grad_norm": 0.9504902958869934,
"learning_rate": 0.001,
"loss": 2.6894,
"step": 33700
},
{
"epoch": 10.92436974789916,
"grad_norm": 1.0889633893966675,
"learning_rate": 0.001,
"loss": 2.693,
"step": 33800
},
{
"epoch": 10.956690368455074,
"grad_norm": 1.2490198612213135,
"learning_rate": 0.001,
"loss": 2.6913,
"step": 33900
},
{
"epoch": 10.989010989010989,
"grad_norm": 1.2247567176818848,
"learning_rate": 0.001,
"loss": 2.6774,
"step": 34000
},
{
"epoch": 11.021331609566904,
"grad_norm": 1.030630350112915,
"learning_rate": 0.001,
"loss": 2.6152,
"step": 34100
},
{
"epoch": 11.053652230122818,
"grad_norm": 1.1876599788665771,
"learning_rate": 0.001,
"loss": 2.5308,
"step": 34200
},
{
"epoch": 11.085972850678733,
"grad_norm": 0.9506245255470276,
"learning_rate": 0.001,
"loss": 2.5633,
"step": 34300
},
{
"epoch": 11.118293471234647,
"grad_norm": 1.082385540008545,
"learning_rate": 0.001,
"loss": 2.5854,
"step": 34400
},
{
"epoch": 11.150614091790562,
"grad_norm": 1.3474210500717163,
"learning_rate": 0.001,
"loss": 2.5664,
"step": 34500
},
{
"epoch": 11.182934712346476,
"grad_norm": 1.0339075326919556,
"learning_rate": 0.001,
"loss": 2.5783,
"step": 34600
},
{
"epoch": 11.215255332902391,
"grad_norm": 1.16727876663208,
"learning_rate": 0.001,
"loss": 2.5977,
"step": 34700
},
{
"epoch": 11.247575953458306,
"grad_norm": 1.0603053569793701,
"learning_rate": 0.001,
"loss": 2.5766,
"step": 34800
},
{
"epoch": 11.279896574014222,
"grad_norm": 1.1884040832519531,
"learning_rate": 0.001,
"loss": 2.5973,
"step": 34900
},
{
"epoch": 11.312217194570136,
"grad_norm": 0.9387410283088684,
"learning_rate": 0.001,
"loss": 2.5945,
"step": 35000
},
{
"epoch": 11.344537815126051,
"grad_norm": 1.3718771934509277,
"learning_rate": 0.001,
"loss": 2.6044,
"step": 35100
},
{
"epoch": 11.376858435681966,
"grad_norm": 1.0469406843185425,
"learning_rate": 0.001,
"loss": 2.5997,
"step": 35200
},
{
"epoch": 11.40917905623788,
"grad_norm": 1.1330077648162842,
"learning_rate": 0.001,
"loss": 2.6206,
"step": 35300
},
{
"epoch": 11.441499676793795,
"grad_norm": 1.0808653831481934,
"learning_rate": 0.001,
"loss": 2.6335,
"step": 35400
},
{
"epoch": 11.47382029734971,
"grad_norm": 0.8717126846313477,
"learning_rate": 0.001,
"loss": 2.6172,
"step": 35500
},
{
"epoch": 11.506140917905624,
"grad_norm": 0.9696609973907471,
"learning_rate": 0.001,
"loss": 2.6145,
"step": 35600
},
{
"epoch": 11.538461538461538,
"grad_norm": 1.1971484422683716,
"learning_rate": 0.001,
"loss": 2.6337,
"step": 35700
},
{
"epoch": 11.570782159017453,
"grad_norm": 1.0605714321136475,
"learning_rate": 0.001,
"loss": 2.6077,
"step": 35800
},
{
"epoch": 11.603102779573367,
"grad_norm": 0.9503991603851318,
"learning_rate": 0.001,
"loss": 2.6299,
"step": 35900
},
{
"epoch": 11.635423400129282,
"grad_norm": 1.060072660446167,
"learning_rate": 0.001,
"loss": 2.6293,
"step": 36000
},
{
"epoch": 11.667744020685197,
"grad_norm": 0.9725481867790222,
"learning_rate": 0.001,
"loss": 2.6324,
"step": 36100
},
{
"epoch": 11.700064641241111,
"grad_norm": 1.1967412233352661,
"learning_rate": 0.001,
"loss": 2.6564,
"step": 36200
},
{
"epoch": 11.732385261797026,
"grad_norm": 1.2135505676269531,
"learning_rate": 0.001,
"loss": 2.6465,
"step": 36300
},
{
"epoch": 11.764705882352942,
"grad_norm": 1.0049983263015747,
"learning_rate": 0.001,
"loss": 2.647,
"step": 36400
},
{
"epoch": 11.797026502908857,
"grad_norm": 1.028984785079956,
"learning_rate": 0.001,
"loss": 2.6212,
"step": 36500
},
{
"epoch": 11.829347123464771,
"grad_norm": 1.2043495178222656,
"learning_rate": 0.001,
"loss": 2.6208,
"step": 36600
},
{
"epoch": 11.861667744020686,
"grad_norm": 1.2476329803466797,
"learning_rate": 0.001,
"loss": 2.6416,
"step": 36700
},
{
"epoch": 11.8939883645766,
"grad_norm": 1.2056334018707275,
"learning_rate": 0.001,
"loss": 2.6308,
"step": 36800
},
{
"epoch": 11.926308985132515,
"grad_norm": 1.0555031299591064,
"learning_rate": 0.001,
"loss": 2.656,
"step": 36900
},
{
"epoch": 11.95862960568843,
"grad_norm": 1.0049588680267334,
"learning_rate": 0.001,
"loss": 2.6562,
"step": 37000
},
{
"epoch": 11.990950226244344,
"grad_norm": 1.3405330181121826,
"learning_rate": 0.001,
"loss": 2.6406,
"step": 37100
},
{
"epoch": 12.023270846800258,
"grad_norm": 1.007863998413086,
"learning_rate": 0.001,
"loss": 2.5602,
"step": 37200
},
{
"epoch": 12.055591467356173,
"grad_norm": 1.3099013566970825,
"learning_rate": 0.001,
"loss": 2.53,
"step": 37300
},
{
"epoch": 12.087912087912088,
"grad_norm": 0.9795656800270081,
"learning_rate": 0.001,
"loss": 2.5177,
"step": 37400
},
{
"epoch": 12.120232708468002,
"grad_norm": 1.1093167066574097,
"learning_rate": 0.001,
"loss": 2.5147,
"step": 37500
},
{
"epoch": 12.152553329023917,
"grad_norm": 1.1975780725479126,
"learning_rate": 0.001,
"loss": 2.5217,
"step": 37600
},
{
"epoch": 12.184873949579831,
"grad_norm": 1.0998467206954956,
"learning_rate": 0.001,
"loss": 2.5313,
"step": 37700
},
{
"epoch": 12.217194570135746,
"grad_norm": 0.9653919339179993,
"learning_rate": 0.001,
"loss": 2.5522,
"step": 37800
},
{
"epoch": 12.24951519069166,
"grad_norm": 0.8531848788261414,
"learning_rate": 0.001,
"loss": 2.5295,
"step": 37900
},
{
"epoch": 12.281835811247577,
"grad_norm": 1.1443430185317993,
"learning_rate": 0.001,
"loss": 2.5352,
"step": 38000
},
{
"epoch": 12.314156431803491,
"grad_norm": 0.8375819325447083,
"learning_rate": 0.001,
"loss": 2.5298,
"step": 38100
},
{
"epoch": 12.346477052359406,
"grad_norm": 1.1427043676376343,
"learning_rate": 0.001,
"loss": 2.5667,
"step": 38200
},
{
"epoch": 12.37879767291532,
"grad_norm": 0.99153733253479,
"learning_rate": 0.001,
"loss": 2.5647,
"step": 38300
},
{
"epoch": 12.411118293471235,
"grad_norm": 1.1989197731018066,
"learning_rate": 0.001,
"loss": 2.5734,
"step": 38400
},
{
"epoch": 12.44343891402715,
"grad_norm": 1.2347819805145264,
"learning_rate": 0.001,
"loss": 2.5789,
"step": 38500
},
{
"epoch": 12.475759534583064,
"grad_norm": 1.221445083618164,
"learning_rate": 0.001,
"loss": 2.574,
"step": 38600
},
{
"epoch": 12.508080155138979,
"grad_norm": 1.026776909828186,
"learning_rate": 0.001,
"loss": 2.56,
"step": 38700
},
{
"epoch": 12.540400775694893,
"grad_norm": 1.066969633102417,
"learning_rate": 0.001,
"loss": 2.596,
"step": 38800
},
{
"epoch": 12.572721396250808,
"grad_norm": 0.8862924575805664,
"learning_rate": 0.001,
"loss": 2.5773,
"step": 38900
},
{
"epoch": 12.605042016806722,
"grad_norm": 0.9672336578369141,
"learning_rate": 0.001,
"loss": 2.578,
"step": 39000
},
{
"epoch": 12.637362637362637,
"grad_norm": 0.9717584252357483,
"learning_rate": 0.001,
"loss": 2.58,
"step": 39100
},
{
"epoch": 12.669683257918551,
"grad_norm": 1.0868942737579346,
"learning_rate": 0.001,
"loss": 2.5766,
"step": 39200
},
{
"epoch": 12.702003878474466,
"grad_norm": 1.1065813302993774,
"learning_rate": 0.001,
"loss": 2.5872,
"step": 39300
},
{
"epoch": 12.73432449903038,
"grad_norm": 0.9647271633148193,
"learning_rate": 0.001,
"loss": 2.5938,
"step": 39400
},
{
"epoch": 12.766645119586297,
"grad_norm": 1.1909931898117065,
"learning_rate": 0.001,
"loss": 2.5888,
"step": 39500
},
{
"epoch": 12.798965740142211,
"grad_norm": 0.6993247866630554,
"learning_rate": 0.001,
"loss": 2.6035,
"step": 39600
},
{
"epoch": 12.831286360698126,
"grad_norm": 1.0692033767700195,
"learning_rate": 0.001,
"loss": 2.5981,
"step": 39700
},
{
"epoch": 12.86360698125404,
"grad_norm": 1.1344980001449585,
"learning_rate": 0.001,
"loss": 2.5783,
"step": 39800
},
{
"epoch": 12.895927601809955,
"grad_norm": 0.979433000087738,
"learning_rate": 0.001,
"loss": 2.5834,
"step": 39900
},
{
"epoch": 12.92824822236587,
"grad_norm": 1.1205140352249146,
"learning_rate": 0.001,
"loss": 2.5822,
"step": 40000
},
{
"epoch": 12.960568842921784,
"grad_norm": 1.1560449600219727,
"learning_rate": 0.001,
"loss": 2.5763,
"step": 40100
},
{
"epoch": 12.992889463477699,
"grad_norm": 1.485905408859253,
"learning_rate": 0.001,
"loss": 2.5956,
"step": 40200
},
{
"epoch": 13.025210084033613,
"grad_norm": 0.9580528736114502,
"learning_rate": 0.001,
"loss": 2.5061,
"step": 40300
},
{
"epoch": 13.057530704589528,
"grad_norm": 0.9847084879875183,
"learning_rate": 0.001,
"loss": 2.4753,
"step": 40400
},
{
"epoch": 13.089851325145442,
"grad_norm": 0.7392004132270813,
"learning_rate": 0.001,
"loss": 2.4638,
"step": 40500
},
{
"epoch": 13.122171945701357,
"grad_norm": 0.9915220737457275,
"learning_rate": 0.001,
"loss": 2.4544,
"step": 40600
},
{
"epoch": 13.154492566257272,
"grad_norm": 0.9892643690109253,
"learning_rate": 0.001,
"loss": 2.4818,
"step": 40700
},
{
"epoch": 13.186813186813186,
"grad_norm": 1.1685428619384766,
"learning_rate": 0.001,
"loss": 2.479,
"step": 40800
},
{
"epoch": 13.2191338073691,
"grad_norm": 1.2040324211120605,
"learning_rate": 0.001,
"loss": 2.4951,
"step": 40900
},
{
"epoch": 13.251454427925015,
"grad_norm": 0.8940082788467407,
"learning_rate": 0.001,
"loss": 2.4716,
"step": 41000
},
{
"epoch": 13.283775048480932,
"grad_norm": 0.9844041466712952,
"learning_rate": 0.001,
"loss": 2.5108,
"step": 41100
},
{
"epoch": 13.316095669036846,
"grad_norm": 1.133620023727417,
"learning_rate": 0.001,
"loss": 2.497,
"step": 41200
},
{
"epoch": 13.34841628959276,
"grad_norm": 0.7823963165283203,
"learning_rate": 0.001,
"loss": 2.5064,
"step": 41300
},
{
"epoch": 13.380736910148675,
"grad_norm": 0.8864171504974365,
"learning_rate": 0.001,
"loss": 2.5129,
"step": 41400
},
{
"epoch": 13.41305753070459,
"grad_norm": 1.1657073497772217,
"learning_rate": 0.001,
"loss": 2.4961,
"step": 41500
},
{
"epoch": 13.445378151260504,
"grad_norm": 0.902041494846344,
"learning_rate": 0.001,
"loss": 2.5202,
"step": 41600
},
{
"epoch": 13.477698771816419,
"grad_norm": 0.9899479150772095,
"learning_rate": 0.001,
"loss": 2.5137,
"step": 41700
},
{
"epoch": 13.510019392372334,
"grad_norm": 0.9798682332038879,
"learning_rate": 0.001,
"loss": 2.5318,
"step": 41800
},
{
"epoch": 13.542340012928248,
"grad_norm": 1.3434364795684814,
"learning_rate": 0.001,
"loss": 2.5318,
"step": 41900
},
{
"epoch": 13.574660633484163,
"grad_norm": 0.7659310102462769,
"learning_rate": 0.001,
"loss": 2.5224,
"step": 42000
},
{
"epoch": 13.606981254040077,
"grad_norm": 0.9587400555610657,
"learning_rate": 0.001,
"loss": 2.5561,
"step": 42100
},
{
"epoch": 13.639301874595992,
"grad_norm": 1.20130455493927,
"learning_rate": 0.001,
"loss": 2.5346,
"step": 42200
},
{
"epoch": 13.671622495151906,
"grad_norm": 0.8852421045303345,
"learning_rate": 0.001,
"loss": 2.5409,
"step": 42300
},
{
"epoch": 13.70394311570782,
"grad_norm": 1.105299949645996,
"learning_rate": 0.001,
"loss": 2.5341,
"step": 42400
},
{
"epoch": 13.736263736263737,
"grad_norm": 1.2417206764221191,
"learning_rate": 0.001,
"loss": 2.5453,
"step": 42500
},
{
"epoch": 13.768584356819652,
"grad_norm": 1.169151782989502,
"learning_rate": 0.001,
"loss": 2.5559,
"step": 42600
},
{
"epoch": 13.800904977375566,
"grad_norm": 1.2130471467971802,
"learning_rate": 0.001,
"loss": 2.5499,
"step": 42700
},
{
"epoch": 13.83322559793148,
"grad_norm": 0.7814033627510071,
"learning_rate": 0.001,
"loss": 2.5647,
"step": 42800
},
{
"epoch": 13.865546218487395,
"grad_norm": 1.0768963098526,
"learning_rate": 0.001,
"loss": 2.5672,
"step": 42900
},
{
"epoch": 13.89786683904331,
"grad_norm": 1.283257246017456,
"learning_rate": 0.001,
"loss": 2.5308,
"step": 43000
},
{
"epoch": 13.930187459599225,
"grad_norm": 0.986774742603302,
"learning_rate": 0.001,
"loss": 2.5567,
"step": 43100
},
{
"epoch": 13.96250808015514,
"grad_norm": 1.0447334051132202,
"learning_rate": 0.001,
"loss": 2.5611,
"step": 43200
},
{
"epoch": 13.994828700711054,
"grad_norm": 1.1663225889205933,
"learning_rate": 0.001,
"loss": 2.5407,
"step": 43300
},
{
"epoch": 14.027149321266968,
"grad_norm": 0.9095755219459534,
"learning_rate": 0.001,
"loss": 2.451,
"step": 43400
},
{
"epoch": 14.059469941822883,
"grad_norm": 1.277373194694519,
"learning_rate": 0.001,
"loss": 2.4433,
"step": 43500
},
{
"epoch": 14.091790562378797,
"grad_norm": 1.4866878986358643,
"learning_rate": 0.001,
"loss": 2.4439,
"step": 43600
},
{
"epoch": 14.124111182934712,
"grad_norm": 0.9406186938285828,
"learning_rate": 0.001,
"loss": 2.4218,
"step": 43700
},
{
"epoch": 14.156431803490626,
"grad_norm": 1.0313225984573364,
"learning_rate": 0.001,
"loss": 2.4476,
"step": 43800
},
{
"epoch": 14.188752424046541,
"grad_norm": 1.0816972255706787,
"learning_rate": 0.001,
"loss": 2.4564,
"step": 43900
},
{
"epoch": 14.221073044602456,
"grad_norm": 0.8819558620452881,
"learning_rate": 0.001,
"loss": 2.447,
"step": 44000
},
{
"epoch": 14.25339366515837,
"grad_norm": 1.2793830633163452,
"learning_rate": 0.001,
"loss": 2.429,
"step": 44100
},
{
"epoch": 14.285714285714286,
"grad_norm": 0.8246544003486633,
"learning_rate": 0.001,
"loss": 2.4459,
"step": 44200
},
{
"epoch": 14.318034906270201,
"grad_norm": 1.0620863437652588,
"learning_rate": 0.001,
"loss": 2.4579,
"step": 44300
},
{
"epoch": 14.350355526826116,
"grad_norm": 1.105036735534668,
"learning_rate": 0.001,
"loss": 2.4666,
"step": 44400
},
{
"epoch": 14.38267614738203,
"grad_norm": 1.5515886545181274,
"learning_rate": 0.001,
"loss": 2.4825,
"step": 44500
},
{
"epoch": 14.414996767937945,
"grad_norm": 1.372827172279358,
"learning_rate": 0.001,
"loss": 2.4767,
"step": 44600
},
{
"epoch": 14.44731738849386,
"grad_norm": 1.1592376232147217,
"learning_rate": 0.001,
"loss": 2.4683,
"step": 44700
},
{
"epoch": 14.479638009049774,
"grad_norm": 0.930735170841217,
"learning_rate": 0.001,
"loss": 2.48,
"step": 44800
},
{
"epoch": 14.511958629605688,
"grad_norm": 0.8660281300544739,
"learning_rate": 0.001,
"loss": 2.4721,
"step": 44900
},
{
"epoch": 14.544279250161603,
"grad_norm": 0.9301630854606628,
"learning_rate": 0.001,
"loss": 2.4801,
"step": 45000
},
{
"epoch": 14.576599870717518,
"grad_norm": 1.1060216426849365,
"learning_rate": 0.001,
"loss": 2.4854,
"step": 45100
},
{
"epoch": 14.608920491273432,
"grad_norm": 0.9152036309242249,
"learning_rate": 0.001,
"loss": 2.4801,
"step": 45200
},
{
"epoch": 14.641241111829347,
"grad_norm": 0.9849165678024292,
"learning_rate": 0.001,
"loss": 2.4937,
"step": 45300
},
{
"epoch": 14.673561732385261,
"grad_norm": 1.0002423524856567,
"learning_rate": 0.001,
"loss": 2.5082,
"step": 45400
},
{
"epoch": 14.705882352941176,
"grad_norm": 1.0424295663833618,
"learning_rate": 0.001,
"loss": 2.5022,
"step": 45500
},
{
"epoch": 14.738202973497092,
"grad_norm": 1.7369807958602905,
"learning_rate": 0.001,
"loss": 2.5166,
"step": 45600
},
{
"epoch": 14.770523594053007,
"grad_norm": 1.1243922710418701,
"learning_rate": 0.001,
"loss": 2.4914,
"step": 45700
},
{
"epoch": 14.802844214608921,
"grad_norm": 1.3372899293899536,
"learning_rate": 0.001,
"loss": 2.5016,
"step": 45800
},
{
"epoch": 14.835164835164836,
"grad_norm": 1.1591620445251465,
"learning_rate": 0.001,
"loss": 2.4887,
"step": 45900
},
{
"epoch": 14.86748545572075,
"grad_norm": 0.9072688221931458,
"learning_rate": 0.001,
"loss": 2.5055,
"step": 46000
},
{
"epoch": 14.899806076276665,
"grad_norm": 1.1313115358352661,
"learning_rate": 0.001,
"loss": 2.4902,
"step": 46100
},
{
"epoch": 14.93212669683258,
"grad_norm": 1.0370368957519531,
"learning_rate": 0.001,
"loss": 2.519,
"step": 46200
},
{
"epoch": 14.964447317388494,
"grad_norm": 1.276435375213623,
"learning_rate": 0.001,
"loss": 2.5295,
"step": 46300
},
{
"epoch": 14.996767937944409,
"grad_norm": 1.1384437084197998,
"learning_rate": 0.001,
"loss": 2.4979,
"step": 46400
},
{
"epoch": 15.029088558500323,
"grad_norm": 1.1041855812072754,
"learning_rate": 0.001,
"loss": 2.3842,
"step": 46500
},
{
"epoch": 15.061409179056238,
"grad_norm": 1.1491726636886597,
"learning_rate": 0.001,
"loss": 2.3783,
"step": 46600
},
{
"epoch": 15.093729799612152,
"grad_norm": 1.016539216041565,
"learning_rate": 0.001,
"loss": 2.3901,
"step": 46700
},
{
"epoch": 15.126050420168067,
"grad_norm": 0.9626669883728027,
"learning_rate": 0.001,
"loss": 2.3968,
"step": 46800
},
{
"epoch": 15.158371040723981,
"grad_norm": 1.133621335029602,
"learning_rate": 0.001,
"loss": 2.4057,
"step": 46900
},
{
"epoch": 15.190691661279896,
"grad_norm": 0.9811501502990723,
"learning_rate": 0.001,
"loss": 2.3961,
"step": 47000
},
{
"epoch": 15.22301228183581,
"grad_norm": 0.891275942325592,
"learning_rate": 0.001,
"loss": 2.4038,
"step": 47100
},
{
"epoch": 15.255332902391725,
"grad_norm": 1.2464264631271362,
"learning_rate": 0.001,
"loss": 2.3949,
"step": 47200
},
{
"epoch": 15.287653522947641,
"grad_norm": 1.0720608234405518,
"learning_rate": 0.001,
"loss": 2.4127,
"step": 47300
},
{
"epoch": 15.319974143503556,
"grad_norm": 1.0240511894226074,
"learning_rate": 0.001,
"loss": 2.4289,
"step": 47400
},
{
"epoch": 15.35229476405947,
"grad_norm": 0.9802812933921814,
"learning_rate": 0.001,
"loss": 2.4235,
"step": 47500
},
{
"epoch": 15.384615384615385,
"grad_norm": 1.444469928741455,
"learning_rate": 0.001,
"loss": 2.4092,
"step": 47600
},
{
"epoch": 15.4169360051713,
"grad_norm": 1.0440921783447266,
"learning_rate": 0.001,
"loss": 2.4263,
"step": 47700
},
{
"epoch": 15.449256625727214,
"grad_norm": 1.1413503885269165,
"learning_rate": 0.001,
"loss": 2.4231,
"step": 47800
},
{
"epoch": 15.481577246283129,
"grad_norm": 1.2500303983688354,
"learning_rate": 0.001,
"loss": 2.4231,
"step": 47900
},
{
"epoch": 15.513897866839043,
"grad_norm": 1.2014676332473755,
"learning_rate": 0.001,
"loss": 2.4698,
"step": 48000
},
{
"epoch": 15.546218487394958,
"grad_norm": 0.983052134513855,
"learning_rate": 0.001,
"loss": 2.4436,
"step": 48100
},
{
"epoch": 15.578539107950872,
"grad_norm": 1.1271555423736572,
"learning_rate": 0.001,
"loss": 2.4446,
"step": 48200
},
{
"epoch": 15.610859728506787,
"grad_norm": 1.2637511491775513,
"learning_rate": 0.001,
"loss": 2.4577,
"step": 48300
},
{
"epoch": 15.643180349062701,
"grad_norm": 1.056535005569458,
"learning_rate": 0.001,
"loss": 2.4348,
"step": 48400
},
{
"epoch": 15.675500969618616,
"grad_norm": 1.1217752695083618,
"learning_rate": 0.001,
"loss": 2.4583,
"step": 48500
},
{
"epoch": 15.70782159017453,
"grad_norm": 1.128398060798645,
"learning_rate": 0.001,
"loss": 2.45,
"step": 48600
},
{
"epoch": 15.740142210730447,
"grad_norm": 1.2292331457138062,
"learning_rate": 0.001,
"loss": 2.4616,
"step": 48700
},
{
"epoch": 15.772462831286362,
"grad_norm": 1.3311684131622314,
"learning_rate": 0.001,
"loss": 2.4588,
"step": 48800
},
{
"epoch": 15.804783451842276,
"grad_norm": 1.1817072629928589,
"learning_rate": 0.001,
"loss": 2.4483,
"step": 48900
},
{
"epoch": 15.83710407239819,
"grad_norm": 1.0943495035171509,
"learning_rate": 0.001,
"loss": 2.4488,
"step": 49000
},
{
"epoch": 15.869424692954105,
"grad_norm": 0.9920291900634766,
"learning_rate": 0.001,
"loss": 2.4668,
"step": 49100
},
{
"epoch": 15.90174531351002,
"grad_norm": 1.0489287376403809,
"learning_rate": 0.001,
"loss": 2.4693,
"step": 49200
},
{
"epoch": 15.934065934065934,
"grad_norm": 1.1572085618972778,
"learning_rate": 0.001,
"loss": 2.4717,
"step": 49300
},
{
"epoch": 15.966386554621849,
"grad_norm": 1.326959490776062,
"learning_rate": 0.001,
"loss": 2.4671,
"step": 49400
},
{
"epoch": 15.998707175177763,
"grad_norm": 1.4123055934906006,
"learning_rate": 0.001,
"loss": 2.4818,
"step": 49500
},
{
"epoch": 16.031027795733678,
"grad_norm": 1.2705885171890259,
"learning_rate": 0.001,
"loss": 2.33,
"step": 49600
},
{
"epoch": 16.063348416289593,
"grad_norm": 1.2567096948623657,
"learning_rate": 0.001,
"loss": 2.3412,
"step": 49700
},
{
"epoch": 16.095669036845507,
"grad_norm": 1.5731749534606934,
"learning_rate": 0.001,
"loss": 2.3512,
"step": 49800
},
{
"epoch": 16.12798965740142,
"grad_norm": 1.4122164249420166,
"learning_rate": 0.001,
"loss": 2.3338,
"step": 49900
},
{
"epoch": 16.160310277957336,
"grad_norm": 1.4363222122192383,
"learning_rate": 0.001,
"loss": 2.3657,
"step": 50000
},
{
"epoch": 16.19263089851325,
"grad_norm": 1.142403483390808,
"learning_rate": 0.001,
"loss": 2.3692,
"step": 50100
},
{
"epoch": 16.224951519069165,
"grad_norm": 1.4952629804611206,
"learning_rate": 0.001,
"loss": 2.3821,
"step": 50200
},
{
"epoch": 16.25727213962508,
"grad_norm": 1.2392936944961548,
"learning_rate": 0.001,
"loss": 2.3703,
"step": 50300
},
{
"epoch": 16.289592760180994,
"grad_norm": 1.2250338792800903,
"learning_rate": 0.001,
"loss": 2.3694,
"step": 50400
},
{
"epoch": 16.32191338073691,
"grad_norm": 1.552544116973877,
"learning_rate": 0.001,
"loss": 2.3612,
"step": 50500
},
{
"epoch": 16.354234001292824,
"grad_norm": 1.299159288406372,
"learning_rate": 0.001,
"loss": 2.3811,
"step": 50600
},
{
"epoch": 16.386554621848738,
"grad_norm": 1.6203311681747437,
"learning_rate": 0.001,
"loss": 2.3887,
"step": 50700
},
{
"epoch": 16.418875242404653,
"grad_norm": 1.1839911937713623,
"learning_rate": 0.001,
"loss": 2.379,
"step": 50800
},
{
"epoch": 16.451195862960567,
"grad_norm": 1.0918307304382324,
"learning_rate": 0.001,
"loss": 2.3992,
"step": 50900
},
{
"epoch": 16.483516483516482,
"grad_norm": 1.2795827388763428,
"learning_rate": 0.001,
"loss": 2.3864,
"step": 51000
},
{
"epoch": 16.5158371040724,
"grad_norm": 1.7874277830123901,
"learning_rate": 0.001,
"loss": 2.3989,
"step": 51100
},
{
"epoch": 16.548157724628314,
"grad_norm": 1.17491614818573,
"learning_rate": 0.001,
"loss": 2.3982,
"step": 51200
},
{
"epoch": 16.58047834518423,
"grad_norm": 1.3821208477020264,
"learning_rate": 0.001,
"loss": 2.4158,
"step": 51300
},
{
"epoch": 16.612798965740144,
"grad_norm": 1.196932077407837,
"learning_rate": 0.001,
"loss": 2.4133,
"step": 51400
},
{
"epoch": 16.645119586296058,
"grad_norm": 1.2657791376113892,
"learning_rate": 0.001,
"loss": 2.4135,
"step": 51500
},
{
"epoch": 16.677440206851973,
"grad_norm": 1.3523930311203003,
"learning_rate": 0.001,
"loss": 2.4179,
"step": 51600
},
{
"epoch": 16.709760827407887,
"grad_norm": 1.7938930988311768,
"learning_rate": 0.001,
"loss": 2.431,
"step": 51700
},
{
"epoch": 16.742081447963802,
"grad_norm": 1.237819790840149,
"learning_rate": 0.001,
"loss": 2.4016,
"step": 51800
},
{
"epoch": 16.774402068519716,
"grad_norm": 1.070419192314148,
"learning_rate": 0.001,
"loss": 2.4219,
"step": 51900
},
{
"epoch": 16.80672268907563,
"grad_norm": 1.1201379299163818,
"learning_rate": 0.001,
"loss": 2.402,
"step": 52000
},
{
"epoch": 16.839043309631545,
"grad_norm": 1.8362864255905151,
"learning_rate": 0.001,
"loss": 2.434,
"step": 52100
},
{
"epoch": 16.87136393018746,
"grad_norm": 1.5529481172561646,
"learning_rate": 0.001,
"loss": 2.4161,
"step": 52200
},
{
"epoch": 16.903684550743375,
"grad_norm": 1.161968469619751,
"learning_rate": 0.001,
"loss": 2.4176,
"step": 52300
},
{
"epoch": 16.93600517129929,
"grad_norm": 1.3376487493515015,
"learning_rate": 0.001,
"loss": 2.4329,
"step": 52400
},
{
"epoch": 16.968325791855204,
"grad_norm": 1.5274125337600708,
"learning_rate": 0.001,
"loss": 2.4224,
"step": 52500
},
{
"epoch": 17.00064641241112,
"grad_norm": 1.3048698902130127,
"learning_rate": 0.001,
"loss": 2.4184,
"step": 52600
},
{
"epoch": 17.032967032967033,
"grad_norm": 0.9883232116699219,
"learning_rate": 0.001,
"loss": 2.2798,
"step": 52700
},
{
"epoch": 17.065287653522947,
"grad_norm": 1.5406153202056885,
"learning_rate": 0.001,
"loss": 2.3052,
"step": 52800
},
{
"epoch": 17.097608274078862,
"grad_norm": 1.5757269859313965,
"learning_rate": 0.001,
"loss": 2.2945,
"step": 52900
},
{
"epoch": 17.129928894634777,
"grad_norm": 1.1663100719451904,
"learning_rate": 0.001,
"loss": 2.3198,
"step": 53000
},
{
"epoch": 17.16224951519069,
"grad_norm": 1.7342582941055298,
"learning_rate": 0.001,
"loss": 2.3012,
"step": 53100
},
{
"epoch": 17.194570135746606,
"grad_norm": 1.9971946477890015,
"learning_rate": 0.001,
"loss": 2.3041,
"step": 53200
},
{
"epoch": 17.22689075630252,
"grad_norm": 1.5045979022979736,
"learning_rate": 0.001,
"loss": 2.3184,
"step": 53300
},
{
"epoch": 17.259211376858435,
"grad_norm": 1.5043833255767822,
"learning_rate": 0.001,
"loss": 2.3472,
"step": 53400
},
{
"epoch": 17.29153199741435,
"grad_norm": 1.8031812906265259,
"learning_rate": 0.001,
"loss": 2.3234,
"step": 53500
},
{
"epoch": 17.323852617970264,
"grad_norm": 1.3667677640914917,
"learning_rate": 0.001,
"loss": 2.354,
"step": 53600
},
{
"epoch": 17.35617323852618,
"grad_norm": 1.7247684001922607,
"learning_rate": 0.001,
"loss": 2.3296,
"step": 53700
},
{
"epoch": 17.388493859082093,
"grad_norm": 1.4302070140838623,
"learning_rate": 0.001,
"loss": 2.3398,
"step": 53800
},
{
"epoch": 17.420814479638008,
"grad_norm": 1.4096903800964355,
"learning_rate": 0.001,
"loss": 2.3537,
"step": 53900
},
{
"epoch": 17.453135100193922,
"grad_norm": 1.7663718461990356,
"learning_rate": 0.001,
"loss": 2.3767,
"step": 54000
},
{
"epoch": 17.485455720749837,
"grad_norm": 1.3540042638778687,
"learning_rate": 0.001,
"loss": 2.3635,
"step": 54100
},
{
"epoch": 17.517776341305755,
"grad_norm": 1.4982707500457764,
"learning_rate": 0.001,
"loss": 2.3645,
"step": 54200
},
{
"epoch": 17.55009696186167,
"grad_norm": 1.199114203453064,
"learning_rate": 0.001,
"loss": 2.3378,
"step": 54300
},
{
"epoch": 17.582417582417584,
"grad_norm": 1.439750075340271,
"learning_rate": 0.001,
"loss": 2.3431,
"step": 54400
},
{
"epoch": 17.6147382029735,
"grad_norm": 1.6232794523239136,
"learning_rate": 0.001,
"loss": 2.3878,
"step": 54500
},
{
"epoch": 17.647058823529413,
"grad_norm": 1.4514551162719727,
"learning_rate": 0.001,
"loss": 2.3807,
"step": 54600
},
{
"epoch": 17.679379444085328,
"grad_norm": 1.4568724632263184,
"learning_rate": 0.001,
"loss": 2.3868,
"step": 54700
},
{
"epoch": 17.711700064641242,
"grad_norm": 1.4354010820388794,
"learning_rate": 0.001,
"loss": 2.3818,
"step": 54800
},
{
"epoch": 17.744020685197157,
"grad_norm": 1.4518014192581177,
"learning_rate": 0.001,
"loss": 2.3937,
"step": 54900
},
{
"epoch": 17.77634130575307,
"grad_norm": 1.2550804615020752,
"learning_rate": 0.001,
"loss": 2.3716,
"step": 55000
},
{
"epoch": 17.808661926308986,
"grad_norm": 1.4425424337387085,
"learning_rate": 0.001,
"loss": 2.3821,
"step": 55100
},
{
"epoch": 17.8409825468649,
"grad_norm": 1.390568494796753,
"learning_rate": 0.001,
"loss": 2.3795,
"step": 55200
},
{
"epoch": 17.873303167420815,
"grad_norm": 0.9332477450370789,
"learning_rate": 0.001,
"loss": 2.4018,
"step": 55300
},
{
"epoch": 17.90562378797673,
"grad_norm": 1.5412309169769287,
"learning_rate": 0.001,
"loss": 2.3888,
"step": 55400
},
{
"epoch": 17.937944408532644,
"grad_norm": 1.3093478679656982,
"learning_rate": 0.001,
"loss": 2.3945,
"step": 55500
},
{
"epoch": 17.97026502908856,
"grad_norm": 1.901310682296753,
"learning_rate": 0.001,
"loss": 2.4127,
"step": 55600
},
{
"epoch": 18.002585649644473,
"grad_norm": 1.3532161712646484,
"learning_rate": 0.001,
"loss": 2.4015,
"step": 55700
},
{
"epoch": 18.034906270200388,
"grad_norm": 1.2525259256362915,
"learning_rate": 0.001,
"loss": 2.2381,
"step": 55800
},
{
"epoch": 18.067226890756302,
"grad_norm": 1.2603477239608765,
"learning_rate": 0.001,
"loss": 2.2665,
"step": 55900
},
{
"epoch": 18.099547511312217,
"grad_norm": 1.0956732034683228,
"learning_rate": 0.001,
"loss": 2.2624,
"step": 56000
},
{
"epoch": 18.13186813186813,
"grad_norm": 1.2281423807144165,
"learning_rate": 0.001,
"loss": 2.2891,
"step": 56100
},
{
"epoch": 18.164188752424046,
"grad_norm": 0.9731630682945251,
"learning_rate": 0.001,
"loss": 2.2886,
"step": 56200
},
{
"epoch": 18.19650937297996,
"grad_norm": 1.124035120010376,
"learning_rate": 0.001,
"loss": 2.2811,
"step": 56300
},
{
"epoch": 18.228829993535875,
"grad_norm": 0.942611813545227,
"learning_rate": 0.001,
"loss": 2.3107,
"step": 56400
},
{
"epoch": 18.26115061409179,
"grad_norm": 1.277614712715149,
"learning_rate": 0.001,
"loss": 2.279,
"step": 56500
},
{
"epoch": 18.293471234647704,
"grad_norm": 1.4019912481307983,
"learning_rate": 0.001,
"loss": 2.3058,
"step": 56600
},
{
"epoch": 18.32579185520362,
"grad_norm": 1.1190897226333618,
"learning_rate": 0.001,
"loss": 2.3308,
"step": 56700
},
{
"epoch": 18.358112475759533,
"grad_norm": 1.0097274780273438,
"learning_rate": 0.001,
"loss": 2.2934,
"step": 56800
},
{
"epoch": 18.390433096315448,
"grad_norm": 1.0941705703735352,
"learning_rate": 0.001,
"loss": 2.3182,
"step": 56900
},
{
"epoch": 18.422753716871362,
"grad_norm": 1.2601319551467896,
"learning_rate": 0.001,
"loss": 2.3211,
"step": 57000
},
{
"epoch": 18.455074337427277,
"grad_norm": 1.3086024522781372,
"learning_rate": 0.001,
"loss": 2.3223,
"step": 57100
},
{
"epoch": 18.48739495798319,
"grad_norm": 1.0803699493408203,
"learning_rate": 0.001,
"loss": 2.3251,
"step": 57200
},
{
"epoch": 18.51971557853911,
"grad_norm": 1.0027995109558105,
"learning_rate": 0.001,
"loss": 2.3212,
"step": 57300
},
{
"epoch": 18.552036199095024,
"grad_norm": 1.0489871501922607,
"learning_rate": 0.001,
"loss": 2.3274,
"step": 57400
},
{
"epoch": 18.58435681965094,
"grad_norm": 1.2849969863891602,
"learning_rate": 0.001,
"loss": 2.3188,
"step": 57500
},
{
"epoch": 18.616677440206853,
"grad_norm": 1.116881251335144,
"learning_rate": 0.001,
"loss": 2.3224,
"step": 57600
},
{
"epoch": 18.648998060762768,
"grad_norm": 1.5892244577407837,
"learning_rate": 0.001,
"loss": 2.33,
"step": 57700
},
{
"epoch": 18.681318681318682,
"grad_norm": 1.2354624271392822,
"learning_rate": 0.001,
"loss": 2.3352,
"step": 57800
},
{
"epoch": 18.713639301874597,
"grad_norm": 1.3601680994033813,
"learning_rate": 0.001,
"loss": 2.3471,
"step": 57900
},
{
"epoch": 18.74595992243051,
"grad_norm": 1.171592116355896,
"learning_rate": 0.001,
"loss": 2.3425,
"step": 58000
},
{
"epoch": 18.778280542986426,
"grad_norm": 1.2357901334762573,
"learning_rate": 0.001,
"loss": 2.3452,
"step": 58100
},
{
"epoch": 18.81060116354234,
"grad_norm": 1.2789829969406128,
"learning_rate": 0.001,
"loss": 2.3234,
"step": 58200
},
{
"epoch": 18.842921784098255,
"grad_norm": 1.4807987213134766,
"learning_rate": 0.001,
"loss": 2.342,
"step": 58300
},
{
"epoch": 18.87524240465417,
"grad_norm": 1.4539543390274048,
"learning_rate": 0.001,
"loss": 2.3604,
"step": 58400
},
{
"epoch": 18.907563025210084,
"grad_norm": 1.1573584079742432,
"learning_rate": 0.001,
"loss": 2.3858,
"step": 58500
},
{
"epoch": 18.939883645766,
"grad_norm": 1.3071974515914917,
"learning_rate": 0.001,
"loss": 2.3529,
"step": 58600
},
{
"epoch": 18.972204266321913,
"grad_norm": 1.4395296573638916,
"learning_rate": 0.001,
"loss": 2.3575,
"step": 58700
},
{
"epoch": 19.004524886877828,
"grad_norm": 1.0174015760421753,
"learning_rate": 0.001,
"loss": 2.361,
"step": 58800
},
{
"epoch": 19.036845507433743,
"grad_norm": 1.0709799528121948,
"learning_rate": 0.001,
"loss": 2.2061,
"step": 58900
},
{
"epoch": 19.069166127989657,
"grad_norm": 0.8764005303382874,
"learning_rate": 0.001,
"loss": 2.2141,
"step": 59000
},
{
"epoch": 19.10148674854557,
"grad_norm": 1.2594536542892456,
"learning_rate": 0.001,
"loss": 2.2382,
"step": 59100
},
{
"epoch": 19.133807369101486,
"grad_norm": 1.1866902112960815,
"learning_rate": 0.001,
"loss": 2.2445,
"step": 59200
},
{
"epoch": 19.1661279896574,
"grad_norm": 0.8631852269172668,
"learning_rate": 0.001,
"loss": 2.2552,
"step": 59300
},
{
"epoch": 19.198448610213315,
"grad_norm": 1.157242774963379,
"learning_rate": 0.001,
"loss": 2.2657,
"step": 59400
},
{
"epoch": 19.23076923076923,
"grad_norm": 1.1106163263320923,
"learning_rate": 0.001,
"loss": 2.2542,
"step": 59500
},
{
"epoch": 19.263089851325145,
"grad_norm": 1.3358241319656372,
"learning_rate": 0.001,
"loss": 2.2752,
"step": 59600
},
{
"epoch": 19.29541047188106,
"grad_norm": 1.267776370048523,
"learning_rate": 0.001,
"loss": 2.2703,
"step": 59700
},
{
"epoch": 19.327731092436974,
"grad_norm": 1.3746817111968994,
"learning_rate": 0.001,
"loss": 2.2788,
"step": 59800
},
{
"epoch": 19.360051712992888,
"grad_norm": 1.1478110551834106,
"learning_rate": 0.001,
"loss": 2.2755,
"step": 59900
},
{
"epoch": 19.392372333548803,
"grad_norm": 1.1277999877929688,
"learning_rate": 0.001,
"loss": 2.2905,
"step": 60000
},
{
"epoch": 19.424692954104717,
"grad_norm": 1.1398897171020508,
"learning_rate": 0.001,
"loss": 2.2888,
"step": 60100
},
{
"epoch": 19.457013574660632,
"grad_norm": 1.531358003616333,
"learning_rate": 0.001,
"loss": 2.2732,
"step": 60200
},
{
"epoch": 19.489334195216546,
"grad_norm": 1.4275100231170654,
"learning_rate": 0.001,
"loss": 2.2957,
"step": 60300
},
{
"epoch": 19.521654815772465,
"grad_norm": 1.1466721296310425,
"learning_rate": 0.001,
"loss": 2.2934,
"step": 60400
},
{
"epoch": 19.55397543632838,
"grad_norm": 1.0285093784332275,
"learning_rate": 0.001,
"loss": 2.2973,
"step": 60500
},
{
"epoch": 19.586296056884294,
"grad_norm": 1.2563018798828125,
"learning_rate": 0.001,
"loss": 2.294,
"step": 60600
},
{
"epoch": 19.618616677440208,
"grad_norm": 1.1977994441986084,
"learning_rate": 0.001,
"loss": 2.2995,
"step": 60700
},
{
"epoch": 19.650937297996123,
"grad_norm": 0.9878324866294861,
"learning_rate": 0.001,
"loss": 2.3079,
"step": 60800
},
{
"epoch": 19.683257918552037,
"grad_norm": 1.6006507873535156,
"learning_rate": 0.001,
"loss": 2.2877,
"step": 60900
},
{
"epoch": 19.715578539107952,
"grad_norm": 1.2420533895492554,
"learning_rate": 0.001,
"loss": 2.3072,
"step": 61000
},
{
"epoch": 19.747899159663866,
"grad_norm": 1.2223336696624756,
"learning_rate": 0.001,
"loss": 2.3157,
"step": 61100
},
{
"epoch": 19.78021978021978,
"grad_norm": 1.2445229291915894,
"learning_rate": 0.001,
"loss": 2.3121,
"step": 61200
},
{
"epoch": 19.812540400775696,
"grad_norm": 1.8219184875488281,
"learning_rate": 0.001,
"loss": 2.3058,
"step": 61300
},
{
"epoch": 19.84486102133161,
"grad_norm": 1.043376088142395,
"learning_rate": 0.001,
"loss": 2.3184,
"step": 61400
},
{
"epoch": 19.877181641887525,
"grad_norm": 1.5648939609527588,
"learning_rate": 0.001,
"loss": 2.3238,
"step": 61500
},
{
"epoch": 19.90950226244344,
"grad_norm": 1.4817874431610107,
"learning_rate": 0.001,
"loss": 2.3302,
"step": 61600
},
{
"epoch": 19.941822882999354,
"grad_norm": 0.971615731716156,
"learning_rate": 0.001,
"loss": 2.3164,
"step": 61700
},
{
"epoch": 19.97414350355527,
"grad_norm": 0.9662232398986816,
"learning_rate": 0.001,
"loss": 2.3247,
"step": 61800
},
{
"epoch": 20.006464124111183,
"grad_norm": 1.4232537746429443,
"learning_rate": 0.001,
"loss": 2.3,
"step": 61900
},
{
"epoch": 20.038784744667097,
"grad_norm": 1.1577904224395752,
"learning_rate": 0.001,
"loss": 2.1962,
"step": 62000
},
{
"epoch": 20.071105365223012,
"grad_norm": 1.4552325010299683,
"learning_rate": 0.001,
"loss": 2.212,
"step": 62100
},
{
"epoch": 20.103425985778927,
"grad_norm": 1.3533803224563599,
"learning_rate": 0.001,
"loss": 2.1985,
"step": 62200
},
{
"epoch": 20.13574660633484,
"grad_norm": 1.1538723707199097,
"learning_rate": 0.001,
"loss": 2.2143,
"step": 62300
},
{
"epoch": 20.168067226890756,
"grad_norm": 0.9910172820091248,
"learning_rate": 0.001,
"loss": 2.2179,
"step": 62400
},
{
"epoch": 20.20038784744667,
"grad_norm": 1.3591842651367188,
"learning_rate": 0.001,
"loss": 2.2123,
"step": 62500
},
{
"epoch": 20.232708468002585,
"grad_norm": 1.0117110013961792,
"learning_rate": 0.001,
"loss": 2.2186,
"step": 62600
},
{
"epoch": 20.2650290885585,
"grad_norm": 1.2286127805709839,
"learning_rate": 0.001,
"loss": 2.214,
"step": 62700
},
{
"epoch": 20.297349709114414,
"grad_norm": 1.016156554222107,
"learning_rate": 0.001,
"loss": 2.2558,
"step": 62800
},
{
"epoch": 20.32967032967033,
"grad_norm": 0.9660277962684631,
"learning_rate": 0.001,
"loss": 2.2326,
"step": 62900
},
{
"epoch": 20.361990950226243,
"grad_norm": 0.8462151288986206,
"learning_rate": 0.001,
"loss": 2.2672,
"step": 63000
},
{
"epoch": 20.394311570782158,
"grad_norm": 1.1386162042617798,
"learning_rate": 0.001,
"loss": 2.2394,
"step": 63100
},
{
"epoch": 20.426632191338072,
"grad_norm": 1.1773128509521484,
"learning_rate": 0.001,
"loss": 2.2426,
"step": 63200
},
{
"epoch": 20.458952811893987,
"grad_norm": 0.9726552963256836,
"learning_rate": 0.001,
"loss": 2.2426,
"step": 63300
},
{
"epoch": 20.4912734324499,
"grad_norm": 1.0654945373535156,
"learning_rate": 0.001,
"loss": 2.2582,
"step": 63400
},
{
"epoch": 20.52359405300582,
"grad_norm": 1.2010523080825806,
"learning_rate": 0.001,
"loss": 2.2476,
"step": 63500
},
{
"epoch": 20.555914673561734,
"grad_norm": 1.315597414970398,
"learning_rate": 0.001,
"loss": 2.2736,
"step": 63600
},
{
"epoch": 20.58823529411765,
"grad_norm": 0.8830787539482117,
"learning_rate": 0.001,
"loss": 2.2654,
"step": 63700
},
{
"epoch": 20.620555914673563,
"grad_norm": 1.2181121110916138,
"learning_rate": 0.001,
"loss": 2.2586,
"step": 63800
},
{
"epoch": 20.652876535229478,
"grad_norm": 1.2166556119918823,
"learning_rate": 0.001,
"loss": 2.2465,
"step": 63900
},
{
"epoch": 20.685197155785392,
"grad_norm": 1.32142174243927,
"learning_rate": 0.001,
"loss": 2.2658,
"step": 64000
},
{
"epoch": 20.717517776341307,
"grad_norm": 1.0378272533416748,
"learning_rate": 0.001,
"loss": 2.2781,
"step": 64100
},
{
"epoch": 20.74983839689722,
"grad_norm": 1.0165777206420898,
"learning_rate": 0.001,
"loss": 2.2877,
"step": 64200
},
{
"epoch": 20.782159017453136,
"grad_norm": 1.2062556743621826,
"learning_rate": 0.001,
"loss": 2.2772,
"step": 64300
},
{
"epoch": 20.81447963800905,
"grad_norm": 1.2662969827651978,
"learning_rate": 0.001,
"loss": 2.2942,
"step": 64400
},
{
"epoch": 20.846800258564965,
"grad_norm": 1.2018824815750122,
"learning_rate": 0.001,
"loss": 2.2825,
"step": 64500
},
{
"epoch": 20.87912087912088,
"grad_norm": 1.2240667343139648,
"learning_rate": 0.001,
"loss": 2.3111,
"step": 64600
},
{
"epoch": 20.911441499676794,
"grad_norm": 1.0020350217819214,
"learning_rate": 0.001,
"loss": 2.2985,
"step": 64700
},
{
"epoch": 20.94376212023271,
"grad_norm": 1.0527266263961792,
"learning_rate": 0.001,
"loss": 2.3037,
"step": 64800
},
{
"epoch": 20.976082740788623,
"grad_norm": 1.2772159576416016,
"learning_rate": 0.001,
"loss": 2.3089,
"step": 64900
},
{
"epoch": 21.008403361344538,
"grad_norm": 1.1994928121566772,
"learning_rate": 0.001,
"loss": 2.2483,
"step": 65000
},
{
"epoch": 21.040723981900452,
"grad_norm": 1.1781855821609497,
"learning_rate": 0.001,
"loss": 2.169,
"step": 65100
},
{
"epoch": 21.073044602456367,
"grad_norm": 1.084336519241333,
"learning_rate": 0.001,
"loss": 2.1714,
"step": 65200
},
{
"epoch": 21.10536522301228,
"grad_norm": 1.272647738456726,
"learning_rate": 0.001,
"loss": 2.187,
"step": 65300
},
{
"epoch": 21.137685843568196,
"grad_norm": 1.1202809810638428,
"learning_rate": 0.001,
"loss": 2.1938,
"step": 65400
},
{
"epoch": 21.17000646412411,
"grad_norm": 1.111128330230713,
"learning_rate": 0.001,
"loss": 2.1839,
"step": 65500
},
{
"epoch": 21.202327084680025,
"grad_norm": 1.043950080871582,
"learning_rate": 0.001,
"loss": 2.1939,
"step": 65600
},
{
"epoch": 21.23464770523594,
"grad_norm": 0.9977900981903076,
"learning_rate": 0.001,
"loss": 2.1967,
"step": 65700
},
{
"epoch": 21.266968325791854,
"grad_norm": 1.4740365743637085,
"learning_rate": 0.001,
"loss": 2.1907,
"step": 65800
},
{
"epoch": 21.29928894634777,
"grad_norm": 1.181904673576355,
"learning_rate": 0.001,
"loss": 2.2121,
"step": 65900
},
{
"epoch": 21.331609566903683,
"grad_norm": 1.1287751197814941,
"learning_rate": 0.001,
"loss": 2.2222,
"step": 66000
},
{
"epoch": 21.363930187459598,
"grad_norm": 1.1852822303771973,
"learning_rate": 0.001,
"loss": 2.2022,
"step": 66100
},
{
"epoch": 21.396250808015512,
"grad_norm": 1.437569499015808,
"learning_rate": 0.001,
"loss": 2.204,
"step": 66200
},
{
"epoch": 21.428571428571427,
"grad_norm": 1.4531433582305908,
"learning_rate": 0.001,
"loss": 2.2141,
"step": 66300
},
{
"epoch": 21.46089204912734,
"grad_norm": 1.152618169784546,
"learning_rate": 0.001,
"loss": 2.2234,
"step": 66400
},
{
"epoch": 21.49321266968326,
"grad_norm": 1.0934582948684692,
"learning_rate": 0.001,
"loss": 2.2,
"step": 66500
},
{
"epoch": 21.525533290239174,
"grad_norm": 1.0906548500061035,
"learning_rate": 0.001,
"loss": 2.1932,
"step": 66600
},
{
"epoch": 21.55785391079509,
"grad_norm": 1.158033847808838,
"learning_rate": 0.001,
"loss": 2.2246,
"step": 66700
},
{
"epoch": 21.590174531351003,
"grad_norm": 0.9389932751655579,
"learning_rate": 0.001,
"loss": 2.236,
"step": 66800
},
{
"epoch": 21.622495151906918,
"grad_norm": 1.1929905414581299,
"learning_rate": 0.001,
"loss": 2.2367,
"step": 66900
},
{
"epoch": 21.654815772462833,
"grad_norm": 1.5029479265213013,
"learning_rate": 0.001,
"loss": 2.2448,
"step": 67000
},
{
"epoch": 21.687136393018747,
"grad_norm": 1.0972460508346558,
"learning_rate": 0.001,
"loss": 2.2474,
"step": 67100
},
{
"epoch": 21.71945701357466,
"grad_norm": 1.2588858604431152,
"learning_rate": 0.001,
"loss": 2.2643,
"step": 67200
},
{
"epoch": 21.751777634130576,
"grad_norm": 0.977533221244812,
"learning_rate": 0.001,
"loss": 2.2572,
"step": 67300
},
{
"epoch": 21.78409825468649,
"grad_norm": 1.1182975769042969,
"learning_rate": 0.001,
"loss": 2.2725,
"step": 67400
},
{
"epoch": 21.816418875242405,
"grad_norm": 1.0630615949630737,
"learning_rate": 0.001,
"loss": 2.278,
"step": 67500
},
{
"epoch": 21.84873949579832,
"grad_norm": 1.2644822597503662,
"learning_rate": 0.001,
"loss": 2.2597,
"step": 67600
},
{
"epoch": 21.881060116354234,
"grad_norm": 1.5932047367095947,
"learning_rate": 0.001,
"loss": 2.2562,
"step": 67700
},
{
"epoch": 21.91338073691015,
"grad_norm": 1.0320112705230713,
"learning_rate": 0.001,
"loss": 2.2469,
"step": 67800
},
{
"epoch": 21.945701357466064,
"grad_norm": 0.8472131490707397,
"learning_rate": 0.001,
"loss": 2.2598,
"step": 67900
},
{
"epoch": 21.978021978021978,
"grad_norm": 1.2792388200759888,
"learning_rate": 0.001,
"loss": 2.2624,
"step": 68000
},
{
"epoch": 22.010342598577893,
"grad_norm": 1.1014400720596313,
"learning_rate": 0.001,
"loss": 2.2049,
"step": 68100
},
{
"epoch": 22.042663219133807,
"grad_norm": 1.2339328527450562,
"learning_rate": 0.001,
"loss": 2.1335,
"step": 68200
},
{
"epoch": 22.07498383968972,
"grad_norm": 1.1862666606903076,
"learning_rate": 0.001,
"loss": 2.1206,
"step": 68300
},
{
"epoch": 22.107304460245636,
"grad_norm": 1.3025742769241333,
"learning_rate": 0.001,
"loss": 2.142,
"step": 68400
},
{
"epoch": 22.13962508080155,
"grad_norm": 1.1897183656692505,
"learning_rate": 0.001,
"loss": 2.1484,
"step": 68500
},
{
"epoch": 22.171945701357465,
"grad_norm": 1.1720412969589233,
"learning_rate": 0.001,
"loss": 2.1559,
"step": 68600
},
{
"epoch": 22.20426632191338,
"grad_norm": 1.0910555124282837,
"learning_rate": 0.001,
"loss": 2.1694,
"step": 68700
},
{
"epoch": 22.236586942469295,
"grad_norm": 1.062313437461853,
"learning_rate": 0.001,
"loss": 2.156,
"step": 68800
},
{
"epoch": 22.26890756302521,
"grad_norm": 1.5841485261917114,
"learning_rate": 0.001,
"loss": 2.1473,
"step": 68900
},
{
"epoch": 22.301228183581124,
"grad_norm": 1.4714710712432861,
"learning_rate": 0.001,
"loss": 2.1686,
"step": 69000
},
{
"epoch": 22.33354880413704,
"grad_norm": 1.0223160982131958,
"learning_rate": 0.001,
"loss": 2.1496,
"step": 69100
},
{
"epoch": 22.365869424692953,
"grad_norm": 1.3570698499679565,
"learning_rate": 0.001,
"loss": 2.2032,
"step": 69200
},
{
"epoch": 22.398190045248867,
"grad_norm": 0.9952728152275085,
"learning_rate": 0.001,
"loss": 2.2026,
"step": 69300
},
{
"epoch": 22.430510665804782,
"grad_norm": 1.4773149490356445,
"learning_rate": 0.001,
"loss": 2.1917,
"step": 69400
},
{
"epoch": 22.462831286360696,
"grad_norm": 0.9691696166992188,
"learning_rate": 0.001,
"loss": 2.1851,
"step": 69500
},
{
"epoch": 22.49515190691661,
"grad_norm": 1.2015392780303955,
"learning_rate": 0.001,
"loss": 2.2027,
"step": 69600
},
{
"epoch": 22.52747252747253,
"grad_norm": 1.2544416189193726,
"learning_rate": 0.001,
"loss": 2.1899,
"step": 69700
},
{
"epoch": 22.559793148028444,
"grad_norm": 1.2024792432785034,
"learning_rate": 0.001,
"loss": 2.1959,
"step": 69800
},
{
"epoch": 22.59211376858436,
"grad_norm": 1.0212665796279907,
"learning_rate": 0.001,
"loss": 2.202,
"step": 69900
},
{
"epoch": 22.624434389140273,
"grad_norm": 0.8795430064201355,
"learning_rate": 0.001,
"loss": 2.2077,
"step": 70000
},
{
"epoch": 22.656755009696187,
"grad_norm": 1.485749363899231,
"learning_rate": 0.001,
"loss": 2.2058,
"step": 70100
},
{
"epoch": 22.689075630252102,
"grad_norm": 0.9236786365509033,
"learning_rate": 0.001,
"loss": 2.2291,
"step": 70200
},
{
"epoch": 22.721396250808017,
"grad_norm": 0.9541992545127869,
"learning_rate": 0.001,
"loss": 2.2224,
"step": 70300
},
{
"epoch": 22.75371687136393,
"grad_norm": 1.2166469097137451,
"learning_rate": 0.001,
"loss": 2.2151,
"step": 70400
},
{
"epoch": 22.786037491919846,
"grad_norm": 1.2344387769699097,
"learning_rate": 0.001,
"loss": 2.2264,
"step": 70500
},
{
"epoch": 22.81835811247576,
"grad_norm": 1.089577078819275,
"learning_rate": 0.001,
"loss": 2.2196,
"step": 70600
},
{
"epoch": 22.850678733031675,
"grad_norm": 0.9872061014175415,
"learning_rate": 0.001,
"loss": 2.2146,
"step": 70700
},
{
"epoch": 22.88299935358759,
"grad_norm": 1.1960519552230835,
"learning_rate": 0.001,
"loss": 2.2381,
"step": 70800
},
{
"epoch": 22.915319974143504,
"grad_norm": 1.0115134716033936,
"learning_rate": 0.001,
"loss": 2.2227,
"step": 70900
},
{
"epoch": 22.94764059469942,
"grad_norm": 1.0959324836730957,
"learning_rate": 0.001,
"loss": 2.2206,
"step": 71000
},
{
"epoch": 22.979961215255333,
"grad_norm": 1.0546246767044067,
"learning_rate": 0.001,
"loss": 2.2363,
"step": 71100
},
{
"epoch": 23.012281835811248,
"grad_norm": 1.3446316719055176,
"learning_rate": 0.001,
"loss": 2.1532,
"step": 71200
},
{
"epoch": 23.044602456367162,
"grad_norm": 1.4532158374786377,
"learning_rate": 0.001,
"loss": 2.1056,
"step": 71300
},
{
"epoch": 23.076923076923077,
"grad_norm": 1.0634416341781616,
"learning_rate": 0.001,
"loss": 2.0932,
"step": 71400
},
{
"epoch": 23.10924369747899,
"grad_norm": 1.5012799501419067,
"learning_rate": 0.001,
"loss": 2.1111,
"step": 71500
},
{
"epoch": 23.141564318034906,
"grad_norm": 1.7863409519195557,
"learning_rate": 0.001,
"loss": 2.127,
"step": 71600
},
{
"epoch": 23.17388493859082,
"grad_norm": 1.26145339012146,
"learning_rate": 0.001,
"loss": 2.1239,
"step": 71700
},
{
"epoch": 23.206205559146735,
"grad_norm": 1.1627092361450195,
"learning_rate": 0.001,
"loss": 2.1377,
"step": 71800
},
{
"epoch": 23.23852617970265,
"grad_norm": 1.429238200187683,
"learning_rate": 0.001,
"loss": 2.1297,
"step": 71900
},
{
"epoch": 23.270846800258564,
"grad_norm": 1.6770697832107544,
"learning_rate": 0.001,
"loss": 2.1312,
"step": 72000
},
{
"epoch": 23.30316742081448,
"grad_norm": 1.1480836868286133,
"learning_rate": 0.001,
"loss": 2.1498,
"step": 72100
},
{
"epoch": 23.335488041370393,
"grad_norm": 1.19061279296875,
"learning_rate": 0.001,
"loss": 2.1345,
"step": 72200
},
{
"epoch": 23.367808661926308,
"grad_norm": 1.359846591949463,
"learning_rate": 0.001,
"loss": 2.1504,
"step": 72300
},
{
"epoch": 23.400129282482222,
"grad_norm": 1.1585800647735596,
"learning_rate": 0.001,
"loss": 2.1543,
"step": 72400
},
{
"epoch": 23.432449903038137,
"grad_norm": 1.0728919506072998,
"learning_rate": 0.001,
"loss": 2.1411,
"step": 72500
},
{
"epoch": 23.46477052359405,
"grad_norm": 1.321189522743225,
"learning_rate": 0.001,
"loss": 2.1564,
"step": 72600
},
{
"epoch": 23.49709114414997,
"grad_norm": 1.283086895942688,
"learning_rate": 0.001,
"loss": 2.1647,
"step": 72700
},
{
"epoch": 23.529411764705884,
"grad_norm": 1.1694728136062622,
"learning_rate": 0.001,
"loss": 2.1748,
"step": 72800
},
{
"epoch": 23.5617323852618,
"grad_norm": 1.2256524562835693,
"learning_rate": 0.001,
"loss": 2.1584,
"step": 72900
},
{
"epoch": 23.594053005817713,
"grad_norm": 1.2081398963928223,
"learning_rate": 0.001,
"loss": 2.1738,
"step": 73000
},
{
"epoch": 23.626373626373628,
"grad_norm": 1.0685648918151855,
"learning_rate": 0.001,
"loss": 2.1762,
"step": 73100
},
{
"epoch": 23.658694246929542,
"grad_norm": 1.532578468322754,
"learning_rate": 0.001,
"loss": 2.1979,
"step": 73200
},
{
"epoch": 23.691014867485457,
"grad_norm": 1.161739468574524,
"learning_rate": 0.001,
"loss": 2.1834,
"step": 73300
},
{
"epoch": 23.72333548804137,
"grad_norm": 1.1709414720535278,
"learning_rate": 0.001,
"loss": 2.1978,
"step": 73400
},
{
"epoch": 23.755656108597286,
"grad_norm": 1.1965583562850952,
"learning_rate": 0.001,
"loss": 2.1906,
"step": 73500
},
{
"epoch": 23.7879767291532,
"grad_norm": 1.279596209526062,
"learning_rate": 0.001,
"loss": 2.2134,
"step": 73600
},
{
"epoch": 23.820297349709115,
"grad_norm": 1.1345198154449463,
"learning_rate": 0.001,
"loss": 2.1758,
"step": 73700
},
{
"epoch": 23.85261797026503,
"grad_norm": 1.4025368690490723,
"learning_rate": 0.001,
"loss": 2.1885,
"step": 73800
},
{
"epoch": 23.884938590820944,
"grad_norm": 1.2485953569412231,
"learning_rate": 0.001,
"loss": 2.1859,
"step": 73900
},
{
"epoch": 23.91725921137686,
"grad_norm": 1.6559557914733887,
"learning_rate": 0.001,
"loss": 2.2217,
"step": 74000
},
{
"epoch": 23.949579831932773,
"grad_norm": 1.0934584140777588,
"learning_rate": 0.001,
"loss": 2.2114,
"step": 74100
},
{
"epoch": 23.981900452488688,
"grad_norm": 1.1533219814300537,
"learning_rate": 0.001,
"loss": 2.1982,
"step": 74200
},
{
"epoch": 24.014221073044602,
"grad_norm": 1.422975778579712,
"learning_rate": 0.001,
"loss": 2.1479,
"step": 74300
},
{
"epoch": 24.046541693600517,
"grad_norm": 1.4203413724899292,
"learning_rate": 0.001,
"loss": 2.0792,
"step": 74400
},
{
"epoch": 24.07886231415643,
"grad_norm": 1.7462531328201294,
"learning_rate": 0.001,
"loss": 2.0772,
"step": 74500
},
{
"epoch": 24.111182934712346,
"grad_norm": 1.1144509315490723,
"learning_rate": 0.001,
"loss": 2.0748,
"step": 74600
},
{
"epoch": 24.14350355526826,
"grad_norm": 1.5747896432876587,
"learning_rate": 0.001,
"loss": 2.0976,
"step": 74700
},
{
"epoch": 24.175824175824175,
"grad_norm": 1.4975110292434692,
"learning_rate": 0.001,
"loss": 2.1269,
"step": 74800
},
{
"epoch": 24.20814479638009,
"grad_norm": 1.0805388689041138,
"learning_rate": 0.001,
"loss": 2.0752,
"step": 74900
},
{
"epoch": 24.240465416936004,
"grad_norm": 1.448410153388977,
"learning_rate": 0.001,
"loss": 2.1049,
"step": 75000
},
{
"epoch": 24.27278603749192,
"grad_norm": 1.7119464874267578,
"learning_rate": 0.001,
"loss": 2.1199,
"step": 75100
},
{
"epoch": 24.305106658047833,
"grad_norm": 1.3084365129470825,
"learning_rate": 0.001,
"loss": 2.1226,
"step": 75200
},
{
"epoch": 24.337427278603748,
"grad_norm": 1.5231938362121582,
"learning_rate": 0.001,
"loss": 2.1286,
"step": 75300
},
{
"epoch": 24.369747899159663,
"grad_norm": 1.41864812374115,
"learning_rate": 0.001,
"loss": 2.1104,
"step": 75400
},
{
"epoch": 24.402068519715577,
"grad_norm": 1.1968578100204468,
"learning_rate": 0.001,
"loss": 2.1167,
"step": 75500
},
{
"epoch": 24.43438914027149,
"grad_norm": 1.3750704526901245,
"learning_rate": 0.001,
"loss": 2.1236,
"step": 75600
},
{
"epoch": 24.466709760827406,
"grad_norm": 1.3365634679794312,
"learning_rate": 0.001,
"loss": 2.1039,
"step": 75700
},
{
"epoch": 24.49903038138332,
"grad_norm": 1.244064450263977,
"learning_rate": 0.001,
"loss": 2.1359,
"step": 75800
},
{
"epoch": 24.53135100193924,
"grad_norm": 1.1887582540512085,
"learning_rate": 0.001,
"loss": 2.1147,
"step": 75900
},
{
"epoch": 24.563671622495153,
"grad_norm": 1.277971625328064,
"learning_rate": 0.001,
"loss": 2.152,
"step": 76000
},
{
"epoch": 24.595992243051068,
"grad_norm": 1.5758848190307617,
"learning_rate": 0.001,
"loss": 2.1391,
"step": 76100
},
{
"epoch": 24.628312863606983,
"grad_norm": 1.458289623260498,
"learning_rate": 0.001,
"loss": 2.1403,
"step": 76200
},
{
"epoch": 24.660633484162897,
"grad_norm": 1.6823402643203735,
"learning_rate": 0.001,
"loss": 2.1554,
"step": 76300
},
{
"epoch": 24.69295410471881,
"grad_norm": 1.0953534841537476,
"learning_rate": 0.001,
"loss": 2.1609,
"step": 76400
},
{
"epoch": 24.725274725274726,
"grad_norm": 1.429518699645996,
"learning_rate": 0.001,
"loss": 2.1535,
"step": 76500
},
{
"epoch": 24.75759534583064,
"grad_norm": 1.6915266513824463,
"learning_rate": 0.001,
"loss": 2.1738,
"step": 76600
},
{
"epoch": 24.789915966386555,
"grad_norm": 1.172033667564392,
"learning_rate": 0.001,
"loss": 2.1533,
"step": 76700
},
{
"epoch": 24.82223658694247,
"grad_norm": 1.1926536560058594,
"learning_rate": 0.001,
"loss": 2.1614,
"step": 76800
},
{
"epoch": 24.854557207498384,
"grad_norm": 1.4337327480316162,
"learning_rate": 0.001,
"loss": 2.1611,
"step": 76900
},
{
"epoch": 24.8868778280543,
"grad_norm": 1.811618685722351,
"learning_rate": 0.001,
"loss": 2.2001,
"step": 77000
},
{
"epoch": 24.919198448610214,
"grad_norm": 1.5144684314727783,
"learning_rate": 0.001,
"loss": 2.1766,
"step": 77100
},
{
"epoch": 24.951519069166128,
"grad_norm": 1.5109102725982666,
"learning_rate": 0.001,
"loss": 2.182,
"step": 77200
},
{
"epoch": 24.983839689722043,
"grad_norm": 1.4338077306747437,
"learning_rate": 0.001,
"loss": 2.1794,
"step": 77300
},
{
"epoch": 25.016160310277957,
"grad_norm": 1.8659993410110474,
"learning_rate": 0.001,
"loss": 2.0663,
"step": 77400
},
{
"epoch": 25.048480930833872,
"grad_norm": 1.850515365600586,
"learning_rate": 0.001,
"loss": 2.0864,
"step": 77500
},
{
"epoch": 25.080801551389786,
"grad_norm": 2.3665661811828613,
"learning_rate": 0.001,
"loss": 2.0442,
"step": 77600
},
{
"epoch": 25.1131221719457,
"grad_norm": 1.9216943979263306,
"learning_rate": 0.001,
"loss": 2.0473,
"step": 77700
},
{
"epoch": 25.145442792501616,
"grad_norm": 2.2430880069732666,
"learning_rate": 0.001,
"loss": 2.0672,
"step": 77800
},
{
"epoch": 25.17776341305753,
"grad_norm": 1.4772371053695679,
"learning_rate": 0.001,
"loss": 2.093,
"step": 77900
},
{
"epoch": 25.210084033613445,
"grad_norm": 1.9768728017807007,
"learning_rate": 0.001,
"loss": 2.0858,
"step": 78000
},
{
"epoch": 25.24240465416936,
"grad_norm": 2.0959818363189697,
"learning_rate": 0.001,
"loss": 2.0782,
"step": 78100
},
{
"epoch": 25.274725274725274,
"grad_norm": 1.9896551370620728,
"learning_rate": 0.001,
"loss": 2.0687,
"step": 78200
},
{
"epoch": 25.30704589528119,
"grad_norm": 1.7270411252975464,
"learning_rate": 0.001,
"loss": 2.0984,
"step": 78300
},
{
"epoch": 25.339366515837103,
"grad_norm": 1.8653007745742798,
"learning_rate": 0.001,
"loss": 2.0979,
"step": 78400
},
{
"epoch": 25.371687136393017,
"grad_norm": 1.811274528503418,
"learning_rate": 0.001,
"loss": 2.0887,
"step": 78500
},
{
"epoch": 25.404007756948932,
"grad_norm": 2.3261477947235107,
"learning_rate": 0.001,
"loss": 2.0987,
"step": 78600
},
{
"epoch": 25.436328377504847,
"grad_norm": 2.384524345397949,
"learning_rate": 0.001,
"loss": 2.0856,
"step": 78700
},
{
"epoch": 25.46864899806076,
"grad_norm": 1.8256169557571411,
"learning_rate": 0.001,
"loss": 2.104,
"step": 78800
},
{
"epoch": 25.50096961861668,
"grad_norm": 2.1942951679229736,
"learning_rate": 0.001,
"loss": 2.0823,
"step": 78900
},
{
"epoch": 25.533290239172594,
"grad_norm": 2.347743272781372,
"learning_rate": 0.001,
"loss": 2.1366,
"step": 79000
},
{
"epoch": 25.56561085972851,
"grad_norm": 1.6978386640548706,
"learning_rate": 0.001,
"loss": 2.1115,
"step": 79100
},
{
"epoch": 25.597931480284423,
"grad_norm": 2.3107948303222656,
"learning_rate": 0.001,
"loss": 2.1255,
"step": 79200
},
{
"epoch": 25.630252100840337,
"grad_norm": 2.433992624282837,
"learning_rate": 0.001,
"loss": 2.0966,
"step": 79300
},
{
"epoch": 25.662572721396252,
"grad_norm": 2.4844846725463867,
"learning_rate": 0.001,
"loss": 2.1251,
"step": 79400
},
{
"epoch": 25.694893341952167,
"grad_norm": 1.811493992805481,
"learning_rate": 0.001,
"loss": 2.1263,
"step": 79500
},
{
"epoch": 25.72721396250808,
"grad_norm": 1.8551549911499023,
"learning_rate": 0.001,
"loss": 2.1271,
"step": 79600
},
{
"epoch": 25.759534583063996,
"grad_norm": 1.6955463886260986,
"learning_rate": 0.001,
"loss": 2.1166,
"step": 79700
},
{
"epoch": 25.79185520361991,
"grad_norm": 1.900719404220581,
"learning_rate": 0.001,
"loss": 2.1364,
"step": 79800
},
{
"epoch": 25.824175824175825,
"grad_norm": 1.699873447418213,
"learning_rate": 0.001,
"loss": 2.1179,
"step": 79900
},
{
"epoch": 25.85649644473174,
"grad_norm": 1.9742425680160522,
"learning_rate": 0.001,
"loss": 2.1492,
"step": 80000
},
{
"epoch": 25.888817065287654,
"grad_norm": 2.201303243637085,
"learning_rate": 0.001,
"loss": 2.1364,
"step": 80100
},
{
"epoch": 25.92113768584357,
"grad_norm": 1.975517749786377,
"learning_rate": 0.001,
"loss": 2.1468,
"step": 80200
},
{
"epoch": 25.953458306399483,
"grad_norm": 1.7931538820266724,
"learning_rate": 0.001,
"loss": 2.1414,
"step": 80300
},
{
"epoch": 25.985778926955398,
"grad_norm": 1.9484506845474243,
"learning_rate": 0.001,
"loss": 2.1401,
"step": 80400
},
{
"epoch": 26.018099547511312,
"grad_norm": 1.0984433889389038,
"learning_rate": 0.001,
"loss": 2.087,
"step": 80500
},
{
"epoch": 26.050420168067227,
"grad_norm": 1.3689485788345337,
"learning_rate": 0.001,
"loss": 2.013,
"step": 80600
},
{
"epoch": 26.08274078862314,
"grad_norm": 1.1524888277053833,
"learning_rate": 0.001,
"loss": 2.0192,
"step": 80700
},
{
"epoch": 26.115061409179056,
"grad_norm": 1.5100384950637817,
"learning_rate": 0.001,
"loss": 2.0432,
"step": 80800
},
{
"epoch": 26.14738202973497,
"grad_norm": 1.5345035791397095,
"learning_rate": 0.001,
"loss": 2.0432,
"step": 80900
},
{
"epoch": 26.179702650290885,
"grad_norm": 1.632036805152893,
"learning_rate": 0.001,
"loss": 2.0369,
"step": 81000
},
{
"epoch": 26.2120232708468,
"grad_norm": 1.320082426071167,
"learning_rate": 0.001,
"loss": 2.0383,
"step": 81100
},
{
"epoch": 26.244343891402714,
"grad_norm": 1.2324976921081543,
"learning_rate": 0.001,
"loss": 2.0511,
"step": 81200
},
{
"epoch": 26.27666451195863,
"grad_norm": 1.0281955003738403,
"learning_rate": 0.001,
"loss": 2.0687,
"step": 81300
},
{
"epoch": 26.308985132514543,
"grad_norm": 1.2883223295211792,
"learning_rate": 0.001,
"loss": 2.0653,
"step": 81400
},
{
"epoch": 26.341305753070458,
"grad_norm": 1.3196618556976318,
"learning_rate": 0.001,
"loss": 2.0713,
"step": 81500
},
{
"epoch": 26.373626373626372,
"grad_norm": 1.4835927486419678,
"learning_rate": 0.001,
"loss": 2.0645,
"step": 81600
},
{
"epoch": 26.405946994182287,
"grad_norm": 1.370380163192749,
"learning_rate": 0.001,
"loss": 2.0777,
"step": 81700
},
{
"epoch": 26.4382676147382,
"grad_norm": 1.5035107135772705,
"learning_rate": 0.001,
"loss": 2.0521,
"step": 81800
},
{
"epoch": 26.470588235294116,
"grad_norm": 1.375370740890503,
"learning_rate": 0.001,
"loss": 2.064,
"step": 81900
},
{
"epoch": 26.50290885585003,
"grad_norm": 1.3118560314178467,
"learning_rate": 0.001,
"loss": 2.0782,
"step": 82000
},
{
"epoch": 26.53522947640595,
"grad_norm": 1.2957309484481812,
"learning_rate": 0.001,
"loss": 2.0736,
"step": 82100
},
{
"epoch": 26.567550096961863,
"grad_norm": 2.023144245147705,
"learning_rate": 0.001,
"loss": 2.0761,
"step": 82200
},
{
"epoch": 26.599870717517778,
"grad_norm": 1.7916072607040405,
"learning_rate": 0.001,
"loss": 2.0932,
"step": 82300
},
{
"epoch": 26.632191338073692,
"grad_norm": 1.3292901515960693,
"learning_rate": 0.001,
"loss": 2.0916,
"step": 82400
},
{
"epoch": 26.664511958629607,
"grad_norm": 1.1266815662384033,
"learning_rate": 0.001,
"loss": 2.097,
"step": 82500
},
{
"epoch": 26.69683257918552,
"grad_norm": 1.7361462116241455,
"learning_rate": 0.001,
"loss": 2.0913,
"step": 82600
},
{
"epoch": 26.729153199741436,
"grad_norm": 1.340409278869629,
"learning_rate": 0.001,
"loss": 2.1019,
"step": 82700
},
{
"epoch": 26.76147382029735,
"grad_norm": 1.3386613130569458,
"learning_rate": 0.001,
"loss": 2.0988,
"step": 82800
},
{
"epoch": 26.793794440853265,
"grad_norm": 1.2218043804168701,
"learning_rate": 0.001,
"loss": 2.1205,
"step": 82900
},
{
"epoch": 26.82611506140918,
"grad_norm": 1.316157341003418,
"learning_rate": 0.001,
"loss": 2.1338,
"step": 83000
},
{
"epoch": 26.858435681965094,
"grad_norm": 1.1175379753112793,
"learning_rate": 0.001,
"loss": 2.1506,
"step": 83100
},
{
"epoch": 26.89075630252101,
"grad_norm": 1.4221986532211304,
"learning_rate": 0.001,
"loss": 2.133,
"step": 83200
},
{
"epoch": 26.923076923076923,
"grad_norm": 1.2862468957901,
"learning_rate": 0.001,
"loss": 2.1089,
"step": 83300
},
{
"epoch": 26.955397543632838,
"grad_norm": 1.1285076141357422,
"learning_rate": 0.001,
"loss": 2.1232,
"step": 83400
},
{
"epoch": 26.987718164188752,
"grad_norm": 1.4763456583023071,
"learning_rate": 0.001,
"loss": 2.1217,
"step": 83500
},
{
"epoch": 27.020038784744667,
"grad_norm": 1.0230133533477783,
"learning_rate": 0.001,
"loss": 2.0579,
"step": 83600
},
{
"epoch": 27.05235940530058,
"grad_norm": 1.219338297843933,
"learning_rate": 0.001,
"loss": 1.9788,
"step": 83700
},
{
"epoch": 27.084680025856496,
"grad_norm": 1.1852729320526123,
"learning_rate": 0.001,
"loss": 1.9838,
"step": 83800
},
{
"epoch": 27.11700064641241,
"grad_norm": 1.1403017044067383,
"learning_rate": 0.001,
"loss": 2.0111,
"step": 83900
},
{
"epoch": 27.149321266968325,
"grad_norm": 1.154348611831665,
"learning_rate": 0.001,
"loss": 2.0075,
"step": 84000
},
{
"epoch": 27.18164188752424,
"grad_norm": 0.9911405444145203,
"learning_rate": 0.001,
"loss": 2.002,
"step": 84100
},
{
"epoch": 27.213962508080154,
"grad_norm": 0.9690923690795898,
"learning_rate": 0.001,
"loss": 2.0306,
"step": 84200
},
{
"epoch": 27.24628312863607,
"grad_norm": 1.1412653923034668,
"learning_rate": 0.001,
"loss": 2.0068,
"step": 84300
},
{
"epoch": 27.278603749191983,
"grad_norm": 1.4080849885940552,
"learning_rate": 0.001,
"loss": 2.054,
"step": 84400
},
{
"epoch": 27.310924369747898,
"grad_norm": 1.2841688394546509,
"learning_rate": 0.001,
"loss": 2.0296,
"step": 84500
},
{
"epoch": 27.343244990303813,
"grad_norm": 0.8256431221961975,
"learning_rate": 0.001,
"loss": 2.0577,
"step": 84600
},
{
"epoch": 27.375565610859727,
"grad_norm": 1.0703890323638916,
"learning_rate": 0.001,
"loss": 2.041,
"step": 84700
},
{
"epoch": 27.40788623141564,
"grad_norm": 1.0905351638793945,
"learning_rate": 0.001,
"loss": 2.0437,
"step": 84800
},
{
"epoch": 27.440206851971556,
"grad_norm": 1.6662650108337402,
"learning_rate": 0.001,
"loss": 2.0565,
"step": 84900
},
{
"epoch": 27.47252747252747,
"grad_norm": 1.1492462158203125,
"learning_rate": 0.001,
"loss": 2.0672,
"step": 85000
},
{
"epoch": 27.50484809308339,
"grad_norm": 1.1019383668899536,
"learning_rate": 0.001,
"loss": 2.0605,
"step": 85100
},
{
"epoch": 27.537168713639304,
"grad_norm": 1.5522205829620361,
"learning_rate": 0.001,
"loss": 2.0688,
"step": 85200
},
{
"epoch": 27.569489334195218,
"grad_norm": 1.4154222011566162,
"learning_rate": 0.001,
"loss": 2.0649,
"step": 85300
},
{
"epoch": 27.601809954751133,
"grad_norm": 1.4599612951278687,
"learning_rate": 0.001,
"loss": 2.0614,
"step": 85400
},
{
"epoch": 27.634130575307047,
"grad_norm": 1.1189088821411133,
"learning_rate": 0.001,
"loss": 2.0691,
"step": 85500
},
{
"epoch": 27.66645119586296,
"grad_norm": 1.3691092729568481,
"learning_rate": 0.001,
"loss": 2.0871,
"step": 85600
},
{
"epoch": 27.698771816418876,
"grad_norm": 1.2873777151107788,
"learning_rate": 0.001,
"loss": 2.0733,
"step": 85700
},
{
"epoch": 27.73109243697479,
"grad_norm": 1.0803229808807373,
"learning_rate": 0.001,
"loss": 2.0736,
"step": 85800
},
{
"epoch": 27.763413057530705,
"grad_norm": 1.1693120002746582,
"learning_rate": 0.001,
"loss": 2.072,
"step": 85900
},
{
"epoch": 27.79573367808662,
"grad_norm": 1.6821506023406982,
"learning_rate": 0.001,
"loss": 2.0589,
"step": 86000
},
{
"epoch": 27.828054298642535,
"grad_norm": 1.1503660678863525,
"learning_rate": 0.001,
"loss": 2.0804,
"step": 86100
},
{
"epoch": 27.86037491919845,
"grad_norm": 1.145229458808899,
"learning_rate": 0.001,
"loss": 2.0892,
"step": 86200
},
{
"epoch": 27.892695539754364,
"grad_norm": 1.2940752506256104,
"learning_rate": 0.001,
"loss": 2.0863,
"step": 86300
},
{
"epoch": 27.92501616031028,
"grad_norm": 1.2379393577575684,
"learning_rate": 0.001,
"loss": 2.0928,
"step": 86400
},
{
"epoch": 27.957336780866193,
"grad_norm": 1.2661943435668945,
"learning_rate": 0.001,
"loss": 2.1085,
"step": 86500
},
{
"epoch": 27.989657401422107,
"grad_norm": 1.5016684532165527,
"learning_rate": 0.001,
"loss": 2.1017,
"step": 86600
},
{
"epoch": 28.021978021978022,
"grad_norm": 1.120667576789856,
"learning_rate": 0.001,
"loss": 2.0434,
"step": 86700
},
{
"epoch": 28.054298642533936,
"grad_norm": 1.0933500528335571,
"learning_rate": 0.001,
"loss": 1.9616,
"step": 86800
},
{
"epoch": 28.08661926308985,
"grad_norm": 1.3572559356689453,
"learning_rate": 0.001,
"loss": 1.9725,
"step": 86900
},
{
"epoch": 28.118939883645766,
"grad_norm": 1.2044649124145508,
"learning_rate": 0.001,
"loss": 1.9898,
"step": 87000
},
{
"epoch": 28.15126050420168,
"grad_norm": 0.969914972782135,
"learning_rate": 0.001,
"loss": 1.974,
"step": 87100
},
{
"epoch": 28.183581124757595,
"grad_norm": 1.0677287578582764,
"learning_rate": 0.001,
"loss": 1.9967,
"step": 87200
},
{
"epoch": 28.21590174531351,
"grad_norm": 1.2415043115615845,
"learning_rate": 0.001,
"loss": 1.9916,
"step": 87300
},
{
"epoch": 28.248222365869424,
"grad_norm": 1.1492565870285034,
"learning_rate": 0.001,
"loss": 1.9952,
"step": 87400
},
{
"epoch": 28.28054298642534,
"grad_norm": 1.4052455425262451,
"learning_rate": 0.001,
"loss": 2.0111,
"step": 87500
},
{
"epoch": 28.312863606981253,
"grad_norm": 1.0275617837905884,
"learning_rate": 0.001,
"loss": 2.0237,
"step": 87600
},
{
"epoch": 28.345184227537167,
"grad_norm": 1.3079900741577148,
"learning_rate": 0.001,
"loss": 1.9931,
"step": 87700
},
{
"epoch": 28.377504848093082,
"grad_norm": 1.058862566947937,
"learning_rate": 0.001,
"loss": 2.0121,
"step": 87800
},
{
"epoch": 28.409825468648997,
"grad_norm": 1.0807355642318726,
"learning_rate": 0.001,
"loss": 2.0252,
"step": 87900
},
{
"epoch": 28.44214608920491,
"grad_norm": 1.1212176084518433,
"learning_rate": 0.001,
"loss": 2.0239,
"step": 88000
},
{
"epoch": 28.474466709760826,
"grad_norm": 1.1459327936172485,
"learning_rate": 0.001,
"loss": 2.0089,
"step": 88100
},
{
"epoch": 28.50678733031674,
"grad_norm": 0.971630334854126,
"learning_rate": 0.001,
"loss": 2.028,
"step": 88200
},
{
"epoch": 28.53910795087266,
"grad_norm": 1.1253451108932495,
"learning_rate": 0.001,
"loss": 2.0251,
"step": 88300
},
{
"epoch": 28.571428571428573,
"grad_norm": 1.2817078828811646,
"learning_rate": 0.001,
"loss": 2.0279,
"step": 88400
},
{
"epoch": 28.603749191984488,
"grad_norm": 1.3347026109695435,
"learning_rate": 0.001,
"loss": 2.0498,
"step": 88500
},
{
"epoch": 28.636069812540402,
"grad_norm": 1.1375293731689453,
"learning_rate": 0.001,
"loss": 2.027,
"step": 88600
},
{
"epoch": 28.668390433096317,
"grad_norm": 0.9918130040168762,
"learning_rate": 0.001,
"loss": 2.0635,
"step": 88700
},
{
"epoch": 28.70071105365223,
"grad_norm": 1.0592323541641235,
"learning_rate": 0.001,
"loss": 2.0386,
"step": 88800
},
{
"epoch": 28.733031674208146,
"grad_norm": 0.9870603680610657,
"learning_rate": 0.001,
"loss": 2.0377,
"step": 88900
},
{
"epoch": 28.76535229476406,
"grad_norm": 1.2788349390029907,
"learning_rate": 0.001,
"loss": 2.0706,
"step": 89000
},
{
"epoch": 28.797672915319975,
"grad_norm": 1.0546563863754272,
"learning_rate": 0.001,
"loss": 2.0529,
"step": 89100
},
{
"epoch": 28.82999353587589,
"grad_norm": 1.1603606939315796,
"learning_rate": 0.001,
"loss": 2.0695,
"step": 89200
},
{
"epoch": 28.862314156431804,
"grad_norm": 1.1769486665725708,
"learning_rate": 0.001,
"loss": 2.0801,
"step": 89300
},
{
"epoch": 28.89463477698772,
"grad_norm": 1.2034118175506592,
"learning_rate": 0.001,
"loss": 2.0698,
"step": 89400
},
{
"epoch": 28.926955397543633,
"grad_norm": 1.1783450841903687,
"learning_rate": 0.001,
"loss": 2.0857,
"step": 89500
},
{
"epoch": 28.959276018099548,
"grad_norm": 1.0825036764144897,
"learning_rate": 0.001,
"loss": 2.0764,
"step": 89600
},
{
"epoch": 28.991596638655462,
"grad_norm": 1.1143277883529663,
"learning_rate": 0.001,
"loss": 2.0862,
"step": 89700
},
{
"epoch": 29.023917259211377,
"grad_norm": 1.0811305046081543,
"learning_rate": 0.001,
"loss": 1.9729,
"step": 89800
},
{
"epoch": 29.05623787976729,
"grad_norm": 1.3418818712234497,
"learning_rate": 0.001,
"loss": 1.9414,
"step": 89900
},
{
"epoch": 29.088558500323206,
"grad_norm": 1.2749602794647217,
"learning_rate": 0.001,
"loss": 1.9419,
"step": 90000
},
{
"epoch": 29.12087912087912,
"grad_norm": 1.0318279266357422,
"learning_rate": 0.001,
"loss": 1.9554,
"step": 90100
},
{
"epoch": 29.153199741435035,
"grad_norm": 1.573747992515564,
"learning_rate": 0.001,
"loss": 1.9641,
"step": 90200
},
{
"epoch": 29.18552036199095,
"grad_norm": 1.30129873752594,
"learning_rate": 0.001,
"loss": 1.9745,
"step": 90300
},
{
"epoch": 29.217840982546864,
"grad_norm": 1.1000499725341797,
"learning_rate": 0.001,
"loss": 1.9569,
"step": 90400
},
{
"epoch": 29.25016160310278,
"grad_norm": 1.1825950145721436,
"learning_rate": 0.001,
"loss": 2.0055,
"step": 90500
},
{
"epoch": 29.282482223658693,
"grad_norm": 1.5407135486602783,
"learning_rate": 0.001,
"loss": 1.9929,
"step": 90600
},
{
"epoch": 29.314802844214608,
"grad_norm": 1.4850648641586304,
"learning_rate": 0.001,
"loss": 1.9763,
"step": 90700
},
{
"epoch": 29.347123464770522,
"grad_norm": 1.007767915725708,
"learning_rate": 0.001,
"loss": 1.9844,
"step": 90800
},
{
"epoch": 29.379444085326437,
"grad_norm": 1.2812808752059937,
"learning_rate": 0.001,
"loss": 1.9896,
"step": 90900
},
{
"epoch": 29.41176470588235,
"grad_norm": 0.9998975992202759,
"learning_rate": 0.001,
"loss": 1.981,
"step": 91000
},
{
"epoch": 29.444085326438266,
"grad_norm": 1.1367158889770508,
"learning_rate": 0.001,
"loss": 2.0031,
"step": 91100
},
{
"epoch": 29.47640594699418,
"grad_norm": 1.337172508239746,
"learning_rate": 0.001,
"loss": 2.0074,
"step": 91200
},
{
"epoch": 29.5087265675501,
"grad_norm": 1.1325889825820923,
"learning_rate": 0.001,
"loss": 2.0186,
"step": 91300
},
{
"epoch": 29.541047188106013,
"grad_norm": 1.1210311651229858,
"learning_rate": 0.001,
"loss": 2.0318,
"step": 91400
},
{
"epoch": 29.573367808661928,
"grad_norm": 0.9622472524642944,
"learning_rate": 0.001,
"loss": 2.0068,
"step": 91500
},
{
"epoch": 29.605688429217842,
"grad_norm": 1.2913539409637451,
"learning_rate": 0.001,
"loss": 2.0198,
"step": 91600
},
{
"epoch": 29.638009049773757,
"grad_norm": 1.1396842002868652,
"learning_rate": 0.001,
"loss": 2.0323,
"step": 91700
},
{
"epoch": 29.67032967032967,
"grad_norm": 0.9822863340377808,
"learning_rate": 0.001,
"loss": 2.0295,
"step": 91800
},
{
"epoch": 29.702650290885586,
"grad_norm": 1.2664222717285156,
"learning_rate": 0.001,
"loss": 2.0386,
"step": 91900
},
{
"epoch": 29.7349709114415,
"grad_norm": 1.1091785430908203,
"learning_rate": 0.001,
"loss": 2.0186,
"step": 92000
},
{
"epoch": 29.767291531997415,
"grad_norm": 1.1308799982070923,
"learning_rate": 0.001,
"loss": 2.0459,
"step": 92100
},
{
"epoch": 29.79961215255333,
"grad_norm": 1.1163250207901,
"learning_rate": 0.001,
"loss": 2.0276,
"step": 92200
},
{
"epoch": 29.831932773109244,
"grad_norm": 1.494959831237793,
"learning_rate": 0.001,
"loss": 2.0295,
"step": 92300
},
{
"epoch": 29.86425339366516,
"grad_norm": 1.1304702758789062,
"learning_rate": 0.001,
"loss": 2.0515,
"step": 92400
},
{
"epoch": 29.896574014221073,
"grad_norm": 1.09246027469635,
"learning_rate": 0.001,
"loss": 2.0398,
"step": 92500
},
{
"epoch": 29.928894634776988,
"grad_norm": 0.9423359036445618,
"learning_rate": 0.001,
"loss": 2.0507,
"step": 92600
},
{
"epoch": 29.961215255332903,
"grad_norm": 1.193638801574707,
"learning_rate": 0.001,
"loss": 2.0531,
"step": 92700
},
{
"epoch": 29.993535875888817,
"grad_norm": 1.414298415184021,
"learning_rate": 0.001,
"loss": 2.0616,
"step": 92800
},
{
"epoch": 30.02585649644473,
"grad_norm": 1.048582673072815,
"learning_rate": 0.001,
"loss": 1.9437,
"step": 92900
},
{
"epoch": 30.058177117000646,
"grad_norm": 1.5626100301742554,
"learning_rate": 0.001,
"loss": 1.9247,
"step": 93000
},
{
"epoch": 30.09049773755656,
"grad_norm": 1.3486158847808838,
"learning_rate": 0.001,
"loss": 1.9305,
"step": 93100
},
{
"epoch": 30.122818358112475,
"grad_norm": 1.2358057498931885,
"learning_rate": 0.001,
"loss": 1.9336,
"step": 93200
},
{
"epoch": 30.15513897866839,
"grad_norm": 1.462971568107605,
"learning_rate": 0.001,
"loss": 1.9462,
"step": 93300
},
{
"epoch": 30.187459599224304,
"grad_norm": 1.156144142150879,
"learning_rate": 0.001,
"loss": 1.9496,
"step": 93400
},
{
"epoch": 30.21978021978022,
"grad_norm": 1.0619349479675293,
"learning_rate": 0.001,
"loss": 1.953,
"step": 93500
},
{
"epoch": 30.252100840336134,
"grad_norm": 1.0924150943756104,
"learning_rate": 0.001,
"loss": 1.9482,
"step": 93600
},
{
"epoch": 30.284421460892048,
"grad_norm": 1.5991413593292236,
"learning_rate": 0.001,
"loss": 1.9786,
"step": 93700
},
{
"epoch": 30.316742081447963,
"grad_norm": 1.2270259857177734,
"learning_rate": 0.001,
"loss": 1.9567,
"step": 93800
},
{
"epoch": 30.349062702003877,
"grad_norm": 1.2201322317123413,
"learning_rate": 0.001,
"loss": 1.9651,
"step": 93900
},
{
"epoch": 30.381383322559792,
"grad_norm": 1.1978286504745483,
"learning_rate": 0.001,
"loss": 1.9742,
"step": 94000
},
{
"epoch": 30.413703943115706,
"grad_norm": 1.507910966873169,
"learning_rate": 0.001,
"loss": 1.9589,
"step": 94100
},
{
"epoch": 30.44602456367162,
"grad_norm": 1.3228987455368042,
"learning_rate": 0.001,
"loss": 1.9881,
"step": 94200
},
{
"epoch": 30.478345184227535,
"grad_norm": 1.2880733013153076,
"learning_rate": 0.001,
"loss": 1.9973,
"step": 94300
},
{
"epoch": 30.51066580478345,
"grad_norm": 1.4545131921768188,
"learning_rate": 0.001,
"loss": 1.9971,
"step": 94400
},
{
"epoch": 30.542986425339368,
"grad_norm": 1.2463194131851196,
"learning_rate": 0.001,
"loss": 1.9901,
"step": 94500
},
{
"epoch": 30.575307045895283,
"grad_norm": 1.2758846282958984,
"learning_rate": 0.001,
"loss": 1.999,
"step": 94600
},
{
"epoch": 30.607627666451197,
"grad_norm": 1.151174545288086,
"learning_rate": 0.001,
"loss": 1.9899,
"step": 94700
},
{
"epoch": 30.639948287007112,
"grad_norm": 1.1927931308746338,
"learning_rate": 0.001,
"loss": 1.9985,
"step": 94800
},
{
"epoch": 30.672268907563026,
"grad_norm": 1.2777458429336548,
"learning_rate": 0.001,
"loss": 2.0183,
"step": 94900
},
{
"epoch": 30.70458952811894,
"grad_norm": 1.053712010383606,
"learning_rate": 0.001,
"loss": 2.0123,
"step": 95000
},
{
"epoch": 30.736910148674855,
"grad_norm": 1.2185966968536377,
"learning_rate": 0.001,
"loss": 1.9986,
"step": 95100
},
{
"epoch": 30.76923076923077,
"grad_norm": 1.239749550819397,
"learning_rate": 0.001,
"loss": 2.0213,
"step": 95200
},
{
"epoch": 30.801551389786685,
"grad_norm": 1.4070818424224854,
"learning_rate": 0.001,
"loss": 2.0101,
"step": 95300
},
{
"epoch": 30.8338720103426,
"grad_norm": 0.9187280535697937,
"learning_rate": 0.001,
"loss": 2.0059,
"step": 95400
},
{
"epoch": 30.866192630898514,
"grad_norm": 1.2038640975952148,
"learning_rate": 0.001,
"loss": 2.0322,
"step": 95500
},
{
"epoch": 30.89851325145443,
"grad_norm": 1.2154532670974731,
"learning_rate": 0.001,
"loss": 1.9982,
"step": 95600
},
{
"epoch": 30.930833872010343,
"grad_norm": 1.2142574787139893,
"learning_rate": 0.001,
"loss": 2.0343,
"step": 95700
},
{
"epoch": 30.963154492566257,
"grad_norm": 1.26584792137146,
"learning_rate": 0.001,
"loss": 2.0022,
"step": 95800
},
{
"epoch": 30.995475113122172,
"grad_norm": 1.661458134651184,
"learning_rate": 0.001,
"loss": 2.0102,
"step": 95900
},
{
"epoch": 31.027795733678087,
"grad_norm": 1.824419617652893,
"learning_rate": 0.001,
"loss": 1.9388,
"step": 96000
},
{
"epoch": 31.060116354234,
"grad_norm": 1.2426424026489258,
"learning_rate": 0.001,
"loss": 1.8816,
"step": 96100
},
{
"epoch": 31.092436974789916,
"grad_norm": 1.540233850479126,
"learning_rate": 0.001,
"loss": 1.8989,
"step": 96200
},
{
"epoch": 31.12475759534583,
"grad_norm": 1.3419034481048584,
"learning_rate": 0.001,
"loss": 1.9082,
"step": 96300
},
{
"epoch": 31.157078215901745,
"grad_norm": 1.418953537940979,
"learning_rate": 0.001,
"loss": 1.9132,
"step": 96400
},
{
"epoch": 31.18939883645766,
"grad_norm": 1.3207831382751465,
"learning_rate": 0.001,
"loss": 1.9249,
"step": 96500
},
{
"epoch": 31.221719457013574,
"grad_norm": 1.145176887512207,
"learning_rate": 0.001,
"loss": 1.9144,
"step": 96600
},
{
"epoch": 31.25404007756949,
"grad_norm": 1.091209888458252,
"learning_rate": 0.001,
"loss": 1.9258,
"step": 96700
},
{
"epoch": 31.286360698125403,
"grad_norm": 1.2040692567825317,
"learning_rate": 0.001,
"loss": 1.9303,
"step": 96800
},
{
"epoch": 31.318681318681318,
"grad_norm": 1.393153429031372,
"learning_rate": 0.001,
"loss": 1.9554,
"step": 96900
},
{
"epoch": 31.351001939237232,
"grad_norm": 1.298164963722229,
"learning_rate": 0.001,
"loss": 1.9668,
"step": 97000
},
{
"epoch": 31.383322559793147,
"grad_norm": 1.1720830202102661,
"learning_rate": 0.001,
"loss": 1.9576,
"step": 97100
},
{
"epoch": 31.41564318034906,
"grad_norm": 1.2951114177703857,
"learning_rate": 0.001,
"loss": 1.9484,
"step": 97200
},
{
"epoch": 31.447963800904976,
"grad_norm": 1.4327456951141357,
"learning_rate": 0.001,
"loss": 1.9666,
"step": 97300
},
{
"epoch": 31.48028442146089,
"grad_norm": 1.3483418226242065,
"learning_rate": 0.001,
"loss": 1.9593,
"step": 97400
},
{
"epoch": 31.51260504201681,
"grad_norm": 1.2708185911178589,
"learning_rate": 0.001,
"loss": 1.9783,
"step": 97500
},
{
"epoch": 31.544925662572723,
"grad_norm": 1.3904316425323486,
"learning_rate": 0.001,
"loss": 1.9749,
"step": 97600
},
{
"epoch": 31.577246283128638,
"grad_norm": 1.4212653636932373,
"learning_rate": 0.001,
"loss": 1.9597,
"step": 97700
},
{
"epoch": 31.609566903684552,
"grad_norm": 1.4417407512664795,
"learning_rate": 0.001,
"loss": 1.9821,
"step": 97800
},
{
"epoch": 31.641887524240467,
"grad_norm": 1.3898744583129883,
"learning_rate": 0.001,
"loss": 1.9808,
"step": 97900
},
{
"epoch": 31.67420814479638,
"grad_norm": 1.6071312427520752,
"learning_rate": 0.001,
"loss": 1.9841,
"step": 98000
},
{
"epoch": 31.706528765352296,
"grad_norm": 1.970450758934021,
"learning_rate": 0.001,
"loss": 1.9785,
"step": 98100
},
{
"epoch": 31.73884938590821,
"grad_norm": 1.5174766778945923,
"learning_rate": 0.001,
"loss": 1.9835,
"step": 98200
},
{
"epoch": 31.771170006464125,
"grad_norm": 1.4182029962539673,
"learning_rate": 0.001,
"loss": 1.989,
"step": 98300
},
{
"epoch": 31.80349062702004,
"grad_norm": 1.5139251947402954,
"learning_rate": 0.001,
"loss": 2.0098,
"step": 98400
},
{
"epoch": 31.835811247575954,
"grad_norm": 1.2674909830093384,
"learning_rate": 0.001,
"loss": 1.9917,
"step": 98500
},
{
"epoch": 31.86813186813187,
"grad_norm": 1.0950101613998413,
"learning_rate": 0.001,
"loss": 2.0219,
"step": 98600
},
{
"epoch": 31.900452488687783,
"grad_norm": 1.2220189571380615,
"learning_rate": 0.001,
"loss": 1.9903,
"step": 98700
},
{
"epoch": 31.932773109243698,
"grad_norm": 1.0769373178482056,
"learning_rate": 0.001,
"loss": 2.0018,
"step": 98800
},
{
"epoch": 31.965093729799612,
"grad_norm": 1.163196325302124,
"learning_rate": 0.001,
"loss": 1.9916,
"step": 98900
},
{
"epoch": 31.997414350355527,
"grad_norm": 1.32037353515625,
"learning_rate": 0.001,
"loss": 1.9994,
"step": 99000
},
{
"epoch": 32.02973497091144,
"grad_norm": 1.383773684501648,
"learning_rate": 0.001,
"loss": 1.9098,
"step": 99100
},
{
"epoch": 32.062055591467356,
"grad_norm": 1.3243025541305542,
"learning_rate": 0.001,
"loss": 1.8755,
"step": 99200
},
{
"epoch": 32.09437621202327,
"grad_norm": 1.4425158500671387,
"learning_rate": 0.001,
"loss": 1.9061,
"step": 99300
},
{
"epoch": 32.126696832579185,
"grad_norm": 1.6512792110443115,
"learning_rate": 0.001,
"loss": 1.8809,
"step": 99400
},
{
"epoch": 32.1590174531351,
"grad_norm": 1.532712697982788,
"learning_rate": 0.001,
"loss": 1.8984,
"step": 99500
},
{
"epoch": 32.191338073691014,
"grad_norm": 1.6402521133422852,
"learning_rate": 0.001,
"loss": 1.921,
"step": 99600
},
{
"epoch": 32.22365869424693,
"grad_norm": 1.362164855003357,
"learning_rate": 0.001,
"loss": 1.914,
"step": 99700
},
{
"epoch": 32.25597931480284,
"grad_norm": 1.405976414680481,
"learning_rate": 0.001,
"loss": 1.9047,
"step": 99800
},
{
"epoch": 32.28829993535876,
"grad_norm": 1.1720060110092163,
"learning_rate": 0.001,
"loss": 1.906,
"step": 99900
},
{
"epoch": 32.32062055591467,
"grad_norm": 1.6015681028366089,
"learning_rate": 0.001,
"loss": 1.9081,
"step": 100000
},
{
"epoch": 32.35294117647059,
"grad_norm": 1.649999976158142,
"learning_rate": 0.001,
"loss": 1.9329,
"step": 100100
},
{
"epoch": 32.3852617970265,
"grad_norm": 1.2213903665542603,
"learning_rate": 0.001,
"loss": 1.9216,
"step": 100200
},
{
"epoch": 32.417582417582416,
"grad_norm": 1.0760034322738647,
"learning_rate": 0.001,
"loss": 1.9397,
"step": 100300
},
{
"epoch": 32.44990303813833,
"grad_norm": 1.4536710977554321,
"learning_rate": 0.001,
"loss": 1.9281,
"step": 100400
},
{
"epoch": 32.482223658694245,
"grad_norm": 1.237513780593872,
"learning_rate": 0.001,
"loss": 1.9482,
"step": 100500
},
{
"epoch": 32.51454427925016,
"grad_norm": 1.0938454866409302,
"learning_rate": 0.001,
"loss": 1.9532,
"step": 100600
},
{
"epoch": 32.546864899806074,
"grad_norm": 1.4232382774353027,
"learning_rate": 0.001,
"loss": 1.922,
"step": 100700
},
{
"epoch": 32.57918552036199,
"grad_norm": 1.4068869352340698,
"learning_rate": 0.001,
"loss": 1.9716,
"step": 100800
},
{
"epoch": 32.6115061409179,
"grad_norm": 1.3143997192382812,
"learning_rate": 0.001,
"loss": 1.9597,
"step": 100900
},
{
"epoch": 32.64382676147382,
"grad_norm": 1.4237006902694702,
"learning_rate": 0.001,
"loss": 1.9761,
"step": 101000
},
{
"epoch": 32.67614738202973,
"grad_norm": 1.2957416772842407,
"learning_rate": 0.001,
"loss": 1.9494,
"step": 101100
},
{
"epoch": 32.70846800258565,
"grad_norm": 1.2689306735992432,
"learning_rate": 0.001,
"loss": 1.9631,
"step": 101200
},
{
"epoch": 32.74078862314156,
"grad_norm": 1.3117979764938354,
"learning_rate": 0.001,
"loss": 1.9657,
"step": 101300
},
{
"epoch": 32.773109243697476,
"grad_norm": 1.3686469793319702,
"learning_rate": 0.001,
"loss": 1.9603,
"step": 101400
},
{
"epoch": 32.80542986425339,
"grad_norm": 2.118864059448242,
"learning_rate": 0.001,
"loss": 1.9698,
"step": 101500
},
{
"epoch": 32.837750484809305,
"grad_norm": 1.3416787385940552,
"learning_rate": 0.001,
"loss": 1.979,
"step": 101600
},
{
"epoch": 32.87007110536522,
"grad_norm": 1.4968751668930054,
"learning_rate": 0.001,
"loss": 1.9813,
"step": 101700
},
{
"epoch": 32.902391725921134,
"grad_norm": 1.3256723880767822,
"learning_rate": 0.001,
"loss": 1.9688,
"step": 101800
},
{
"epoch": 32.93471234647705,
"grad_norm": 1.2774500846862793,
"learning_rate": 0.001,
"loss": 1.958,
"step": 101900
},
{
"epoch": 32.967032967032964,
"grad_norm": 1.463400959968567,
"learning_rate": 0.001,
"loss": 1.9767,
"step": 102000
},
{
"epoch": 32.999353587588885,
"grad_norm": 1.4615466594696045,
"learning_rate": 0.001,
"loss": 1.982,
"step": 102100
},
{
"epoch": 33.0316742081448,
"grad_norm": 1.7073653936386108,
"learning_rate": 0.001,
"loss": 1.8432,
"step": 102200
},
{
"epoch": 33.063994828700714,
"grad_norm": 1.9279892444610596,
"learning_rate": 0.001,
"loss": 1.8569,
"step": 102300
},
{
"epoch": 33.09631544925663,
"grad_norm": 1.6899241209030151,
"learning_rate": 0.001,
"loss": 1.8696,
"step": 102400
},
{
"epoch": 33.12863606981254,
"grad_norm": 2.205573797225952,
"learning_rate": 0.001,
"loss": 1.8868,
"step": 102500
},
{
"epoch": 33.16095669036846,
"grad_norm": 1.977630615234375,
"learning_rate": 0.001,
"loss": 1.8899,
"step": 102600
},
{
"epoch": 33.19327731092437,
"grad_norm": 1.4682167768478394,
"learning_rate": 0.001,
"loss": 1.9016,
"step": 102700
},
{
"epoch": 33.22559793148029,
"grad_norm": 1.5439730882644653,
"learning_rate": 0.001,
"loss": 1.8952,
"step": 102800
},
{
"epoch": 33.2579185520362,
"grad_norm": 1.789674162864685,
"learning_rate": 0.001,
"loss": 1.865,
"step": 102900
},
{
"epoch": 33.290239172592116,
"grad_norm": 1.6647350788116455,
"learning_rate": 0.001,
"loss": 1.9034,
"step": 103000
},
{
"epoch": 33.32255979314803,
"grad_norm": 1.892222285270691,
"learning_rate": 0.001,
"loss": 1.8859,
"step": 103100
},
{
"epoch": 33.354880413703945,
"grad_norm": 1.7519506216049194,
"learning_rate": 0.001,
"loss": 1.8986,
"step": 103200
},
{
"epoch": 33.38720103425986,
"grad_norm": 2.073690176010132,
"learning_rate": 0.001,
"loss": 1.9166,
"step": 103300
},
{
"epoch": 33.419521654815775,
"grad_norm": 1.3731398582458496,
"learning_rate": 0.001,
"loss": 1.9031,
"step": 103400
},
{
"epoch": 33.45184227537169,
"grad_norm": 1.480362892150879,
"learning_rate": 0.001,
"loss": 1.8979,
"step": 103500
},
{
"epoch": 33.484162895927604,
"grad_norm": 1.6137398481369019,
"learning_rate": 0.001,
"loss": 1.8968,
"step": 103600
},
{
"epoch": 33.51648351648352,
"grad_norm": 1.6744515895843506,
"learning_rate": 0.001,
"loss": 1.9318,
"step": 103700
},
{
"epoch": 33.54880413703943,
"grad_norm": 1.352559208869934,
"learning_rate": 0.001,
"loss": 1.9356,
"step": 103800
},
{
"epoch": 33.58112475759535,
"grad_norm": 1.5185978412628174,
"learning_rate": 0.001,
"loss": 1.9304,
"step": 103900
},
{
"epoch": 33.61344537815126,
"grad_norm": 1.8947575092315674,
"learning_rate": 0.001,
"loss": 1.9289,
"step": 104000
},
{
"epoch": 33.645765998707176,
"grad_norm": 1.6168104410171509,
"learning_rate": 0.001,
"loss": 1.9526,
"step": 104100
},
{
"epoch": 33.67808661926309,
"grad_norm": 1.509453296661377,
"learning_rate": 0.001,
"loss": 1.9467,
"step": 104200
},
{
"epoch": 33.710407239819006,
"grad_norm": 1.50039541721344,
"learning_rate": 0.001,
"loss": 1.94,
"step": 104300
},
{
"epoch": 33.74272786037492,
"grad_norm": 1.4825176000595093,
"learning_rate": 0.001,
"loss": 1.9455,
"step": 104400
},
{
"epoch": 33.775048480930835,
"grad_norm": 1.9427762031555176,
"learning_rate": 0.001,
"loss": 1.9586,
"step": 104500
},
{
"epoch": 33.80736910148675,
"grad_norm": 1.5415905714035034,
"learning_rate": 0.001,
"loss": 1.9532,
"step": 104600
},
{
"epoch": 33.839689722042664,
"grad_norm": 1.4916168451309204,
"learning_rate": 0.001,
"loss": 1.9705,
"step": 104700
},
{
"epoch": 33.87201034259858,
"grad_norm": 1.6510076522827148,
"learning_rate": 0.001,
"loss": 1.9516,
"step": 104800
},
{
"epoch": 33.90433096315449,
"grad_norm": 1.460364818572998,
"learning_rate": 0.001,
"loss": 1.9801,
"step": 104900
},
{
"epoch": 33.93665158371041,
"grad_norm": 1.7680420875549316,
"learning_rate": 0.001,
"loss": 1.9787,
"step": 105000
},
{
"epoch": 33.96897220426632,
"grad_norm": 1.6624025106430054,
"learning_rate": 0.001,
"loss": 1.9594,
"step": 105100
},
{
"epoch": 34.00129282482224,
"grad_norm": 1.4129273891448975,
"learning_rate": 0.001,
"loss": 1.9536,
"step": 105200
},
{
"epoch": 34.03361344537815,
"grad_norm": 1.3180772066116333,
"learning_rate": 0.001,
"loss": 1.8311,
"step": 105300
},
{
"epoch": 34.065934065934066,
"grad_norm": 1.8205658197402954,
"learning_rate": 0.001,
"loss": 1.8407,
"step": 105400
},
{
"epoch": 34.09825468648998,
"grad_norm": 1.8987846374511719,
"learning_rate": 0.001,
"loss": 1.848,
"step": 105500
},
{
"epoch": 34.130575307045895,
"grad_norm": 1.5684449672698975,
"learning_rate": 0.001,
"loss": 1.8598,
"step": 105600
},
{
"epoch": 34.16289592760181,
"grad_norm": 1.8653782606124878,
"learning_rate": 0.001,
"loss": 1.8451,
"step": 105700
},
{
"epoch": 34.195216548157724,
"grad_norm": 1.5234085321426392,
"learning_rate": 0.001,
"loss": 1.846,
"step": 105800
},
{
"epoch": 34.22753716871364,
"grad_norm": 1.7665132284164429,
"learning_rate": 0.001,
"loss": 1.8603,
"step": 105900
},
{
"epoch": 34.25985778926955,
"grad_norm": 1.7553211450576782,
"learning_rate": 0.001,
"loss": 1.8822,
"step": 106000
},
{
"epoch": 34.29217840982547,
"grad_norm": 1.3010718822479248,
"learning_rate": 0.001,
"loss": 1.8718,
"step": 106100
},
{
"epoch": 34.32449903038138,
"grad_norm": 1.834821343421936,
"learning_rate": 0.001,
"loss": 1.878,
"step": 106200
},
{
"epoch": 34.3568196509373,
"grad_norm": 2.108060598373413,
"learning_rate": 0.001,
"loss": 1.89,
"step": 106300
},
{
"epoch": 34.38914027149321,
"grad_norm": 1.9852526187896729,
"learning_rate": 0.001,
"loss": 1.8889,
"step": 106400
},
{
"epoch": 34.421460892049126,
"grad_norm": 1.7558857202529907,
"learning_rate": 0.001,
"loss": 1.895,
"step": 106500
},
{
"epoch": 34.45378151260504,
"grad_norm": 1.578678846359253,
"learning_rate": 0.001,
"loss": 1.904,
"step": 106600
},
{
"epoch": 34.486102133160955,
"grad_norm": 1.799863576889038,
"learning_rate": 0.001,
"loss": 1.9032,
"step": 106700
},
{
"epoch": 34.51842275371687,
"grad_norm": 1.9496479034423828,
"learning_rate": 0.001,
"loss": 1.8775,
"step": 106800
},
{
"epoch": 34.550743374272784,
"grad_norm": 1.9154863357543945,
"learning_rate": 0.001,
"loss": 1.919,
"step": 106900
},
{
"epoch": 34.5830639948287,
"grad_norm": 1.5972155332565308,
"learning_rate": 0.001,
"loss": 1.8896,
"step": 107000
},
{
"epoch": 34.61538461538461,
"grad_norm": 1.8452078104019165,
"learning_rate": 0.001,
"loss": 1.9033,
"step": 107100
},
{
"epoch": 34.64770523594053,
"grad_norm": 1.7549279928207397,
"learning_rate": 0.001,
"loss": 1.9244,
"step": 107200
},
{
"epoch": 34.68002585649644,
"grad_norm": 1.4020614624023438,
"learning_rate": 0.001,
"loss": 1.9236,
"step": 107300
},
{
"epoch": 34.71234647705236,
"grad_norm": 1.3533083200454712,
"learning_rate": 0.001,
"loss": 1.9217,
"step": 107400
},
{
"epoch": 34.74466709760827,
"grad_norm": 1.2460354566574097,
"learning_rate": 0.001,
"loss": 1.9412,
"step": 107500
},
{
"epoch": 34.776987718164186,
"grad_norm": 2.3600213527679443,
"learning_rate": 0.001,
"loss": 1.9324,
"step": 107600
},
{
"epoch": 34.8093083387201,
"grad_norm": 1.997406244277954,
"learning_rate": 0.001,
"loss": 1.9437,
"step": 107700
},
{
"epoch": 34.841628959276015,
"grad_norm": 1.8307485580444336,
"learning_rate": 0.001,
"loss": 1.9441,
"step": 107800
},
{
"epoch": 34.87394957983193,
"grad_norm": 2.2610950469970703,
"learning_rate": 0.001,
"loss": 1.9382,
"step": 107900
},
{
"epoch": 34.906270200387844,
"grad_norm": 2.170487880706787,
"learning_rate": 0.001,
"loss": 1.9417,
"step": 108000
},
{
"epoch": 34.93859082094376,
"grad_norm": 1.4013649225234985,
"learning_rate": 0.001,
"loss": 1.8065,
"step": 108100
},
{
"epoch": 34.97091144149967,
"grad_norm": 1.6289832592010498,
"learning_rate": 0.001,
"loss": 1.8117,
"step": 108200
},
{
"epoch": 35.003232062055595,
"grad_norm": 1.270293951034546,
"learning_rate": 0.001,
"loss": 1.8191,
"step": 108300
},
{
"epoch": 35.03555268261151,
"grad_norm": 1.2765839099884033,
"learning_rate": 0.001,
"loss": 1.8251,
"step": 108400
},
{
"epoch": 35.067873303167424,
"grad_norm": 1.5106779336929321,
"learning_rate": 0.001,
"loss": 1.8396,
"step": 108500
},
{
"epoch": 35.10019392372334,
"grad_norm": 1.2737852334976196,
"learning_rate": 0.001,
"loss": 1.8212,
"step": 108600
},
{
"epoch": 35.13251454427925,
"grad_norm": 1.17478609085083,
"learning_rate": 0.001,
"loss": 1.826,
"step": 108700
},
{
"epoch": 35.16483516483517,
"grad_norm": 1.2254098653793335,
"learning_rate": 0.001,
"loss": 1.8582,
"step": 108800
},
{
"epoch": 35.19715578539108,
"grad_norm": 1.1669076681137085,
"learning_rate": 0.001,
"loss": 1.8635,
"step": 108900
},
{
"epoch": 35.229476405947,
"grad_norm": 1.737324833869934,
"learning_rate": 0.001,
"loss": 1.8179,
"step": 109000
},
{
"epoch": 35.26179702650291,
"grad_norm": 1.4550049304962158,
"learning_rate": 0.001,
"loss": 1.8437,
"step": 109100
},
{
"epoch": 35.294117647058826,
"grad_norm": 1.1262215375900269,
"learning_rate": 0.001,
"loss": 1.8429,
"step": 109200
},
{
"epoch": 35.32643826761474,
"grad_norm": 1.0687785148620605,
"learning_rate": 0.001,
"loss": 1.8831,
"step": 109300
},
{
"epoch": 35.358758888170655,
"grad_norm": 1.1845074892044067,
"learning_rate": 0.001,
"loss": 1.8829,
"step": 109400
},
{
"epoch": 35.39107950872657,
"grad_norm": 1.4229836463928223,
"learning_rate": 0.001,
"loss": 1.8915,
"step": 109500
},
{
"epoch": 35.423400129282484,
"grad_norm": 1.925002932548523,
"learning_rate": 0.001,
"loss": 1.8926,
"step": 109600
},
{
"epoch": 35.4557207498384,
"grad_norm": 1.5750701427459717,
"learning_rate": 0.001,
"loss": 1.8884,
"step": 109700
},
{
"epoch": 35.48804137039431,
"grad_norm": 1.014614224433899,
"learning_rate": 0.001,
"loss": 1.8861,
"step": 109800
},
{
"epoch": 35.52036199095023,
"grad_norm": 1.449761986732483,
"learning_rate": 0.001,
"loss": 1.8849,
"step": 109900
},
{
"epoch": 35.55268261150614,
"grad_norm": 1.0549952983856201,
"learning_rate": 0.001,
"loss": 1.8771,
"step": 110000
},
{
"epoch": 35.58500323206206,
"grad_norm": 0.9580369591712952,
"learning_rate": 0.001,
"loss": 1.8883,
"step": 110100
},
{
"epoch": 35.61732385261797,
"grad_norm": 1.1707701683044434,
"learning_rate": 0.001,
"loss": 1.9099,
"step": 110200
},
{
"epoch": 35.649644473173886,
"grad_norm": 1.1806001663208008,
"learning_rate": 0.001,
"loss": 1.8985,
"step": 110300
},
{
"epoch": 35.6819650937298,
"grad_norm": 1.0377438068389893,
"learning_rate": 0.001,
"loss": 1.8998,
"step": 110400
},
{
"epoch": 35.714285714285715,
"grad_norm": 1.778603196144104,
"learning_rate": 0.001,
"loss": 1.9012,
"step": 110500
},
{
"epoch": 35.74660633484163,
"grad_norm": 0.9975597858428955,
"learning_rate": 0.001,
"loss": 1.9266,
"step": 110600
},
{
"epoch": 35.778926955397544,
"grad_norm": 1.165048360824585,
"learning_rate": 0.001,
"loss": 1.8915,
"step": 110700
},
{
"epoch": 35.81124757595346,
"grad_norm": 1.077488899230957,
"learning_rate": 0.001,
"loss": 1.9096,
"step": 110800
},
{
"epoch": 35.84356819650937,
"grad_norm": 1.1906706094741821,
"learning_rate": 0.001,
"loss": 1.9161,
"step": 110900
},
{
"epoch": 35.87588881706529,
"grad_norm": 1.0711023807525635,
"learning_rate": 0.001,
"loss": 1.9245,
"step": 111000
},
{
"epoch": 35.9082094376212,
"grad_norm": 1.271654486656189,
"learning_rate": 0.001,
"loss": 1.9274,
"step": 111100
},
{
"epoch": 35.94053005817712,
"grad_norm": 1.110924243927002,
"learning_rate": 0.001,
"loss": 1.9286,
"step": 111200
},
{
"epoch": 35.97285067873303,
"grad_norm": 1.2917808294296265,
"learning_rate": 0.001,
"loss": 1.9163,
"step": 111300
},
{
"epoch": 36.005171299288946,
"grad_norm": 1.2950485944747925,
"learning_rate": 0.001,
"loss": 1.9129,
"step": 111400
},
{
"epoch": 36.03749191984486,
"grad_norm": 1.4829621315002441,
"learning_rate": 0.001,
"loss": 1.7886,
"step": 111500
},
{
"epoch": 36.069812540400775,
"grad_norm": 1.2834502458572388,
"learning_rate": 0.001,
"loss": 1.8144,
"step": 111600
},
{
"epoch": 36.10213316095669,
"grad_norm": 1.270350456237793,
"learning_rate": 0.001,
"loss": 1.8301,
"step": 111700
},
{
"epoch": 36.134453781512605,
"grad_norm": 1.4357999563217163,
"learning_rate": 0.001,
"loss": 1.7967,
"step": 111800
},
{
"epoch": 36.16677440206852,
"grad_norm": 1.1246198415756226,
"learning_rate": 0.001,
"loss": 1.8218,
"step": 111900
},
{
"epoch": 36.199095022624434,
"grad_norm": 1.2654744386672974,
"learning_rate": 0.001,
"loss": 1.8145,
"step": 112000
},
{
"epoch": 36.23141564318035,
"grad_norm": 1.8354674577713013,
"learning_rate": 0.001,
"loss": 1.8153,
"step": 112100
},
{
"epoch": 36.26373626373626,
"grad_norm": 1.1133300065994263,
"learning_rate": 0.001,
"loss": 1.8435,
"step": 112200
},
{
"epoch": 36.29605688429218,
"grad_norm": 1.1699156761169434,
"learning_rate": 0.001,
"loss": 1.8281,
"step": 112300
},
{
"epoch": 36.32837750484809,
"grad_norm": 1.5078686475753784,
"learning_rate": 0.001,
"loss": 1.8332,
"step": 112400
},
{
"epoch": 36.36069812540401,
"grad_norm": 1.1657894849777222,
"learning_rate": 0.001,
"loss": 1.8448,
"step": 112500
},
{
"epoch": 36.39301874595992,
"grad_norm": 1.599133014678955,
"learning_rate": 0.001,
"loss": 1.8624,
"step": 112600
},
{
"epoch": 36.425339366515836,
"grad_norm": 1.1557692289352417,
"learning_rate": 0.001,
"loss": 1.8584,
"step": 112700
},
{
"epoch": 36.45765998707175,
"grad_norm": 0.9912543892860413,
"learning_rate": 0.001,
"loss": 1.8578,
"step": 112800
},
{
"epoch": 36.489980607627665,
"grad_norm": 1.2750211954116821,
"learning_rate": 0.001,
"loss": 1.8695,
"step": 112900
},
{
"epoch": 36.52230122818358,
"grad_norm": 1.4301689863204956,
"learning_rate": 0.001,
"loss": 1.8802,
"step": 113000
},
{
"epoch": 36.554621848739494,
"grad_norm": 1.2405816316604614,
"learning_rate": 0.001,
"loss": 1.884,
"step": 113100
},
{
"epoch": 36.58694246929541,
"grad_norm": 1.1132670640945435,
"learning_rate": 0.001,
"loss": 1.8758,
"step": 113200
},
{
"epoch": 36.61926308985132,
"grad_norm": 1.3236029148101807,
"learning_rate": 0.001,
"loss": 1.8611,
"step": 113300
},
{
"epoch": 36.65158371040724,
"grad_norm": 1.155246615409851,
"learning_rate": 0.001,
"loss": 1.8855,
"step": 113400
},
{
"epoch": 36.68390433096315,
"grad_norm": 1.4745430946350098,
"learning_rate": 0.001,
"loss": 1.8873,
"step": 113500
},
{
"epoch": 36.71622495151907,
"grad_norm": 1.5691783428192139,
"learning_rate": 0.001,
"loss": 1.9005,
"step": 113600
},
{
"epoch": 36.74854557207498,
"grad_norm": 1.1504912376403809,
"learning_rate": 0.001,
"loss": 1.8877,
"step": 113700
},
{
"epoch": 36.780866192630896,
"grad_norm": 1.3625767230987549,
"learning_rate": 0.001,
"loss": 1.904,
"step": 113800
},
{
"epoch": 36.81318681318681,
"grad_norm": 1.1113677024841309,
"learning_rate": 0.001,
"loss": 1.89,
"step": 113900
},
{
"epoch": 36.845507433742725,
"grad_norm": 1.145803451538086,
"learning_rate": 0.001,
"loss": 1.9003,
"step": 114000
},
{
"epoch": 36.87782805429864,
"grad_norm": 1.5965522527694702,
"learning_rate": 0.001,
"loss": 1.8836,
"step": 114100
},
{
"epoch": 36.910148674854554,
"grad_norm": 1.327394723892212,
"learning_rate": 0.001,
"loss": 1.9113,
"step": 114200
},
{
"epoch": 36.94246929541047,
"grad_norm": 1.6013472080230713,
"learning_rate": 0.001,
"loss": 1.9,
"step": 114300
},
{
"epoch": 36.97478991596638,
"grad_norm": 1.7597237825393677,
"learning_rate": 0.001,
"loss": 1.8883,
"step": 114400
},
{
"epoch": 37.007110536522305,
"grad_norm": 1.0151056051254272,
"learning_rate": 0.001,
"loss": 1.8983,
"step": 114500
},
{
"epoch": 37.03943115707822,
"grad_norm": 1.1664142608642578,
"learning_rate": 0.001,
"loss": 1.782,
"step": 114600
},
{
"epoch": 37.071751777634134,
"grad_norm": 1.137633204460144,
"learning_rate": 0.001,
"loss": 1.7748,
"step": 114700
},
{
"epoch": 37.10407239819005,
"grad_norm": 1.2252521514892578,
"learning_rate": 0.001,
"loss": 1.7822,
"step": 114800
},
{
"epoch": 37.13639301874596,
"grad_norm": 1.2115730047225952,
"learning_rate": 0.001,
"loss": 1.7875,
"step": 114900
},
{
"epoch": 37.16871363930188,
"grad_norm": 1.4415392875671387,
"learning_rate": 0.001,
"loss": 1.7997,
"step": 115000
},
{
"epoch": 37.20103425985779,
"grad_norm": 1.3861497640609741,
"learning_rate": 0.001,
"loss": 1.8012,
"step": 115100
},
{
"epoch": 37.23335488041371,
"grad_norm": 1.2090829610824585,
"learning_rate": 0.001,
"loss": 1.8125,
"step": 115200
},
{
"epoch": 37.26567550096962,
"grad_norm": 1.1123391389846802,
"learning_rate": 0.001,
"loss": 1.8151,
"step": 115300
},
{
"epoch": 37.297996121525536,
"grad_norm": 1.5135940313339233,
"learning_rate": 0.001,
"loss": 1.8508,
"step": 115400
},
{
"epoch": 37.33031674208145,
"grad_norm": 1.0719391107559204,
"learning_rate": 0.001,
"loss": 1.8327,
"step": 115500
},
{
"epoch": 37.362637362637365,
"grad_norm": 1.1940752267837524,
"learning_rate": 0.001,
"loss": 1.8109,
"step": 115600
},
{
"epoch": 37.39495798319328,
"grad_norm": 1.0901352167129517,
"learning_rate": 0.001,
"loss": 1.8367,
"step": 115700
},
{
"epoch": 37.427278603749194,
"grad_norm": 1.2462451457977295,
"learning_rate": 0.001,
"loss": 1.8481,
"step": 115800
},
{
"epoch": 37.45959922430511,
"grad_norm": 1.1731107234954834,
"learning_rate": 0.001,
"loss": 1.8475,
"step": 115900
},
{
"epoch": 37.49191984486102,
"grad_norm": 1.084913969039917,
"learning_rate": 0.001,
"loss": 1.8554,
"step": 116000
},
{
"epoch": 37.52424046541694,
"grad_norm": 1.2160426378250122,
"learning_rate": 0.001,
"loss": 1.8617,
"step": 116100
},
{
"epoch": 37.55656108597285,
"grad_norm": 1.1056900024414062,
"learning_rate": 0.001,
"loss": 1.8533,
"step": 116200
},
{
"epoch": 37.58888170652877,
"grad_norm": 1.3891226053237915,
"learning_rate": 0.001,
"loss": 1.8457,
"step": 116300
},
{
"epoch": 37.62120232708468,
"grad_norm": 1.1189597845077515,
"learning_rate": 0.001,
"loss": 1.842,
"step": 116400
},
{
"epoch": 37.653522947640596,
"grad_norm": 1.3444385528564453,
"learning_rate": 0.001,
"loss": 1.8537,
"step": 116500
},
{
"epoch": 37.68584356819651,
"grad_norm": 1.1547706127166748,
"learning_rate": 0.001,
"loss": 1.854,
"step": 116600
},
{
"epoch": 37.718164188752425,
"grad_norm": 1.2527415752410889,
"learning_rate": 0.001,
"loss": 1.8603,
"step": 116700
},
{
"epoch": 37.75048480930834,
"grad_norm": 1.585039496421814,
"learning_rate": 0.001,
"loss": 1.8622,
"step": 116800
},
{
"epoch": 37.782805429864254,
"grad_norm": 1.0463566780090332,
"learning_rate": 0.001,
"loss": 1.8865,
"step": 116900
},
{
"epoch": 37.81512605042017,
"grad_norm": 1.2690690755844116,
"learning_rate": 0.001,
"loss": 1.8914,
"step": 117000
},
{
"epoch": 37.84744667097608,
"grad_norm": 1.0178471803665161,
"learning_rate": 0.001,
"loss": 1.879,
"step": 117100
},
{
"epoch": 37.879767291532,
"grad_norm": 1.3066000938415527,
"learning_rate": 0.001,
"loss": 1.8612,
"step": 117200
},
{
"epoch": 37.91208791208791,
"grad_norm": 1.1154258251190186,
"learning_rate": 0.001,
"loss": 1.8915,
"step": 117300
},
{
"epoch": 37.94440853264383,
"grad_norm": 1.3577524423599243,
"learning_rate": 0.001,
"loss": 1.887,
"step": 117400
},
{
"epoch": 37.97672915319974,
"grad_norm": 1.2028940916061401,
"learning_rate": 0.001,
"loss": 1.8889,
"step": 117500
},
{
"epoch": 38.009049773755656,
"grad_norm": 1.2554447650909424,
"learning_rate": 0.001,
"loss": 1.8535,
"step": 117600
},
{
"epoch": 38.04137039431157,
"grad_norm": 0.9973150491714478,
"learning_rate": 0.001,
"loss": 1.7474,
"step": 117700
},
{
"epoch": 38.073691014867485,
"grad_norm": 1.3402920961380005,
"learning_rate": 0.001,
"loss": 1.768,
"step": 117800
},
{
"epoch": 38.1060116354234,
"grad_norm": 1.1451469659805298,
"learning_rate": 0.001,
"loss": 1.7855,
"step": 117900
},
{
"epoch": 38.138332255979314,
"grad_norm": 1.3159552812576294,
"learning_rate": 0.001,
"loss": 1.7795,
"step": 118000
},
{
"epoch": 38.17065287653523,
"grad_norm": 1.3436837196350098,
"learning_rate": 0.001,
"loss": 1.7757,
"step": 118100
},
{
"epoch": 38.20297349709114,
"grad_norm": 1.199753999710083,
"learning_rate": 0.001,
"loss": 1.795,
"step": 118200
},
{
"epoch": 38.23529411764706,
"grad_norm": 1.554911494255066,
"learning_rate": 0.001,
"loss": 1.7991,
"step": 118300
},
{
"epoch": 38.26761473820297,
"grad_norm": 1.1167523860931396,
"learning_rate": 0.001,
"loss": 1.7872,
"step": 118400
},
{
"epoch": 38.29993535875889,
"grad_norm": 1.311654806137085,
"learning_rate": 0.001,
"loss": 1.8169,
"step": 118500
},
{
"epoch": 38.3322559793148,
"grad_norm": 1.6712486743927002,
"learning_rate": 0.001,
"loss": 1.8145,
"step": 118600
},
{
"epoch": 38.364576599870716,
"grad_norm": 0.9774516820907593,
"learning_rate": 0.001,
"loss": 1.8278,
"step": 118700
},
{
"epoch": 38.39689722042663,
"grad_norm": 1.1913068294525146,
"learning_rate": 0.001,
"loss": 1.8286,
"step": 118800
},
{
"epoch": 38.429217840982545,
"grad_norm": 1.3465591669082642,
"learning_rate": 0.001,
"loss": 1.8082,
"step": 118900
},
{
"epoch": 38.46153846153846,
"grad_norm": 1.1644989252090454,
"learning_rate": 0.001,
"loss": 1.8158,
"step": 119000
},
{
"epoch": 38.493859082094374,
"grad_norm": 1.0077077150344849,
"learning_rate": 0.001,
"loss": 1.8106,
"step": 119100
},
{
"epoch": 38.52617970265029,
"grad_norm": 1.3327784538269043,
"learning_rate": 0.001,
"loss": 1.8498,
"step": 119200
},
{
"epoch": 38.558500323206204,
"grad_norm": 1.1973927021026611,
"learning_rate": 0.001,
"loss": 1.8168,
"step": 119300
},
{
"epoch": 38.59082094376212,
"grad_norm": 1.050500750541687,
"learning_rate": 0.001,
"loss": 1.846,
"step": 119400
},
{
"epoch": 38.62314156431803,
"grad_norm": 1.1645978689193726,
"learning_rate": 0.001,
"loss": 1.8453,
"step": 119500
},
{
"epoch": 38.65546218487395,
"grad_norm": 1.3637398481369019,
"learning_rate": 0.001,
"loss": 1.8463,
"step": 119600
},
{
"epoch": 38.68778280542986,
"grad_norm": 1.3179012537002563,
"learning_rate": 0.001,
"loss": 1.8473,
"step": 119700
},
{
"epoch": 38.720103425985776,
"grad_norm": 1.071566104888916,
"learning_rate": 0.001,
"loss": 1.8554,
"step": 119800
},
{
"epoch": 38.75242404654169,
"grad_norm": 1.2979928255081177,
"learning_rate": 0.001,
"loss": 1.8318,
"step": 119900
},
{
"epoch": 38.784744667097605,
"grad_norm": 1.385211706161499,
"learning_rate": 0.001,
"loss": 1.869,
"step": 120000
},
{
"epoch": 38.81706528765352,
"grad_norm": 1.2491968870162964,
"learning_rate": 0.001,
"loss": 1.8605,
"step": 120100
},
{
"epoch": 38.849385908209435,
"grad_norm": 1.3451728820800781,
"learning_rate": 0.001,
"loss": 1.8739,
"step": 120200
},
{
"epoch": 38.88170652876535,
"grad_norm": 1.1662501096725464,
"learning_rate": 0.001,
"loss": 1.8527,
"step": 120300
},
{
"epoch": 38.914027149321264,
"grad_norm": 1.403456449508667,
"learning_rate": 0.001,
"loss": 1.8517,
"step": 120400
},
{
"epoch": 38.94634776987718,
"grad_norm": 1.1871743202209473,
"learning_rate": 0.001,
"loss": 1.8573,
"step": 120500
},
{
"epoch": 38.97866839043309,
"grad_norm": 1.348589539527893,
"learning_rate": 0.001,
"loss": 1.8732,
"step": 120600
},
{
"epoch": 39.010989010989015,
"grad_norm": 1.171324372291565,
"learning_rate": 0.001,
"loss": 1.8387,
"step": 120700
},
{
"epoch": 39.04330963154493,
"grad_norm": 1.1314347982406616,
"learning_rate": 0.001,
"loss": 1.7288,
"step": 120800
},
{
"epoch": 39.075630252100844,
"grad_norm": 1.217405915260315,
"learning_rate": 0.001,
"loss": 1.7486,
"step": 120900
},
{
"epoch": 39.10795087265676,
"grad_norm": 1.478462815284729,
"learning_rate": 0.001,
"loss": 1.7444,
"step": 121000
},
{
"epoch": 39.14027149321267,
"grad_norm": 1.2993769645690918,
"learning_rate": 0.001,
"loss": 1.7543,
"step": 121100
},
{
"epoch": 39.17259211376859,
"grad_norm": 1.2720222473144531,
"learning_rate": 0.001,
"loss": 1.7821,
"step": 121200
},
{
"epoch": 39.2049127343245,
"grad_norm": 1.4102205038070679,
"learning_rate": 0.001,
"loss": 1.7989,
"step": 121300
},
{
"epoch": 39.237233354880416,
"grad_norm": 1.1800966262817383,
"learning_rate": 0.001,
"loss": 1.7867,
"step": 121400
},
{
"epoch": 39.26955397543633,
"grad_norm": 1.4887824058532715,
"learning_rate": 0.001,
"loss": 1.7867,
"step": 121500
},
{
"epoch": 39.301874595992246,
"grad_norm": 1.2865124940872192,
"learning_rate": 0.001,
"loss": 1.7895,
"step": 121600
},
{
"epoch": 39.33419521654816,
"grad_norm": 1.3429749011993408,
"learning_rate": 0.001,
"loss": 1.7979,
"step": 121700
},
{
"epoch": 39.366515837104075,
"grad_norm": 1.4717779159545898,
"learning_rate": 0.001,
"loss": 1.771,
"step": 121800
},
{
"epoch": 39.39883645765999,
"grad_norm": 1.2592474222183228,
"learning_rate": 0.001,
"loss": 1.7973,
"step": 121900
},
{
"epoch": 39.431157078215904,
"grad_norm": 1.437862753868103,
"learning_rate": 0.001,
"loss": 1.7797,
"step": 122000
},
{
"epoch": 39.46347769877182,
"grad_norm": 1.138623595237732,
"learning_rate": 0.001,
"loss": 1.81,
"step": 122100
},
{
"epoch": 39.49579831932773,
"grad_norm": 1.2658520936965942,
"learning_rate": 0.001,
"loss": 1.8067,
"step": 122200
},
{
"epoch": 39.52811893988365,
"grad_norm": 1.3263376951217651,
"learning_rate": 0.001,
"loss": 1.8373,
"step": 122300
},
{
"epoch": 39.56043956043956,
"grad_norm": 1.1942200660705566,
"learning_rate": 0.001,
"loss": 1.8216,
"step": 122400
},
{
"epoch": 39.59276018099548,
"grad_norm": 1.4261666536331177,
"learning_rate": 0.001,
"loss": 1.8351,
"step": 122500
},
{
"epoch": 39.62508080155139,
"grad_norm": 1.150886058807373,
"learning_rate": 0.001,
"loss": 1.8424,
"step": 122600
},
{
"epoch": 39.657401422107306,
"grad_norm": 1.138588309288025,
"learning_rate": 0.001,
"loss": 1.7967,
"step": 122700
},
{
"epoch": 39.68972204266322,
"grad_norm": 1.2674330472946167,
"learning_rate": 0.001,
"loss": 1.8387,
"step": 122800
},
{
"epoch": 39.722042663219135,
"grad_norm": 1.2652851343154907,
"learning_rate": 0.001,
"loss": 1.8355,
"step": 122900
},
{
"epoch": 39.75436328377505,
"grad_norm": 1.4298686981201172,
"learning_rate": 0.001,
"loss": 1.8304,
"step": 123000
},
{
"epoch": 39.786683904330964,
"grad_norm": 1.1751641035079956,
"learning_rate": 0.001,
"loss": 1.8263,
"step": 123100
},
{
"epoch": 39.81900452488688,
"grad_norm": 1.1655226945877075,
"learning_rate": 0.001,
"loss": 1.8414,
"step": 123200
},
{
"epoch": 39.85132514544279,
"grad_norm": 1.0795261859893799,
"learning_rate": 0.001,
"loss": 1.8327,
"step": 123300
},
{
"epoch": 39.88364576599871,
"grad_norm": 1.1121963262557983,
"learning_rate": 0.001,
"loss": 1.8496,
"step": 123400
},
{
"epoch": 39.91596638655462,
"grad_norm": 1.2073920965194702,
"learning_rate": 0.001,
"loss": 1.8319,
"step": 123500
},
{
"epoch": 39.94828700711054,
"grad_norm": 1.1852302551269531,
"learning_rate": 0.001,
"loss": 1.8667,
"step": 123600
},
{
"epoch": 39.98060762766645,
"grad_norm": 1.0930063724517822,
"learning_rate": 0.001,
"loss": 1.8584,
"step": 123700
},
{
"epoch": 40.012928248222366,
"grad_norm": 1.1705491542816162,
"learning_rate": 0.001,
"loss": 1.7917,
"step": 123800
},
{
"epoch": 40.04524886877828,
"grad_norm": 1.2436919212341309,
"learning_rate": 0.001,
"loss": 1.7282,
"step": 123900
},
{
"epoch": 40.077569489334195,
"grad_norm": 1.3184545040130615,
"learning_rate": 0.001,
"loss": 1.7347,
"step": 124000
},
{
"epoch": 40.10989010989011,
"grad_norm": 1.3338842391967773,
"learning_rate": 0.001,
"loss": 1.7409,
"step": 124100
},
{
"epoch": 40.142210730446024,
"grad_norm": 1.5150983333587646,
"learning_rate": 0.001,
"loss": 1.741,
"step": 124200
},
{
"epoch": 40.17453135100194,
"grad_norm": 1.5849483013153076,
"learning_rate": 0.001,
"loss": 1.7584,
"step": 124300
},
{
"epoch": 40.20685197155785,
"grad_norm": 1.0649653673171997,
"learning_rate": 0.001,
"loss": 1.7502,
"step": 124400
},
{
"epoch": 40.23917259211377,
"grad_norm": 1.420655608177185,
"learning_rate": 0.001,
"loss": 1.7463,
"step": 124500
},
{
"epoch": 40.27149321266968,
"grad_norm": 1.3057841062545776,
"learning_rate": 0.001,
"loss": 1.7717,
"step": 124600
},
{
"epoch": 40.3038138332256,
"grad_norm": 1.3937748670578003,
"learning_rate": 0.001,
"loss": 1.7604,
"step": 124700
},
{
"epoch": 40.33613445378151,
"grad_norm": 1.790507435798645,
"learning_rate": 0.001,
"loss": 1.7711,
"step": 124800
},
{
"epoch": 40.368455074337426,
"grad_norm": 1.1703600883483887,
"learning_rate": 0.001,
"loss": 1.7776,
"step": 124900
},
{
"epoch": 40.40077569489334,
"grad_norm": 1.289276123046875,
"learning_rate": 0.001,
"loss": 1.7834,
"step": 125000
},
{
"epoch": 40.433096315449255,
"grad_norm": 1.3727136850357056,
"learning_rate": 0.001,
"loss": 1.8043,
"step": 125100
},
{
"epoch": 40.46541693600517,
"grad_norm": 1.207993507385254,
"learning_rate": 0.001,
"loss": 1.7791,
"step": 125200
},
{
"epoch": 40.497737556561084,
"grad_norm": 1.7379050254821777,
"learning_rate": 0.001,
"loss": 1.7926,
"step": 125300
},
{
"epoch": 40.530058177117,
"grad_norm": 1.3238317966461182,
"learning_rate": 0.001,
"loss": 1.8091,
"step": 125400
},
{
"epoch": 40.56237879767291,
"grad_norm": 1.2644617557525635,
"learning_rate": 0.001,
"loss": 1.8055,
"step": 125500
},
{
"epoch": 40.59469941822883,
"grad_norm": 1.5530192852020264,
"learning_rate": 0.001,
"loss": 1.8002,
"step": 125600
},
{
"epoch": 40.62702003878474,
"grad_norm": 1.3474758863449097,
"learning_rate": 0.001,
"loss": 1.8054,
"step": 125700
},
{
"epoch": 40.65934065934066,
"grad_norm": 1.2770376205444336,
"learning_rate": 0.001,
"loss": 1.7988,
"step": 125800
},
{
"epoch": 40.69166127989657,
"grad_norm": 1.4564834833145142,
"learning_rate": 0.001,
"loss": 1.8026,
"step": 125900
},
{
"epoch": 40.723981900452486,
"grad_norm": 1.1911042928695679,
"learning_rate": 0.001,
"loss": 1.8408,
"step": 126000
},
{
"epoch": 40.7563025210084,
"grad_norm": 1.2096915245056152,
"learning_rate": 0.001,
"loss": 1.8205,
"step": 126100
},
{
"epoch": 40.788623141564315,
"grad_norm": 1.3765822649002075,
"learning_rate": 0.001,
"loss": 1.8151,
"step": 126200
},
{
"epoch": 40.82094376212023,
"grad_norm": 1.6281813383102417,
"learning_rate": 0.001,
"loss": 1.829,
"step": 126300
},
{
"epoch": 40.853264382676144,
"grad_norm": 1.737960696220398,
"learning_rate": 0.001,
"loss": 1.8347,
"step": 126400
},
{
"epoch": 40.88558500323206,
"grad_norm": 1.4104831218719482,
"learning_rate": 0.001,
"loss": 1.8304,
"step": 126500
},
{
"epoch": 40.91790562378797,
"grad_norm": 1.5236070156097412,
"learning_rate": 0.001,
"loss": 1.839,
"step": 126600
},
{
"epoch": 40.95022624434389,
"grad_norm": 1.216387391090393,
"learning_rate": 0.001,
"loss": 1.8372,
"step": 126700
},
{
"epoch": 40.9825468648998,
"grad_norm": 1.1370539665222168,
"learning_rate": 0.001,
"loss": 1.8326,
"step": 126800
},
{
"epoch": 41.014867485455724,
"grad_norm": 1.4220774173736572,
"learning_rate": 0.001,
"loss": 1.7687,
"step": 126900
},
{
"epoch": 41.04718810601164,
"grad_norm": 1.4660181999206543,
"learning_rate": 0.001,
"loss": 1.712,
"step": 127000
},
{
"epoch": 41.07950872656755,
"grad_norm": 1.775199294090271,
"learning_rate": 0.001,
"loss": 1.7265,
"step": 127100
},
{
"epoch": 41.11182934712347,
"grad_norm": 1.3170700073242188,
"learning_rate": 0.001,
"loss": 1.7168,
"step": 127200
},
{
"epoch": 41.14414996767938,
"grad_norm": 1.3832727670669556,
"learning_rate": 0.001,
"loss": 1.7168,
"step": 127300
},
{
"epoch": 41.1764705882353,
"grad_norm": 1.2850581407546997,
"learning_rate": 0.001,
"loss": 1.725,
"step": 127400
},
{
"epoch": 41.20879120879121,
"grad_norm": 1.934682011604309,
"learning_rate": 0.001,
"loss": 1.7353,
"step": 127500
},
{
"epoch": 41.241111829347126,
"grad_norm": 1.7525626420974731,
"learning_rate": 0.001,
"loss": 1.7417,
"step": 127600
},
{
"epoch": 41.27343244990304,
"grad_norm": 1.6928536891937256,
"learning_rate": 0.001,
"loss": 1.7562,
"step": 127700
},
{
"epoch": 41.305753070458955,
"grad_norm": 1.5103766918182373,
"learning_rate": 0.001,
"loss": 1.7574,
"step": 127800
},
{
"epoch": 41.33807369101487,
"grad_norm": 1.3830496072769165,
"learning_rate": 0.001,
"loss": 1.7741,
"step": 127900
},
{
"epoch": 41.370394311570784,
"grad_norm": 1.5678727626800537,
"learning_rate": 0.001,
"loss": 1.7421,
"step": 128000
},
{
"epoch": 41.4027149321267,
"grad_norm": 1.347474217414856,
"learning_rate": 0.001,
"loss": 1.768,
"step": 128100
},
{
"epoch": 41.43503555268261,
"grad_norm": 1.4597406387329102,
"learning_rate": 0.001,
"loss": 1.7628,
"step": 128200
},
{
"epoch": 41.46735617323853,
"grad_norm": 1.6288129091262817,
"learning_rate": 0.001,
"loss": 1.7674,
"step": 128300
},
{
"epoch": 41.49967679379444,
"grad_norm": 1.61184561252594,
"learning_rate": 0.001,
"loss": 1.7732,
"step": 128400
},
{
"epoch": 41.53199741435036,
"grad_norm": 1.2897828817367554,
"learning_rate": 0.001,
"loss": 1.7847,
"step": 128500
},
{
"epoch": 41.56431803490627,
"grad_norm": 1.4574735164642334,
"learning_rate": 0.001,
"loss": 1.7764,
"step": 128600
},
{
"epoch": 41.596638655462186,
"grad_norm": 1.6599246263504028,
"learning_rate": 0.001,
"loss": 1.8019,
"step": 128700
},
{
"epoch": 41.6289592760181,
"grad_norm": 1.3278685808181763,
"learning_rate": 0.001,
"loss": 1.7833,
"step": 128800
},
{
"epoch": 41.661279896574015,
"grad_norm": 1.7833093404769897,
"learning_rate": 0.001,
"loss": 1.8139,
"step": 128900
},
{
"epoch": 41.69360051712993,
"grad_norm": 1.8515368700027466,
"learning_rate": 0.001,
"loss": 1.7861,
"step": 129000
},
{
"epoch": 41.725921137685845,
"grad_norm": 1.4936338663101196,
"learning_rate": 0.001,
"loss": 1.7915,
"step": 129100
},
{
"epoch": 41.75824175824176,
"grad_norm": 1.5401523113250732,
"learning_rate": 0.001,
"loss": 1.8067,
"step": 129200
},
{
"epoch": 41.790562378797674,
"grad_norm": 1.4799628257751465,
"learning_rate": 0.001,
"loss": 1.8145,
"step": 129300
},
{
"epoch": 41.82288299935359,
"grad_norm": 1.9319769144058228,
"learning_rate": 0.001,
"loss": 1.8139,
"step": 129400
},
{
"epoch": 41.8552036199095,
"grad_norm": 1.4047659635543823,
"learning_rate": 0.001,
"loss": 1.8053,
"step": 129500
},
{
"epoch": 41.88752424046542,
"grad_norm": 1.699422836303711,
"learning_rate": 0.001,
"loss": 1.8142,
"step": 129600
},
{
"epoch": 41.91984486102133,
"grad_norm": 1.6171603202819824,
"learning_rate": 0.001,
"loss": 1.806,
"step": 129700
},
{
"epoch": 41.95216548157725,
"grad_norm": 1.5447322130203247,
"learning_rate": 0.001,
"loss": 1.823,
"step": 129800
},
{
"epoch": 41.98448610213316,
"grad_norm": 1.7629934549331665,
"learning_rate": 0.001,
"loss": 1.838,
"step": 129900
},
{
"epoch": 42.016806722689076,
"grad_norm": 1.6176948547363281,
"learning_rate": 0.001,
"loss": 1.7306,
"step": 130000
},
{
"epoch": 42.04912734324499,
"grad_norm": 2.1387195587158203,
"learning_rate": 0.001,
"loss": 1.6773,
"step": 130100
},
{
"epoch": 42.081447963800905,
"grad_norm": 1.9669650793075562,
"learning_rate": 0.001,
"loss": 1.7198,
"step": 130200
},
{
"epoch": 42.11376858435682,
"grad_norm": 1.9538878202438354,
"learning_rate": 0.001,
"loss": 1.6987,
"step": 130300
},
{
"epoch": 42.146089204912734,
"grad_norm": 2.1951351165771484,
"learning_rate": 0.001,
"loss": 1.7034,
"step": 130400
},
{
"epoch": 42.17840982546865,
"grad_norm": 2.5766024589538574,
"learning_rate": 0.001,
"loss": 1.7365,
"step": 130500
},
{
"epoch": 42.21073044602456,
"grad_norm": 1.9949456453323364,
"learning_rate": 0.001,
"loss": 1.7106,
"step": 130600
},
{
"epoch": 42.24305106658048,
"grad_norm": 2.021125078201294,
"learning_rate": 0.001,
"loss": 1.714,
"step": 130700
},
{
"epoch": 42.27537168713639,
"grad_norm": 1.7046053409576416,
"learning_rate": 0.001,
"loss": 1.712,
"step": 130800
},
{
"epoch": 42.30769230769231,
"grad_norm": 2.0010311603546143,
"learning_rate": 0.001,
"loss": 1.7228,
"step": 130900
},
{
"epoch": 42.34001292824822,
"grad_norm": 1.3972162008285522,
"learning_rate": 0.001,
"loss": 1.7467,
"step": 131000
},
{
"epoch": 42.372333548804136,
"grad_norm": 1.5133181810379028,
"learning_rate": 0.001,
"loss": 1.7408,
"step": 131100
},
{
"epoch": 42.40465416936005,
"grad_norm": 1.9723691940307617,
"learning_rate": 0.001,
"loss": 1.7605,
"step": 131200
},
{
"epoch": 42.436974789915965,
"grad_norm": 1.8973649740219116,
"learning_rate": 0.001,
"loss": 1.7396,
"step": 131300
},
{
"epoch": 42.46929541047188,
"grad_norm": 1.744896650314331,
"learning_rate": 0.001,
"loss": 1.7506,
"step": 131400
},
{
"epoch": 42.501616031027794,
"grad_norm": 1.9301925897598267,
"learning_rate": 0.001,
"loss": 1.7605,
"step": 131500
},
{
"epoch": 42.53393665158371,
"grad_norm": 2.0170085430145264,
"learning_rate": 0.001,
"loss": 1.752,
"step": 131600
},
{
"epoch": 42.56625727213962,
"grad_norm": 1.1445215940475464,
"learning_rate": 0.001,
"loss": 1.7764,
"step": 131700
},
{
"epoch": 42.59857789269554,
"grad_norm": 2.092996597290039,
"learning_rate": 0.001,
"loss": 1.7804,
"step": 131800
},
{
"epoch": 42.63089851325145,
"grad_norm": 2.0091073513031006,
"learning_rate": 0.001,
"loss": 1.7648,
"step": 131900
},
{
"epoch": 42.66321913380737,
"grad_norm": 1.9225186109542847,
"learning_rate": 0.001,
"loss": 1.7727,
"step": 132000
},
{
"epoch": 42.69553975436328,
"grad_norm": 1.7107021808624268,
"learning_rate": 0.001,
"loss": 1.775,
"step": 132100
},
{
"epoch": 42.727860374919196,
"grad_norm": 1.3927497863769531,
"learning_rate": 0.001,
"loss": 1.7671,
"step": 132200
},
{
"epoch": 42.76018099547511,
"grad_norm": 1.5459237098693848,
"learning_rate": 0.001,
"loss": 1.8023,
"step": 132300
},
{
"epoch": 42.792501616031025,
"grad_norm": 1.3331663608551025,
"learning_rate": 0.001,
"loss": 1.7908,
"step": 132400
},
{
"epoch": 42.82482223658694,
"grad_norm": 2.1158242225646973,
"learning_rate": 0.001,
"loss": 1.8026,
"step": 132500
},
{
"epoch": 42.857142857142854,
"grad_norm": 1.8317747116088867,
"learning_rate": 0.001,
"loss": 1.8011,
"step": 132600
},
{
"epoch": 42.88946347769877,
"grad_norm": 1.4998470544815063,
"learning_rate": 0.001,
"loss": 1.7998,
"step": 132700
},
{
"epoch": 42.92178409825468,
"grad_norm": 1.4395893812179565,
"learning_rate": 0.001,
"loss": 1.8224,
"step": 132800
},
{
"epoch": 42.9541047188106,
"grad_norm": 1.49153470993042,
"learning_rate": 0.001,
"loss": 1.8059,
"step": 132900
},
{
"epoch": 42.98642533936652,
"grad_norm": 1.8367140293121338,
"learning_rate": 0.001,
"loss": 1.8171,
"step": 133000
},
{
"epoch": 43.018745959922434,
"grad_norm": 1.2948594093322754,
"learning_rate": 0.001,
"loss": 1.754,
"step": 133100
},
{
"epoch": 43.05106658047835,
"grad_norm": 1.4209692478179932,
"learning_rate": 0.001,
"loss": 1.6638,
"step": 133200
},
{
"epoch": 43.08338720103426,
"grad_norm": 1.572913408279419,
"learning_rate": 0.001,
"loss": 1.6966,
"step": 133300
},
{
"epoch": 43.11570782159018,
"grad_norm": 1.2720844745635986,
"learning_rate": 0.001,
"loss": 1.6996,
"step": 133400
},
{
"epoch": 43.14802844214609,
"grad_norm": 1.0568156242370605,
"learning_rate": 0.001,
"loss": 1.6909,
"step": 133500
},
{
"epoch": 43.18034906270201,
"grad_norm": 1.1196587085723877,
"learning_rate": 0.001,
"loss": 1.6917,
"step": 133600
},
{
"epoch": 43.21266968325792,
"grad_norm": 1.3726385831832886,
"learning_rate": 0.001,
"loss": 1.698,
"step": 133700
},
{
"epoch": 43.244990303813836,
"grad_norm": 1.357574462890625,
"learning_rate": 0.001,
"loss": 1.7049,
"step": 133800
},
{
"epoch": 43.27731092436975,
"grad_norm": 1.2265974283218384,
"learning_rate": 0.001,
"loss": 1.7088,
"step": 133900
},
{
"epoch": 43.309631544925665,
"grad_norm": 1.5340291261672974,
"learning_rate": 0.001,
"loss": 1.7125,
"step": 134000
},
{
"epoch": 43.34195216548158,
"grad_norm": 1.5428131818771362,
"learning_rate": 0.001,
"loss": 1.7163,
"step": 134100
},
{
"epoch": 43.374272786037494,
"grad_norm": 1.3394761085510254,
"learning_rate": 0.001,
"loss": 1.7525,
"step": 134200
},
{
"epoch": 43.40659340659341,
"grad_norm": 1.1329821348190308,
"learning_rate": 0.001,
"loss": 1.7229,
"step": 134300
},
{
"epoch": 43.43891402714932,
"grad_norm": 1.2983161211013794,
"learning_rate": 0.001,
"loss": 1.7067,
"step": 134400
},
{
"epoch": 43.47123464770524,
"grad_norm": 1.276678204536438,
"learning_rate": 0.001,
"loss": 1.7471,
"step": 134500
},
{
"epoch": 43.50355526826115,
"grad_norm": 1.3433374166488647,
"learning_rate": 0.001,
"loss": 1.7431,
"step": 134600
},
{
"epoch": 43.53587588881707,
"grad_norm": 1.0081208944320679,
"learning_rate": 0.001,
"loss": 1.7449,
"step": 134700
},
{
"epoch": 43.56819650937298,
"grad_norm": 1.333104133605957,
"learning_rate": 0.001,
"loss": 1.7442,
"step": 134800
},
{
"epoch": 43.600517129928896,
"grad_norm": 1.4379981756210327,
"learning_rate": 0.001,
"loss": 1.7416,
"step": 134900
},
{
"epoch": 43.63283775048481,
"grad_norm": 1.1848903894424438,
"learning_rate": 0.001,
"loss": 1.765,
"step": 135000
},
{
"epoch": 43.665158371040725,
"grad_norm": 1.187063217163086,
"learning_rate": 0.001,
"loss": 1.7603,
"step": 135100
},
{
"epoch": 43.69747899159664,
"grad_norm": 1.11912202835083,
"learning_rate": 0.001,
"loss": 1.7573,
"step": 135200
},
{
"epoch": 43.729799612152554,
"grad_norm": 1.25471031665802,
"learning_rate": 0.001,
"loss": 1.7711,
"step": 135300
},
{
"epoch": 43.76212023270847,
"grad_norm": 1.2159885168075562,
"learning_rate": 0.001,
"loss": 1.7716,
"step": 135400
},
{
"epoch": 43.79444085326438,
"grad_norm": 1.2494391202926636,
"learning_rate": 0.001,
"loss": 1.7711,
"step": 135500
},
{
"epoch": 43.8267614738203,
"grad_norm": 1.2417315244674683,
"learning_rate": 0.001,
"loss": 1.7871,
"step": 135600
},
{
"epoch": 43.85908209437621,
"grad_norm": 1.1563262939453125,
"learning_rate": 0.001,
"loss": 1.7767,
"step": 135700
},
{
"epoch": 43.89140271493213,
"grad_norm": 1.0287482738494873,
"learning_rate": 0.001,
"loss": 1.7849,
"step": 135800
},
{
"epoch": 43.92372333548804,
"grad_norm": 1.4275707006454468,
"learning_rate": 0.001,
"loss": 1.8058,
"step": 135900
},
{
"epoch": 43.956043956043956,
"grad_norm": 1.2920149564743042,
"learning_rate": 0.001,
"loss": 1.7951,
"step": 136000
},
{
"epoch": 43.98836457659987,
"grad_norm": 1.296828269958496,
"learning_rate": 0.001,
"loss": 1.7996,
"step": 136100
},
{
"epoch": 44.020685197155785,
"grad_norm": 1.3403992652893066,
"learning_rate": 0.001,
"loss": 1.7191,
"step": 136200
},
{
"epoch": 44.0530058177117,
"grad_norm": 1.2069450616836548,
"learning_rate": 0.001,
"loss": 1.6725,
"step": 136300
},
{
"epoch": 44.085326438267614,
"grad_norm": 1.4134432077407837,
"learning_rate": 0.001,
"loss": 1.6671,
"step": 136400
},
{
"epoch": 44.11764705882353,
"grad_norm": 1.1776862144470215,
"learning_rate": 0.001,
"loss": 1.6715,
"step": 136500
},
{
"epoch": 44.14996767937944,
"grad_norm": 1.1428087949752808,
"learning_rate": 0.001,
"loss": 1.6761,
"step": 136600
},
{
"epoch": 44.18228829993536,
"grad_norm": 1.2049680948257446,
"learning_rate": 0.001,
"loss": 1.6872,
"step": 136700
},
{
"epoch": 44.21460892049127,
"grad_norm": 1.144275188446045,
"learning_rate": 0.001,
"loss": 1.7005,
"step": 136800
},
{
"epoch": 44.24692954104719,
"grad_norm": 1.3733307123184204,
"learning_rate": 0.001,
"loss": 1.7016,
"step": 136900
},
{
"epoch": 44.2792501616031,
"grad_norm": 1.1372228860855103,
"learning_rate": 0.001,
"loss": 1.7033,
"step": 137000
},
{
"epoch": 44.311570782159016,
"grad_norm": 1.2623342275619507,
"learning_rate": 0.001,
"loss": 1.7136,
"step": 137100
},
{
"epoch": 44.34389140271493,
"grad_norm": 1.043894648551941,
"learning_rate": 0.001,
"loss": 1.7116,
"step": 137200
},
{
"epoch": 44.376212023270845,
"grad_norm": 1.0046995878219604,
"learning_rate": 0.001,
"loss": 1.7213,
"step": 137300
},
{
"epoch": 44.40853264382676,
"grad_norm": 1.1621452569961548,
"learning_rate": 0.001,
"loss": 1.7161,
"step": 137400
},
{
"epoch": 44.440853264382675,
"grad_norm": 1.4386694431304932,
"learning_rate": 0.001,
"loss": 1.7227,
"step": 137500
},
{
"epoch": 44.47317388493859,
"grad_norm": 1.2840430736541748,
"learning_rate": 0.001,
"loss": 1.7289,
"step": 137600
},
{
"epoch": 44.505494505494504,
"grad_norm": 1.3832279443740845,
"learning_rate": 0.001,
"loss": 1.7046,
"step": 137700
},
{
"epoch": 44.53781512605042,
"grad_norm": 1.3804891109466553,
"learning_rate": 0.001,
"loss": 1.7302,
"step": 137800
},
{
"epoch": 44.57013574660633,
"grad_norm": 1.189717411994934,
"learning_rate": 0.001,
"loss": 1.7289,
"step": 137900
},
{
"epoch": 44.60245636716225,
"grad_norm": 2.2020466327667236,
"learning_rate": 0.001,
"loss": 1.7504,
"step": 138000
},
{
"epoch": 44.63477698771816,
"grad_norm": 1.2542753219604492,
"learning_rate": 0.001,
"loss": 1.7345,
"step": 138100
},
{
"epoch": 44.66709760827408,
"grad_norm": 1.3898725509643555,
"learning_rate": 0.001,
"loss": 1.7473,
"step": 138200
},
{
"epoch": 44.69941822882999,
"grad_norm": 1.042006015777588,
"learning_rate": 0.001,
"loss": 1.7503,
"step": 138300
},
{
"epoch": 44.731738849385906,
"grad_norm": 0.9393309950828552,
"learning_rate": 0.001,
"loss": 1.7488,
"step": 138400
},
{
"epoch": 44.76405946994182,
"grad_norm": 1.120185136795044,
"learning_rate": 0.001,
"loss": 1.7621,
"step": 138500
},
{
"epoch": 44.796380090497735,
"grad_norm": 1.320479154586792,
"learning_rate": 0.001,
"loss": 1.7632,
"step": 138600
},
{
"epoch": 44.82870071105365,
"grad_norm": 1.2445160150527954,
"learning_rate": 0.001,
"loss": 1.755,
"step": 138700
},
{
"epoch": 44.861021331609564,
"grad_norm": 1.3491894006729126,
"learning_rate": 0.001,
"loss": 1.778,
"step": 138800
},
{
"epoch": 44.89334195216548,
"grad_norm": 1.1886285543441772,
"learning_rate": 0.001,
"loss": 1.7729,
"step": 138900
},
{
"epoch": 44.92566257272139,
"grad_norm": 1.2313807010650635,
"learning_rate": 0.001,
"loss": 1.7762,
"step": 139000
},
{
"epoch": 44.95798319327731,
"grad_norm": 1.4140355587005615,
"learning_rate": 0.001,
"loss": 1.7579,
"step": 139100
},
{
"epoch": 44.99030381383322,
"grad_norm": 1.0410194396972656,
"learning_rate": 0.001,
"loss": 1.7332,
"step": 139200
},
{
"epoch": 45.022624434389144,
"grad_norm": 1.048113226890564,
"learning_rate": 0.001,
"loss": 1.7027,
"step": 139300
},
{
"epoch": 45.05494505494506,
"grad_norm": 1.1873400211334229,
"learning_rate": 0.001,
"loss": 1.6334,
"step": 139400
},
{
"epoch": 45.08726567550097,
"grad_norm": 1.2329055070877075,
"learning_rate": 0.001,
"loss": 1.6492,
"step": 139500
},
{
"epoch": 45.11958629605689,
"grad_norm": 1.0407984256744385,
"learning_rate": 0.001,
"loss": 1.6427,
"step": 139600
},
{
"epoch": 45.1519069166128,
"grad_norm": 1.0917448997497559,
"learning_rate": 0.001,
"loss": 1.658,
"step": 139700
},
{
"epoch": 45.18422753716872,
"grad_norm": 1.1773816347122192,
"learning_rate": 0.001,
"loss": 1.6732,
"step": 139800
},
{
"epoch": 45.21654815772463,
"grad_norm": 1.658249020576477,
"learning_rate": 0.001,
"loss": 1.6625,
"step": 139900
},
{
"epoch": 45.248868778280546,
"grad_norm": 1.5720107555389404,
"learning_rate": 0.001,
"loss": 1.6774,
"step": 140000
},
{
"epoch": 45.28118939883646,
"grad_norm": 1.2399346828460693,
"learning_rate": 0.001,
"loss": 1.6753,
"step": 140100
},
{
"epoch": 45.313510019392375,
"grad_norm": 1.329980731010437,
"learning_rate": 0.001,
"loss": 1.6896,
"step": 140200
},
{
"epoch": 45.34583063994829,
"grad_norm": 1.1705814599990845,
"learning_rate": 0.001,
"loss": 1.6716,
"step": 140300
},
{
"epoch": 45.378151260504204,
"grad_norm": 1.4165809154510498,
"learning_rate": 0.001,
"loss": 1.7108,
"step": 140400
},
{
"epoch": 45.41047188106012,
"grad_norm": 1.368194341659546,
"learning_rate": 0.001,
"loss": 1.6997,
"step": 140500
},
{
"epoch": 45.44279250161603,
"grad_norm": 1.1229863166809082,
"learning_rate": 0.001,
"loss": 1.6974,
"step": 140600
},
{
"epoch": 45.47511312217195,
"grad_norm": 1.2703062295913696,
"learning_rate": 0.001,
"loss": 1.705,
"step": 140700
},
{
"epoch": 45.50743374272786,
"grad_norm": 1.3791265487670898,
"learning_rate": 0.001,
"loss": 1.6941,
"step": 140800
},
{
"epoch": 45.53975436328378,
"grad_norm": 1.1580239534378052,
"learning_rate": 0.001,
"loss": 1.7248,
"step": 140900
},
{
"epoch": 45.57207498383969,
"grad_norm": 1.0476043224334717,
"learning_rate": 0.001,
"loss": 1.7077,
"step": 141000
},
{
"epoch": 45.604395604395606,
"grad_norm": 1.4689886569976807,
"learning_rate": 0.001,
"loss": 1.724,
"step": 141100
},
{
"epoch": 45.63671622495152,
"grad_norm": 1.0328713655471802,
"learning_rate": 0.001,
"loss": 1.7425,
"step": 141200
},
{
"epoch": 45.669036845507435,
"grad_norm": 1.0610711574554443,
"learning_rate": 0.001,
"loss": 1.7307,
"step": 141300
},
{
"epoch": 45.70135746606335,
"grad_norm": 1.0974621772766113,
"learning_rate": 0.001,
"loss": 1.7307,
"step": 141400
},
{
"epoch": 45.733678086619264,
"grad_norm": 1.216745138168335,
"learning_rate": 0.001,
"loss": 1.7432,
"step": 141500
},
{
"epoch": 45.76599870717518,
"grad_norm": 1.0733321905136108,
"learning_rate": 0.001,
"loss": 1.7484,
"step": 141600
},
{
"epoch": 45.79831932773109,
"grad_norm": 0.9696832895278931,
"learning_rate": 0.001,
"loss": 1.751,
"step": 141700
},
{
"epoch": 45.83063994828701,
"grad_norm": 1.1975910663604736,
"learning_rate": 0.001,
"loss": 1.7506,
"step": 141800
},
{
"epoch": 45.86296056884292,
"grad_norm": 1.1205062866210938,
"learning_rate": 0.001,
"loss": 1.7508,
"step": 141900
},
{
"epoch": 45.89528118939884,
"grad_norm": 1.140873908996582,
"learning_rate": 0.001,
"loss": 1.7555,
"step": 142000
},
{
"epoch": 45.92760180995475,
"grad_norm": 0.9974058866500854,
"learning_rate": 0.001,
"loss": 1.7575,
"step": 142100
},
{
"epoch": 45.959922430510666,
"grad_norm": 1.1679439544677734,
"learning_rate": 0.001,
"loss": 1.7463,
"step": 142200
},
{
"epoch": 45.99224305106658,
"grad_norm": 1.2680093050003052,
"learning_rate": 0.001,
"loss": 1.7786,
"step": 142300
},
{
"epoch": 46.024563671622495,
"grad_norm": 1.1620368957519531,
"learning_rate": 0.001,
"loss": 1.6726,
"step": 142400
},
{
"epoch": 46.05688429217841,
"grad_norm": 1.236384391784668,
"learning_rate": 0.001,
"loss": 1.6274,
"step": 142500
},
{
"epoch": 46.089204912734324,
"grad_norm": 1.2753865718841553,
"learning_rate": 0.001,
"loss": 1.6372,
"step": 142600
},
{
"epoch": 46.12152553329024,
"grad_norm": 1.1083111763000488,
"learning_rate": 0.001,
"loss": 1.64,
"step": 142700
},
{
"epoch": 46.15384615384615,
"grad_norm": 1.5008981227874756,
"learning_rate": 0.001,
"loss": 1.6584,
"step": 142800
},
{
"epoch": 46.18616677440207,
"grad_norm": 1.4679440259933472,
"learning_rate": 0.001,
"loss": 1.6597,
"step": 142900
},
{
"epoch": 46.21848739495798,
"grad_norm": 1.2379117012023926,
"learning_rate": 0.001,
"loss": 1.6758,
"step": 143000
},
{
"epoch": 46.2508080155139,
"grad_norm": 1.652895450592041,
"learning_rate": 0.001,
"loss": 1.659,
"step": 143100
},
{
"epoch": 46.28312863606981,
"grad_norm": 1.1448554992675781,
"learning_rate": 0.001,
"loss": 1.6725,
"step": 143200
},
{
"epoch": 46.315449256625726,
"grad_norm": 1.1704193353652954,
"learning_rate": 0.001,
"loss": 1.6663,
"step": 143300
},
{
"epoch": 46.34776987718164,
"grad_norm": 1.1541613340377808,
"learning_rate": 0.001,
"loss": 1.6694,
"step": 143400
},
{
"epoch": 46.380090497737555,
"grad_norm": 0.9936338067054749,
"learning_rate": 0.001,
"loss": 1.6858,
"step": 143500
},
{
"epoch": 46.41241111829347,
"grad_norm": 1.7548086643218994,
"learning_rate": 0.001,
"loss": 1.6764,
"step": 143600
},
{
"epoch": 46.444731738849384,
"grad_norm": 1.2922335863113403,
"learning_rate": 0.001,
"loss": 1.6929,
"step": 143700
},
{
"epoch": 46.4770523594053,
"grad_norm": 1.2852715253829956,
"learning_rate": 0.001,
"loss": 1.7032,
"step": 143800
},
{
"epoch": 46.50937297996121,
"grad_norm": 1.1478217840194702,
"learning_rate": 0.001,
"loss": 1.6902,
"step": 143900
},
{
"epoch": 46.54169360051713,
"grad_norm": 1.0429531335830688,
"learning_rate": 0.001,
"loss": 1.6988,
"step": 144000
},
{
"epoch": 46.57401422107304,
"grad_norm": 1.299400806427002,
"learning_rate": 0.001,
"loss": 1.6928,
"step": 144100
},
{
"epoch": 46.60633484162896,
"grad_norm": 1.503138542175293,
"learning_rate": 0.001,
"loss": 1.7166,
"step": 144200
},
{
"epoch": 46.63865546218487,
"grad_norm": 1.1313776969909668,
"learning_rate": 0.001,
"loss": 1.7079,
"step": 144300
},
{
"epoch": 46.670976082740786,
"grad_norm": 1.2164987325668335,
"learning_rate": 0.001,
"loss": 1.7295,
"step": 144400
},
{
"epoch": 46.7032967032967,
"grad_norm": 1.297400712966919,
"learning_rate": 0.001,
"loss": 1.7195,
"step": 144500
},
{
"epoch": 46.735617323852615,
"grad_norm": 1.4451942443847656,
"learning_rate": 0.001,
"loss": 1.7135,
"step": 144600
},
{
"epoch": 46.76793794440853,
"grad_norm": 1.3415268659591675,
"learning_rate": 0.001,
"loss": 1.7186,
"step": 144700
},
{
"epoch": 46.800258564964444,
"grad_norm": 1.1677778959274292,
"learning_rate": 0.001,
"loss": 1.7394,
"step": 144800
},
{
"epoch": 46.83257918552036,
"grad_norm": 1.1284695863723755,
"learning_rate": 0.001,
"loss": 1.7275,
"step": 144900
},
{
"epoch": 46.864899806076274,
"grad_norm": 1.3730621337890625,
"learning_rate": 0.001,
"loss": 1.7432,
"step": 145000
},
{
"epoch": 46.89722042663219,
"grad_norm": 1.3622809648513794,
"learning_rate": 0.001,
"loss": 1.7449,
"step": 145100
},
{
"epoch": 46.9295410471881,
"grad_norm": 0.9944179654121399,
"learning_rate": 0.001,
"loss": 1.7164,
"step": 145200
},
{
"epoch": 46.96186166774402,
"grad_norm": 1.303714394569397,
"learning_rate": 0.001,
"loss": 1.744,
"step": 145300
},
{
"epoch": 46.99418228829994,
"grad_norm": 1.3328958749771118,
"learning_rate": 0.001,
"loss": 1.749,
"step": 145400
},
{
"epoch": 47.02650290885585,
"grad_norm": 1.3161877393722534,
"learning_rate": 0.001,
"loss": 1.649,
"step": 145500
},
{
"epoch": 47.05882352941177,
"grad_norm": 1.5366662740707397,
"learning_rate": 0.001,
"loss": 1.6058,
"step": 145600
},
{
"epoch": 47.09114414996768,
"grad_norm": 1.253591537475586,
"learning_rate": 0.001,
"loss": 1.6181,
"step": 145700
},
{
"epoch": 47.1234647705236,
"grad_norm": 1.2932579517364502,
"learning_rate": 0.001,
"loss": 1.624,
"step": 145800
},
{
"epoch": 47.15578539107951,
"grad_norm": 1.3068318367004395,
"learning_rate": 0.001,
"loss": 1.6277,
"step": 145900
},
{
"epoch": 47.188106011635426,
"grad_norm": 1.383806586265564,
"learning_rate": 0.001,
"loss": 1.6314,
"step": 146000
},
{
"epoch": 47.22042663219134,
"grad_norm": 1.0713759660720825,
"learning_rate": 0.001,
"loss": 1.6558,
"step": 146100
},
{
"epoch": 47.252747252747255,
"grad_norm": 1.3981571197509766,
"learning_rate": 0.001,
"loss": 1.6294,
"step": 146200
},
{
"epoch": 47.28506787330317,
"grad_norm": 1.1140717267990112,
"learning_rate": 0.001,
"loss": 1.6651,
"step": 146300
},
{
"epoch": 47.317388493859085,
"grad_norm": 1.1881879568099976,
"learning_rate": 0.001,
"loss": 1.658,
"step": 146400
},
{
"epoch": 47.349709114415,
"grad_norm": 1.2144122123718262,
"learning_rate": 0.001,
"loss": 1.668,
"step": 146500
},
{
"epoch": 47.382029734970914,
"grad_norm": 1.7134535312652588,
"learning_rate": 0.001,
"loss": 1.68,
"step": 146600
},
{
"epoch": 47.41435035552683,
"grad_norm": 1.2519067525863647,
"learning_rate": 0.001,
"loss": 1.6727,
"step": 146700
},
{
"epoch": 47.44667097608274,
"grad_norm": 1.1829216480255127,
"learning_rate": 0.001,
"loss": 1.6554,
"step": 146800
},
{
"epoch": 47.47899159663866,
"grad_norm": 1.0220831632614136,
"learning_rate": 0.001,
"loss": 1.6635,
"step": 146900
},
{
"epoch": 47.51131221719457,
"grad_norm": 1.3235119581222534,
"learning_rate": 0.001,
"loss": 1.6951,
"step": 147000
},
{
"epoch": 47.543632837750486,
"grad_norm": 1.3594244718551636,
"learning_rate": 0.001,
"loss": 1.6987,
"step": 147100
},
{
"epoch": 47.5759534583064,
"grad_norm": 1.788058876991272,
"learning_rate": 0.001,
"loss": 1.6945,
"step": 147200
},
{
"epoch": 47.608274078862316,
"grad_norm": 1.1875792741775513,
"learning_rate": 0.001,
"loss": 1.7033,
"step": 147300
},
{
"epoch": 47.64059469941823,
"grad_norm": 1.5549057722091675,
"learning_rate": 0.001,
"loss": 1.6989,
"step": 147400
},
{
"epoch": 47.672915319974145,
"grad_norm": 1.230398416519165,
"learning_rate": 0.001,
"loss": 1.7024,
"step": 147500
},
{
"epoch": 47.70523594053006,
"grad_norm": 1.6183626651763916,
"learning_rate": 0.001,
"loss": 1.7055,
"step": 147600
},
{
"epoch": 47.737556561085974,
"grad_norm": 1.8326687812805176,
"learning_rate": 0.001,
"loss": 1.7056,
"step": 147700
},
{
"epoch": 47.76987718164189,
"grad_norm": 1.1874862909317017,
"learning_rate": 0.001,
"loss": 1.7156,
"step": 147800
},
{
"epoch": 47.8021978021978,
"grad_norm": 1.2575957775115967,
"learning_rate": 0.001,
"loss": 1.7264,
"step": 147900
},
{
"epoch": 47.83451842275372,
"grad_norm": 1.193088412284851,
"learning_rate": 0.001,
"loss": 1.7184,
"step": 148000
},
{
"epoch": 47.86683904330963,
"grad_norm": 1.1475036144256592,
"learning_rate": 0.001,
"loss": 1.7198,
"step": 148100
},
{
"epoch": 47.89915966386555,
"grad_norm": 1.2878127098083496,
"learning_rate": 0.001,
"loss": 1.7383,
"step": 148200
},
{
"epoch": 47.93148028442146,
"grad_norm": 1.2342934608459473,
"learning_rate": 0.001,
"loss": 1.7097,
"step": 148300
},
{
"epoch": 47.963800904977376,
"grad_norm": 1.435988187789917,
"learning_rate": 0.001,
"loss": 1.7289,
"step": 148400
},
{
"epoch": 47.99612152553329,
"grad_norm": 1.3129411935806274,
"learning_rate": 0.001,
"loss": 1.7033,
"step": 148500
},
{
"epoch": 48.028442146089205,
"grad_norm": 1.4182060956954956,
"learning_rate": 0.001,
"loss": 1.608,
"step": 148600
},
{
"epoch": 48.06076276664512,
"grad_norm": 1.6356170177459717,
"learning_rate": 0.001,
"loss": 1.6059,
"step": 148700
},
{
"epoch": 48.093083387201034,
"grad_norm": 1.415715217590332,
"learning_rate": 0.001,
"loss": 1.6117,
"step": 148800
},
{
"epoch": 48.12540400775695,
"grad_norm": 1.3101872205734253,
"learning_rate": 0.001,
"loss": 1.6278,
"step": 148900
},
{
"epoch": 48.15772462831286,
"grad_norm": 1.1667542457580566,
"learning_rate": 0.001,
"loss": 1.6146,
"step": 149000
},
{
"epoch": 48.19004524886878,
"grad_norm": 1.8620003461837769,
"learning_rate": 0.001,
"loss": 1.6246,
"step": 149100
},
{
"epoch": 48.22236586942469,
"grad_norm": 1.1716053485870361,
"learning_rate": 0.001,
"loss": 1.6139,
"step": 149200
},
{
"epoch": 48.25468648998061,
"grad_norm": 1.7759467363357544,
"learning_rate": 0.001,
"loss": 1.6344,
"step": 149300
},
{
"epoch": 48.28700711053652,
"grad_norm": 1.3979485034942627,
"learning_rate": 0.001,
"loss": 1.651,
"step": 149400
},
{
"epoch": 48.319327731092436,
"grad_norm": 1.4052960872650146,
"learning_rate": 0.001,
"loss": 1.6206,
"step": 149500
},
{
"epoch": 48.35164835164835,
"grad_norm": 1.207869291305542,
"learning_rate": 0.001,
"loss": 1.6584,
"step": 149600
},
{
"epoch": 48.383968972204265,
"grad_norm": 1.6692159175872803,
"learning_rate": 0.001,
"loss": 1.6559,
"step": 149700
},
{
"epoch": 48.41628959276018,
"grad_norm": 1.7472057342529297,
"learning_rate": 0.001,
"loss": 1.643,
"step": 149800
},
{
"epoch": 48.448610213316094,
"grad_norm": 1.3058013916015625,
"learning_rate": 0.001,
"loss": 1.6692,
"step": 149900
},
{
"epoch": 48.48093083387201,
"grad_norm": 1.4042518138885498,
"learning_rate": 0.001,
"loss": 1.6509,
"step": 150000
},
{
"epoch": 48.51325145442792,
"grad_norm": 2.1124401092529297,
"learning_rate": 0.001,
"loss": 1.6736,
"step": 150100
},
{
"epoch": 48.54557207498384,
"grad_norm": 1.3654600381851196,
"learning_rate": 0.001,
"loss": 1.6823,
"step": 150200
},
{
"epoch": 48.57789269553975,
"grad_norm": 1.5814836025238037,
"learning_rate": 0.001,
"loss": 1.6768,
"step": 150300
},
{
"epoch": 48.61021331609567,
"grad_norm": 1.3242175579071045,
"learning_rate": 0.001,
"loss": 1.6928,
"step": 150400
},
{
"epoch": 48.64253393665158,
"grad_norm": 1.6400493383407593,
"learning_rate": 0.001,
"loss": 1.6863,
"step": 150500
},
{
"epoch": 48.674854557207496,
"grad_norm": 1.2978423833847046,
"learning_rate": 0.001,
"loss": 1.682,
"step": 150600
},
{
"epoch": 48.70717517776341,
"grad_norm": 1.3399165868759155,
"learning_rate": 0.001,
"loss": 1.6951,
"step": 150700
},
{
"epoch": 48.739495798319325,
"grad_norm": 1.3431782722473145,
"learning_rate": 0.001,
"loss": 1.6975,
"step": 150800
},
{
"epoch": 48.77181641887524,
"grad_norm": 1.300646424293518,
"learning_rate": 0.001,
"loss": 1.6944,
"step": 150900
},
{
"epoch": 48.804137039431154,
"grad_norm": 1.6648505926132202,
"learning_rate": 0.001,
"loss": 1.6911,
"step": 151000
},
{
"epoch": 48.83645765998707,
"grad_norm": 1.2654876708984375,
"learning_rate": 0.001,
"loss": 1.7074,
"step": 151100
},
{
"epoch": 48.86877828054298,
"grad_norm": 1.3668768405914307,
"learning_rate": 0.001,
"loss": 1.7168,
"step": 151200
},
{
"epoch": 48.9010989010989,
"grad_norm": 1.6690231561660767,
"learning_rate": 0.001,
"loss": 1.7055,
"step": 151300
},
{
"epoch": 48.93341952165481,
"grad_norm": 1.5245463848114014,
"learning_rate": 0.001,
"loss": 1.7168,
"step": 151400
},
{
"epoch": 48.96574014221073,
"grad_norm": 1.3595877885818481,
"learning_rate": 0.001,
"loss": 1.7119,
"step": 151500
},
{
"epoch": 48.99806076276664,
"grad_norm": 1.5084638595581055,
"learning_rate": 0.001,
"loss": 1.6959,
"step": 151600
},
{
"epoch": 49.03038138332256,
"grad_norm": 1.4485028982162476,
"learning_rate": 0.001,
"loss": 1.5905,
"step": 151700
},
{
"epoch": 49.06270200387848,
"grad_norm": 1.3076313734054565,
"learning_rate": 0.001,
"loss": 1.6,
"step": 151800
},
{
"epoch": 49.09502262443439,
"grad_norm": 1.7316114902496338,
"learning_rate": 0.001,
"loss": 1.6014,
"step": 151900
},
{
"epoch": 49.12734324499031,
"grad_norm": 1.5655585527420044,
"learning_rate": 0.001,
"loss": 1.6008,
"step": 152000
},
{
"epoch": 49.15966386554622,
"grad_norm": 1.3464454412460327,
"learning_rate": 0.001,
"loss": 1.6198,
"step": 152100
},
{
"epoch": 49.191984486102136,
"grad_norm": 1.3522858619689941,
"learning_rate": 0.001,
"loss": 1.6008,
"step": 152200
},
{
"epoch": 49.22430510665805,
"grad_norm": 2.0585334300994873,
"learning_rate": 0.001,
"loss": 1.6151,
"step": 152300
},
{
"epoch": 49.256625727213965,
"grad_norm": 1.9595283269882202,
"learning_rate": 0.001,
"loss": 1.6309,
"step": 152400
},
{
"epoch": 49.28894634776988,
"grad_norm": 1.6998517513275146,
"learning_rate": 0.001,
"loss": 1.6264,
"step": 152500
},
{
"epoch": 49.321266968325794,
"grad_norm": 1.498993992805481,
"learning_rate": 0.001,
"loss": 1.6317,
"step": 152600
},
{
"epoch": 49.35358758888171,
"grad_norm": 2.1215360164642334,
"learning_rate": 0.001,
"loss": 1.6553,
"step": 152700
},
{
"epoch": 49.38590820943762,
"grad_norm": 1.4123033285140991,
"learning_rate": 0.001,
"loss": 1.6303,
"step": 152800
},
{
"epoch": 49.41822882999354,
"grad_norm": 1.7786338329315186,
"learning_rate": 0.001,
"loss": 1.6451,
"step": 152900
},
{
"epoch": 49.45054945054945,
"grad_norm": 1.9890562295913696,
"learning_rate": 0.001,
"loss": 1.6549,
"step": 153000
},
{
"epoch": 49.48287007110537,
"grad_norm": 1.3054035902023315,
"learning_rate": 0.001,
"loss": 1.648,
"step": 153100
},
{
"epoch": 49.51519069166128,
"grad_norm": 1.3741906881332397,
"learning_rate": 0.001,
"loss": 1.6617,
"step": 153200
},
{
"epoch": 49.547511312217196,
"grad_norm": 1.2835386991500854,
"learning_rate": 0.001,
"loss": 1.6662,
"step": 153300
},
{
"epoch": 49.57983193277311,
"grad_norm": 1.8139817714691162,
"learning_rate": 0.001,
"loss": 1.6708,
"step": 153400
},
{
"epoch": 49.612152553329025,
"grad_norm": 1.3035625219345093,
"learning_rate": 0.001,
"loss": 1.6728,
"step": 153500
},
{
"epoch": 49.64447317388494,
"grad_norm": 1.3546018600463867,
"learning_rate": 0.001,
"loss": 1.6712,
"step": 153600
},
{
"epoch": 49.676793794440854,
"grad_norm": 1.3287252187728882,
"learning_rate": 0.001,
"loss": 1.6543,
"step": 153700
},
{
"epoch": 49.70911441499677,
"grad_norm": 1.3180716037750244,
"learning_rate": 0.001,
"loss": 1.6752,
"step": 153800
},
{
"epoch": 49.74143503555268,
"grad_norm": 1.4854649305343628,
"learning_rate": 0.001,
"loss": 1.6772,
"step": 153900
},
{
"epoch": 49.7737556561086,
"grad_norm": 1.431453824043274,
"learning_rate": 0.001,
"loss": 1.6771,
"step": 154000
},
{
"epoch": 49.80607627666451,
"grad_norm": 1.555336833000183,
"learning_rate": 0.001,
"loss": 1.6712,
"step": 154100
},
{
"epoch": 49.83839689722043,
"grad_norm": 1.423966646194458,
"learning_rate": 0.001,
"loss": 1.681,
"step": 154200
},
{
"epoch": 49.87071751777634,
"grad_norm": 1.4483835697174072,
"learning_rate": 0.001,
"loss": 1.6988,
"step": 154300
},
{
"epoch": 49.903038138332256,
"grad_norm": 1.2455962896347046,
"learning_rate": 0.001,
"loss": 1.6905,
"step": 154400
},
{
"epoch": 49.93535875888817,
"grad_norm": 1.520082712173462,
"learning_rate": 0.001,
"loss": 1.6931,
"step": 154500
},
{
"epoch": 49.967679379444085,
"grad_norm": 1.745983362197876,
"learning_rate": 0.001,
"loss": 1.7124,
"step": 154600
},
{
"epoch": 50.0,
"grad_norm": 2.6711509227752686,
"learning_rate": 0.001,
"loss": 1.6643,
"step": 154700
},
{
"epoch": 50.032320620555915,
"grad_norm": 2.0904746055603027,
"learning_rate": 0.001,
"loss": 1.553,
"step": 154800
},
{
"epoch": 50.06464124111183,
"grad_norm": 2.082350254058838,
"learning_rate": 0.001,
"loss": 1.5871,
"step": 154900
},
{
"epoch": 50.096961861667744,
"grad_norm": 1.6668823957443237,
"learning_rate": 0.001,
"loss": 1.5962,
"step": 155000
},
{
"epoch": 50.12928248222366,
"grad_norm": 2.0948734283447266,
"learning_rate": 0.001,
"loss": 1.5955,
"step": 155100
},
{
"epoch": 50.16160310277957,
"grad_norm": 1.8577214479446411,
"learning_rate": 0.001,
"loss": 1.5987,
"step": 155200
},
{
"epoch": 50.19392372333549,
"grad_norm": 2.1729252338409424,
"learning_rate": 0.001,
"loss": 1.6153,
"step": 155300
},
{
"epoch": 50.2262443438914,
"grad_norm": 2.4875996112823486,
"learning_rate": 0.001,
"loss": 1.6038,
"step": 155400
},
{
"epoch": 50.25856496444732,
"grad_norm": 2.2401797771453857,
"learning_rate": 0.001,
"loss": 1.6118,
"step": 155500
},
{
"epoch": 50.29088558500323,
"grad_norm": 1.8535085916519165,
"learning_rate": 0.001,
"loss": 1.6258,
"step": 155600
},
{
"epoch": 50.323206205559146,
"grad_norm": 1.8018325567245483,
"learning_rate": 0.001,
"loss": 1.6138,
"step": 155700
},
{
"epoch": 50.35552682611506,
"grad_norm": 1.9164797067642212,
"learning_rate": 0.001,
"loss": 1.6396,
"step": 155800
},
{
"epoch": 50.387847446670975,
"grad_norm": 1.7693629264831543,
"learning_rate": 0.001,
"loss": 1.6316,
"step": 155900
},
{
"epoch": 50.42016806722689,
"grad_norm": 2.478234052658081,
"learning_rate": 0.001,
"loss": 1.648,
"step": 156000
},
{
"epoch": 50.452488687782804,
"grad_norm": 1.620030164718628,
"learning_rate": 0.001,
"loss": 1.6249,
"step": 156100
},
{
"epoch": 50.48480930833872,
"grad_norm": 1.9655663967132568,
"learning_rate": 0.001,
"loss": 1.6238,
"step": 156200
},
{
"epoch": 50.51712992889463,
"grad_norm": 2.349524736404419,
"learning_rate": 0.001,
"loss": 1.6398,
"step": 156300
},
{
"epoch": 50.54945054945055,
"grad_norm": 2.469360828399658,
"learning_rate": 0.001,
"loss": 1.6381,
"step": 156400
},
{
"epoch": 50.58177117000646,
"grad_norm": 1.8590805530548096,
"learning_rate": 0.001,
"loss": 1.6407,
"step": 156500
},
{
"epoch": 50.61409179056238,
"grad_norm": 1.8361051082611084,
"learning_rate": 0.001,
"loss": 1.6545,
"step": 156600
},
{
"epoch": 50.64641241111829,
"grad_norm": 2.1574673652648926,
"learning_rate": 0.001,
"loss": 1.6414,
"step": 156700
},
{
"epoch": 50.678733031674206,
"grad_norm": 2.5325615406036377,
"learning_rate": 0.001,
"loss": 1.6423,
"step": 156800
},
{
"epoch": 50.71105365223012,
"grad_norm": 1.9068058729171753,
"learning_rate": 0.001,
"loss": 1.6399,
"step": 156900
},
{
"epoch": 50.743374272786035,
"grad_norm": 2.3132662773132324,
"learning_rate": 0.001,
"loss": 1.6825,
"step": 157000
},
{
"epoch": 50.77569489334195,
"grad_norm": 1.922098994255066,
"learning_rate": 0.001,
"loss": 1.6573,
"step": 157100
},
{
"epoch": 50.808015513897864,
"grad_norm": 1.8278084993362427,
"learning_rate": 0.001,
"loss": 1.6581,
"step": 157200
},
{
"epoch": 50.84033613445378,
"grad_norm": 1.8328584432601929,
"learning_rate": 0.001,
"loss": 1.6749,
"step": 157300
},
{
"epoch": 50.87265675500969,
"grad_norm": 2.7844550609588623,
"learning_rate": 0.001,
"loss": 1.6836,
"step": 157400
},
{
"epoch": 50.90497737556561,
"grad_norm": 2.194227457046509,
"learning_rate": 0.001,
"loss": 1.6729,
"step": 157500
},
{
"epoch": 50.93729799612152,
"grad_norm": 1.7657898664474487,
"learning_rate": 0.001,
"loss": 1.6867,
"step": 157600
},
{
"epoch": 50.96961861667744,
"grad_norm": 2.189462423324585,
"learning_rate": 0.001,
"loss": 1.6741,
"step": 157700
},
{
"epoch": 51.00193923723336,
"grad_norm": 1.5788180828094482,
"learning_rate": 0.001,
"loss": 1.6987,
"step": 157800
},
{
"epoch": 51.03425985778927,
"grad_norm": 1.0550611019134521,
"learning_rate": 0.001,
"loss": 1.5532,
"step": 157900
},
{
"epoch": 51.06658047834519,
"grad_norm": 1.3609521389007568,
"learning_rate": 0.001,
"loss": 1.5526,
"step": 158000
},
{
"epoch": 51.0989010989011,
"grad_norm": 1.516286015510559,
"learning_rate": 0.001,
"loss": 1.5752,
"step": 158100
},
{
"epoch": 51.13122171945702,
"grad_norm": 1.8906481266021729,
"learning_rate": 0.001,
"loss": 1.589,
"step": 158200
},
{
"epoch": 51.16354234001293,
"grad_norm": 1.6561570167541504,
"learning_rate": 0.001,
"loss": 1.5793,
"step": 158300
},
{
"epoch": 51.195862960568846,
"grad_norm": 0.9634937644004822,
"learning_rate": 0.001,
"loss": 1.5692,
"step": 158400
},
{
"epoch": 51.22818358112476,
"grad_norm": 1.2189708948135376,
"learning_rate": 0.001,
"loss": 1.6067,
"step": 158500
},
{
"epoch": 51.260504201680675,
"grad_norm": 1.3045414686203003,
"learning_rate": 0.001,
"loss": 1.6058,
"step": 158600
},
{
"epoch": 51.29282482223659,
"grad_norm": 1.2715264558792114,
"learning_rate": 0.001,
"loss": 1.6221,
"step": 158700
},
{
"epoch": 51.325145442792504,
"grad_norm": 1.441200613975525,
"learning_rate": 0.001,
"loss": 1.6022,
"step": 158800
},
{
"epoch": 51.35746606334842,
"grad_norm": 1.6127737760543823,
"learning_rate": 0.001,
"loss": 1.6066,
"step": 158900
},
{
"epoch": 51.38978668390433,
"grad_norm": 1.4911373853683472,
"learning_rate": 0.001,
"loss": 1.6233,
"step": 159000
},
{
"epoch": 51.42210730446025,
"grad_norm": 1.4280664920806885,
"learning_rate": 0.001,
"loss": 1.6093,
"step": 159100
},
{
"epoch": 51.45442792501616,
"grad_norm": 1.3522392511367798,
"learning_rate": 0.001,
"loss": 1.6406,
"step": 159200
},
{
"epoch": 51.48674854557208,
"grad_norm": 1.2031400203704834,
"learning_rate": 0.001,
"loss": 1.6281,
"step": 159300
},
{
"epoch": 51.51906916612799,
"grad_norm": 1.4021563529968262,
"learning_rate": 0.001,
"loss": 1.6064,
"step": 159400
},
{
"epoch": 51.551389786683906,
"grad_norm": 2.3829009532928467,
"learning_rate": 0.001,
"loss": 1.6471,
"step": 159500
},
{
"epoch": 51.58371040723982,
"grad_norm": 1.9554762840270996,
"learning_rate": 0.001,
"loss": 1.6421,
"step": 159600
},
{
"epoch": 51.616031027795735,
"grad_norm": 1.420657992362976,
"learning_rate": 0.001,
"loss": 1.6307,
"step": 159700
},
{
"epoch": 51.64835164835165,
"grad_norm": 2.0701277256011963,
"learning_rate": 0.001,
"loss": 1.6466,
"step": 159800
},
{
"epoch": 51.680672268907564,
"grad_norm": 1.3747299909591675,
"learning_rate": 0.001,
"loss": 1.6531,
"step": 159900
},
{
"epoch": 51.71299288946348,
"grad_norm": 1.4589117765426636,
"learning_rate": 0.001,
"loss": 1.6541,
"step": 160000
},
{
"epoch": 51.74531351001939,
"grad_norm": 1.1903269290924072,
"learning_rate": 0.001,
"loss": 1.6564,
"step": 160100
},
{
"epoch": 51.77763413057531,
"grad_norm": 1.5743082761764526,
"learning_rate": 0.001,
"loss": 1.6554,
"step": 160200
},
{
"epoch": 51.80995475113122,
"grad_norm": 1.3097198009490967,
"learning_rate": 0.001,
"loss": 1.6583,
"step": 160300
},
{
"epoch": 51.84227537168714,
"grad_norm": 1.162923812866211,
"learning_rate": 0.001,
"loss": 1.654,
"step": 160400
},
{
"epoch": 51.87459599224305,
"grad_norm": 1.5432562828063965,
"learning_rate": 0.001,
"loss": 1.6637,
"step": 160500
},
{
"epoch": 51.906916612798966,
"grad_norm": 1.384651780128479,
"learning_rate": 0.001,
"loss": 1.6712,
"step": 160600
},
{
"epoch": 51.93923723335488,
"grad_norm": 1.2560335397720337,
"learning_rate": 0.001,
"loss": 1.6662,
"step": 160700
},
{
"epoch": 51.971557853910795,
"grad_norm": 1.276965618133545,
"learning_rate": 0.001,
"loss": 1.6721,
"step": 160800
},
{
"epoch": 52.00387847446671,
"grad_norm": 1.3887721300125122,
"learning_rate": 0.001,
"loss": 1.6759,
"step": 160900
},
{
"epoch": 52.036199095022624,
"grad_norm": 1.037302851676941,
"learning_rate": 0.001,
"loss": 1.5608,
"step": 161000
},
{
"epoch": 52.06851971557854,
"grad_norm": 1.4275438785552979,
"learning_rate": 0.001,
"loss": 1.5684,
"step": 161100
},
{
"epoch": 52.10084033613445,
"grad_norm": 1.3897730112075806,
"learning_rate": 0.001,
"loss": 1.5754,
"step": 161200
},
{
"epoch": 52.13316095669037,
"grad_norm": 1.288751482963562,
"learning_rate": 0.001,
"loss": 1.5666,
"step": 161300
},
{
"epoch": 52.16548157724628,
"grad_norm": 1.2900019884109497,
"learning_rate": 0.001,
"loss": 1.5694,
"step": 161400
},
{
"epoch": 52.1978021978022,
"grad_norm": 1.6506409645080566,
"learning_rate": 0.001,
"loss": 1.5767,
"step": 161500
},
{
"epoch": 52.23012281835811,
"grad_norm": 1.238895058631897,
"learning_rate": 0.001,
"loss": 1.5911,
"step": 161600
},
{
"epoch": 52.262443438914026,
"grad_norm": 1.1252110004425049,
"learning_rate": 0.001,
"loss": 1.5879,
"step": 161700
},
{
"epoch": 52.29476405946994,
"grad_norm": 1.4504222869873047,
"learning_rate": 0.001,
"loss": 1.592,
"step": 161800
},
{
"epoch": 52.327084680025855,
"grad_norm": 1.192259430885315,
"learning_rate": 0.001,
"loss": 1.586,
"step": 161900
},
{
"epoch": 52.35940530058177,
"grad_norm": 1.1508797407150269,
"learning_rate": 0.001,
"loss": 1.6021,
"step": 162000
},
{
"epoch": 52.391725921137684,
"grad_norm": 1.1547602415084839,
"learning_rate": 0.001,
"loss": 1.6134,
"step": 162100
},
{
"epoch": 52.4240465416936,
"grad_norm": 1.4814950227737427,
"learning_rate": 0.001,
"loss": 1.6135,
"step": 162200
},
{
"epoch": 52.456367162249514,
"grad_norm": 1.1219741106033325,
"learning_rate": 0.001,
"loss": 1.6145,
"step": 162300
},
{
"epoch": 52.48868778280543,
"grad_norm": 1.1321628093719482,
"learning_rate": 0.001,
"loss": 1.6198,
"step": 162400
},
{
"epoch": 52.52100840336134,
"grad_norm": 1.065873622894287,
"learning_rate": 0.001,
"loss": 1.6202,
"step": 162500
},
{
"epoch": 52.55332902391726,
"grad_norm": 1.4122074842453003,
"learning_rate": 0.001,
"loss": 1.6145,
"step": 162600
},
{
"epoch": 52.58564964447317,
"grad_norm": 1.0316053628921509,
"learning_rate": 0.001,
"loss": 1.6188,
"step": 162700
},
{
"epoch": 52.617970265029086,
"grad_norm": 1.2382274866104126,
"learning_rate": 0.001,
"loss": 1.6304,
"step": 162800
},
{
"epoch": 52.650290885585,
"grad_norm": 1.294029712677002,
"learning_rate": 0.001,
"loss": 1.6291,
"step": 162900
},
{
"epoch": 52.682611506140915,
"grad_norm": 2.0511300563812256,
"learning_rate": 0.001,
"loss": 1.6369,
"step": 163000
},
{
"epoch": 52.71493212669683,
"grad_norm": 1.2690668106079102,
"learning_rate": 0.001,
"loss": 1.6476,
"step": 163100
},
{
"epoch": 52.747252747252745,
"grad_norm": 1.251131534576416,
"learning_rate": 0.001,
"loss": 1.6257,
"step": 163200
},
{
"epoch": 52.77957336780866,
"grad_norm": 1.4768463373184204,
"learning_rate": 0.001,
"loss": 1.6246,
"step": 163300
},
{
"epoch": 52.811893988364574,
"grad_norm": 1.0410164594650269,
"learning_rate": 0.001,
"loss": 1.6217,
"step": 163400
},
{
"epoch": 52.84421460892049,
"grad_norm": 1.4737279415130615,
"learning_rate": 0.001,
"loss": 1.6426,
"step": 163500
},
{
"epoch": 52.8765352294764,
"grad_norm": 1.5894403457641602,
"learning_rate": 0.001,
"loss": 1.6764,
"step": 163600
},
{
"epoch": 52.90885585003232,
"grad_norm": 1.1352916955947876,
"learning_rate": 0.001,
"loss": 1.6566,
"step": 163700
},
{
"epoch": 52.94117647058823,
"grad_norm": 1.047954797744751,
"learning_rate": 0.001,
"loss": 1.6635,
"step": 163800
},
{
"epoch": 52.97349709114415,
"grad_norm": 1.1823827028274536,
"learning_rate": 0.001,
"loss": 1.6613,
"step": 163900
},
{
"epoch": 53.00581771170007,
"grad_norm": 1.3463735580444336,
"learning_rate": 0.001,
"loss": 1.6492,
"step": 164000
},
{
"epoch": 53.03813833225598,
"grad_norm": 1.5756522417068481,
"learning_rate": 0.001,
"loss": 1.5503,
"step": 164100
},
{
"epoch": 53.0704589528119,
"grad_norm": 1.174526572227478,
"learning_rate": 0.001,
"loss": 1.5504,
"step": 164200
},
{
"epoch": 53.10277957336781,
"grad_norm": 1.1576006412506104,
"learning_rate": 0.001,
"loss": 1.5194,
"step": 164300
},
{
"epoch": 53.135100193923726,
"grad_norm": 1.4668391942977905,
"learning_rate": 0.001,
"loss": 1.5494,
"step": 164400
},
{
"epoch": 53.16742081447964,
"grad_norm": 1.3181508779525757,
"learning_rate": 0.001,
"loss": 1.5556,
"step": 164500
},
{
"epoch": 53.199741435035556,
"grad_norm": 1.3680332899093628,
"learning_rate": 0.001,
"loss": 1.567,
"step": 164600
},
{
"epoch": 53.23206205559147,
"grad_norm": 1.7144526243209839,
"learning_rate": 0.001,
"loss": 1.5625,
"step": 164700
},
{
"epoch": 53.264382676147385,
"grad_norm": 1.8993232250213623,
"learning_rate": 0.001,
"loss": 1.5719,
"step": 164800
},
{
"epoch": 53.2967032967033,
"grad_norm": 1.384782075881958,
"learning_rate": 0.001,
"loss": 1.5853,
"step": 164900
},
{
"epoch": 53.329023917259214,
"grad_norm": 1.3512181043624878,
"learning_rate": 0.001,
"loss": 1.5777,
"step": 165000
},
{
"epoch": 53.36134453781513,
"grad_norm": 1.3165514469146729,
"learning_rate": 0.001,
"loss": 1.5714,
"step": 165100
},
{
"epoch": 53.39366515837104,
"grad_norm": 1.1895434856414795,
"learning_rate": 0.001,
"loss": 1.6086,
"step": 165200
},
{
"epoch": 53.42598577892696,
"grad_norm": 1.21664297580719,
"learning_rate": 0.001,
"loss": 1.5901,
"step": 165300
},
{
"epoch": 53.45830639948287,
"grad_norm": 1.3109116554260254,
"learning_rate": 0.001,
"loss": 1.6006,
"step": 165400
},
{
"epoch": 53.49062702003879,
"grad_norm": 1.4079968929290771,
"learning_rate": 0.001,
"loss": 1.5986,
"step": 165500
},
{
"epoch": 53.5229476405947,
"grad_norm": 1.7659581899642944,
"learning_rate": 0.001,
"loss": 1.6027,
"step": 165600
},
{
"epoch": 53.555268261150616,
"grad_norm": 1.2824375629425049,
"learning_rate": 0.001,
"loss": 1.6005,
"step": 165700
},
{
"epoch": 53.58758888170653,
"grad_norm": 1.8572561740875244,
"learning_rate": 0.001,
"loss": 1.5877,
"step": 165800
},
{
"epoch": 53.619909502262445,
"grad_norm": 1.0824077129364014,
"learning_rate": 0.001,
"loss": 1.621,
"step": 165900
},
{
"epoch": 53.65223012281836,
"grad_norm": 1.4150506258010864,
"learning_rate": 0.001,
"loss": 1.6302,
"step": 166000
},
{
"epoch": 53.684550743374274,
"grad_norm": 1.3486123085021973,
"learning_rate": 0.001,
"loss": 1.6216,
"step": 166100
},
{
"epoch": 53.71687136393019,
"grad_norm": 1.5042097568511963,
"learning_rate": 0.001,
"loss": 1.6177,
"step": 166200
},
{
"epoch": 53.7491919844861,
"grad_norm": 1.6768954992294312,
"learning_rate": 0.001,
"loss": 1.6379,
"step": 166300
},
{
"epoch": 53.78151260504202,
"grad_norm": 1.474862813949585,
"learning_rate": 0.001,
"loss": 1.6434,
"step": 166400
},
{
"epoch": 53.81383322559793,
"grad_norm": 1.4645066261291504,
"learning_rate": 0.001,
"loss": 1.6422,
"step": 166500
},
{
"epoch": 53.84615384615385,
"grad_norm": 1.3194541931152344,
"learning_rate": 0.001,
"loss": 1.6393,
"step": 166600
},
{
"epoch": 53.87847446670976,
"grad_norm": 1.619032859802246,
"learning_rate": 0.001,
"loss": 1.6559,
"step": 166700
},
{
"epoch": 53.910795087265676,
"grad_norm": 1.0101556777954102,
"learning_rate": 0.001,
"loss": 1.6184,
"step": 166800
},
{
"epoch": 53.94311570782159,
"grad_norm": 1.505044937133789,
"learning_rate": 0.001,
"loss": 1.6517,
"step": 166900
},
{
"epoch": 53.975436328377505,
"grad_norm": 1.1020125150680542,
"learning_rate": 0.001,
"loss": 1.6597,
"step": 167000
},
{
"epoch": 54.00775694893342,
"grad_norm": 1.0444972515106201,
"learning_rate": 0.001,
"loss": 1.6331,
"step": 167100
},
{
"epoch": 54.040077569489334,
"grad_norm": 1.1941118240356445,
"learning_rate": 0.001,
"loss": 1.5218,
"step": 167200
},
{
"epoch": 54.07239819004525,
"grad_norm": 1.714586615562439,
"learning_rate": 0.001,
"loss": 1.5285,
"step": 167300
},
{
"epoch": 54.10471881060116,
"grad_norm": 1.3004491329193115,
"learning_rate": 0.001,
"loss": 1.5373,
"step": 167400
},
{
"epoch": 54.13703943115708,
"grad_norm": 1.0127760171890259,
"learning_rate": 0.001,
"loss": 1.523,
"step": 167500
},
{
"epoch": 54.16936005171299,
"grad_norm": 1.298446536064148,
"learning_rate": 0.001,
"loss": 1.5464,
"step": 167600
},
{
"epoch": 54.20168067226891,
"grad_norm": 1.4641162157058716,
"learning_rate": 0.001,
"loss": 1.5587,
"step": 167700
},
{
"epoch": 54.23400129282482,
"grad_norm": 1.3359525203704834,
"learning_rate": 0.001,
"loss": 1.5341,
"step": 167800
},
{
"epoch": 54.266321913380736,
"grad_norm": 1.6306580305099487,
"learning_rate": 0.001,
"loss": 1.5609,
"step": 167900
},
{
"epoch": 54.29864253393665,
"grad_norm": 1.177018404006958,
"learning_rate": 0.001,
"loss": 1.5955,
"step": 168000
},
{
"epoch": 54.330963154492565,
"grad_norm": 1.1682850122451782,
"learning_rate": 0.001,
"loss": 1.5336,
"step": 168100
},
{
"epoch": 54.36328377504848,
"grad_norm": 1.604606032371521,
"learning_rate": 0.001,
"loss": 1.528,
"step": 168200
},
{
"epoch": 54.395604395604394,
"grad_norm": 1.5719093084335327,
"learning_rate": 0.001,
"loss": 1.5373,
"step": 168300
},
{
"epoch": 54.42792501616031,
"grad_norm": 1.1435718536376953,
"learning_rate": 0.001,
"loss": 1.5452,
"step": 168400
},
{
"epoch": 54.46024563671622,
"grad_norm": 1.2969406843185425,
"learning_rate": 0.001,
"loss": 1.5434,
"step": 168500
},
{
"epoch": 54.49256625727214,
"grad_norm": 1.4800134897232056,
"learning_rate": 0.001,
"loss": 1.5581,
"step": 168600
},
{
"epoch": 54.52488687782805,
"grad_norm": 1.5861899852752686,
"learning_rate": 0.001,
"loss": 1.5544,
"step": 168700
},
{
"epoch": 54.55720749838397,
"grad_norm": 1.827258825302124,
"learning_rate": 0.001,
"loss": 1.5638,
"step": 168800
},
{
"epoch": 54.58952811893988,
"grad_norm": 1.4785209894180298,
"learning_rate": 0.001,
"loss": 1.5648,
"step": 168900
},
{
"epoch": 54.621848739495796,
"grad_norm": 1.2366305589675903,
"learning_rate": 0.001,
"loss": 1.5773,
"step": 169000
},
{
"epoch": 54.65416936005171,
"grad_norm": 1.2721750736236572,
"learning_rate": 0.001,
"loss": 1.5731,
"step": 169100
},
{
"epoch": 54.686489980607625,
"grad_norm": 2.044588565826416,
"learning_rate": 0.001,
"loss": 1.583,
"step": 169200
},
{
"epoch": 54.71881060116354,
"grad_norm": 1.2052690982818604,
"learning_rate": 0.001,
"loss": 1.5782,
"step": 169300
},
{
"epoch": 54.751131221719454,
"grad_norm": 1.4870338439941406,
"learning_rate": 0.001,
"loss": 1.5743,
"step": 169400
},
{
"epoch": 54.78345184227537,
"grad_norm": 1.3881982564926147,
"learning_rate": 0.001,
"loss": 1.5693,
"step": 169500
},
{
"epoch": 54.81577246283128,
"grad_norm": 1.4930682182312012,
"learning_rate": 0.001,
"loss": 1.5715,
"step": 169600
},
{
"epoch": 54.8480930833872,
"grad_norm": 1.5199006795883179,
"learning_rate": 0.001,
"loss": 1.6108,
"step": 169700
},
{
"epoch": 54.88041370394311,
"grad_norm": 1.5097509622573853,
"learning_rate": 0.001,
"loss": 1.6136,
"step": 169800
},
{
"epoch": 54.91273432449903,
"grad_norm": 1.0916470289230347,
"learning_rate": 0.001,
"loss": 1.6138,
"step": 169900
},
{
"epoch": 54.94505494505494,
"grad_norm": 1.1135693788528442,
"learning_rate": 0.001,
"loss": 1.5827,
"step": 170000
},
{
"epoch": 54.977375565610856,
"grad_norm": 1.2118571996688843,
"learning_rate": 0.001,
"loss": 1.5861,
"step": 170100
},
{
"epoch": 55.00969618616678,
"grad_norm": 1.3085315227508545,
"learning_rate": 0.001,
"loss": 1.5589,
"step": 170200
},
{
"epoch": 55.04201680672269,
"grad_norm": 1.7613391876220703,
"learning_rate": 0.001,
"loss": 1.5128,
"step": 170300
},
{
"epoch": 55.07433742727861,
"grad_norm": 1.2778468132019043,
"learning_rate": 0.001,
"loss": 1.5314,
"step": 170400
},
{
"epoch": 55.10665804783452,
"grad_norm": 1.1575835943222046,
"learning_rate": 0.001,
"loss": 1.5342,
"step": 170500
},
{
"epoch": 55.138978668390436,
"grad_norm": 1.1837750673294067,
"learning_rate": 0.001,
"loss": 1.5435,
"step": 170600
},
{
"epoch": 55.17129928894635,
"grad_norm": 1.5179579257965088,
"learning_rate": 0.001,
"loss": 1.5354,
"step": 170700
},
{
"epoch": 55.203619909502265,
"grad_norm": 1.809478998184204,
"learning_rate": 0.001,
"loss": 1.5514,
"step": 170800
},
{
"epoch": 55.23594053005818,
"grad_norm": 1.2741111516952515,
"learning_rate": 0.001,
"loss": 1.5437,
"step": 170900
},
{
"epoch": 55.268261150614094,
"grad_norm": 1.5548871755599976,
"learning_rate": 0.001,
"loss": 1.5718,
"step": 171000
},
{
"epoch": 55.30058177117001,
"grad_norm": 1.6229969263076782,
"learning_rate": 0.001,
"loss": 1.5407,
"step": 171100
},
{
"epoch": 55.33290239172592,
"grad_norm": 1.4185291528701782,
"learning_rate": 0.001,
"loss": 1.5669,
"step": 171200
},
{
"epoch": 55.36522301228184,
"grad_norm": 1.3026727437973022,
"learning_rate": 0.001,
"loss": 1.5595,
"step": 171300
},
{
"epoch": 55.39754363283775,
"grad_norm": 1.5108344554901123,
"learning_rate": 0.001,
"loss": 1.5661,
"step": 171400
},
{
"epoch": 55.42986425339367,
"grad_norm": 1.2846623659133911,
"learning_rate": 0.001,
"loss": 1.5709,
"step": 171500
},
{
"epoch": 55.46218487394958,
"grad_norm": 1.586911916732788,
"learning_rate": 0.001,
"loss": 1.5867,
"step": 171600
},
{
"epoch": 55.494505494505496,
"grad_norm": 1.2302652597427368,
"learning_rate": 0.001,
"loss": 1.5926,
"step": 171700
},
{
"epoch": 55.52682611506141,
"grad_norm": 1.1358070373535156,
"learning_rate": 0.001,
"loss": 1.5928,
"step": 171800
},
{
"epoch": 55.559146735617325,
"grad_norm": 1.2191767692565918,
"learning_rate": 0.001,
"loss": 1.5836,
"step": 171900
},
{
"epoch": 55.59146735617324,
"grad_norm": 1.3371460437774658,
"learning_rate": 0.001,
"loss": 1.5994,
"step": 172000
},
{
"epoch": 55.623787976729155,
"grad_norm": 1.3227031230926514,
"learning_rate": 0.001,
"loss": 1.5941,
"step": 172100
},
{
"epoch": 55.65610859728507,
"grad_norm": 1.6309515237808228,
"learning_rate": 0.001,
"loss": 1.5912,
"step": 172200
},
{
"epoch": 55.688429217840984,
"grad_norm": 1.4074634313583374,
"learning_rate": 0.001,
"loss": 1.6138,
"step": 172300
},
{
"epoch": 55.7207498383969,
"grad_norm": 1.3193225860595703,
"learning_rate": 0.001,
"loss": 1.5881,
"step": 172400
},
{
"epoch": 55.75307045895281,
"grad_norm": 1.2992820739746094,
"learning_rate": 0.001,
"loss": 1.6143,
"step": 172500
},
{
"epoch": 55.78539107950873,
"grad_norm": 1.7191874980926514,
"learning_rate": 0.001,
"loss": 1.5931,
"step": 172600
},
{
"epoch": 55.81771170006464,
"grad_norm": 1.2515835762023926,
"learning_rate": 0.001,
"loss": 1.6379,
"step": 172700
},
{
"epoch": 55.85003232062056,
"grad_norm": 1.8979182243347168,
"learning_rate": 0.001,
"loss": 1.6126,
"step": 172800
},
{
"epoch": 55.88235294117647,
"grad_norm": 1.4136550426483154,
"learning_rate": 0.001,
"loss": 1.6263,
"step": 172900
},
{
"epoch": 55.914673561732386,
"grad_norm": 1.3156025409698486,
"learning_rate": 0.001,
"loss": 1.6165,
"step": 173000
},
{
"epoch": 55.9469941822883,
"grad_norm": 1.2711496353149414,
"learning_rate": 0.001,
"loss": 1.6249,
"step": 173100
},
{
"epoch": 55.979314802844215,
"grad_norm": 1.2456592321395874,
"learning_rate": 0.001,
"loss": 1.6369,
"step": 173200
},
{
"epoch": 56.01163542340013,
"grad_norm": 1.3482497930526733,
"learning_rate": 0.001,
"loss": 1.5842,
"step": 173300
},
{
"epoch": 56.043956043956044,
"grad_norm": 1.467483639717102,
"learning_rate": 0.001,
"loss": 1.5107,
"step": 173400
},
{
"epoch": 56.07627666451196,
"grad_norm": 1.3834830522537231,
"learning_rate": 0.001,
"loss": 1.5098,
"step": 173500
},
{
"epoch": 56.10859728506787,
"grad_norm": 1.6721076965332031,
"learning_rate": 0.001,
"loss": 1.5163,
"step": 173600
},
{
"epoch": 56.14091790562379,
"grad_norm": 1.8128893375396729,
"learning_rate": 0.001,
"loss": 1.5265,
"step": 173700
},
{
"epoch": 56.1732385261797,
"grad_norm": 1.1749082803726196,
"learning_rate": 0.001,
"loss": 1.5164,
"step": 173800
},
{
"epoch": 56.20555914673562,
"grad_norm": 1.9532384872436523,
"learning_rate": 0.001,
"loss": 1.5205,
"step": 173900
},
{
"epoch": 56.23787976729153,
"grad_norm": 1.53546142578125,
"learning_rate": 0.001,
"loss": 1.5736,
"step": 174000
},
{
"epoch": 56.270200387847446,
"grad_norm": 1.26724374294281,
"learning_rate": 0.001,
"loss": 1.5604,
"step": 174100
},
{
"epoch": 56.30252100840336,
"grad_norm": 1.3619731664657593,
"learning_rate": 0.001,
"loss": 1.5542,
"step": 174200
},
{
"epoch": 56.334841628959275,
"grad_norm": 1.371666669845581,
"learning_rate": 0.001,
"loss": 1.5291,
"step": 174300
},
{
"epoch": 56.36716224951519,
"grad_norm": 1.5174932479858398,
"learning_rate": 0.001,
"loss": 1.5356,
"step": 174400
},
{
"epoch": 56.399482870071104,
"grad_norm": 1.4292333126068115,
"learning_rate": 0.001,
"loss": 1.5622,
"step": 174500
},
{
"epoch": 56.43180349062702,
"grad_norm": 1.2732679843902588,
"learning_rate": 0.001,
"loss": 1.5667,
"step": 174600
},
{
"epoch": 56.46412411118293,
"grad_norm": 1.2781318426132202,
"learning_rate": 0.001,
"loss": 1.577,
"step": 174700
},
{
"epoch": 56.49644473173885,
"grad_norm": 1.1026822328567505,
"learning_rate": 0.001,
"loss": 1.5847,
"step": 174800
},
{
"epoch": 56.52876535229476,
"grad_norm": 1.706931471824646,
"learning_rate": 0.001,
"loss": 1.5817,
"step": 174900
},
{
"epoch": 56.56108597285068,
"grad_norm": 1.4177536964416504,
"learning_rate": 0.001,
"loss": 1.5695,
"step": 175000
},
{
"epoch": 56.59340659340659,
"grad_norm": 1.6883102655410767,
"learning_rate": 0.001,
"loss": 1.5738,
"step": 175100
},
{
"epoch": 56.625727213962506,
"grad_norm": 1.265228509902954,
"learning_rate": 0.001,
"loss": 1.5656,
"step": 175200
},
{
"epoch": 56.65804783451842,
"grad_norm": 1.3572072982788086,
"learning_rate": 0.001,
"loss": 1.5745,
"step": 175300
},
{
"epoch": 56.690368455074335,
"grad_norm": 1.1190592050552368,
"learning_rate": 0.001,
"loss": 1.5859,
"step": 175400
},
{
"epoch": 56.72268907563025,
"grad_norm": 2.085348129272461,
"learning_rate": 0.001,
"loss": 1.5849,
"step": 175500
},
{
"epoch": 56.755009696186164,
"grad_norm": 1.3309807777404785,
"learning_rate": 0.001,
"loss": 1.5928,
"step": 175600
},
{
"epoch": 56.78733031674208,
"grad_norm": 1.3919912576675415,
"learning_rate": 0.001,
"loss": 1.5954,
"step": 175700
},
{
"epoch": 56.81965093729799,
"grad_norm": 1.5010066032409668,
"learning_rate": 0.001,
"loss": 1.606,
"step": 175800
},
{
"epoch": 56.85197155785391,
"grad_norm": 1.517170786857605,
"learning_rate": 0.001,
"loss": 1.6159,
"step": 175900
},
{
"epoch": 56.88429217840982,
"grad_norm": 1.292891502380371,
"learning_rate": 0.001,
"loss": 1.6238,
"step": 176000
},
{
"epoch": 56.91661279896574,
"grad_norm": 1.3492282629013062,
"learning_rate": 0.001,
"loss": 1.6104,
"step": 176100
},
{
"epoch": 56.94893341952165,
"grad_norm": 1.4147794246673584,
"learning_rate": 0.001,
"loss": 1.6052,
"step": 176200
},
{
"epoch": 56.981254040077566,
"grad_norm": 1.486971378326416,
"learning_rate": 0.001,
"loss": 1.6105,
"step": 176300
},
{
"epoch": 57.01357466063349,
"grad_norm": 2.073136806488037,
"learning_rate": 0.001,
"loss": 1.5454,
"step": 176400
},
{
"epoch": 57.0458952811894,
"grad_norm": 1.979223370552063,
"learning_rate": 0.001,
"loss": 1.5025,
"step": 176500
},
{
"epoch": 57.07821590174532,
"grad_norm": 1.523560643196106,
"learning_rate": 0.001,
"loss": 1.5058,
"step": 176600
},
{
"epoch": 57.11053652230123,
"grad_norm": 1.6766302585601807,
"learning_rate": 0.001,
"loss": 1.4926,
"step": 176700
},
{
"epoch": 57.142857142857146,
"grad_norm": 1.478427767753601,
"learning_rate": 0.001,
"loss": 1.5075,
"step": 176800
},
{
"epoch": 57.17517776341306,
"grad_norm": 1.240274429321289,
"learning_rate": 0.001,
"loss": 1.5094,
"step": 176900
},
{
"epoch": 57.207498383968975,
"grad_norm": 1.593769907951355,
"learning_rate": 0.001,
"loss": 1.5314,
"step": 177000
},
{
"epoch": 57.23981900452489,
"grad_norm": 1.3210577964782715,
"learning_rate": 0.001,
"loss": 1.5274,
"step": 177100
},
{
"epoch": 57.272139625080804,
"grad_norm": 1.8289737701416016,
"learning_rate": 0.001,
"loss": 1.5262,
"step": 177200
},
{
"epoch": 57.30446024563672,
"grad_norm": 1.3610612154006958,
"learning_rate": 0.001,
"loss": 1.5401,
"step": 177300
},
{
"epoch": 57.33678086619263,
"grad_norm": 1.648084282875061,
"learning_rate": 0.001,
"loss": 1.5175,
"step": 177400
},
{
"epoch": 57.36910148674855,
"grad_norm": 2.194225549697876,
"learning_rate": 0.001,
"loss": 1.5409,
"step": 177500
},
{
"epoch": 57.40142210730446,
"grad_norm": 1.3858141899108887,
"learning_rate": 0.001,
"loss": 1.5468,
"step": 177600
},
{
"epoch": 57.43374272786038,
"grad_norm": 1.6206876039505005,
"learning_rate": 0.001,
"loss": 1.5523,
"step": 177700
},
{
"epoch": 57.46606334841629,
"grad_norm": 1.4267843961715698,
"learning_rate": 0.001,
"loss": 1.5486,
"step": 177800
},
{
"epoch": 57.498383968972206,
"grad_norm": 1.886731505393982,
"learning_rate": 0.001,
"loss": 1.5495,
"step": 177900
},
{
"epoch": 57.53070458952812,
"grad_norm": 1.8021095991134644,
"learning_rate": 0.001,
"loss": 1.548,
"step": 178000
},
{
"epoch": 57.563025210084035,
"grad_norm": 2.0562520027160645,
"learning_rate": 0.001,
"loss": 1.5729,
"step": 178100
},
{
"epoch": 57.59534583063995,
"grad_norm": 1.6516581773757935,
"learning_rate": 0.001,
"loss": 1.558,
"step": 178200
},
{
"epoch": 57.627666451195864,
"grad_norm": 1.4823943376541138,
"learning_rate": 0.001,
"loss": 1.5836,
"step": 178300
},
{
"epoch": 57.65998707175178,
"grad_norm": 1.4865487813949585,
"learning_rate": 0.001,
"loss": 1.5569,
"step": 178400
},
{
"epoch": 57.69230769230769,
"grad_norm": 1.2166651487350464,
"learning_rate": 0.001,
"loss": 1.5945,
"step": 178500
},
{
"epoch": 57.72462831286361,
"grad_norm": 1.7272731065750122,
"learning_rate": 0.001,
"loss": 1.5892,
"step": 178600
},
{
"epoch": 57.75694893341952,
"grad_norm": 1.4788447618484497,
"learning_rate": 0.001,
"loss": 1.5716,
"step": 178700
},
{
"epoch": 57.78926955397544,
"grad_norm": 1.2383166551589966,
"learning_rate": 0.001,
"loss": 1.5927,
"step": 178800
},
{
"epoch": 57.82159017453135,
"grad_norm": 1.577024221420288,
"learning_rate": 0.001,
"loss": 1.5945,
"step": 178900
},
{
"epoch": 57.853910795087266,
"grad_norm": 1.381188988685608,
"learning_rate": 0.001,
"loss": 1.5729,
"step": 179000
},
{
"epoch": 57.88623141564318,
"grad_norm": 1.726085901260376,
"learning_rate": 0.001,
"loss": 1.5904,
"step": 179100
},
{
"epoch": 57.918552036199095,
"grad_norm": 1.6200309991836548,
"learning_rate": 0.001,
"loss": 1.5998,
"step": 179200
},
{
"epoch": 57.95087265675501,
"grad_norm": 1.248194694519043,
"learning_rate": 0.001,
"loss": 1.5982,
"step": 179300
},
{
"epoch": 57.983193277310924,
"grad_norm": 1.5726380348205566,
"learning_rate": 0.001,
"loss": 1.6043,
"step": 179400
},
{
"epoch": 58.01551389786684,
"grad_norm": 1.6587324142456055,
"learning_rate": 0.001,
"loss": 1.5207,
"step": 179500
},
{
"epoch": 58.04783451842275,
"grad_norm": 1.759029746055603,
"learning_rate": 0.001,
"loss": 1.4772,
"step": 179600
},
{
"epoch": 58.08015513897867,
"grad_norm": 1.9677127599716187,
"learning_rate": 0.001,
"loss": 1.4735,
"step": 179700
},
{
"epoch": 58.11247575953458,
"grad_norm": 1.4482223987579346,
"learning_rate": 0.001,
"loss": 1.4874,
"step": 179800
},
{
"epoch": 58.1447963800905,
"grad_norm": 2.5144407749176025,
"learning_rate": 0.001,
"loss": 1.5055,
"step": 179900
},
{
"epoch": 58.17711700064641,
"grad_norm": 1.5867522954940796,
"learning_rate": 0.001,
"loss": 1.517,
"step": 180000
},
{
"epoch": 58.209437621202326,
"grad_norm": 1.4128469228744507,
"learning_rate": 0.001,
"loss": 1.5021,
"step": 180100
},
{
"epoch": 58.24175824175824,
"grad_norm": 1.7925270795822144,
"learning_rate": 0.001,
"loss": 1.5166,
"step": 180200
},
{
"epoch": 58.274078862314155,
"grad_norm": 1.7557506561279297,
"learning_rate": 0.001,
"loss": 1.5265,
"step": 180300
},
{
"epoch": 58.30639948287007,
"grad_norm": 1.9686635732650757,
"learning_rate": 0.001,
"loss": 1.5217,
"step": 180400
},
{
"epoch": 58.338720103425985,
"grad_norm": 1.915852427482605,
"learning_rate": 0.001,
"loss": 1.5237,
"step": 180500
},
{
"epoch": 58.3710407239819,
"grad_norm": 1.4966778755187988,
"learning_rate": 0.001,
"loss": 1.529,
"step": 180600
},
{
"epoch": 58.403361344537814,
"grad_norm": 1.871946930885315,
"learning_rate": 0.001,
"loss": 1.5394,
"step": 180700
},
{
"epoch": 58.43568196509373,
"grad_norm": 1.6767258644104004,
"learning_rate": 0.001,
"loss": 1.5508,
"step": 180800
},
{
"epoch": 58.46800258564964,
"grad_norm": 2.0740795135498047,
"learning_rate": 0.001,
"loss": 1.5408,
"step": 180900
},
{
"epoch": 58.50032320620556,
"grad_norm": 1.4852629899978638,
"learning_rate": 0.001,
"loss": 1.5535,
"step": 181000
},
{
"epoch": 58.53264382676147,
"grad_norm": 1.4267995357513428,
"learning_rate": 0.001,
"loss": 1.5487,
"step": 181100
},
{
"epoch": 58.56496444731739,
"grad_norm": 1.6049513816833496,
"learning_rate": 0.001,
"loss": 1.5493,
"step": 181200
},
{
"epoch": 58.5972850678733,
"grad_norm": 1.7209135293960571,
"learning_rate": 0.001,
"loss": 1.5486,
"step": 181300
},
{
"epoch": 58.629605688429216,
"grad_norm": 1.9109245538711548,
"learning_rate": 0.001,
"loss": 1.5505,
"step": 181400
},
{
"epoch": 58.66192630898513,
"grad_norm": 1.649224877357483,
"learning_rate": 0.001,
"loss": 1.5649,
"step": 181500
},
{
"epoch": 58.694246929541045,
"grad_norm": 1.5617544651031494,
"learning_rate": 0.001,
"loss": 1.5855,
"step": 181600
},
{
"epoch": 58.72656755009696,
"grad_norm": 1.6211555004119873,
"learning_rate": 0.001,
"loss": 1.5817,
"step": 181700
},
{
"epoch": 58.758888170652874,
"grad_norm": 1.4985055923461914,
"learning_rate": 0.001,
"loss": 1.5625,
"step": 181800
},
{
"epoch": 58.79120879120879,
"grad_norm": 1.8693698644638062,
"learning_rate": 0.001,
"loss": 1.5846,
"step": 181900
},
{
"epoch": 58.8235294117647,
"grad_norm": 1.8782329559326172,
"learning_rate": 0.001,
"loss": 1.5818,
"step": 182000
},
{
"epoch": 58.85585003232062,
"grad_norm": 1.9878121614456177,
"learning_rate": 0.001,
"loss": 1.5731,
"step": 182100
},
{
"epoch": 58.88817065287653,
"grad_norm": 1.6412078142166138,
"learning_rate": 0.001,
"loss": 1.5694,
"step": 182200
},
{
"epoch": 58.92049127343245,
"grad_norm": 1.4442908763885498,
"learning_rate": 0.001,
"loss": 1.5963,
"step": 182300
},
{
"epoch": 58.95281189398836,
"grad_norm": 2.419037342071533,
"learning_rate": 0.001,
"loss": 1.5871,
"step": 182400
},
{
"epoch": 58.985132514544276,
"grad_norm": 1.5002714395523071,
"learning_rate": 0.001,
"loss": 1.586,
"step": 182500
},
{
"epoch": 59.0174531351002,
"grad_norm": 1.559140682220459,
"learning_rate": 0.001,
"loss": 1.5192,
"step": 182600
},
{
"epoch": 59.04977375565611,
"grad_norm": 1.5507798194885254,
"learning_rate": 0.001,
"loss": 1.4667,
"step": 182700
},
{
"epoch": 59.08209437621203,
"grad_norm": 1.871140480041504,
"learning_rate": 0.001,
"loss": 1.4922,
"step": 182800
},
{
"epoch": 59.11441499676794,
"grad_norm": 2.057178258895874,
"learning_rate": 0.001,
"loss": 1.4773,
"step": 182900
},
{
"epoch": 59.146735617323856,
"grad_norm": 1.649742841720581,
"learning_rate": 0.001,
"loss": 1.4915,
"step": 183000
},
{
"epoch": 59.17905623787977,
"grad_norm": 1.779327392578125,
"learning_rate": 0.001,
"loss": 1.5015,
"step": 183100
},
{
"epoch": 59.211376858435685,
"grad_norm": 2.852686882019043,
"learning_rate": 0.001,
"loss": 1.5007,
"step": 183200
},
{
"epoch": 59.2436974789916,
"grad_norm": 1.2723041772842407,
"learning_rate": 0.001,
"loss": 1.5016,
"step": 183300
},
{
"epoch": 59.276018099547514,
"grad_norm": 1.511348843574524,
"learning_rate": 0.001,
"loss": 1.5084,
"step": 183400
},
{
"epoch": 59.30833872010343,
"grad_norm": 1.63169264793396,
"learning_rate": 0.001,
"loss": 1.5054,
"step": 183500
},
{
"epoch": 59.34065934065934,
"grad_norm": 1.4756476879119873,
"learning_rate": 0.001,
"loss": 1.5189,
"step": 183600
},
{
"epoch": 59.37297996121526,
"grad_norm": 1.4257620573043823,
"learning_rate": 0.001,
"loss": 1.5226,
"step": 183700
},
{
"epoch": 59.40530058177117,
"grad_norm": 1.4187026023864746,
"learning_rate": 0.001,
"loss": 1.5252,
"step": 183800
},
{
"epoch": 59.43762120232709,
"grad_norm": 2.708465576171875,
"learning_rate": 0.001,
"loss": 1.5502,
"step": 183900
},
{
"epoch": 59.469941822883,
"grad_norm": 2.4164412021636963,
"learning_rate": 0.001,
"loss": 1.5345,
"step": 184000
},
{
"epoch": 59.502262443438916,
"grad_norm": 1.3273594379425049,
"learning_rate": 0.001,
"loss": 1.5464,
"step": 184100
},
{
"epoch": 59.53458306399483,
"grad_norm": 2.1284921169281006,
"learning_rate": 0.001,
"loss": 1.5395,
"step": 184200
},
{
"epoch": 59.566903684550745,
"grad_norm": 1.217075228691101,
"learning_rate": 0.001,
"loss": 1.5427,
"step": 184300
},
{
"epoch": 59.59922430510666,
"grad_norm": 1.549274206161499,
"learning_rate": 0.001,
"loss": 1.5516,
"step": 184400
},
{
"epoch": 59.631544925662574,
"grad_norm": 2.0977909564971924,
"learning_rate": 0.001,
"loss": 1.5503,
"step": 184500
},
{
"epoch": 59.66386554621849,
"grad_norm": 1.6122702360153198,
"learning_rate": 0.001,
"loss": 1.5425,
"step": 184600
},
{
"epoch": 59.6961861667744,
"grad_norm": 2.3751943111419678,
"learning_rate": 0.001,
"loss": 1.5548,
"step": 184700
},
{
"epoch": 59.72850678733032,
"grad_norm": 1.7273507118225098,
"learning_rate": 0.001,
"loss": 1.5665,
"step": 184800
},
{
"epoch": 59.76082740788623,
"grad_norm": 1.4305226802825928,
"learning_rate": 0.001,
"loss": 1.5712,
"step": 184900
},
{
"epoch": 59.79314802844215,
"grad_norm": 1.3184654712677002,
"learning_rate": 0.001,
"loss": 1.5644,
"step": 185000
},
{
"epoch": 59.82546864899806,
"grad_norm": 1.6823676824569702,
"learning_rate": 0.001,
"loss": 1.5641,
"step": 185100
},
{
"epoch": 59.857789269553976,
"grad_norm": 1.5022037029266357,
"learning_rate": 0.001,
"loss": 1.554,
"step": 185200
},
{
"epoch": 59.89010989010989,
"grad_norm": 2.0881659984588623,
"learning_rate": 0.001,
"loss": 1.5629,
"step": 185300
},
{
"epoch": 59.922430510665805,
"grad_norm": 1.4326902627944946,
"learning_rate": 0.001,
"loss": 1.566,
"step": 185400
},
{
"epoch": 59.95475113122172,
"grad_norm": 1.388048768043518,
"learning_rate": 0.001,
"loss": 1.5739,
"step": 185500
},
{
"epoch": 59.987071751777634,
"grad_norm": 1.365792155265808,
"learning_rate": 0.001,
"loss": 1.5773,
"step": 185600
},
{
"epoch": 60.01939237233355,
"grad_norm": 1.231660008430481,
"learning_rate": 0.001,
"loss": 1.5027,
"step": 185700
},
{
"epoch": 60.05171299288946,
"grad_norm": 1.6712000370025635,
"learning_rate": 0.001,
"loss": 1.4663,
"step": 185800
},
{
"epoch": 60.08403361344538,
"grad_norm": 1.950838327407837,
"learning_rate": 0.001,
"loss": 1.4832,
"step": 185900
},
{
"epoch": 60.11635423400129,
"grad_norm": 1.4073816537857056,
"learning_rate": 0.001,
"loss": 1.4664,
"step": 186000
},
{
"epoch": 60.14867485455721,
"grad_norm": 1.3018288612365723,
"learning_rate": 0.001,
"loss": 1.4688,
"step": 186100
},
{
"epoch": 60.18099547511312,
"grad_norm": 1.6453629732131958,
"learning_rate": 0.001,
"loss": 1.4869,
"step": 186200
},
{
"epoch": 60.213316095669036,
"grad_norm": 1.3400012254714966,
"learning_rate": 0.001,
"loss": 1.4764,
"step": 186300
},
{
"epoch": 60.24563671622495,
"grad_norm": 1.837446689605713,
"learning_rate": 0.001,
"loss": 1.4799,
"step": 186400
},
{
"epoch": 60.277957336780865,
"grad_norm": 2.0825252532958984,
"learning_rate": 0.001,
"loss": 1.5098,
"step": 186500
},
{
"epoch": 60.31027795733678,
"grad_norm": 1.915043592453003,
"learning_rate": 0.001,
"loss": 1.5119,
"step": 186600
},
{
"epoch": 60.342598577892694,
"grad_norm": 2.5546317100524902,
"learning_rate": 0.001,
"loss": 1.507,
"step": 186700
},
{
"epoch": 60.37491919844861,
"grad_norm": 1.6669049263000488,
"learning_rate": 0.001,
"loss": 1.4984,
"step": 186800
},
{
"epoch": 60.40723981900452,
"grad_norm": 1.3404308557510376,
"learning_rate": 0.001,
"loss": 1.5175,
"step": 186900
},
{
"epoch": 60.43956043956044,
"grad_norm": 1.329489827156067,
"learning_rate": 0.001,
"loss": 1.5093,
"step": 187000
},
{
"epoch": 60.47188106011635,
"grad_norm": 1.6057754755020142,
"learning_rate": 0.001,
"loss": 1.5041,
"step": 187100
},
{
"epoch": 60.50420168067227,
"grad_norm": 1.8171005249023438,
"learning_rate": 0.001,
"loss": 1.5264,
"step": 187200
},
{
"epoch": 60.53652230122818,
"grad_norm": 1.7487467527389526,
"learning_rate": 0.001,
"loss": 1.5163,
"step": 187300
},
{
"epoch": 60.568842921784096,
"grad_norm": 1.189156413078308,
"learning_rate": 0.001,
"loss": 1.5579,
"step": 187400
},
{
"epoch": 60.60116354234001,
"grad_norm": 0.9593215584754944,
"learning_rate": 0.001,
"loss": 1.5416,
"step": 187500
},
{
"epoch": 60.633484162895925,
"grad_norm": 1.8685455322265625,
"learning_rate": 0.001,
"loss": 1.5164,
"step": 187600
},
{
"epoch": 60.66580478345184,
"grad_norm": 1.5182610750198364,
"learning_rate": 0.001,
"loss": 1.553,
"step": 187700
},
{
"epoch": 60.698125404007754,
"grad_norm": 2.1702611446380615,
"learning_rate": 0.001,
"loss": 1.5432,
"step": 187800
},
{
"epoch": 60.73044602456367,
"grad_norm": 1.8959705829620361,
"learning_rate": 0.001,
"loss": 1.543,
"step": 187900
},
{
"epoch": 60.762766645119584,
"grad_norm": 1.1643645763397217,
"learning_rate": 0.001,
"loss": 1.551,
"step": 188000
},
{
"epoch": 60.7950872656755,
"grad_norm": 1.3722823858261108,
"learning_rate": 0.001,
"loss": 1.5532,
"step": 188100
},
{
"epoch": 60.82740788623141,
"grad_norm": 1.4278934001922607,
"learning_rate": 0.001,
"loss": 1.5529,
"step": 188200
},
{
"epoch": 60.85972850678733,
"grad_norm": 1.2416459321975708,
"learning_rate": 0.001,
"loss": 1.5702,
"step": 188300
},
{
"epoch": 60.89204912734324,
"grad_norm": 1.896902322769165,
"learning_rate": 0.001,
"loss": 1.568,
"step": 188400
},
{
"epoch": 60.924369747899156,
"grad_norm": 2.145780086517334,
"learning_rate": 0.001,
"loss": 1.5714,
"step": 188500
},
{
"epoch": 60.95669036845507,
"grad_norm": 1.3877936601638794,
"learning_rate": 0.001,
"loss": 1.5771,
"step": 188600
},
{
"epoch": 60.98901098901099,
"grad_norm": 1.6279957294464111,
"learning_rate": 0.001,
"loss": 1.562,
"step": 188700
},
{
"epoch": 61.02133160956691,
"grad_norm": 1.2547106742858887,
"learning_rate": 0.001,
"loss": 1.4953,
"step": 188800
},
{
"epoch": 61.05365223012282,
"grad_norm": 1.4344184398651123,
"learning_rate": 0.001,
"loss": 1.4481,
"step": 188900
},
{
"epoch": 61.085972850678736,
"grad_norm": 1.5340839624404907,
"learning_rate": 0.001,
"loss": 1.4535,
"step": 189000
},
{
"epoch": 61.11829347123465,
"grad_norm": 1.5223768949508667,
"learning_rate": 0.001,
"loss": 1.4769,
"step": 189100
},
{
"epoch": 61.150614091790565,
"grad_norm": 1.7879722118377686,
"learning_rate": 0.001,
"loss": 1.4481,
"step": 189200
},
{
"epoch": 61.18293471234648,
"grad_norm": 1.2013511657714844,
"learning_rate": 0.001,
"loss": 1.4667,
"step": 189300
},
{
"epoch": 61.215255332902395,
"grad_norm": 1.1109360456466675,
"learning_rate": 0.001,
"loss": 1.4677,
"step": 189400
},
{
"epoch": 61.24757595345831,
"grad_norm": 1.8947685956954956,
"learning_rate": 0.001,
"loss": 1.4805,
"step": 189500
},
{
"epoch": 61.279896574014224,
"grad_norm": 1.7237753868103027,
"learning_rate": 0.001,
"loss": 1.4705,
"step": 189600
},
{
"epoch": 61.31221719457014,
"grad_norm": 1.3296040296554565,
"learning_rate": 0.001,
"loss": 1.515,
"step": 189700
},
{
"epoch": 61.34453781512605,
"grad_norm": 1.1296497583389282,
"learning_rate": 0.001,
"loss": 1.4985,
"step": 189800
},
{
"epoch": 61.37685843568197,
"grad_norm": 1.102357268333435,
"learning_rate": 0.001,
"loss": 1.4837,
"step": 189900
},
{
"epoch": 61.40917905623788,
"grad_norm": 2.1110856533050537,
"learning_rate": 0.001,
"loss": 1.5152,
"step": 190000
},
{
"epoch": 61.441499676793796,
"grad_norm": 1.3354947566986084,
"learning_rate": 0.001,
"loss": 1.4995,
"step": 190100
},
{
"epoch": 61.47382029734971,
"grad_norm": 1.379221796989441,
"learning_rate": 0.001,
"loss": 1.5061,
"step": 190200
},
{
"epoch": 61.506140917905626,
"grad_norm": 1.2813968658447266,
"learning_rate": 0.001,
"loss": 1.5274,
"step": 190300
},
{
"epoch": 61.53846153846154,
"grad_norm": 1.1499813795089722,
"learning_rate": 0.001,
"loss": 1.5124,
"step": 190400
},
{
"epoch": 61.570782159017455,
"grad_norm": 1.5203831195831299,
"learning_rate": 0.001,
"loss": 1.5279,
"step": 190500
},
{
"epoch": 61.60310277957337,
"grad_norm": 1.4393309354782104,
"learning_rate": 0.001,
"loss": 1.5155,
"step": 190600
},
{
"epoch": 61.635423400129284,
"grad_norm": 1.2175180912017822,
"learning_rate": 0.001,
"loss": 1.5136,
"step": 190700
},
{
"epoch": 61.6677440206852,
"grad_norm": 1.224487543106079,
"learning_rate": 0.001,
"loss": 1.5324,
"step": 190800
},
{
"epoch": 61.70006464124111,
"grad_norm": 1.1817762851715088,
"learning_rate": 0.001,
"loss": 1.5256,
"step": 190900
},
{
"epoch": 61.73238526179703,
"grad_norm": 1.865926742553711,
"learning_rate": 0.001,
"loss": 1.5273,
"step": 191000
},
{
"epoch": 61.76470588235294,
"grad_norm": 1.322381854057312,
"learning_rate": 0.001,
"loss": 1.5356,
"step": 191100
},
{
"epoch": 61.79702650290886,
"grad_norm": 1.1925381422042847,
"learning_rate": 0.001,
"loss": 1.5348,
"step": 191200
},
{
"epoch": 61.82934712346477,
"grad_norm": 1.3930214643478394,
"learning_rate": 0.001,
"loss": 1.5581,
"step": 191300
},
{
"epoch": 61.861667744020686,
"grad_norm": 1.4687880277633667,
"learning_rate": 0.001,
"loss": 1.5541,
"step": 191400
},
{
"epoch": 61.8939883645766,
"grad_norm": 1.4731204509735107,
"learning_rate": 0.001,
"loss": 1.5672,
"step": 191500
},
{
"epoch": 61.926308985132515,
"grad_norm": 1.1631594896316528,
"learning_rate": 0.001,
"loss": 1.5517,
"step": 191600
},
{
"epoch": 61.95862960568843,
"grad_norm": 1.4175986051559448,
"learning_rate": 0.001,
"loss": 1.5528,
"step": 191700
},
{
"epoch": 61.990950226244344,
"grad_norm": 1.2329492568969727,
"learning_rate": 0.001,
"loss": 1.5772,
"step": 191800
},
{
"epoch": 62.02327084680026,
"grad_norm": 1.1341323852539062,
"learning_rate": 0.001,
"loss": 1.499,
"step": 191900
},
{
"epoch": 62.05559146735617,
"grad_norm": 1.2522056102752686,
"learning_rate": 0.001,
"loss": 1.4384,
"step": 192000
},
{
"epoch": 62.08791208791209,
"grad_norm": 1.2916074991226196,
"learning_rate": 0.001,
"loss": 1.4542,
"step": 192100
},
{
"epoch": 62.120232708468,
"grad_norm": 1.459539771080017,
"learning_rate": 0.001,
"loss": 1.4547,
"step": 192200
},
{
"epoch": 62.15255332902392,
"grad_norm": 1.1526257991790771,
"learning_rate": 0.001,
"loss": 1.4492,
"step": 192300
},
{
"epoch": 62.18487394957983,
"grad_norm": 1.9539614915847778,
"learning_rate": 0.001,
"loss": 1.4636,
"step": 192400
},
{
"epoch": 62.217194570135746,
"grad_norm": 1.6022253036499023,
"learning_rate": 0.001,
"loss": 1.466,
"step": 192500
},
{
"epoch": 62.24951519069166,
"grad_norm": 1.05202317237854,
"learning_rate": 0.001,
"loss": 1.4728,
"step": 192600
},
{
"epoch": 62.281835811247575,
"grad_norm": 1.318872094154358,
"learning_rate": 0.001,
"loss": 1.4751,
"step": 192700
},
{
"epoch": 62.31415643180349,
"grad_norm": 1.1124187707901,
"learning_rate": 0.001,
"loss": 1.4825,
"step": 192800
},
{
"epoch": 62.346477052359404,
"grad_norm": 1.4929132461547852,
"learning_rate": 0.001,
"loss": 1.4673,
"step": 192900
},
{
"epoch": 62.37879767291532,
"grad_norm": 1.4624375104904175,
"learning_rate": 0.001,
"loss": 1.5004,
"step": 193000
},
{
"epoch": 62.41111829347123,
"grad_norm": 1.5032709836959839,
"learning_rate": 0.001,
"loss": 1.5019,
"step": 193100
},
{
"epoch": 62.44343891402715,
"grad_norm": 1.1061334609985352,
"learning_rate": 0.001,
"loss": 1.4808,
"step": 193200
},
{
"epoch": 62.47575953458306,
"grad_norm": 1.3267040252685547,
"learning_rate": 0.001,
"loss": 1.4962,
"step": 193300
},
{
"epoch": 62.50808015513898,
"grad_norm": 1.327292799949646,
"learning_rate": 0.001,
"loss": 1.4837,
"step": 193400
},
{
"epoch": 62.54040077569489,
"grad_norm": 1.1017119884490967,
"learning_rate": 0.001,
"loss": 1.5183,
"step": 193500
},
{
"epoch": 62.572721396250806,
"grad_norm": 1.0475913286209106,
"learning_rate": 0.001,
"loss": 1.5137,
"step": 193600
},
{
"epoch": 62.60504201680672,
"grad_norm": 1.450590968132019,
"learning_rate": 0.001,
"loss": 1.5141,
"step": 193700
},
{
"epoch": 62.637362637362635,
"grad_norm": 1.103518009185791,
"learning_rate": 0.001,
"loss": 1.5265,
"step": 193800
},
{
"epoch": 62.66968325791855,
"grad_norm": 1.4655108451843262,
"learning_rate": 0.001,
"loss": 1.5153,
"step": 193900
},
{
"epoch": 62.702003878474464,
"grad_norm": 1.503920316696167,
"learning_rate": 0.001,
"loss": 1.5253,
"step": 194000
},
{
"epoch": 62.73432449903038,
"grad_norm": 1.3186782598495483,
"learning_rate": 0.001,
"loss": 1.5283,
"step": 194100
},
{
"epoch": 62.76664511958629,
"grad_norm": 1.0896756649017334,
"learning_rate": 0.001,
"loss": 1.5223,
"step": 194200
},
{
"epoch": 62.79896574014221,
"grad_norm": 1.3912789821624756,
"learning_rate": 0.001,
"loss": 1.5605,
"step": 194300
},
{
"epoch": 62.83128636069812,
"grad_norm": 1.2818633317947388,
"learning_rate": 0.001,
"loss": 1.5416,
"step": 194400
},
{
"epoch": 62.86360698125404,
"grad_norm": 1.3630318641662598,
"learning_rate": 0.001,
"loss": 1.5222,
"step": 194500
},
{
"epoch": 62.89592760180995,
"grad_norm": 1.393178939819336,
"learning_rate": 0.001,
"loss": 1.5248,
"step": 194600
},
{
"epoch": 62.928248222365866,
"grad_norm": 1.1949397325515747,
"learning_rate": 0.001,
"loss": 1.5621,
"step": 194700
},
{
"epoch": 62.96056884292178,
"grad_norm": 1.152626395225525,
"learning_rate": 0.001,
"loss": 1.5515,
"step": 194800
},
{
"epoch": 62.992889463477695,
"grad_norm": 1.4040319919586182,
"learning_rate": 0.001,
"loss": 1.5468,
"step": 194900
},
{
"epoch": 63.02521008403362,
"grad_norm": 1.6285364627838135,
"learning_rate": 0.001,
"loss": 1.4447,
"step": 195000
},
{
"epoch": 63.05753070458953,
"grad_norm": 1.2859936952590942,
"learning_rate": 0.001,
"loss": 1.4452,
"step": 195100
},
{
"epoch": 63.089851325145446,
"grad_norm": 1.6732817888259888,
"learning_rate": 0.001,
"loss": 1.4485,
"step": 195200
},
{
"epoch": 63.12217194570136,
"grad_norm": 1.6121927499771118,
"learning_rate": 0.001,
"loss": 1.4583,
"step": 195300
},
{
"epoch": 63.154492566257275,
"grad_norm": 1.7589558362960815,
"learning_rate": 0.001,
"loss": 1.4371,
"step": 195400
},
{
"epoch": 63.18681318681319,
"grad_norm": 1.545434832572937,
"learning_rate": 0.001,
"loss": 1.452,
"step": 195500
},
{
"epoch": 63.219133807369104,
"grad_norm": 1.0917245149612427,
"learning_rate": 0.001,
"loss": 1.455,
"step": 195600
},
{
"epoch": 63.25145442792502,
"grad_norm": 1.555050015449524,
"learning_rate": 0.001,
"loss": 1.4625,
"step": 195700
},
{
"epoch": 63.28377504848093,
"grad_norm": 1.6527012586593628,
"learning_rate": 0.001,
"loss": 1.4467,
"step": 195800
},
{
"epoch": 63.31609566903685,
"grad_norm": 1.6350040435791016,
"learning_rate": 0.001,
"loss": 1.4798,
"step": 195900
},
{
"epoch": 63.34841628959276,
"grad_norm": 1.0728449821472168,
"learning_rate": 0.001,
"loss": 1.4849,
"step": 196000
},
{
"epoch": 63.38073691014868,
"grad_norm": 1.18773353099823,
"learning_rate": 0.001,
"loss": 1.4705,
"step": 196100
},
{
"epoch": 63.41305753070459,
"grad_norm": 1.3900010585784912,
"learning_rate": 0.001,
"loss": 1.4979,
"step": 196200
},
{
"epoch": 63.445378151260506,
"grad_norm": 1.216658592224121,
"learning_rate": 0.001,
"loss": 1.4812,
"step": 196300
},
{
"epoch": 63.47769877181642,
"grad_norm": 1.284078598022461,
"learning_rate": 0.001,
"loss": 1.4921,
"step": 196400
},
{
"epoch": 63.510019392372335,
"grad_norm": 1.2323185205459595,
"learning_rate": 0.001,
"loss": 1.4968,
"step": 196500
},
{
"epoch": 63.54234001292825,
"grad_norm": 1.0883482694625854,
"learning_rate": 0.001,
"loss": 1.4875,
"step": 196600
},
{
"epoch": 63.574660633484164,
"grad_norm": 1.1361480951309204,
"learning_rate": 0.001,
"loss": 1.4957,
"step": 196700
},
{
"epoch": 63.60698125404008,
"grad_norm": 1.162347435951233,
"learning_rate": 0.001,
"loss": 1.4854,
"step": 196800
},
{
"epoch": 63.63930187459599,
"grad_norm": 1.204131841659546,
"learning_rate": 0.001,
"loss": 1.5198,
"step": 196900
},
{
"epoch": 63.67162249515191,
"grad_norm": 1.4061599969863892,
"learning_rate": 0.001,
"loss": 1.5245,
"step": 197000
},
{
"epoch": 63.70394311570782,
"grad_norm": 1.4909995794296265,
"learning_rate": 0.001,
"loss": 1.5155,
"step": 197100
},
{
"epoch": 63.73626373626374,
"grad_norm": 1.213577389717102,
"learning_rate": 0.001,
"loss": 1.5164,
"step": 197200
},
{
"epoch": 63.76858435681965,
"grad_norm": 1.3162715435028076,
"learning_rate": 0.001,
"loss": 1.5039,
"step": 197300
},
{
"epoch": 63.800904977375566,
"grad_norm": 1.358174443244934,
"learning_rate": 0.001,
"loss": 1.5151,
"step": 197400
},
{
"epoch": 63.83322559793148,
"grad_norm": 1.3790055513381958,
"learning_rate": 0.001,
"loss": 1.5168,
"step": 197500
},
{
"epoch": 63.865546218487395,
"grad_norm": 1.2166327238082886,
"learning_rate": 0.001,
"loss": 1.551,
"step": 197600
},
{
"epoch": 63.89786683904331,
"grad_norm": 1.145918846130371,
"learning_rate": 0.001,
"loss": 1.5283,
"step": 197700
},
{
"epoch": 63.930187459599225,
"grad_norm": 1.4355617761611938,
"learning_rate": 0.001,
"loss": 1.5269,
"step": 197800
},
{
"epoch": 63.96250808015514,
"grad_norm": 1.4867405891418457,
"learning_rate": 0.001,
"loss": 1.5495,
"step": 197900
},
{
"epoch": 63.994828700711054,
"grad_norm": 1.1840007305145264,
"learning_rate": 0.001,
"loss": 1.5293,
"step": 198000
},
{
"epoch": 64.02714932126698,
"grad_norm": 1.58708655834198,
"learning_rate": 0.001,
"loss": 1.4324,
"step": 198100
},
{
"epoch": 64.05946994182288,
"grad_norm": 1.692739725112915,
"learning_rate": 0.001,
"loss": 1.4223,
"step": 198200
},
{
"epoch": 64.0917905623788,
"grad_norm": 1.8461979627609253,
"learning_rate": 0.001,
"loss": 1.4266,
"step": 198300
},
{
"epoch": 64.12411118293471,
"grad_norm": 1.447763442993164,
"learning_rate": 0.001,
"loss": 1.4439,
"step": 198400
},
{
"epoch": 64.15643180349063,
"grad_norm": 1.5018121004104614,
"learning_rate": 0.001,
"loss": 1.4366,
"step": 198500
},
{
"epoch": 64.18875242404654,
"grad_norm": 1.3584603071212769,
"learning_rate": 0.001,
"loss": 1.4343,
"step": 198600
},
{
"epoch": 64.22107304460246,
"grad_norm": 1.3046512603759766,
"learning_rate": 0.001,
"loss": 1.4636,
"step": 198700
},
{
"epoch": 64.25339366515837,
"grad_norm": 1.353569746017456,
"learning_rate": 0.001,
"loss": 1.4494,
"step": 198800
},
{
"epoch": 64.28571428571429,
"grad_norm": 1.2177481651306152,
"learning_rate": 0.001,
"loss": 1.4551,
"step": 198900
},
{
"epoch": 64.3180349062702,
"grad_norm": 1.4385019540786743,
"learning_rate": 0.001,
"loss": 1.4552,
"step": 199000
},
{
"epoch": 64.35035552682612,
"grad_norm": 1.4048755168914795,
"learning_rate": 0.001,
"loss": 1.4605,
"step": 199100
},
{
"epoch": 64.38267614738203,
"grad_norm": 1.3325035572052002,
"learning_rate": 0.001,
"loss": 1.4711,
"step": 199200
},
{
"epoch": 64.41499676793795,
"grad_norm": 1.636793851852417,
"learning_rate": 0.001,
"loss": 1.4672,
"step": 199300
},
{
"epoch": 64.44731738849386,
"grad_norm": 1.5934114456176758,
"learning_rate": 0.001,
"loss": 1.4687,
"step": 199400
},
{
"epoch": 64.47963800904978,
"grad_norm": 1.354085087776184,
"learning_rate": 0.001,
"loss": 1.4892,
"step": 199500
},
{
"epoch": 64.51195862960569,
"grad_norm": 1.6613529920578003,
"learning_rate": 0.001,
"loss": 1.4896,
"step": 199600
},
{
"epoch": 64.54427925016161,
"grad_norm": 1.1835780143737793,
"learning_rate": 0.001,
"loss": 1.4928,
"step": 199700
},
{
"epoch": 64.57659987071752,
"grad_norm": 1.3278464078903198,
"learning_rate": 0.001,
"loss": 1.4949,
"step": 199800
},
{
"epoch": 64.60892049127344,
"grad_norm": 1.313319444656372,
"learning_rate": 0.001,
"loss": 1.5086,
"step": 199900
},
{
"epoch": 64.64124111182934,
"grad_norm": 1.815351963043213,
"learning_rate": 0.001,
"loss": 1.4891,
"step": 200000
},
{
"epoch": 64.67356173238527,
"grad_norm": 1.5486884117126465,
"learning_rate": 0.001,
"loss": 1.4982,
"step": 200100
},
{
"epoch": 64.70588235294117,
"grad_norm": 1.261666178703308,
"learning_rate": 0.001,
"loss": 1.5102,
"step": 200200
},
{
"epoch": 64.7382029734971,
"grad_norm": 1.3538340330123901,
"learning_rate": 0.001,
"loss": 1.5032,
"step": 200300
},
{
"epoch": 64.770523594053,
"grad_norm": 1.0698812007904053,
"learning_rate": 0.001,
"loss": 1.5122,
"step": 200400
},
{
"epoch": 64.80284421460892,
"grad_norm": 1.4403045177459717,
"learning_rate": 0.001,
"loss": 1.5234,
"step": 200500
},
{
"epoch": 64.83516483516483,
"grad_norm": 1.8430064916610718,
"learning_rate": 0.001,
"loss": 1.519,
"step": 200600
},
{
"epoch": 64.86748545572075,
"grad_norm": 1.4815309047698975,
"learning_rate": 0.001,
"loss": 1.5213,
"step": 200700
},
{
"epoch": 64.89980607627666,
"grad_norm": 1.6491549015045166,
"learning_rate": 0.001,
"loss": 1.5104,
"step": 200800
},
{
"epoch": 64.93212669683258,
"grad_norm": 1.305277705192566,
"learning_rate": 0.001,
"loss": 1.5074,
"step": 200900
},
{
"epoch": 64.96444731738849,
"grad_norm": 1.182571291923523,
"learning_rate": 0.001,
"loss": 1.5405,
"step": 201000
},
{
"epoch": 64.99676793794441,
"grad_norm": 1.2062500715255737,
"learning_rate": 0.001,
"loss": 1.5109,
"step": 201100
},
{
"epoch": 65.02908855850032,
"grad_norm": 2.032524347305298,
"learning_rate": 0.001,
"loss": 1.429,
"step": 201200
},
{
"epoch": 65.06140917905624,
"grad_norm": 1.1922607421875,
"learning_rate": 0.001,
"loss": 1.4019,
"step": 201300
},
{
"epoch": 65.09372979961215,
"grad_norm": 1.6347752809524536,
"learning_rate": 0.001,
"loss": 1.4173,
"step": 201400
},
{
"epoch": 65.12605042016807,
"grad_norm": 1.301655650138855,
"learning_rate": 0.001,
"loss": 1.429,
"step": 201500
},
{
"epoch": 65.15837104072398,
"grad_norm": 1.3937475681304932,
"learning_rate": 0.001,
"loss": 1.4282,
"step": 201600
},
{
"epoch": 65.1906916612799,
"grad_norm": 1.3848620653152466,
"learning_rate": 0.001,
"loss": 1.4286,
"step": 201700
},
{
"epoch": 65.2230122818358,
"grad_norm": 1.374640941619873,
"learning_rate": 0.001,
"loss": 1.4316,
"step": 201800
},
{
"epoch": 65.25533290239173,
"grad_norm": 1.748427391052246,
"learning_rate": 0.001,
"loss": 1.4465,
"step": 201900
},
{
"epoch": 65.28765352294764,
"grad_norm": 1.5066227912902832,
"learning_rate": 0.001,
"loss": 1.4581,
"step": 202000
},
{
"epoch": 65.31997414350356,
"grad_norm": 1.4027007818222046,
"learning_rate": 0.001,
"loss": 1.4358,
"step": 202100
},
{
"epoch": 65.35229476405947,
"grad_norm": 1.8810714483261108,
"learning_rate": 0.001,
"loss": 1.4577,
"step": 202200
},
{
"epoch": 65.38461538461539,
"grad_norm": 2.0307705402374268,
"learning_rate": 0.001,
"loss": 1.4538,
"step": 202300
},
{
"epoch": 65.4169360051713,
"grad_norm": 1.4029256105422974,
"learning_rate": 0.001,
"loss": 1.4623,
"step": 202400
},
{
"epoch": 65.44925662572722,
"grad_norm": 1.3962090015411377,
"learning_rate": 0.001,
"loss": 1.4685,
"step": 202500
},
{
"epoch": 65.48157724628312,
"grad_norm": 1.1838845014572144,
"learning_rate": 0.001,
"loss": 1.4675,
"step": 202600
},
{
"epoch": 65.51389786683905,
"grad_norm": 1.4567015171051025,
"learning_rate": 0.001,
"loss": 1.4733,
"step": 202700
},
{
"epoch": 65.54621848739495,
"grad_norm": 1.7105584144592285,
"learning_rate": 0.001,
"loss": 1.4828,
"step": 202800
},
{
"epoch": 65.57853910795087,
"grad_norm": 1.3521935939788818,
"learning_rate": 0.001,
"loss": 1.4747,
"step": 202900
},
{
"epoch": 65.61085972850678,
"grad_norm": 2.0361671447753906,
"learning_rate": 0.001,
"loss": 1.4699,
"step": 203000
},
{
"epoch": 65.6431803490627,
"grad_norm": 1.2892478704452515,
"learning_rate": 0.001,
"loss": 1.5083,
"step": 203100
},
{
"epoch": 65.67550096961861,
"grad_norm": 1.7287880182266235,
"learning_rate": 0.001,
"loss": 1.5021,
"step": 203200
},
{
"epoch": 65.70782159017453,
"grad_norm": 1.4580110311508179,
"learning_rate": 0.001,
"loss": 1.4984,
"step": 203300
},
{
"epoch": 65.74014221073044,
"grad_norm": 1.4175513982772827,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 203400
},
{
"epoch": 65.77246283128636,
"grad_norm": 1.3390719890594482,
"learning_rate": 0.001,
"loss": 1.5015,
"step": 203500
},
{
"epoch": 65.80478345184227,
"grad_norm": 1.7799029350280762,
"learning_rate": 0.001,
"loss": 1.4917,
"step": 203600
},
{
"epoch": 65.83710407239819,
"grad_norm": 1.4678008556365967,
"learning_rate": 0.001,
"loss": 1.5114,
"step": 203700
},
{
"epoch": 65.8694246929541,
"grad_norm": 1.3278785943984985,
"learning_rate": 0.001,
"loss": 1.5124,
"step": 203800
},
{
"epoch": 65.90174531351002,
"grad_norm": 1.6712968349456787,
"learning_rate": 0.001,
"loss": 1.4969,
"step": 203900
},
{
"epoch": 65.93406593406593,
"grad_norm": 1.1765999794006348,
"learning_rate": 0.001,
"loss": 1.5144,
"step": 204000
},
{
"epoch": 65.96638655462185,
"grad_norm": 1.3417824506759644,
"learning_rate": 0.001,
"loss": 1.5187,
"step": 204100
},
{
"epoch": 65.99870717517777,
"grad_norm": 1.395086646080017,
"learning_rate": 0.001,
"loss": 1.5194,
"step": 204200
},
{
"epoch": 66.03102779573368,
"grad_norm": 2.1266183853149414,
"learning_rate": 0.001,
"loss": 1.3944,
"step": 204300
},
{
"epoch": 66.0633484162896,
"grad_norm": 2.038113594055176,
"learning_rate": 0.001,
"loss": 1.3975,
"step": 204400
},
{
"epoch": 66.0956690368455,
"grad_norm": 1.7127610445022583,
"learning_rate": 0.001,
"loss": 1.3932,
"step": 204500
},
{
"epoch": 66.12798965740143,
"grad_norm": 1.584177017211914,
"learning_rate": 0.001,
"loss": 1.437,
"step": 204600
},
{
"epoch": 66.16031027795734,
"grad_norm": 1.5602288246154785,
"learning_rate": 0.001,
"loss": 1.4035,
"step": 204700
},
{
"epoch": 66.19263089851326,
"grad_norm": 1.9326955080032349,
"learning_rate": 0.001,
"loss": 1.4268,
"step": 204800
},
{
"epoch": 66.22495151906917,
"grad_norm": 1.6825189590454102,
"learning_rate": 0.001,
"loss": 1.4475,
"step": 204900
},
{
"epoch": 66.25727213962509,
"grad_norm": 1.5411744117736816,
"learning_rate": 0.001,
"loss": 1.4382,
"step": 205000
},
{
"epoch": 66.289592760181,
"grad_norm": 1.7852213382720947,
"learning_rate": 0.001,
"loss": 1.4155,
"step": 205100
},
{
"epoch": 66.32191338073692,
"grad_norm": 1.9252405166625977,
"learning_rate": 0.001,
"loss": 1.4432,
"step": 205200
},
{
"epoch": 66.35423400129282,
"grad_norm": 1.3320082426071167,
"learning_rate": 0.001,
"loss": 1.4358,
"step": 205300
},
{
"epoch": 66.38655462184875,
"grad_norm": 1.876755714416504,
"learning_rate": 0.001,
"loss": 1.4479,
"step": 205400
},
{
"epoch": 66.41887524240465,
"grad_norm": 1.432909607887268,
"learning_rate": 0.001,
"loss": 1.462,
"step": 205500
},
{
"epoch": 66.45119586296057,
"grad_norm": 1.6559160947799683,
"learning_rate": 0.001,
"loss": 1.4737,
"step": 205600
},
{
"epoch": 66.48351648351648,
"grad_norm": 1.6080788373947144,
"learning_rate": 0.001,
"loss": 1.4749,
"step": 205700
},
{
"epoch": 66.5158371040724,
"grad_norm": 1.3938243389129639,
"learning_rate": 0.001,
"loss": 1.4643,
"step": 205800
},
{
"epoch": 66.54815772462831,
"grad_norm": 1.7196204662322998,
"learning_rate": 0.001,
"loss": 1.4492,
"step": 205900
},
{
"epoch": 66.58047834518423,
"grad_norm": 2.160825490951538,
"learning_rate": 0.001,
"loss": 1.4582,
"step": 206000
},
{
"epoch": 66.61279896574014,
"grad_norm": 1.711246132850647,
"learning_rate": 0.001,
"loss": 1.488,
"step": 206100
},
{
"epoch": 66.64511958629606,
"grad_norm": 1.7977290153503418,
"learning_rate": 0.001,
"loss": 1.4914,
"step": 206200
},
{
"epoch": 66.67744020685197,
"grad_norm": 2.112316608428955,
"learning_rate": 0.001,
"loss": 1.4693,
"step": 206300
},
{
"epoch": 66.70976082740789,
"grad_norm": 1.745553731918335,
"learning_rate": 0.001,
"loss": 1.4814,
"step": 206400
},
{
"epoch": 66.7420814479638,
"grad_norm": 2.0733413696289062,
"learning_rate": 0.001,
"loss": 1.5045,
"step": 206500
},
{
"epoch": 66.77440206851972,
"grad_norm": 1.9156938791275024,
"learning_rate": 0.001,
"loss": 1.49,
"step": 206600
},
{
"epoch": 66.80672268907563,
"grad_norm": 1.8867133855819702,
"learning_rate": 0.001,
"loss": 1.5036,
"step": 206700
},
{
"epoch": 66.83904330963155,
"grad_norm": 1.5210342407226562,
"learning_rate": 0.001,
"loss": 1.4994,
"step": 206800
},
{
"epoch": 66.87136393018746,
"grad_norm": 1.6601982116699219,
"learning_rate": 0.001,
"loss": 1.5036,
"step": 206900
},
{
"epoch": 66.90368455074338,
"grad_norm": 1.545251488685608,
"learning_rate": 0.001,
"loss": 1.484,
"step": 207000
},
{
"epoch": 66.93600517129929,
"grad_norm": 1.7518340349197388,
"learning_rate": 0.001,
"loss": 1.51,
"step": 207100
},
{
"epoch": 66.96832579185521,
"grad_norm": 1.5869966745376587,
"learning_rate": 0.001,
"loss": 1.5062,
"step": 207200
},
{
"epoch": 67.00064641241111,
"grad_norm": 2.387399196624756,
"learning_rate": 0.001,
"loss": 1.4847,
"step": 207300
},
{
"epoch": 67.03296703296704,
"grad_norm": 1.7294135093688965,
"learning_rate": 0.001,
"loss": 1.3812,
"step": 207400
},
{
"epoch": 67.06528765352294,
"grad_norm": 1.832441806793213,
"learning_rate": 0.001,
"loss": 1.3991,
"step": 207500
},
{
"epoch": 67.09760827407887,
"grad_norm": 1.9681521654129028,
"learning_rate": 0.001,
"loss": 1.3971,
"step": 207600
},
{
"epoch": 67.12992889463477,
"grad_norm": 2.1798298358917236,
"learning_rate": 0.001,
"loss": 1.3943,
"step": 207700
},
{
"epoch": 67.1622495151907,
"grad_norm": 2.8551106452941895,
"learning_rate": 0.001,
"loss": 1.4149,
"step": 207800
},
{
"epoch": 67.1945701357466,
"grad_norm": 1.8717961311340332,
"learning_rate": 0.001,
"loss": 1.418,
"step": 207900
},
{
"epoch": 67.22689075630252,
"grad_norm": 1.8349744081497192,
"learning_rate": 0.001,
"loss": 1.4392,
"step": 208000
},
{
"epoch": 67.25921137685843,
"grad_norm": 2.4214842319488525,
"learning_rate": 0.001,
"loss": 1.4163,
"step": 208100
},
{
"epoch": 67.29153199741435,
"grad_norm": 1.9186242818832397,
"learning_rate": 0.001,
"loss": 1.4197,
"step": 208200
},
{
"epoch": 67.32385261797026,
"grad_norm": 2.6227431297302246,
"learning_rate": 0.001,
"loss": 1.4428,
"step": 208300
},
{
"epoch": 67.35617323852618,
"grad_norm": 2.1113786697387695,
"learning_rate": 0.001,
"loss": 1.4546,
"step": 208400
},
{
"epoch": 67.38849385908209,
"grad_norm": 2.537322998046875,
"learning_rate": 0.001,
"loss": 1.4343,
"step": 208500
},
{
"epoch": 67.42081447963801,
"grad_norm": 2.2560858726501465,
"learning_rate": 0.001,
"loss": 1.4427,
"step": 208600
},
{
"epoch": 67.45313510019392,
"grad_norm": 2.499816417694092,
"learning_rate": 0.001,
"loss": 1.4425,
"step": 208700
},
{
"epoch": 67.48545572074984,
"grad_norm": 1.8515291213989258,
"learning_rate": 0.001,
"loss": 1.4456,
"step": 208800
},
{
"epoch": 67.51777634130575,
"grad_norm": 2.8835208415985107,
"learning_rate": 0.001,
"loss": 1.4472,
"step": 208900
},
{
"epoch": 67.55009696186167,
"grad_norm": 2.2833924293518066,
"learning_rate": 0.001,
"loss": 1.4531,
"step": 209000
},
{
"epoch": 67.58241758241758,
"grad_norm": 2.0586676597595215,
"learning_rate": 0.001,
"loss": 1.4724,
"step": 209100
},
{
"epoch": 67.6147382029735,
"grad_norm": 2.410280227661133,
"learning_rate": 0.001,
"loss": 1.4707,
"step": 209200
},
{
"epoch": 67.6470588235294,
"grad_norm": 2.4234213829040527,
"learning_rate": 0.001,
"loss": 1.4729,
"step": 209300
},
{
"epoch": 67.67937944408533,
"grad_norm": 2.246528148651123,
"learning_rate": 0.001,
"loss": 1.4816,
"step": 209400
},
{
"epoch": 67.71170006464124,
"grad_norm": 2.361851692199707,
"learning_rate": 0.001,
"loss": 1.4701,
"step": 209500
},
{
"epoch": 67.74402068519716,
"grad_norm": 1.6785026788711548,
"learning_rate": 0.001,
"loss": 1.4751,
"step": 209600
},
{
"epoch": 67.77634130575306,
"grad_norm": 3.2610843181610107,
"learning_rate": 0.001,
"loss": 1.506,
"step": 209700
},
{
"epoch": 67.80866192630899,
"grad_norm": 2.754026412963867,
"learning_rate": 0.001,
"loss": 1.5054,
"step": 209800
},
{
"epoch": 67.8409825468649,
"grad_norm": 2.0175817012786865,
"learning_rate": 0.001,
"loss": 1.4782,
"step": 209900
},
{
"epoch": 67.87330316742081,
"grad_norm": 2.8812363147735596,
"learning_rate": 0.001,
"loss": 1.4816,
"step": 210000
},
{
"epoch": 67.90562378797672,
"grad_norm": 2.4063074588775635,
"learning_rate": 0.001,
"loss": 1.4866,
"step": 210100
},
{
"epoch": 67.93794440853264,
"grad_norm": 2.533017873764038,
"learning_rate": 0.001,
"loss": 1.4853,
"step": 210200
},
{
"epoch": 67.97026502908855,
"grad_norm": 2.1814818382263184,
"learning_rate": 0.001,
"loss": 1.4818,
"step": 210300
},
{
"epoch": 68.00258564964447,
"grad_norm": 1.7012830972671509,
"learning_rate": 0.001,
"loss": 1.5166,
"step": 210400
},
{
"epoch": 68.0349062702004,
"grad_norm": 1.5861760377883911,
"learning_rate": 0.001,
"loss": 1.3896,
"step": 210500
},
{
"epoch": 68.0672268907563,
"grad_norm": 1.5751943588256836,
"learning_rate": 0.001,
"loss": 1.3889,
"step": 210600
},
{
"epoch": 68.09954751131222,
"grad_norm": 1.440201997756958,
"learning_rate": 0.001,
"loss": 1.3967,
"step": 210700
},
{
"epoch": 68.13186813186813,
"grad_norm": 1.7287768125534058,
"learning_rate": 0.001,
"loss": 1.3819,
"step": 210800
},
{
"epoch": 68.16418875242405,
"grad_norm": 1.773942470550537,
"learning_rate": 0.001,
"loss": 1.3946,
"step": 210900
},
{
"epoch": 68.19650937297996,
"grad_norm": 1.8268699645996094,
"learning_rate": 0.001,
"loss": 1.4166,
"step": 211000
},
{
"epoch": 68.22882999353588,
"grad_norm": 1.7913458347320557,
"learning_rate": 0.001,
"loss": 1.4142,
"step": 211100
},
{
"epoch": 68.26115061409179,
"grad_norm": 1.3699434995651245,
"learning_rate": 0.001,
"loss": 1.4242,
"step": 211200
},
{
"epoch": 68.29347123464771,
"grad_norm": 1.6637346744537354,
"learning_rate": 0.001,
"loss": 1.4192,
"step": 211300
},
{
"epoch": 68.32579185520362,
"grad_norm": 1.4104464054107666,
"learning_rate": 0.001,
"loss": 1.4331,
"step": 211400
},
{
"epoch": 68.35811247575954,
"grad_norm": 1.6352618932724,
"learning_rate": 0.001,
"loss": 1.4321,
"step": 211500
},
{
"epoch": 68.39043309631545,
"grad_norm": 2.0348496437072754,
"learning_rate": 0.001,
"loss": 1.4319,
"step": 211600
},
{
"epoch": 68.42275371687137,
"grad_norm": 1.4709843397140503,
"learning_rate": 0.001,
"loss": 1.4366,
"step": 211700
},
{
"epoch": 68.45507433742728,
"grad_norm": 1.4409093856811523,
"learning_rate": 0.001,
"loss": 1.4518,
"step": 211800
},
{
"epoch": 68.4873949579832,
"grad_norm": 1.572164535522461,
"learning_rate": 0.001,
"loss": 1.4413,
"step": 211900
},
{
"epoch": 68.5197155785391,
"grad_norm": 1.332074761390686,
"learning_rate": 0.001,
"loss": 1.4544,
"step": 212000
},
{
"epoch": 68.55203619909503,
"grad_norm": 1.3531146049499512,
"learning_rate": 0.001,
"loss": 1.4624,
"step": 212100
},
{
"epoch": 68.58435681965094,
"grad_norm": 1.6849377155303955,
"learning_rate": 0.001,
"loss": 1.4565,
"step": 212200
},
{
"epoch": 68.61667744020686,
"grad_norm": 1.0843379497528076,
"learning_rate": 0.001,
"loss": 1.4596,
"step": 212300
},
{
"epoch": 68.64899806076276,
"grad_norm": 1.327228307723999,
"learning_rate": 0.001,
"loss": 1.4611,
"step": 212400
},
{
"epoch": 68.68131868131869,
"grad_norm": 1.573947548866272,
"learning_rate": 0.001,
"loss": 1.4403,
"step": 212500
},
{
"epoch": 68.7136393018746,
"grad_norm": 1.7759371995925903,
"learning_rate": 0.001,
"loss": 1.4639,
"step": 212600
},
{
"epoch": 68.74595992243052,
"grad_norm": 1.9295721054077148,
"learning_rate": 0.001,
"loss": 1.4705,
"step": 212700
},
{
"epoch": 68.77828054298642,
"grad_norm": 1.1141043901443481,
"learning_rate": 0.001,
"loss": 1.4633,
"step": 212800
},
{
"epoch": 68.81060116354234,
"grad_norm": 1.577907919883728,
"learning_rate": 0.001,
"loss": 1.4774,
"step": 212900
},
{
"epoch": 68.84292178409825,
"grad_norm": 1.7735939025878906,
"learning_rate": 0.001,
"loss": 1.4736,
"step": 213000
},
{
"epoch": 68.87524240465417,
"grad_norm": 1.21549391746521,
"learning_rate": 0.001,
"loss": 1.4914,
"step": 213100
},
{
"epoch": 68.90756302521008,
"grad_norm": 1.251387119293213,
"learning_rate": 0.001,
"loss": 1.4656,
"step": 213200
},
{
"epoch": 68.939883645766,
"grad_norm": 1.7132710218429565,
"learning_rate": 0.001,
"loss": 1.4937,
"step": 213300
},
{
"epoch": 68.97220426632191,
"grad_norm": 1.3986599445343018,
"learning_rate": 0.001,
"loss": 1.4654,
"step": 213400
},
{
"epoch": 69.00452488687783,
"grad_norm": 1.4473422765731812,
"learning_rate": 0.001,
"loss": 1.4933,
"step": 213500
},
{
"epoch": 69.03684550743374,
"grad_norm": 1.8489301204681396,
"learning_rate": 0.001,
"loss": 1.3512,
"step": 213600
},
{
"epoch": 69.06916612798966,
"grad_norm": 1.2581794261932373,
"learning_rate": 0.001,
"loss": 1.3798,
"step": 213700
},
{
"epoch": 69.10148674854557,
"grad_norm": 1.275679349899292,
"learning_rate": 0.001,
"loss": 1.3854,
"step": 213800
},
{
"epoch": 69.13380736910149,
"grad_norm": 1.6581004858016968,
"learning_rate": 0.001,
"loss": 1.388,
"step": 213900
},
{
"epoch": 69.1661279896574,
"grad_norm": 1.3569682836532593,
"learning_rate": 0.001,
"loss": 1.4234,
"step": 214000
},
{
"epoch": 69.19844861021332,
"grad_norm": 1.4363735914230347,
"learning_rate": 0.001,
"loss": 1.4007,
"step": 214100
},
{
"epoch": 69.23076923076923,
"grad_norm": 1.223611831665039,
"learning_rate": 0.001,
"loss": 1.396,
"step": 214200
},
{
"epoch": 69.26308985132515,
"grad_norm": 1.415222406387329,
"learning_rate": 0.001,
"loss": 1.4079,
"step": 214300
},
{
"epoch": 69.29541047188106,
"grad_norm": 1.2893728017807007,
"learning_rate": 0.001,
"loss": 1.4062,
"step": 214400
},
{
"epoch": 69.32773109243698,
"grad_norm": 1.2803157567977905,
"learning_rate": 0.001,
"loss": 1.404,
"step": 214500
},
{
"epoch": 69.36005171299288,
"grad_norm": 1.1905156373977661,
"learning_rate": 0.001,
"loss": 1.4154,
"step": 214600
},
{
"epoch": 69.3923723335488,
"grad_norm": 1.5156515836715698,
"learning_rate": 0.001,
"loss": 1.4181,
"step": 214700
},
{
"epoch": 69.42469295410471,
"grad_norm": 1.6935062408447266,
"learning_rate": 0.001,
"loss": 1.424,
"step": 214800
},
{
"epoch": 69.45701357466064,
"grad_norm": 1.687487006187439,
"learning_rate": 0.001,
"loss": 1.4473,
"step": 214900
},
{
"epoch": 69.48933419521654,
"grad_norm": 1.122978687286377,
"learning_rate": 0.001,
"loss": 1.459,
"step": 215000
},
{
"epoch": 69.52165481577246,
"grad_norm": 1.5706843137741089,
"learning_rate": 0.001,
"loss": 1.4469,
"step": 215100
},
{
"epoch": 69.55397543632837,
"grad_norm": 1.6593797206878662,
"learning_rate": 0.001,
"loss": 1.4515,
"step": 215200
},
{
"epoch": 69.5862960568843,
"grad_norm": 1.2683002948760986,
"learning_rate": 0.001,
"loss": 1.453,
"step": 215300
},
{
"epoch": 69.6186166774402,
"grad_norm": 1.5265580415725708,
"learning_rate": 0.001,
"loss": 1.4406,
"step": 215400
},
{
"epoch": 69.65093729799612,
"grad_norm": 1.6200045347213745,
"learning_rate": 0.001,
"loss": 1.4596,
"step": 215500
},
{
"epoch": 69.68325791855203,
"grad_norm": 1.119982123374939,
"learning_rate": 0.001,
"loss": 1.4626,
"step": 215600
},
{
"epoch": 69.71557853910795,
"grad_norm": 1.4945249557495117,
"learning_rate": 0.001,
"loss": 1.4749,
"step": 215700
},
{
"epoch": 69.74789915966386,
"grad_norm": 1.0488057136535645,
"learning_rate": 0.001,
"loss": 1.4622,
"step": 215800
},
{
"epoch": 69.78021978021978,
"grad_norm": 1.492078185081482,
"learning_rate": 0.001,
"loss": 1.4492,
"step": 215900
},
{
"epoch": 69.81254040077569,
"grad_norm": 1.3652523756027222,
"learning_rate": 0.001,
"loss": 1.4524,
"step": 216000
},
{
"epoch": 69.84486102133161,
"grad_norm": 1.3739368915557861,
"learning_rate": 0.001,
"loss": 1.4578,
"step": 216100
},
{
"epoch": 69.87718164188752,
"grad_norm": 1.428552269935608,
"learning_rate": 0.001,
"loss": 1.4608,
"step": 216200
},
{
"epoch": 69.90950226244344,
"grad_norm": 1.40023672580719,
"learning_rate": 0.001,
"loss": 1.4838,
"step": 216300
},
{
"epoch": 69.94182288299935,
"grad_norm": 1.210654616355896,
"learning_rate": 0.001,
"loss": 1.4633,
"step": 216400
},
{
"epoch": 69.97414350355527,
"grad_norm": 1.1459152698516846,
"learning_rate": 0.001,
"loss": 1.4922,
"step": 216500
},
{
"epoch": 70.00646412411119,
"grad_norm": 1.5286470651626587,
"learning_rate": 0.001,
"loss": 1.4832,
"step": 216600
},
{
"epoch": 70.0387847446671,
"grad_norm": 1.279040813446045,
"learning_rate": 0.001,
"loss": 1.3763,
"step": 216700
},
{
"epoch": 70.07110536522302,
"grad_norm": 1.1975349187850952,
"learning_rate": 0.001,
"loss": 1.3691,
"step": 216800
},
{
"epoch": 70.10342598577893,
"grad_norm": 1.497211217880249,
"learning_rate": 0.001,
"loss": 1.3784,
"step": 216900
},
{
"epoch": 70.13574660633485,
"grad_norm": 1.261684536933899,
"learning_rate": 0.001,
"loss": 1.374,
"step": 217000
},
{
"epoch": 70.16806722689076,
"grad_norm": 1.1770519018173218,
"learning_rate": 0.001,
"loss": 1.3965,
"step": 217100
},
{
"epoch": 70.20038784744668,
"grad_norm": 1.1520289182662964,
"learning_rate": 0.001,
"loss": 1.3785,
"step": 217200
},
{
"epoch": 70.23270846800258,
"grad_norm": 2.061236619949341,
"learning_rate": 0.001,
"loss": 1.3857,
"step": 217300
},
{
"epoch": 70.2650290885585,
"grad_norm": 1.5574382543563843,
"learning_rate": 0.001,
"loss": 1.3961,
"step": 217400
},
{
"epoch": 70.29734970911441,
"grad_norm": 1.1871528625488281,
"learning_rate": 0.001,
"loss": 1.4104,
"step": 217500
},
{
"epoch": 70.32967032967034,
"grad_norm": 1.2969461679458618,
"learning_rate": 0.001,
"loss": 1.4099,
"step": 217600
},
{
"epoch": 70.36199095022624,
"grad_norm": 1.206467628479004,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 217700
},
{
"epoch": 70.39431157078216,
"grad_norm": 1.0971918106079102,
"learning_rate": 0.001,
"loss": 1.4138,
"step": 217800
},
{
"epoch": 70.42663219133807,
"grad_norm": 1.7490986585617065,
"learning_rate": 0.001,
"loss": 1.4128,
"step": 217900
},
{
"epoch": 70.458952811894,
"grad_norm": 1.4288859367370605,
"learning_rate": 0.001,
"loss": 1.4226,
"step": 218000
},
{
"epoch": 70.4912734324499,
"grad_norm": 1.3152230978012085,
"learning_rate": 0.001,
"loss": 1.4358,
"step": 218100
},
{
"epoch": 70.52359405300582,
"grad_norm": 1.6234618425369263,
"learning_rate": 0.001,
"loss": 1.4496,
"step": 218200
},
{
"epoch": 70.55591467356173,
"grad_norm": 1.4537832736968994,
"learning_rate": 0.001,
"loss": 1.4439,
"step": 218300
},
{
"epoch": 70.58823529411765,
"grad_norm": 1.4175965785980225,
"learning_rate": 0.001,
"loss": 1.4415,
"step": 218400
},
{
"epoch": 70.62055591467356,
"grad_norm": 1.441012978553772,
"learning_rate": 0.001,
"loss": 1.4401,
"step": 218500
},
{
"epoch": 70.65287653522948,
"grad_norm": 1.5091608762741089,
"learning_rate": 0.001,
"loss": 1.4355,
"step": 218600
},
{
"epoch": 70.68519715578539,
"grad_norm": 1.2717233896255493,
"learning_rate": 0.001,
"loss": 1.4463,
"step": 218700
},
{
"epoch": 70.71751777634131,
"grad_norm": 1.2213460206985474,
"learning_rate": 0.001,
"loss": 1.4381,
"step": 218800
},
{
"epoch": 70.74983839689722,
"grad_norm": 1.2895859479904175,
"learning_rate": 0.001,
"loss": 1.4397,
"step": 218900
},
{
"epoch": 70.78215901745314,
"grad_norm": 1.6317278146743774,
"learning_rate": 0.001,
"loss": 1.449,
"step": 219000
},
{
"epoch": 70.81447963800905,
"grad_norm": 1.1045271158218384,
"learning_rate": 0.001,
"loss": 1.4488,
"step": 219100
},
{
"epoch": 70.84680025856497,
"grad_norm": 1.388588786125183,
"learning_rate": 0.001,
"loss": 1.468,
"step": 219200
},
{
"epoch": 70.87912087912088,
"grad_norm": 1.425180196762085,
"learning_rate": 0.001,
"loss": 1.4633,
"step": 219300
},
{
"epoch": 70.9114414996768,
"grad_norm": 1.0216953754425049,
"learning_rate": 0.001,
"loss": 1.47,
"step": 219400
},
{
"epoch": 70.9437621202327,
"grad_norm": 1.2392832040786743,
"learning_rate": 0.001,
"loss": 1.4775,
"step": 219500
},
{
"epoch": 70.97608274078863,
"grad_norm": 1.483750820159912,
"learning_rate": 0.001,
"loss": 1.4777,
"step": 219600
},
{
"epoch": 71.00840336134453,
"grad_norm": 1.5142189264297485,
"learning_rate": 0.001,
"loss": 1.4498,
"step": 219700
},
{
"epoch": 71.04072398190046,
"grad_norm": 1.4313535690307617,
"learning_rate": 0.001,
"loss": 1.3475,
"step": 219800
},
{
"epoch": 71.07304460245636,
"grad_norm": 1.0934395790100098,
"learning_rate": 0.001,
"loss": 1.3756,
"step": 219900
},
{
"epoch": 71.10536522301229,
"grad_norm": 1.2501411437988281,
"learning_rate": 0.001,
"loss": 1.3796,
"step": 220000
},
{
"epoch": 71.13768584356819,
"grad_norm": 1.2367230653762817,
"learning_rate": 0.001,
"loss": 1.3764,
"step": 220100
},
{
"epoch": 71.17000646412411,
"grad_norm": 1.2580682039260864,
"learning_rate": 0.001,
"loss": 1.3743,
"step": 220200
},
{
"epoch": 71.20232708468002,
"grad_norm": 1.1118855476379395,
"learning_rate": 0.001,
"loss": 1.3991,
"step": 220300
},
{
"epoch": 71.23464770523594,
"grad_norm": 1.0862705707550049,
"learning_rate": 0.001,
"loss": 1.3688,
"step": 220400
},
{
"epoch": 71.26696832579185,
"grad_norm": 1.3011057376861572,
"learning_rate": 0.001,
"loss": 1.3919,
"step": 220500
},
{
"epoch": 71.29928894634777,
"grad_norm": 1.203726053237915,
"learning_rate": 0.001,
"loss": 1.3793,
"step": 220600
},
{
"epoch": 71.33160956690368,
"grad_norm": 1.116129755973816,
"learning_rate": 0.001,
"loss": 1.4003,
"step": 220700
},
{
"epoch": 71.3639301874596,
"grad_norm": 1.25223970413208,
"learning_rate": 0.001,
"loss": 1.4133,
"step": 220800
},
{
"epoch": 71.39625080801551,
"grad_norm": 1.7051457166671753,
"learning_rate": 0.001,
"loss": 1.4078,
"step": 220900
},
{
"epoch": 71.42857142857143,
"grad_norm": 1.403725504875183,
"learning_rate": 0.001,
"loss": 1.4153,
"step": 221000
},
{
"epoch": 71.46089204912734,
"grad_norm": 1.3073865175247192,
"learning_rate": 0.001,
"loss": 1.4148,
"step": 221100
},
{
"epoch": 71.49321266968326,
"grad_norm": 1.7432401180267334,
"learning_rate": 0.001,
"loss": 1.4156,
"step": 221200
},
{
"epoch": 71.52553329023917,
"grad_norm": 1.5628880262374878,
"learning_rate": 0.001,
"loss": 1.4259,
"step": 221300
},
{
"epoch": 71.55785391079509,
"grad_norm": 1.3986612558364868,
"learning_rate": 0.001,
"loss": 1.4384,
"step": 221400
},
{
"epoch": 71.590174531351,
"grad_norm": 1.3593755960464478,
"learning_rate": 0.001,
"loss": 1.4213,
"step": 221500
},
{
"epoch": 71.62249515190692,
"grad_norm": 1.79757821559906,
"learning_rate": 0.001,
"loss": 1.4069,
"step": 221600
},
{
"epoch": 71.65481577246283,
"grad_norm": 1.477777123451233,
"learning_rate": 0.001,
"loss": 1.4294,
"step": 221700
},
{
"epoch": 71.68713639301875,
"grad_norm": 1.86314857006073,
"learning_rate": 0.001,
"loss": 1.4456,
"step": 221800
},
{
"epoch": 71.71945701357465,
"grad_norm": 1.677902102470398,
"learning_rate": 0.001,
"loss": 1.4308,
"step": 221900
},
{
"epoch": 71.75177763413058,
"grad_norm": 1.7817661762237549,
"learning_rate": 0.001,
"loss": 1.458,
"step": 222000
},
{
"epoch": 71.78409825468648,
"grad_norm": 1.5491183996200562,
"learning_rate": 0.001,
"loss": 1.4422,
"step": 222100
},
{
"epoch": 71.8164188752424,
"grad_norm": 1.2707067728042603,
"learning_rate": 0.001,
"loss": 1.4513,
"step": 222200
},
{
"epoch": 71.84873949579831,
"grad_norm": 1.3343786001205444,
"learning_rate": 0.001,
"loss": 1.4563,
"step": 222300
},
{
"epoch": 71.88106011635423,
"grad_norm": 1.7027918100357056,
"learning_rate": 0.001,
"loss": 1.4624,
"step": 222400
},
{
"epoch": 71.91338073691014,
"grad_norm": 1.3434370756149292,
"learning_rate": 0.001,
"loss": 1.454,
"step": 222500
},
{
"epoch": 71.94570135746606,
"grad_norm": 1.5335760116577148,
"learning_rate": 0.001,
"loss": 1.465,
"step": 222600
},
{
"epoch": 71.97802197802197,
"grad_norm": 1.1541085243225098,
"learning_rate": 0.001,
"loss": 1.4689,
"step": 222700
},
{
"epoch": 72.01034259857789,
"grad_norm": 1.7981853485107422,
"learning_rate": 0.001,
"loss": 1.4219,
"step": 222800
},
{
"epoch": 72.04266321913381,
"grad_norm": 1.2448865175247192,
"learning_rate": 0.001,
"loss": 1.349,
"step": 222900
},
{
"epoch": 72.07498383968972,
"grad_norm": 1.3990108966827393,
"learning_rate": 0.001,
"loss": 1.3649,
"step": 223000
},
{
"epoch": 72.10730446024564,
"grad_norm": 1.4207581281661987,
"learning_rate": 0.001,
"loss": 1.3766,
"step": 223100
},
{
"epoch": 72.13962508080155,
"grad_norm": 1.802091360092163,
"learning_rate": 0.001,
"loss": 1.3719,
"step": 223200
},
{
"epoch": 72.17194570135747,
"grad_norm": 1.458859920501709,
"learning_rate": 0.001,
"loss": 1.3653,
"step": 223300
},
{
"epoch": 72.20426632191338,
"grad_norm": 1.418333649635315,
"learning_rate": 0.001,
"loss": 1.3735,
"step": 223400
},
{
"epoch": 72.2365869424693,
"grad_norm": 1.7760887145996094,
"learning_rate": 0.001,
"loss": 1.3786,
"step": 223500
},
{
"epoch": 72.26890756302521,
"grad_norm": 1.5326206684112549,
"learning_rate": 0.001,
"loss": 1.3734,
"step": 223600
},
{
"epoch": 72.30122818358113,
"grad_norm": 1.4824262857437134,
"learning_rate": 0.001,
"loss": 1.4001,
"step": 223700
},
{
"epoch": 72.33354880413704,
"grad_norm": 2.0081357955932617,
"learning_rate": 0.001,
"loss": 1.4037,
"step": 223800
},
{
"epoch": 72.36586942469296,
"grad_norm": 1.2242908477783203,
"learning_rate": 0.001,
"loss": 1.3982,
"step": 223900
},
{
"epoch": 72.39819004524887,
"grad_norm": 1.391268253326416,
"learning_rate": 0.001,
"loss": 1.3912,
"step": 224000
},
{
"epoch": 72.43051066580479,
"grad_norm": 1.7648073434829712,
"learning_rate": 0.001,
"loss": 1.4109,
"step": 224100
},
{
"epoch": 72.4628312863607,
"grad_norm": 1.6145282983779907,
"learning_rate": 0.001,
"loss": 1.3876,
"step": 224200
},
{
"epoch": 72.49515190691662,
"grad_norm": 1.2425663471221924,
"learning_rate": 0.001,
"loss": 1.4089,
"step": 224300
},
{
"epoch": 72.52747252747253,
"grad_norm": 1.5860071182250977,
"learning_rate": 0.001,
"loss": 1.4072,
"step": 224400
},
{
"epoch": 72.55979314802845,
"grad_norm": 1.2650121450424194,
"learning_rate": 0.001,
"loss": 1.4019,
"step": 224500
},
{
"epoch": 72.59211376858435,
"grad_norm": 1.2191665172576904,
"learning_rate": 0.001,
"loss": 1.4349,
"step": 224600
},
{
"epoch": 72.62443438914028,
"grad_norm": 1.093478798866272,
"learning_rate": 0.001,
"loss": 1.4231,
"step": 224700
},
{
"epoch": 72.65675500969618,
"grad_norm": 1.3362218141555786,
"learning_rate": 0.001,
"loss": 1.4204,
"step": 224800
},
{
"epoch": 72.6890756302521,
"grad_norm": 1.4437624216079712,
"learning_rate": 0.001,
"loss": 1.4358,
"step": 224900
},
{
"epoch": 72.72139625080801,
"grad_norm": 1.8382424116134644,
"learning_rate": 0.001,
"loss": 1.4326,
"step": 225000
},
{
"epoch": 72.75371687136393,
"grad_norm": 1.5192632675170898,
"learning_rate": 0.001,
"loss": 1.4329,
"step": 225100
},
{
"epoch": 72.78603749191984,
"grad_norm": 1.560240626335144,
"learning_rate": 0.001,
"loss": 1.4335,
"step": 225200
},
{
"epoch": 72.81835811247576,
"grad_norm": 1.32393479347229,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 225300
},
{
"epoch": 72.85067873303167,
"grad_norm": 1.5922271013259888,
"learning_rate": 0.001,
"loss": 1.4442,
"step": 225400
},
{
"epoch": 72.88299935358759,
"grad_norm": 2.061397075653076,
"learning_rate": 0.001,
"loss": 1.4546,
"step": 225500
},
{
"epoch": 72.9153199741435,
"grad_norm": 1.4817662239074707,
"learning_rate": 0.001,
"loss": 1.4292,
"step": 225600
},
{
"epoch": 72.94764059469942,
"grad_norm": 1.4019739627838135,
"learning_rate": 0.001,
"loss": 1.455,
"step": 225700
},
{
"epoch": 72.97996121525533,
"grad_norm": 1.5779668092727661,
"learning_rate": 0.001,
"loss": 1.463,
"step": 225800
},
{
"epoch": 73.01228183581125,
"grad_norm": 1.6895304918289185,
"learning_rate": 0.001,
"loss": 1.4029,
"step": 225900
},
{
"epoch": 73.04460245636716,
"grad_norm": 1.4958463907241821,
"learning_rate": 0.001,
"loss": 1.332,
"step": 226000
},
{
"epoch": 73.07692307692308,
"grad_norm": 1.3953677415847778,
"learning_rate": 0.001,
"loss": 1.3409,
"step": 226100
},
{
"epoch": 73.10924369747899,
"grad_norm": 1.9095354080200195,
"learning_rate": 0.001,
"loss": 1.342,
"step": 226200
},
{
"epoch": 73.14156431803491,
"grad_norm": 1.3742811679840088,
"learning_rate": 0.001,
"loss": 1.3633,
"step": 226300
},
{
"epoch": 73.17388493859082,
"grad_norm": 1.4097434282302856,
"learning_rate": 0.001,
"loss": 1.3545,
"step": 226400
},
{
"epoch": 73.20620555914674,
"grad_norm": 1.7019058465957642,
"learning_rate": 0.001,
"loss": 1.3666,
"step": 226500
},
{
"epoch": 73.23852617970265,
"grad_norm": 1.4350638389587402,
"learning_rate": 0.001,
"loss": 1.3626,
"step": 226600
},
{
"epoch": 73.27084680025857,
"grad_norm": 1.7013248205184937,
"learning_rate": 0.001,
"loss": 1.3819,
"step": 226700
},
{
"epoch": 73.30316742081448,
"grad_norm": 1.3824927806854248,
"learning_rate": 0.001,
"loss": 1.3804,
"step": 226800
},
{
"epoch": 73.3354880413704,
"grad_norm": 1.9702237844467163,
"learning_rate": 0.001,
"loss": 1.3951,
"step": 226900
},
{
"epoch": 73.3678086619263,
"grad_norm": 1.3919236660003662,
"learning_rate": 0.001,
"loss": 1.4113,
"step": 227000
},
{
"epoch": 73.40012928248223,
"grad_norm": 1.702906847000122,
"learning_rate": 0.001,
"loss": 1.398,
"step": 227100
},
{
"epoch": 73.43244990303813,
"grad_norm": 1.4393773078918457,
"learning_rate": 0.001,
"loss": 1.3926,
"step": 227200
},
{
"epoch": 73.46477052359405,
"grad_norm": 1.297790288925171,
"learning_rate": 0.001,
"loss": 1.4053,
"step": 227300
},
{
"epoch": 73.49709114414996,
"grad_norm": 1.4835476875305176,
"learning_rate": 0.001,
"loss": 1.394,
"step": 227400
},
{
"epoch": 73.52941176470588,
"grad_norm": 1.3250718116760254,
"learning_rate": 0.001,
"loss": 1.405,
"step": 227500
},
{
"epoch": 73.56173238526179,
"grad_norm": 1.4591543674468994,
"learning_rate": 0.001,
"loss": 1.3889,
"step": 227600
},
{
"epoch": 73.59405300581771,
"grad_norm": 1.3447352647781372,
"learning_rate": 0.001,
"loss": 1.4,
"step": 227700
},
{
"epoch": 73.62637362637362,
"grad_norm": 1.2506396770477295,
"learning_rate": 0.001,
"loss": 1.4153,
"step": 227800
},
{
"epoch": 73.65869424692954,
"grad_norm": 1.5525623559951782,
"learning_rate": 0.001,
"loss": 1.4145,
"step": 227900
},
{
"epoch": 73.69101486748545,
"grad_norm": 1.6079643964767456,
"learning_rate": 0.001,
"loss": 1.4278,
"step": 228000
},
{
"epoch": 73.72333548804137,
"grad_norm": 1.2849229574203491,
"learning_rate": 0.001,
"loss": 1.4271,
"step": 228100
},
{
"epoch": 73.75565610859728,
"grad_norm": 1.4724830389022827,
"learning_rate": 0.001,
"loss": 1.4226,
"step": 228200
},
{
"epoch": 73.7879767291532,
"grad_norm": 1.4989345073699951,
"learning_rate": 0.001,
"loss": 1.4348,
"step": 228300
},
{
"epoch": 73.82029734970911,
"grad_norm": 1.5120007991790771,
"learning_rate": 0.001,
"loss": 1.4382,
"step": 228400
},
{
"epoch": 73.85261797026503,
"grad_norm": 1.553991675376892,
"learning_rate": 0.001,
"loss": 1.43,
"step": 228500
},
{
"epoch": 73.88493859082094,
"grad_norm": 1.3788057565689087,
"learning_rate": 0.001,
"loss": 1.4343,
"step": 228600
},
{
"epoch": 73.91725921137686,
"grad_norm": 1.3961942195892334,
"learning_rate": 0.001,
"loss": 1.455,
"step": 228700
},
{
"epoch": 73.94957983193277,
"grad_norm": 1.4251012802124023,
"learning_rate": 0.001,
"loss": 1.4477,
"step": 228800
},
{
"epoch": 73.98190045248869,
"grad_norm": 1.2669397592544556,
"learning_rate": 0.001,
"loss": 1.4479,
"step": 228900
},
{
"epoch": 74.01422107304461,
"grad_norm": 1.5436172485351562,
"learning_rate": 0.001,
"loss": 1.3815,
"step": 229000
},
{
"epoch": 74.04654169360052,
"grad_norm": 1.419766902923584,
"learning_rate": 0.001,
"loss": 1.3326,
"step": 229100
},
{
"epoch": 74.07886231415644,
"grad_norm": 1.6289554834365845,
"learning_rate": 0.001,
"loss": 1.3426,
"step": 229200
},
{
"epoch": 74.11118293471235,
"grad_norm": 1.5455468893051147,
"learning_rate": 0.001,
"loss": 1.3689,
"step": 229300
},
{
"epoch": 74.14350355526827,
"grad_norm": 1.471002221107483,
"learning_rate": 0.001,
"loss": 1.3503,
"step": 229400
},
{
"epoch": 74.17582417582418,
"grad_norm": 1.5914878845214844,
"learning_rate": 0.001,
"loss": 1.3443,
"step": 229500
},
{
"epoch": 74.2081447963801,
"grad_norm": 1.485289454460144,
"learning_rate": 0.001,
"loss": 1.3772,
"step": 229600
},
{
"epoch": 74.240465416936,
"grad_norm": 1.401279091835022,
"learning_rate": 0.001,
"loss": 1.3746,
"step": 229700
},
{
"epoch": 74.27278603749193,
"grad_norm": 1.527219295501709,
"learning_rate": 0.001,
"loss": 1.357,
"step": 229800
},
{
"epoch": 74.30510665804783,
"grad_norm": 1.9624526500701904,
"learning_rate": 0.001,
"loss": 1.3746,
"step": 229900
},
{
"epoch": 74.33742727860376,
"grad_norm": 1.8752304315567017,
"learning_rate": 0.001,
"loss": 1.3811,
"step": 230000
},
{
"epoch": 74.36974789915966,
"grad_norm": 1.5416440963745117,
"learning_rate": 0.001,
"loss": 1.3775,
"step": 230100
},
{
"epoch": 74.40206851971558,
"grad_norm": 2.0629148483276367,
"learning_rate": 0.001,
"loss": 1.3892,
"step": 230200
},
{
"epoch": 74.43438914027149,
"grad_norm": 1.5549070835113525,
"learning_rate": 0.001,
"loss": 1.3995,
"step": 230300
},
{
"epoch": 74.46670976082741,
"grad_norm": 1.8633453845977783,
"learning_rate": 0.001,
"loss": 1.4094,
"step": 230400
},
{
"epoch": 74.49903038138332,
"grad_norm": 1.4920631647109985,
"learning_rate": 0.001,
"loss": 1.3971,
"step": 230500
},
{
"epoch": 74.53135100193924,
"grad_norm": 1.655637502670288,
"learning_rate": 0.001,
"loss": 1.3866,
"step": 230600
},
{
"epoch": 74.56367162249515,
"grad_norm": 1.7276124954223633,
"learning_rate": 0.001,
"loss": 1.3814,
"step": 230700
},
{
"epoch": 74.59599224305107,
"grad_norm": 1.8864883184432983,
"learning_rate": 0.001,
"loss": 1.3954,
"step": 230800
},
{
"epoch": 74.62831286360698,
"grad_norm": 1.5905134677886963,
"learning_rate": 0.001,
"loss": 1.4019,
"step": 230900
},
{
"epoch": 74.6606334841629,
"grad_norm": 1.7310081720352173,
"learning_rate": 0.001,
"loss": 1.4108,
"step": 231000
},
{
"epoch": 74.69295410471881,
"grad_norm": 1.8442275524139404,
"learning_rate": 0.001,
"loss": 1.419,
"step": 231100
},
{
"epoch": 74.72527472527473,
"grad_norm": 2.597527265548706,
"learning_rate": 0.001,
"loss": 1.4086,
"step": 231200
},
{
"epoch": 74.75759534583064,
"grad_norm": 1.941349744796753,
"learning_rate": 0.001,
"loss": 1.4169,
"step": 231300
},
{
"epoch": 74.78991596638656,
"grad_norm": 2.8810555934906006,
"learning_rate": 0.001,
"loss": 1.4267,
"step": 231400
},
{
"epoch": 74.82223658694247,
"grad_norm": 2.6110336780548096,
"learning_rate": 0.001,
"loss": 1.4207,
"step": 231500
},
{
"epoch": 74.85455720749839,
"grad_norm": 2.2146189212799072,
"learning_rate": 0.001,
"loss": 1.4088,
"step": 231600
},
{
"epoch": 74.8868778280543,
"grad_norm": 1.469613790512085,
"learning_rate": 0.001,
"loss": 1.4244,
"step": 231700
},
{
"epoch": 74.91919844861022,
"grad_norm": 1.3704525232315063,
"learning_rate": 0.001,
"loss": 1.4388,
"step": 231800
},
{
"epoch": 74.95151906916612,
"grad_norm": 1.6175132989883423,
"learning_rate": 0.001,
"loss": 1.4224,
"step": 231900
},
{
"epoch": 74.98383968972205,
"grad_norm": 1.592887043952942,
"learning_rate": 0.001,
"loss": 1.4282,
"step": 232000
},
{
"epoch": 75.01616031027795,
"grad_norm": 2.2297322750091553,
"learning_rate": 0.001,
"loss": 1.3406,
"step": 232100
},
{
"epoch": 75.04848093083388,
"grad_norm": 1.9152071475982666,
"learning_rate": 0.001,
"loss": 1.3093,
"step": 232200
},
{
"epoch": 75.08080155138978,
"grad_norm": 2.018739938735962,
"learning_rate": 0.001,
"loss": 1.331,
"step": 232300
},
{
"epoch": 75.1131221719457,
"grad_norm": 2.3672211170196533,
"learning_rate": 0.001,
"loss": 1.3398,
"step": 232400
},
{
"epoch": 75.14544279250161,
"grad_norm": 1.900190830230713,
"learning_rate": 0.001,
"loss": 1.3489,
"step": 232500
},
{
"epoch": 75.17776341305753,
"grad_norm": 1.9358630180358887,
"learning_rate": 0.001,
"loss": 1.3527,
"step": 232600
},
{
"epoch": 75.21008403361344,
"grad_norm": 1.9489713907241821,
"learning_rate": 0.001,
"loss": 1.3546,
"step": 232700
},
{
"epoch": 75.24240465416936,
"grad_norm": 1.9890162944793701,
"learning_rate": 0.001,
"loss": 1.3521,
"step": 232800
},
{
"epoch": 75.27472527472527,
"grad_norm": 2.0825791358947754,
"learning_rate": 0.001,
"loss": 1.3568,
"step": 232900
},
{
"epoch": 75.30704589528119,
"grad_norm": 1.9532067775726318,
"learning_rate": 0.001,
"loss": 1.3546,
"step": 233000
},
{
"epoch": 75.3393665158371,
"grad_norm": 2.3394577503204346,
"learning_rate": 0.001,
"loss": 1.3581,
"step": 233100
},
{
"epoch": 75.37168713639302,
"grad_norm": 1.8234117031097412,
"learning_rate": 0.001,
"loss": 1.381,
"step": 233200
},
{
"epoch": 75.40400775694893,
"grad_norm": 2.1496427059173584,
"learning_rate": 0.001,
"loss": 1.373,
"step": 233300
},
{
"epoch": 75.43632837750485,
"grad_norm": 2.500803232192993,
"learning_rate": 0.001,
"loss": 1.3648,
"step": 233400
},
{
"epoch": 75.46864899806076,
"grad_norm": 2.471842050552368,
"learning_rate": 0.001,
"loss": 1.3808,
"step": 233500
},
{
"epoch": 75.50096961861668,
"grad_norm": 2.3195273876190186,
"learning_rate": 0.001,
"loss": 1.3752,
"step": 233600
},
{
"epoch": 75.53329023917259,
"grad_norm": 1.806527018547058,
"learning_rate": 0.001,
"loss": 1.3975,
"step": 233700
},
{
"epoch": 75.56561085972851,
"grad_norm": 2.4330615997314453,
"learning_rate": 0.001,
"loss": 1.3945,
"step": 233800
},
{
"epoch": 75.59793148028442,
"grad_norm": 2.1549460887908936,
"learning_rate": 0.001,
"loss": 1.3929,
"step": 233900
},
{
"epoch": 75.63025210084034,
"grad_norm": 1.9956578016281128,
"learning_rate": 0.001,
"loss": 1.4234,
"step": 234000
},
{
"epoch": 75.66257272139624,
"grad_norm": 1.9170476198196411,
"learning_rate": 0.001,
"loss": 1.4015,
"step": 234100
},
{
"epoch": 75.69489334195217,
"grad_norm": 2.0377280712127686,
"learning_rate": 0.001,
"loss": 1.4175,
"step": 234200
},
{
"epoch": 75.72721396250807,
"grad_norm": 2.1174378395080566,
"learning_rate": 0.001,
"loss": 1.4144,
"step": 234300
},
{
"epoch": 75.759534583064,
"grad_norm": 3.720243215560913,
"learning_rate": 0.001,
"loss": 1.4042,
"step": 234400
},
{
"epoch": 75.7918552036199,
"grad_norm": 2.0578911304473877,
"learning_rate": 0.001,
"loss": 1.422,
"step": 234500
},
{
"epoch": 75.82417582417582,
"grad_norm": 2.209416627883911,
"learning_rate": 0.001,
"loss": 1.4224,
"step": 234600
},
{
"epoch": 75.85649644473173,
"grad_norm": 2.499605894088745,
"learning_rate": 0.001,
"loss": 1.4315,
"step": 234700
},
{
"epoch": 75.88881706528765,
"grad_norm": 2.3347866535186768,
"learning_rate": 0.001,
"loss": 1.424,
"step": 234800
},
{
"epoch": 75.92113768584356,
"grad_norm": 2.053255081176758,
"learning_rate": 0.001,
"loss": 1.4121,
"step": 234900
},
{
"epoch": 75.95345830639948,
"grad_norm": 1.9303524494171143,
"learning_rate": 0.001,
"loss": 1.4007,
"step": 235000
},
{
"epoch": 75.98577892695539,
"grad_norm": 2.1097376346588135,
"learning_rate": 0.001,
"loss": 1.431,
"step": 235100
},
{
"epoch": 76.01809954751131,
"grad_norm": 2.1241233348846436,
"learning_rate": 0.001,
"loss": 1.3763,
"step": 235200
},
{
"epoch": 76.05042016806723,
"grad_norm": 1.4251773357391357,
"learning_rate": 0.001,
"loss": 1.3067,
"step": 235300
},
{
"epoch": 76.08274078862314,
"grad_norm": 1.1673078536987305,
"learning_rate": 0.001,
"loss": 1.3261,
"step": 235400
},
{
"epoch": 76.11506140917906,
"grad_norm": 1.216108798980713,
"learning_rate": 0.001,
"loss": 1.3301,
"step": 235500
},
{
"epoch": 76.14738202973497,
"grad_norm": 1.5017114877700806,
"learning_rate": 0.001,
"loss": 1.3236,
"step": 235600
},
{
"epoch": 76.17970265029089,
"grad_norm": 1.7115299701690674,
"learning_rate": 0.001,
"loss": 1.3373,
"step": 235700
},
{
"epoch": 76.2120232708468,
"grad_norm": 1.8459619283676147,
"learning_rate": 0.001,
"loss": 1.3567,
"step": 235800
},
{
"epoch": 76.24434389140272,
"grad_norm": 1.5143654346466064,
"learning_rate": 0.001,
"loss": 1.3208,
"step": 235900
},
{
"epoch": 76.27666451195863,
"grad_norm": 1.234527587890625,
"learning_rate": 0.001,
"loss": 1.3639,
"step": 236000
},
{
"epoch": 76.30898513251455,
"grad_norm": 2.20048451423645,
"learning_rate": 0.001,
"loss": 1.3644,
"step": 236100
},
{
"epoch": 76.34130575307046,
"grad_norm": 2.0350420475006104,
"learning_rate": 0.001,
"loss": 1.3627,
"step": 236200
},
{
"epoch": 76.37362637362638,
"grad_norm": 1.8759989738464355,
"learning_rate": 0.001,
"loss": 1.3517,
"step": 236300
},
{
"epoch": 76.40594699418229,
"grad_norm": 1.4786455631256104,
"learning_rate": 0.001,
"loss": 1.3635,
"step": 236400
},
{
"epoch": 76.43826761473821,
"grad_norm": 1.7183560132980347,
"learning_rate": 0.001,
"loss": 1.3635,
"step": 236500
},
{
"epoch": 76.47058823529412,
"grad_norm": 1.6352123022079468,
"learning_rate": 0.001,
"loss": 1.3656,
"step": 236600
},
{
"epoch": 76.50290885585004,
"grad_norm": 1.6529114246368408,
"learning_rate": 0.001,
"loss": 1.3839,
"step": 236700
},
{
"epoch": 76.53522947640595,
"grad_norm": 1.2477518320083618,
"learning_rate": 0.001,
"loss": 1.3975,
"step": 236800
},
{
"epoch": 76.56755009696187,
"grad_norm": 1.4550848007202148,
"learning_rate": 0.001,
"loss": 1.3762,
"step": 236900
},
{
"epoch": 76.59987071751777,
"grad_norm": 1.314056396484375,
"learning_rate": 0.001,
"loss": 1.3909,
"step": 237000
},
{
"epoch": 76.6321913380737,
"grad_norm": 1.9487625360488892,
"learning_rate": 0.001,
"loss": 1.3814,
"step": 237100
},
{
"epoch": 76.6645119586296,
"grad_norm": 1.8857320547103882,
"learning_rate": 0.001,
"loss": 1.3865,
"step": 237200
},
{
"epoch": 76.69683257918552,
"grad_norm": 1.4241645336151123,
"learning_rate": 0.001,
"loss": 1.3814,
"step": 237300
},
{
"epoch": 76.72915319974143,
"grad_norm": 1.2107256650924683,
"learning_rate": 0.001,
"loss": 1.4093,
"step": 237400
},
{
"epoch": 76.76147382029735,
"grad_norm": 1.8503042459487915,
"learning_rate": 0.001,
"loss": 1.3959,
"step": 237500
},
{
"epoch": 76.79379444085326,
"grad_norm": 1.7110308408737183,
"learning_rate": 0.001,
"loss": 1.4123,
"step": 237600
},
{
"epoch": 76.82611506140918,
"grad_norm": 1.6563066244125366,
"learning_rate": 0.001,
"loss": 1.4211,
"step": 237700
},
{
"epoch": 76.85843568196509,
"grad_norm": 1.4431328773498535,
"learning_rate": 0.001,
"loss": 1.4075,
"step": 237800
},
{
"epoch": 76.89075630252101,
"grad_norm": 1.4936169385910034,
"learning_rate": 0.001,
"loss": 1.4349,
"step": 237900
},
{
"epoch": 76.92307692307692,
"grad_norm": 1.3411835432052612,
"learning_rate": 0.001,
"loss": 1.4276,
"step": 238000
},
{
"epoch": 76.95539754363284,
"grad_norm": 1.8169236183166504,
"learning_rate": 0.001,
"loss": 1.4228,
"step": 238100
},
{
"epoch": 76.98771816418875,
"grad_norm": 1.2901304960250854,
"learning_rate": 0.001,
"loss": 1.4421,
"step": 238200
},
{
"epoch": 77.02003878474467,
"grad_norm": 1.8420506715774536,
"learning_rate": 0.001,
"loss": 1.3608,
"step": 238300
},
{
"epoch": 77.05235940530058,
"grad_norm": 1.7031060457229614,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 238400
},
{
"epoch": 77.0846800258565,
"grad_norm": 1.4739128351211548,
"learning_rate": 0.001,
"loss": 1.3203,
"step": 238500
},
{
"epoch": 77.11700064641241,
"grad_norm": 1.3008060455322266,
"learning_rate": 0.001,
"loss": 1.3136,
"step": 238600
},
{
"epoch": 77.14932126696833,
"grad_norm": 1.3382480144500732,
"learning_rate": 0.001,
"loss": 1.3288,
"step": 238700
},
{
"epoch": 77.18164188752424,
"grad_norm": 1.2809972763061523,
"learning_rate": 0.001,
"loss": 1.329,
"step": 238800
},
{
"epoch": 77.21396250808016,
"grad_norm": 1.763350486755371,
"learning_rate": 0.001,
"loss": 1.3444,
"step": 238900
},
{
"epoch": 77.24628312863607,
"grad_norm": 1.3648512363433838,
"learning_rate": 0.001,
"loss": 1.3482,
"step": 239000
},
{
"epoch": 77.27860374919199,
"grad_norm": 1.4460837841033936,
"learning_rate": 0.001,
"loss": 1.3319,
"step": 239100
},
{
"epoch": 77.3109243697479,
"grad_norm": 1.3852758407592773,
"learning_rate": 0.001,
"loss": 1.3541,
"step": 239200
},
{
"epoch": 77.34324499030382,
"grad_norm": 1.4559884071350098,
"learning_rate": 0.001,
"loss": 1.3509,
"step": 239300
},
{
"epoch": 77.37556561085972,
"grad_norm": 1.448446273803711,
"learning_rate": 0.001,
"loss": 1.3506,
"step": 239400
},
{
"epoch": 77.40788623141565,
"grad_norm": 1.0669159889221191,
"learning_rate": 0.001,
"loss": 1.3502,
"step": 239500
},
{
"epoch": 77.44020685197155,
"grad_norm": 1.2951864004135132,
"learning_rate": 0.001,
"loss": 1.3694,
"step": 239600
},
{
"epoch": 77.47252747252747,
"grad_norm": 1.4366976022720337,
"learning_rate": 0.001,
"loss": 1.3585,
"step": 239700
},
{
"epoch": 77.50484809308338,
"grad_norm": 1.384254813194275,
"learning_rate": 0.001,
"loss": 1.3804,
"step": 239800
},
{
"epoch": 77.5371687136393,
"grad_norm": 1.2596639394760132,
"learning_rate": 0.001,
"loss": 1.3735,
"step": 239900
},
{
"epoch": 77.56948933419521,
"grad_norm": 1.4282643795013428,
"learning_rate": 0.001,
"loss": 1.3781,
"step": 240000
},
{
"epoch": 77.60180995475113,
"grad_norm": 1.3321995735168457,
"learning_rate": 0.001,
"loss": 1.3842,
"step": 240100
},
{
"epoch": 77.63413057530704,
"grad_norm": 1.8192081451416016,
"learning_rate": 0.001,
"loss": 1.3796,
"step": 240200
},
{
"epoch": 77.66645119586296,
"grad_norm": 1.1320068836212158,
"learning_rate": 0.001,
"loss": 1.4096,
"step": 240300
},
{
"epoch": 77.69877181641887,
"grad_norm": 1.0336666107177734,
"learning_rate": 0.001,
"loss": 1.3818,
"step": 240400
},
{
"epoch": 77.73109243697479,
"grad_norm": 1.469718098640442,
"learning_rate": 0.001,
"loss": 1.3892,
"step": 240500
},
{
"epoch": 77.7634130575307,
"grad_norm": 1.5609066486358643,
"learning_rate": 0.001,
"loss": 1.4106,
"step": 240600
},
{
"epoch": 77.79573367808662,
"grad_norm": 1.2797508239746094,
"learning_rate": 0.001,
"loss": 1.4041,
"step": 240700
},
{
"epoch": 77.82805429864253,
"grad_norm": 1.0996882915496826,
"learning_rate": 0.001,
"loss": 1.4071,
"step": 240800
},
{
"epoch": 77.86037491919845,
"grad_norm": 2.0114216804504395,
"learning_rate": 0.001,
"loss": 1.3899,
"step": 240900
},
{
"epoch": 77.89269553975436,
"grad_norm": 1.5275205373764038,
"learning_rate": 0.001,
"loss": 1.3939,
"step": 241000
},
{
"epoch": 77.92501616031028,
"grad_norm": 1.758346676826477,
"learning_rate": 0.001,
"loss": 1.4072,
"step": 241100
},
{
"epoch": 77.95733678086619,
"grad_norm": 1.1122857332229614,
"learning_rate": 0.001,
"loss": 1.4176,
"step": 241200
},
{
"epoch": 77.98965740142211,
"grad_norm": 1.1081218719482422,
"learning_rate": 0.001,
"loss": 1.4043,
"step": 241300
},
{
"epoch": 78.02197802197803,
"grad_norm": 1.153250813484192,
"learning_rate": 0.001,
"loss": 1.3459,
"step": 241400
},
{
"epoch": 78.05429864253394,
"grad_norm": 1.346892237663269,
"learning_rate": 0.001,
"loss": 1.3046,
"step": 241500
},
{
"epoch": 78.08661926308986,
"grad_norm": 1.608567476272583,
"learning_rate": 0.001,
"loss": 1.3043,
"step": 241600
},
{
"epoch": 78.11893988364577,
"grad_norm": 1.1944892406463623,
"learning_rate": 0.001,
"loss": 1.3163,
"step": 241700
},
{
"epoch": 78.15126050420169,
"grad_norm": 1.291203498840332,
"learning_rate": 0.001,
"loss": 1.3322,
"step": 241800
},
{
"epoch": 78.1835811247576,
"grad_norm": 1.618190050125122,
"learning_rate": 0.001,
"loss": 1.3229,
"step": 241900
},
{
"epoch": 78.21590174531352,
"grad_norm": 1.2477359771728516,
"learning_rate": 0.001,
"loss": 1.3373,
"step": 242000
},
{
"epoch": 78.24822236586942,
"grad_norm": 1.1648848056793213,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 242100
},
{
"epoch": 78.28054298642535,
"grad_norm": 1.4763754606246948,
"learning_rate": 0.001,
"loss": 1.3499,
"step": 242200
},
{
"epoch": 78.31286360698125,
"grad_norm": 1.7171794176101685,
"learning_rate": 0.001,
"loss": 1.3437,
"step": 242300
},
{
"epoch": 78.34518422753717,
"grad_norm": 2.0236151218414307,
"learning_rate": 0.001,
"loss": 1.3472,
"step": 242400
},
{
"epoch": 78.37750484809308,
"grad_norm": 1.9753978252410889,
"learning_rate": 0.001,
"loss": 1.3322,
"step": 242500
},
{
"epoch": 78.409825468649,
"grad_norm": 1.5920159816741943,
"learning_rate": 0.001,
"loss": 1.3564,
"step": 242600
},
{
"epoch": 78.44214608920491,
"grad_norm": 1.769425630569458,
"learning_rate": 0.001,
"loss": 1.3438,
"step": 242700
},
{
"epoch": 78.47446670976083,
"grad_norm": 1.9399422407150269,
"learning_rate": 0.001,
"loss": 1.3509,
"step": 242800
},
{
"epoch": 78.50678733031674,
"grad_norm": 1.1298264265060425,
"learning_rate": 0.001,
"loss": 1.3609,
"step": 242900
},
{
"epoch": 78.53910795087266,
"grad_norm": 1.1017903089523315,
"learning_rate": 0.001,
"loss": 1.3655,
"step": 243000
},
{
"epoch": 78.57142857142857,
"grad_norm": 1.3760286569595337,
"learning_rate": 0.001,
"loss": 1.3611,
"step": 243100
},
{
"epoch": 78.60374919198449,
"grad_norm": 1.3502388000488281,
"learning_rate": 0.001,
"loss": 1.355,
"step": 243200
},
{
"epoch": 78.6360698125404,
"grad_norm": 1.5598732233047485,
"learning_rate": 0.001,
"loss": 1.3763,
"step": 243300
},
{
"epoch": 78.66839043309632,
"grad_norm": 1.263742208480835,
"learning_rate": 0.001,
"loss": 1.3643,
"step": 243400
},
{
"epoch": 78.70071105365223,
"grad_norm": 1.0220881700515747,
"learning_rate": 0.001,
"loss": 1.385,
"step": 243500
},
{
"epoch": 78.73303167420815,
"grad_norm": 1.5035992860794067,
"learning_rate": 0.001,
"loss": 1.3765,
"step": 243600
},
{
"epoch": 78.76535229476406,
"grad_norm": 1.5086522102355957,
"learning_rate": 0.001,
"loss": 1.4007,
"step": 243700
},
{
"epoch": 78.79767291531998,
"grad_norm": 1.9201769828796387,
"learning_rate": 0.001,
"loss": 1.3995,
"step": 243800
},
{
"epoch": 78.82999353587589,
"grad_norm": 1.4245063066482544,
"learning_rate": 0.001,
"loss": 1.392,
"step": 243900
},
{
"epoch": 78.86231415643181,
"grad_norm": 1.4617079496383667,
"learning_rate": 0.001,
"loss": 1.3933,
"step": 244000
},
{
"epoch": 78.89463477698771,
"grad_norm": 1.313814401626587,
"learning_rate": 0.001,
"loss": 1.4043,
"step": 244100
},
{
"epoch": 78.92695539754364,
"grad_norm": 1.484602928161621,
"learning_rate": 0.001,
"loss": 1.3941,
"step": 244200
},
{
"epoch": 78.95927601809954,
"grad_norm": 1.5593854188919067,
"learning_rate": 0.001,
"loss": 1.4019,
"step": 244300
},
{
"epoch": 78.99159663865547,
"grad_norm": 1.8733112812042236,
"learning_rate": 0.001,
"loss": 1.4085,
"step": 244400
},
{
"epoch": 79.02391725921137,
"grad_norm": 1.5746532678604126,
"learning_rate": 0.001,
"loss": 1.323,
"step": 244500
},
{
"epoch": 79.0562378797673,
"grad_norm": 1.2936152219772339,
"learning_rate": 0.001,
"loss": 1.2868,
"step": 244600
},
{
"epoch": 79.0885585003232,
"grad_norm": 1.0535131692886353,
"learning_rate": 0.001,
"loss": 1.3059,
"step": 244700
},
{
"epoch": 79.12087912087912,
"grad_norm": 1.2122840881347656,
"learning_rate": 0.001,
"loss": 1.3061,
"step": 244800
},
{
"epoch": 79.15319974143503,
"grad_norm": 1.5595974922180176,
"learning_rate": 0.001,
"loss": 1.3182,
"step": 244900
},
{
"epoch": 79.18552036199095,
"grad_norm": 1.6976802349090576,
"learning_rate": 0.001,
"loss": 1.32,
"step": 245000
},
{
"epoch": 79.21784098254686,
"grad_norm": 1.623854398727417,
"learning_rate": 0.001,
"loss": 1.3113,
"step": 245100
},
{
"epoch": 79.25016160310278,
"grad_norm": 1.300982117652893,
"learning_rate": 0.001,
"loss": 1.3379,
"step": 245200
},
{
"epoch": 79.28248222365869,
"grad_norm": 1.4869391918182373,
"learning_rate": 0.001,
"loss": 1.3246,
"step": 245300
},
{
"epoch": 79.31480284421461,
"grad_norm": 1.595557689666748,
"learning_rate": 0.001,
"loss": 1.3295,
"step": 245400
},
{
"epoch": 79.34712346477052,
"grad_norm": 1.4191237688064575,
"learning_rate": 0.001,
"loss": 1.3443,
"step": 245500
},
{
"epoch": 79.37944408532644,
"grad_norm": 1.151425838470459,
"learning_rate": 0.001,
"loss": 1.3503,
"step": 245600
},
{
"epoch": 79.41176470588235,
"grad_norm": 1.34136164188385,
"learning_rate": 0.001,
"loss": 1.3414,
"step": 245700
},
{
"epoch": 79.44408532643827,
"grad_norm": 1.4046324491500854,
"learning_rate": 0.001,
"loss": 1.3602,
"step": 245800
},
{
"epoch": 79.47640594699418,
"grad_norm": 1.1064730882644653,
"learning_rate": 0.001,
"loss": 1.3608,
"step": 245900
},
{
"epoch": 79.5087265675501,
"grad_norm": 1.5794597864151,
"learning_rate": 0.001,
"loss": 1.3461,
"step": 246000
},
{
"epoch": 79.541047188106,
"grad_norm": 1.1105804443359375,
"learning_rate": 0.001,
"loss": 1.3492,
"step": 246100
},
{
"epoch": 79.57336780866193,
"grad_norm": 1.2005650997161865,
"learning_rate": 0.001,
"loss": 1.3657,
"step": 246200
},
{
"epoch": 79.60568842921784,
"grad_norm": 1.3313239812850952,
"learning_rate": 0.001,
"loss": 1.363,
"step": 246300
},
{
"epoch": 79.63800904977376,
"grad_norm": 1.5635491609573364,
"learning_rate": 0.001,
"loss": 1.3864,
"step": 246400
},
{
"epoch": 79.67032967032966,
"grad_norm": 1.3171753883361816,
"learning_rate": 0.001,
"loss": 1.3719,
"step": 246500
},
{
"epoch": 79.70265029088559,
"grad_norm": 1.2938628196716309,
"learning_rate": 0.001,
"loss": 1.3757,
"step": 246600
},
{
"epoch": 79.7349709114415,
"grad_norm": 1.6659266948699951,
"learning_rate": 0.001,
"loss": 1.3744,
"step": 246700
},
{
"epoch": 79.76729153199742,
"grad_norm": 1.1770553588867188,
"learning_rate": 0.001,
"loss": 1.3683,
"step": 246800
},
{
"epoch": 79.79961215255332,
"grad_norm": 1.9406375885009766,
"learning_rate": 0.001,
"loss": 1.3731,
"step": 246900
},
{
"epoch": 79.83193277310924,
"grad_norm": 1.540042519569397,
"learning_rate": 0.001,
"loss": 1.3846,
"step": 247000
},
{
"epoch": 79.86425339366515,
"grad_norm": 1.4754735231399536,
"learning_rate": 0.001,
"loss": 1.3739,
"step": 247100
},
{
"epoch": 79.89657401422107,
"grad_norm": 1.2748198509216309,
"learning_rate": 0.001,
"loss": 1.4012,
"step": 247200
},
{
"epoch": 79.92889463477698,
"grad_norm": 1.9284954071044922,
"learning_rate": 0.001,
"loss": 1.3883,
"step": 247300
},
{
"epoch": 79.9612152553329,
"grad_norm": 1.2156397104263306,
"learning_rate": 0.001,
"loss": 1.4138,
"step": 247400
},
{
"epoch": 79.99353587588882,
"grad_norm": 1.4663833379745483,
"learning_rate": 0.001,
"loss": 1.3909,
"step": 247500
},
{
"epoch": 80.02585649644473,
"grad_norm": 1.323318362236023,
"learning_rate": 0.001,
"loss": 1.318,
"step": 247600
},
{
"epoch": 80.05817711700065,
"grad_norm": 1.6648008823394775,
"learning_rate": 0.001,
"loss": 1.2818,
"step": 247700
},
{
"epoch": 80.09049773755656,
"grad_norm": 1.5227824449539185,
"learning_rate": 0.001,
"loss": 1.2892,
"step": 247800
},
{
"epoch": 80.12281835811248,
"grad_norm": 1.1148303747177124,
"learning_rate": 0.001,
"loss": 1.3032,
"step": 247900
},
{
"epoch": 80.15513897866839,
"grad_norm": 1.3241313695907593,
"learning_rate": 0.001,
"loss": 1.2977,
"step": 248000
},
{
"epoch": 80.18745959922431,
"grad_norm": 1.196300983428955,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 248100
},
{
"epoch": 80.21978021978022,
"grad_norm": 1.5840294361114502,
"learning_rate": 0.001,
"loss": 1.313,
"step": 248200
},
{
"epoch": 80.25210084033614,
"grad_norm": 1.059472680091858,
"learning_rate": 0.001,
"loss": 1.3158,
"step": 248300
},
{
"epoch": 80.28442146089205,
"grad_norm": 1.0643521547317505,
"learning_rate": 0.001,
"loss": 1.3296,
"step": 248400
},
{
"epoch": 80.31674208144797,
"grad_norm": 1.188053846359253,
"learning_rate": 0.001,
"loss": 1.3315,
"step": 248500
},
{
"epoch": 80.34906270200388,
"grad_norm": 1.8046199083328247,
"learning_rate": 0.001,
"loss": 1.3246,
"step": 248600
},
{
"epoch": 80.3813833225598,
"grad_norm": 1.598333477973938,
"learning_rate": 0.001,
"loss": 1.338,
"step": 248700
},
{
"epoch": 80.4137039431157,
"grad_norm": 2.098092794418335,
"learning_rate": 0.001,
"loss": 1.3444,
"step": 248800
},
{
"epoch": 80.44602456367163,
"grad_norm": 1.6266300678253174,
"learning_rate": 0.001,
"loss": 1.3438,
"step": 248900
},
{
"epoch": 80.47834518422754,
"grad_norm": 1.4667638540267944,
"learning_rate": 0.001,
"loss": 1.3384,
"step": 249000
},
{
"epoch": 80.51066580478346,
"grad_norm": 1.2033751010894775,
"learning_rate": 0.001,
"loss": 1.3351,
"step": 249100
},
{
"epoch": 80.54298642533936,
"grad_norm": 1.3361022472381592,
"learning_rate": 0.001,
"loss": 1.3515,
"step": 249200
},
{
"epoch": 80.57530704589529,
"grad_norm": 1.3528895378112793,
"learning_rate": 0.001,
"loss": 1.3708,
"step": 249300
},
{
"epoch": 80.6076276664512,
"grad_norm": 1.6100021600723267,
"learning_rate": 0.001,
"loss": 1.334,
"step": 249400
},
{
"epoch": 80.63994828700712,
"grad_norm": 1.26712167263031,
"learning_rate": 0.001,
"loss": 1.3523,
"step": 249500
},
{
"epoch": 80.67226890756302,
"grad_norm": 1.1492619514465332,
"learning_rate": 0.001,
"loss": 1.361,
"step": 249600
},
{
"epoch": 80.70458952811894,
"grad_norm": 1.5942678451538086,
"learning_rate": 0.001,
"loss": 1.37,
"step": 249700
},
{
"epoch": 80.73691014867485,
"grad_norm": 1.4381914138793945,
"learning_rate": 0.001,
"loss": 1.3625,
"step": 249800
},
{
"epoch": 80.76923076923077,
"grad_norm": 1.6703159809112549,
"learning_rate": 0.001,
"loss": 1.3819,
"step": 249900
},
{
"epoch": 80.80155138978668,
"grad_norm": 1.9152064323425293,
"learning_rate": 0.001,
"loss": 1.3705,
"step": 250000
},
{
"epoch": 80.8338720103426,
"grad_norm": 1.176517128944397,
"learning_rate": 0.001,
"loss": 1.3655,
"step": 250100
},
{
"epoch": 80.86619263089851,
"grad_norm": 1.8336535692214966,
"learning_rate": 0.001,
"loss": 1.3832,
"step": 250200
},
{
"epoch": 80.89851325145443,
"grad_norm": 1.3208738565444946,
"learning_rate": 0.001,
"loss": 1.3747,
"step": 250300
},
{
"epoch": 80.93083387201034,
"grad_norm": 1.662622094154358,
"learning_rate": 0.001,
"loss": 1.3779,
"step": 250400
},
{
"epoch": 80.96315449256626,
"grad_norm": 1.3853704929351807,
"learning_rate": 0.001,
"loss": 1.4126,
"step": 250500
},
{
"epoch": 80.99547511312217,
"grad_norm": 1.4937704801559448,
"learning_rate": 0.001,
"loss": 1.3745,
"step": 250600
},
{
"epoch": 81.02779573367809,
"grad_norm": 1.3379616737365723,
"learning_rate": 0.001,
"loss": 1.2855,
"step": 250700
},
{
"epoch": 81.060116354234,
"grad_norm": 1.2476041316986084,
"learning_rate": 0.001,
"loss": 1.2857,
"step": 250800
},
{
"epoch": 81.09243697478992,
"grad_norm": 1.768044352531433,
"learning_rate": 0.001,
"loss": 1.2942,
"step": 250900
},
{
"epoch": 81.12475759534583,
"grad_norm": 1.562035322189331,
"learning_rate": 0.001,
"loss": 1.3082,
"step": 251000
},
{
"epoch": 81.15707821590175,
"grad_norm": 1.4280803203582764,
"learning_rate": 0.001,
"loss": 1.301,
"step": 251100
},
{
"epoch": 81.18939883645766,
"grad_norm": 1.9252009391784668,
"learning_rate": 0.001,
"loss": 1.3129,
"step": 251200
},
{
"epoch": 81.22171945701358,
"grad_norm": 1.9205073118209839,
"learning_rate": 0.001,
"loss": 1.3009,
"step": 251300
},
{
"epoch": 81.25404007756948,
"grad_norm": 1.634516954421997,
"learning_rate": 0.001,
"loss": 1.2998,
"step": 251400
},
{
"epoch": 81.2863606981254,
"grad_norm": 1.873108983039856,
"learning_rate": 0.001,
"loss": 1.3192,
"step": 251500
},
{
"epoch": 81.31868131868131,
"grad_norm": 1.3185842037200928,
"learning_rate": 0.001,
"loss": 1.3295,
"step": 251600
},
{
"epoch": 81.35100193923724,
"grad_norm": 1.3746492862701416,
"learning_rate": 0.001,
"loss": 1.3023,
"step": 251700
},
{
"epoch": 81.38332255979314,
"grad_norm": 1.5967165231704712,
"learning_rate": 0.001,
"loss": 1.3211,
"step": 251800
},
{
"epoch": 81.41564318034906,
"grad_norm": 1.6355814933776855,
"learning_rate": 0.001,
"loss": 1.3407,
"step": 251900
},
{
"epoch": 81.44796380090497,
"grad_norm": 1.6571320295333862,
"learning_rate": 0.001,
"loss": 1.3398,
"step": 252000
},
{
"epoch": 81.4802844214609,
"grad_norm": 1.8758339881896973,
"learning_rate": 0.001,
"loss": 1.3558,
"step": 252100
},
{
"epoch": 81.5126050420168,
"grad_norm": 1.5424550771713257,
"learning_rate": 0.001,
"loss": 1.3266,
"step": 252200
},
{
"epoch": 81.54492566257272,
"grad_norm": 1.3402589559555054,
"learning_rate": 0.001,
"loss": 1.3384,
"step": 252300
},
{
"epoch": 81.57724628312863,
"grad_norm": 1.3305882215499878,
"learning_rate": 0.001,
"loss": 1.3437,
"step": 252400
},
{
"epoch": 81.60956690368455,
"grad_norm": 1.331802248954773,
"learning_rate": 0.001,
"loss": 1.3667,
"step": 252500
},
{
"epoch": 81.64188752424046,
"grad_norm": 1.5935205221176147,
"learning_rate": 0.001,
"loss": 1.3452,
"step": 252600
},
{
"epoch": 81.67420814479638,
"grad_norm": 2.076895236968994,
"learning_rate": 0.001,
"loss": 1.351,
"step": 252700
},
{
"epoch": 81.70652876535229,
"grad_norm": 1.6277326345443726,
"learning_rate": 0.001,
"loss": 1.3483,
"step": 252800
},
{
"epoch": 81.73884938590821,
"grad_norm": 2.1291604042053223,
"learning_rate": 0.001,
"loss": 1.3806,
"step": 252900
},
{
"epoch": 81.77117000646412,
"grad_norm": 1.7317148447036743,
"learning_rate": 0.001,
"loss": 1.3652,
"step": 253000
},
{
"epoch": 81.80349062702004,
"grad_norm": 1.4864466190338135,
"learning_rate": 0.001,
"loss": 1.3701,
"step": 253100
},
{
"epoch": 81.83581124757595,
"grad_norm": 1.6385993957519531,
"learning_rate": 0.001,
"loss": 1.3723,
"step": 253200
},
{
"epoch": 81.86813186813187,
"grad_norm": 1.415981411933899,
"learning_rate": 0.001,
"loss": 1.3982,
"step": 253300
},
{
"epoch": 81.90045248868778,
"grad_norm": 1.371760606765747,
"learning_rate": 0.001,
"loss": 1.3739,
"step": 253400
},
{
"epoch": 81.9327731092437,
"grad_norm": 1.423779845237732,
"learning_rate": 0.001,
"loss": 1.3655,
"step": 253500
},
{
"epoch": 81.9650937297996,
"grad_norm": 1.8604999780654907,
"learning_rate": 0.001,
"loss": 1.3714,
"step": 253600
},
{
"epoch": 81.99741435035553,
"grad_norm": 1.4258402585983276,
"learning_rate": 0.001,
"loss": 1.3795,
"step": 253700
},
{
"epoch": 82.02973497091145,
"grad_norm": 1.5627765655517578,
"learning_rate": 0.001,
"loss": 1.2867,
"step": 253800
},
{
"epoch": 82.06205559146736,
"grad_norm": 1.750666618347168,
"learning_rate": 0.001,
"loss": 1.2703,
"step": 253900
},
{
"epoch": 82.09437621202328,
"grad_norm": 1.8301845788955688,
"learning_rate": 0.001,
"loss": 1.2725,
"step": 254000
},
{
"epoch": 82.12669683257919,
"grad_norm": 1.524874210357666,
"learning_rate": 0.001,
"loss": 1.2821,
"step": 254100
},
{
"epoch": 82.1590174531351,
"grad_norm": 1.502784013748169,
"learning_rate": 0.001,
"loss": 1.2918,
"step": 254200
},
{
"epoch": 82.19133807369101,
"grad_norm": 2.3623995780944824,
"learning_rate": 0.001,
"loss": 1.2891,
"step": 254300
},
{
"epoch": 82.22365869424694,
"grad_norm": 1.4277948141098022,
"learning_rate": 0.001,
"loss": 1.2997,
"step": 254400
},
{
"epoch": 82.25597931480284,
"grad_norm": 1.5161558389663696,
"learning_rate": 0.001,
"loss": 1.2971,
"step": 254500
},
{
"epoch": 82.28829993535876,
"grad_norm": 1.5056096315383911,
"learning_rate": 0.001,
"loss": 1.3026,
"step": 254600
},
{
"epoch": 82.32062055591467,
"grad_norm": 1.746509313583374,
"learning_rate": 0.001,
"loss": 1.3057,
"step": 254700
},
{
"epoch": 82.3529411764706,
"grad_norm": 1.566635251045227,
"learning_rate": 0.001,
"loss": 1.3301,
"step": 254800
},
{
"epoch": 82.3852617970265,
"grad_norm": 1.726900577545166,
"learning_rate": 0.001,
"loss": 1.3213,
"step": 254900
},
{
"epoch": 82.41758241758242,
"grad_norm": 1.8667417764663696,
"learning_rate": 0.001,
"loss": 1.3332,
"step": 255000
},
{
"epoch": 82.44990303813833,
"grad_norm": 1.4660007953643799,
"learning_rate": 0.001,
"loss": 1.3243,
"step": 255100
},
{
"epoch": 82.48222365869425,
"grad_norm": 1.8999733924865723,
"learning_rate": 0.001,
"loss": 1.3329,
"step": 255200
},
{
"epoch": 82.51454427925016,
"grad_norm": 1.7823513746261597,
"learning_rate": 0.001,
"loss": 1.3423,
"step": 255300
},
{
"epoch": 82.54686489980608,
"grad_norm": 1.6508938074111938,
"learning_rate": 0.001,
"loss": 1.3479,
"step": 255400
},
{
"epoch": 82.57918552036199,
"grad_norm": 1.566430926322937,
"learning_rate": 0.001,
"loss": 1.3431,
"step": 255500
},
{
"epoch": 82.61150614091791,
"grad_norm": 1.9573032855987549,
"learning_rate": 0.001,
"loss": 1.3428,
"step": 255600
},
{
"epoch": 82.64382676147382,
"grad_norm": 1.2677340507507324,
"learning_rate": 0.001,
"loss": 1.3287,
"step": 255700
},
{
"epoch": 82.67614738202974,
"grad_norm": 1.4905617237091064,
"learning_rate": 0.001,
"loss": 1.342,
"step": 255800
},
{
"epoch": 82.70846800258565,
"grad_norm": 2.4624197483062744,
"learning_rate": 0.001,
"loss": 1.3758,
"step": 255900
},
{
"epoch": 82.74078862314157,
"grad_norm": 1.7902895212173462,
"learning_rate": 0.001,
"loss": 1.3471,
"step": 256000
},
{
"epoch": 82.77310924369748,
"grad_norm": 2.4824066162109375,
"learning_rate": 0.001,
"loss": 1.3586,
"step": 256100
},
{
"epoch": 82.8054298642534,
"grad_norm": 1.6157920360565186,
"learning_rate": 0.001,
"loss": 1.3418,
"step": 256200
},
{
"epoch": 82.8377504848093,
"grad_norm": 1.3979263305664062,
"learning_rate": 0.001,
"loss": 1.365,
"step": 256300
},
{
"epoch": 82.87007110536523,
"grad_norm": 1.5057672262191772,
"learning_rate": 0.001,
"loss": 1.3552,
"step": 256400
},
{
"epoch": 82.90239172592113,
"grad_norm": 1.861106276512146,
"learning_rate": 0.001,
"loss": 1.3552,
"step": 256500
},
{
"epoch": 82.93471234647706,
"grad_norm": 1.6322661638259888,
"learning_rate": 0.001,
"loss": 1.3831,
"step": 256600
},
{
"epoch": 82.96703296703296,
"grad_norm": 1.849718451499939,
"learning_rate": 0.001,
"loss": 1.361,
"step": 256700
},
{
"epoch": 82.99935358758889,
"grad_norm": 1.8086191415786743,
"learning_rate": 0.001,
"loss": 1.3523,
"step": 256800
},
{
"epoch": 83.03167420814479,
"grad_norm": 1.804932951927185,
"learning_rate": 0.001,
"loss": 1.2482,
"step": 256900
},
{
"epoch": 83.06399482870071,
"grad_norm": 2.1396706104278564,
"learning_rate": 0.001,
"loss": 1.2676,
"step": 257000
},
{
"epoch": 83.09631544925662,
"grad_norm": 2.0993313789367676,
"learning_rate": 0.001,
"loss": 1.2772,
"step": 257100
},
{
"epoch": 83.12863606981254,
"grad_norm": 1.9156270027160645,
"learning_rate": 0.001,
"loss": 1.2817,
"step": 257200
},
{
"epoch": 83.16095669036845,
"grad_norm": 1.9919105768203735,
"learning_rate": 0.001,
"loss": 1.2824,
"step": 257300
},
{
"epoch": 83.19327731092437,
"grad_norm": 1.8345844745635986,
"learning_rate": 0.001,
"loss": 1.2966,
"step": 257400
},
{
"epoch": 83.22559793148028,
"grad_norm": 1.9297863245010376,
"learning_rate": 0.001,
"loss": 1.2622,
"step": 257500
},
{
"epoch": 83.2579185520362,
"grad_norm": 2.093437433242798,
"learning_rate": 0.001,
"loss": 1.2927,
"step": 257600
},
{
"epoch": 83.29023917259211,
"grad_norm": 1.6618058681488037,
"learning_rate": 0.001,
"loss": 1.2971,
"step": 257700
},
{
"epoch": 83.32255979314803,
"grad_norm": 1.8699930906295776,
"learning_rate": 0.001,
"loss": 1.3082,
"step": 257800
},
{
"epoch": 83.35488041370394,
"grad_norm": 1.9026910066604614,
"learning_rate": 0.001,
"loss": 1.3041,
"step": 257900
},
{
"epoch": 83.38720103425986,
"grad_norm": 1.6460542678833008,
"learning_rate": 0.001,
"loss": 1.3182,
"step": 258000
},
{
"epoch": 83.41952165481577,
"grad_norm": 1.8697502613067627,
"learning_rate": 0.001,
"loss": 1.325,
"step": 258100
},
{
"epoch": 83.45184227537169,
"grad_norm": 1.919716715812683,
"learning_rate": 0.001,
"loss": 1.3239,
"step": 258200
},
{
"epoch": 83.4841628959276,
"grad_norm": 2.285747766494751,
"learning_rate": 0.001,
"loss": 1.3265,
"step": 258300
},
{
"epoch": 83.51648351648352,
"grad_norm": 1.7424185276031494,
"learning_rate": 0.001,
"loss": 1.3393,
"step": 258400
},
{
"epoch": 83.54880413703943,
"grad_norm": 2.293386697769165,
"learning_rate": 0.001,
"loss": 1.3281,
"step": 258500
},
{
"epoch": 83.58112475759535,
"grad_norm": 1.8465816974639893,
"learning_rate": 0.001,
"loss": 1.3349,
"step": 258600
},
{
"epoch": 83.61344537815125,
"grad_norm": 2.1441986560821533,
"learning_rate": 0.001,
"loss": 1.3436,
"step": 258700
},
{
"epoch": 83.64576599870718,
"grad_norm": 1.6771026849746704,
"learning_rate": 0.001,
"loss": 1.3468,
"step": 258800
},
{
"epoch": 83.67808661926308,
"grad_norm": 2.416293144226074,
"learning_rate": 0.001,
"loss": 1.3501,
"step": 258900
},
{
"epoch": 83.710407239819,
"grad_norm": 1.6904393434524536,
"learning_rate": 0.001,
"loss": 1.3665,
"step": 259000
},
{
"epoch": 83.74272786037491,
"grad_norm": 2.0716259479522705,
"learning_rate": 0.001,
"loss": 1.3492,
"step": 259100
},
{
"epoch": 83.77504848093083,
"grad_norm": 1.6293600797653198,
"learning_rate": 0.001,
"loss": 1.3463,
"step": 259200
},
{
"epoch": 83.80736910148674,
"grad_norm": 1.5275318622589111,
"learning_rate": 0.001,
"loss": 1.3488,
"step": 259300
},
{
"epoch": 83.83968972204266,
"grad_norm": 2.0070884227752686,
"learning_rate": 0.001,
"loss": 1.3611,
"step": 259400
},
{
"epoch": 83.87201034259857,
"grad_norm": 1.89895498752594,
"learning_rate": 0.001,
"loss": 1.3572,
"step": 259500
},
{
"epoch": 83.9043309631545,
"grad_norm": 1.539660930633545,
"learning_rate": 0.001,
"loss": 1.3621,
"step": 259600
},
{
"epoch": 83.9366515837104,
"grad_norm": 2.285750150680542,
"learning_rate": 0.001,
"loss": 1.3636,
"step": 259700
},
{
"epoch": 83.96897220426632,
"grad_norm": 1.6824946403503418,
"learning_rate": 0.001,
"loss": 1.358,
"step": 259800
},
{
"epoch": 84.00129282482224,
"grad_norm": 1.5993847846984863,
"learning_rate": 0.001,
"loss": 1.3656,
"step": 259900
},
{
"epoch": 84.03361344537815,
"grad_norm": 2.223039388656616,
"learning_rate": 0.001,
"loss": 1.2675,
"step": 260000
},
{
"epoch": 84.06593406593407,
"grad_norm": 2.1998283863067627,
"learning_rate": 0.001,
"loss": 1.2566,
"step": 260100
},
{
"epoch": 84.09825468648998,
"grad_norm": 2.2801389694213867,
"learning_rate": 0.001,
"loss": 1.2549,
"step": 260200
},
{
"epoch": 84.1305753070459,
"grad_norm": 2.0500996112823486,
"learning_rate": 0.001,
"loss": 1.2782,
"step": 260300
},
{
"epoch": 84.16289592760181,
"grad_norm": 1.7366375923156738,
"learning_rate": 0.001,
"loss": 1.28,
"step": 260400
},
{
"epoch": 84.19521654815773,
"grad_norm": 1.2978771924972534,
"learning_rate": 0.001,
"loss": 1.2964,
"step": 260500
},
{
"epoch": 84.22753716871364,
"grad_norm": 2.125480890274048,
"learning_rate": 0.001,
"loss": 1.2844,
"step": 260600
},
{
"epoch": 84.25985778926956,
"grad_norm": 1.6512577533721924,
"learning_rate": 0.001,
"loss": 1.2718,
"step": 260700
},
{
"epoch": 84.29217840982547,
"grad_norm": 1.575394868850708,
"learning_rate": 0.001,
"loss": 1.2907,
"step": 260800
},
{
"epoch": 84.32449903038139,
"grad_norm": 1.7736165523529053,
"learning_rate": 0.001,
"loss": 1.3104,
"step": 260900
},
{
"epoch": 84.3568196509373,
"grad_norm": 2.15912127494812,
"learning_rate": 0.001,
"loss": 1.3195,
"step": 261000
},
{
"epoch": 84.38914027149322,
"grad_norm": 1.7736667394638062,
"learning_rate": 0.001,
"loss": 1.2958,
"step": 261100
},
{
"epoch": 84.42146089204913,
"grad_norm": 1.929276943206787,
"learning_rate": 0.001,
"loss": 1.3089,
"step": 261200
},
{
"epoch": 84.45378151260505,
"grad_norm": 1.7246546745300293,
"learning_rate": 0.001,
"loss": 1.3053,
"step": 261300
},
{
"epoch": 84.48610213316095,
"grad_norm": 2.1055355072021484,
"learning_rate": 0.001,
"loss": 1.3053,
"step": 261400
},
{
"epoch": 84.51842275371688,
"grad_norm": 1.7114437818527222,
"learning_rate": 0.001,
"loss": 1.3279,
"step": 261500
},
{
"epoch": 84.55074337427278,
"grad_norm": 1.8844554424285889,
"learning_rate": 0.001,
"loss": 1.3137,
"step": 261600
},
{
"epoch": 84.5830639948287,
"grad_norm": 1.9290236234664917,
"learning_rate": 0.001,
"loss": 1.3142,
"step": 261700
},
{
"epoch": 84.61538461538461,
"grad_norm": 1.9991931915283203,
"learning_rate": 0.001,
"loss": 1.3337,
"step": 261800
},
{
"epoch": 84.64770523594053,
"grad_norm": 1.4212626218795776,
"learning_rate": 0.001,
"loss": 1.3268,
"step": 261900
},
{
"epoch": 84.68002585649644,
"grad_norm": 2.720505952835083,
"learning_rate": 0.001,
"loss": 1.3279,
"step": 262000
},
{
"epoch": 84.71234647705236,
"grad_norm": 3.221327543258667,
"learning_rate": 0.001,
"loss": 1.3298,
"step": 262100
},
{
"epoch": 84.74466709760827,
"grad_norm": 2.4408390522003174,
"learning_rate": 0.001,
"loss": 1.3374,
"step": 262200
},
{
"epoch": 84.7769877181642,
"grad_norm": 1.7664300203323364,
"learning_rate": 0.001,
"loss": 1.3541,
"step": 262300
},
{
"epoch": 84.8093083387201,
"grad_norm": 1.4509869813919067,
"learning_rate": 0.001,
"loss": 1.3454,
"step": 262400
},
{
"epoch": 84.84162895927602,
"grad_norm": 1.7354905605316162,
"learning_rate": 0.001,
"loss": 1.3576,
"step": 262500
},
{
"epoch": 84.87394957983193,
"grad_norm": 1.3726277351379395,
"learning_rate": 0.001,
"loss": 1.3538,
"step": 262600
},
{
"epoch": 84.90627020038785,
"grad_norm": 2.5734338760375977,
"learning_rate": 0.001,
"loss": 1.3558,
"step": 262700
},
{
"epoch": 84.93859082094376,
"grad_norm": 1.7000683546066284,
"learning_rate": 0.001,
"loss": 1.3508,
"step": 262800
},
{
"epoch": 84.97091144149968,
"grad_norm": 2.372432231903076,
"learning_rate": 0.001,
"loss": 1.3745,
"step": 262900
},
{
"epoch": 85.00323206205559,
"grad_norm": 1.355724573135376,
"learning_rate": 0.001,
"loss": 1.3708,
"step": 263000
},
{
"epoch": 85.03555268261151,
"grad_norm": 1.3304284811019897,
"learning_rate": 0.001,
"loss": 1.2543,
"step": 263100
},
{
"epoch": 85.06787330316742,
"grad_norm": 1.490299940109253,
"learning_rate": 0.001,
"loss": 1.2625,
"step": 263200
},
{
"epoch": 85.10019392372334,
"grad_norm": 2.0961129665374756,
"learning_rate": 0.001,
"loss": 1.2652,
"step": 263300
},
{
"epoch": 85.13251454427925,
"grad_norm": 1.4494916200637817,
"learning_rate": 0.001,
"loss": 1.2485,
"step": 263400
},
{
"epoch": 85.16483516483517,
"grad_norm": 1.7496620416641235,
"learning_rate": 0.001,
"loss": 1.2654,
"step": 263500
},
{
"epoch": 85.19715578539108,
"grad_norm": 1.441652774810791,
"learning_rate": 0.001,
"loss": 1.2684,
"step": 263600
},
{
"epoch": 85.229476405947,
"grad_norm": 1.6997289657592773,
"learning_rate": 0.001,
"loss": 1.2826,
"step": 263700
},
{
"epoch": 85.2617970265029,
"grad_norm": 1.629094123840332,
"learning_rate": 0.001,
"loss": 1.3095,
"step": 263800
},
{
"epoch": 85.29411764705883,
"grad_norm": 2.266509771347046,
"learning_rate": 0.001,
"loss": 1.2827,
"step": 263900
},
{
"epoch": 85.32643826761473,
"grad_norm": 1.381280779838562,
"learning_rate": 0.001,
"loss": 1.3009,
"step": 264000
},
{
"epoch": 85.35875888817066,
"grad_norm": 1.1983163356781006,
"learning_rate": 0.001,
"loss": 1.3122,
"step": 264100
},
{
"epoch": 85.39107950872656,
"grad_norm": 1.5109035968780518,
"learning_rate": 0.001,
"loss": 1.3001,
"step": 264200
},
{
"epoch": 85.42340012928248,
"grad_norm": 1.9536347389221191,
"learning_rate": 0.001,
"loss": 1.3082,
"step": 264300
},
{
"epoch": 85.45572074983839,
"grad_norm": 1.9702403545379639,
"learning_rate": 0.001,
"loss": 1.3093,
"step": 264400
},
{
"epoch": 85.48804137039431,
"grad_norm": 1.8216692209243774,
"learning_rate": 0.001,
"loss": 1.2987,
"step": 264500
},
{
"epoch": 85.52036199095022,
"grad_norm": 1.486433506011963,
"learning_rate": 0.001,
"loss": 1.3232,
"step": 264600
},
{
"epoch": 85.55268261150614,
"grad_norm": 1.2404817342758179,
"learning_rate": 0.001,
"loss": 1.327,
"step": 264700
},
{
"epoch": 85.58500323206205,
"grad_norm": 2.0966551303863525,
"learning_rate": 0.001,
"loss": 1.3195,
"step": 264800
},
{
"epoch": 85.61732385261797,
"grad_norm": 2.2200942039489746,
"learning_rate": 0.001,
"loss": 1.3215,
"step": 264900
},
{
"epoch": 85.64964447317388,
"grad_norm": 2.0443599224090576,
"learning_rate": 0.001,
"loss": 1.3119,
"step": 265000
},
{
"epoch": 85.6819650937298,
"grad_norm": 1.4810398817062378,
"learning_rate": 0.001,
"loss": 1.3239,
"step": 265100
},
{
"epoch": 85.71428571428571,
"grad_norm": 2.1817426681518555,
"learning_rate": 0.001,
"loss": 1.3113,
"step": 265200
},
{
"epoch": 85.74660633484163,
"grad_norm": 1.5747994184494019,
"learning_rate": 0.001,
"loss": 1.3488,
"step": 265300
},
{
"epoch": 85.77892695539754,
"grad_norm": 1.737360954284668,
"learning_rate": 0.001,
"loss": 1.3396,
"step": 265400
},
{
"epoch": 85.81124757595346,
"grad_norm": 1.530966877937317,
"learning_rate": 0.001,
"loss": 1.348,
"step": 265500
},
{
"epoch": 85.84356819650937,
"grad_norm": 1.6953054666519165,
"learning_rate": 0.001,
"loss": 1.334,
"step": 265600
},
{
"epoch": 85.87588881706529,
"grad_norm": 1.2459678649902344,
"learning_rate": 0.001,
"loss": 1.3461,
"step": 265700
},
{
"epoch": 85.9082094376212,
"grad_norm": 1.4472323656082153,
"learning_rate": 0.001,
"loss": 1.3435,
"step": 265800
},
{
"epoch": 85.94053005817712,
"grad_norm": 1.6780081987380981,
"learning_rate": 0.001,
"loss": 1.3617,
"step": 265900
},
{
"epoch": 85.97285067873302,
"grad_norm": 1.6231255531311035,
"learning_rate": 0.001,
"loss": 1.3532,
"step": 266000
},
{
"epoch": 86.00517129928895,
"grad_norm": 1.3546642065048218,
"learning_rate": 0.001,
"loss": 1.3457,
"step": 266100
},
{
"epoch": 86.03749191984487,
"grad_norm": 1.5437463521957397,
"learning_rate": 0.001,
"loss": 1.2356,
"step": 266200
},
{
"epoch": 86.06981254040078,
"grad_norm": 1.0911564826965332,
"learning_rate": 0.001,
"loss": 1.251,
"step": 266300
},
{
"epoch": 86.1021331609567,
"grad_norm": 1.3593432903289795,
"learning_rate": 0.001,
"loss": 1.2514,
"step": 266400
},
{
"epoch": 86.1344537815126,
"grad_norm": 1.4442890882492065,
"learning_rate": 0.001,
"loss": 1.275,
"step": 266500
},
{
"epoch": 86.16677440206853,
"grad_norm": 1.0219645500183105,
"learning_rate": 0.001,
"loss": 1.2698,
"step": 266600
},
{
"epoch": 86.19909502262443,
"grad_norm": 1.1881905794143677,
"learning_rate": 0.001,
"loss": 1.2636,
"step": 266700
},
{
"epoch": 86.23141564318036,
"grad_norm": 1.651125192642212,
"learning_rate": 0.001,
"loss": 1.269,
"step": 266800
},
{
"epoch": 86.26373626373626,
"grad_norm": 1.6648918390274048,
"learning_rate": 0.001,
"loss": 1.2792,
"step": 266900
},
{
"epoch": 86.29605688429218,
"grad_norm": 1.610687017440796,
"learning_rate": 0.001,
"loss": 1.2972,
"step": 267000
},
{
"epoch": 86.32837750484809,
"grad_norm": 1.5407851934432983,
"learning_rate": 0.001,
"loss": 1.2782,
"step": 267100
},
{
"epoch": 86.36069812540401,
"grad_norm": 1.479880690574646,
"learning_rate": 0.001,
"loss": 1.2904,
"step": 267200
},
{
"epoch": 86.39301874595992,
"grad_norm": 1.3401010036468506,
"learning_rate": 0.001,
"loss": 1.3039,
"step": 267300
},
{
"epoch": 86.42533936651584,
"grad_norm": 1.4931241273880005,
"learning_rate": 0.001,
"loss": 1.3045,
"step": 267400
},
{
"epoch": 86.45765998707175,
"grad_norm": 1.8216471672058105,
"learning_rate": 0.001,
"loss": 1.2842,
"step": 267500
},
{
"epoch": 86.48998060762767,
"grad_norm": 1.646466851234436,
"learning_rate": 0.001,
"loss": 1.3086,
"step": 267600
},
{
"epoch": 86.52230122818358,
"grad_norm": 1.3056232929229736,
"learning_rate": 0.001,
"loss": 1.313,
"step": 267700
},
{
"epoch": 86.5546218487395,
"grad_norm": 1.3183597326278687,
"learning_rate": 0.001,
"loss": 1.3095,
"step": 267800
},
{
"epoch": 86.58694246929541,
"grad_norm": 1.447995662689209,
"learning_rate": 0.001,
"loss": 1.3204,
"step": 267900
},
{
"epoch": 86.61926308985133,
"grad_norm": 1.9042973518371582,
"learning_rate": 0.001,
"loss": 1.3077,
"step": 268000
},
{
"epoch": 86.65158371040724,
"grad_norm": 1.3480629920959473,
"learning_rate": 0.001,
"loss": 1.3222,
"step": 268100
},
{
"epoch": 86.68390433096316,
"grad_norm": 1.9837242364883423,
"learning_rate": 0.001,
"loss": 1.3353,
"step": 268200
},
{
"epoch": 86.71622495151907,
"grad_norm": 1.5706090927124023,
"learning_rate": 0.001,
"loss": 1.3148,
"step": 268300
},
{
"epoch": 86.74854557207499,
"grad_norm": 1.2803431749343872,
"learning_rate": 0.001,
"loss": 1.3331,
"step": 268400
},
{
"epoch": 86.7808661926309,
"grad_norm": 2.0832996368408203,
"learning_rate": 0.001,
"loss": 1.3261,
"step": 268500
},
{
"epoch": 86.81318681318682,
"grad_norm": 1.6948343515396118,
"learning_rate": 0.001,
"loss": 1.3394,
"step": 268600
},
{
"epoch": 86.84550743374272,
"grad_norm": 1.1426417827606201,
"learning_rate": 0.001,
"loss": 1.3404,
"step": 268700
},
{
"epoch": 86.87782805429865,
"grad_norm": 1.770984172821045,
"learning_rate": 0.001,
"loss": 1.3299,
"step": 268800
},
{
"epoch": 86.91014867485455,
"grad_norm": 1.7489936351776123,
"learning_rate": 0.001,
"loss": 1.3434,
"step": 268900
},
{
"epoch": 86.94246929541048,
"grad_norm": 1.9508732557296753,
"learning_rate": 0.001,
"loss": 1.3357,
"step": 269000
},
{
"epoch": 86.97478991596638,
"grad_norm": 1.5043200254440308,
"learning_rate": 0.001,
"loss": 1.3495,
"step": 269100
},
{
"epoch": 87.0071105365223,
"grad_norm": 1.377152681350708,
"learning_rate": 0.001,
"loss": 1.3393,
"step": 269200
},
{
"epoch": 87.03943115707821,
"grad_norm": 1.2322174310684204,
"learning_rate": 0.001,
"loss": 1.2347,
"step": 269300
},
{
"epoch": 87.07175177763413,
"grad_norm": 1.3193856477737427,
"learning_rate": 0.001,
"loss": 1.2399,
"step": 269400
},
{
"epoch": 87.10407239819004,
"grad_norm": 1.3290382623672485,
"learning_rate": 0.001,
"loss": 1.2389,
"step": 269500
},
{
"epoch": 87.13639301874596,
"grad_norm": 1.1766009330749512,
"learning_rate": 0.001,
"loss": 1.2565,
"step": 269600
},
{
"epoch": 87.16871363930187,
"grad_norm": 1.2276049852371216,
"learning_rate": 0.001,
"loss": 1.2409,
"step": 269700
},
{
"epoch": 87.20103425985779,
"grad_norm": 1.2885748147964478,
"learning_rate": 0.001,
"loss": 1.2617,
"step": 269800
},
{
"epoch": 87.2333548804137,
"grad_norm": 1.0642529726028442,
"learning_rate": 0.001,
"loss": 1.2746,
"step": 269900
},
{
"epoch": 87.26567550096962,
"grad_norm": 1.1524479389190674,
"learning_rate": 0.001,
"loss": 1.2684,
"step": 270000
},
{
"epoch": 87.29799612152553,
"grad_norm": 1.0871683359146118,
"learning_rate": 0.001,
"loss": 1.2798,
"step": 270100
},
{
"epoch": 87.33031674208145,
"grad_norm": 1.0725356340408325,
"learning_rate": 0.001,
"loss": 1.2857,
"step": 270200
},
{
"epoch": 87.36263736263736,
"grad_norm": 1.4500463008880615,
"learning_rate": 0.001,
"loss": 1.2979,
"step": 270300
},
{
"epoch": 87.39495798319328,
"grad_norm": 1.4343501329421997,
"learning_rate": 0.001,
"loss": 1.2943,
"step": 270400
},
{
"epoch": 87.42727860374919,
"grad_norm": 1.3510661125183105,
"learning_rate": 0.001,
"loss": 1.2616,
"step": 270500
},
{
"epoch": 87.45959922430511,
"grad_norm": 1.1200387477874756,
"learning_rate": 0.001,
"loss": 1.3017,
"step": 270600
},
{
"epoch": 87.49191984486102,
"grad_norm": 1.4455077648162842,
"learning_rate": 0.001,
"loss": 1.2979,
"step": 270700
},
{
"epoch": 87.52424046541694,
"grad_norm": 1.2550908327102661,
"learning_rate": 0.001,
"loss": 1.2938,
"step": 270800
},
{
"epoch": 87.55656108597285,
"grad_norm": 1.6386743783950806,
"learning_rate": 0.001,
"loss": 1.2917,
"step": 270900
},
{
"epoch": 87.58888170652877,
"grad_norm": 1.8549699783325195,
"learning_rate": 0.001,
"loss": 1.3039,
"step": 271000
},
{
"epoch": 87.62120232708467,
"grad_norm": 1.521889328956604,
"learning_rate": 0.001,
"loss": 1.3212,
"step": 271100
},
{
"epoch": 87.6535229476406,
"grad_norm": 1.7515789270401,
"learning_rate": 0.001,
"loss": 1.3141,
"step": 271200
},
{
"epoch": 87.6858435681965,
"grad_norm": 1.517162799835205,
"learning_rate": 0.001,
"loss": 1.3393,
"step": 271300
},
{
"epoch": 87.71816418875243,
"grad_norm": 1.5641160011291504,
"learning_rate": 0.001,
"loss": 1.3278,
"step": 271400
},
{
"epoch": 87.75048480930833,
"grad_norm": 1.2876046895980835,
"learning_rate": 0.001,
"loss": 1.3302,
"step": 271500
},
{
"epoch": 87.78280542986425,
"grad_norm": 1.3556506633758545,
"learning_rate": 0.001,
"loss": 1.3299,
"step": 271600
},
{
"epoch": 87.81512605042016,
"grad_norm": 1.5794581174850464,
"learning_rate": 0.001,
"loss": 1.3372,
"step": 271700
},
{
"epoch": 87.84744667097608,
"grad_norm": 1.4244409799575806,
"learning_rate": 0.001,
"loss": 1.3254,
"step": 271800
},
{
"epoch": 87.87976729153199,
"grad_norm": 1.2709298133850098,
"learning_rate": 0.001,
"loss": 1.3121,
"step": 271900
},
{
"epoch": 87.91208791208791,
"grad_norm": 1.087098240852356,
"learning_rate": 0.001,
"loss": 1.3425,
"step": 272000
},
{
"epoch": 87.94440853264382,
"grad_norm": 1.3802369832992554,
"learning_rate": 0.001,
"loss": 1.3494,
"step": 272100
},
{
"epoch": 87.97672915319974,
"grad_norm": 1.3726370334625244,
"learning_rate": 0.001,
"loss": 1.3451,
"step": 272200
},
{
"epoch": 88.00904977375566,
"grad_norm": 1.5868128538131714,
"learning_rate": 0.001,
"loss": 1.3083,
"step": 272300
},
{
"epoch": 88.04137039431157,
"grad_norm": 1.8111586570739746,
"learning_rate": 0.001,
"loss": 1.2413,
"step": 272400
},
{
"epoch": 88.07369101486749,
"grad_norm": 1.7072577476501465,
"learning_rate": 0.001,
"loss": 1.2248,
"step": 272500
},
{
"epoch": 88.1060116354234,
"grad_norm": 1.5115412473678589,
"learning_rate": 0.001,
"loss": 1.2437,
"step": 272600
},
{
"epoch": 88.13833225597932,
"grad_norm": 1.4994944334030151,
"learning_rate": 0.001,
"loss": 1.2623,
"step": 272700
},
{
"epoch": 88.17065287653523,
"grad_norm": 1.8919925689697266,
"learning_rate": 0.001,
"loss": 1.2589,
"step": 272800
},
{
"epoch": 88.20297349709115,
"grad_norm": 1.605226993560791,
"learning_rate": 0.001,
"loss": 1.2458,
"step": 272900
},
{
"epoch": 88.23529411764706,
"grad_norm": 1.906920313835144,
"learning_rate": 0.001,
"loss": 1.2704,
"step": 273000
},
{
"epoch": 88.26761473820298,
"grad_norm": 1.3726482391357422,
"learning_rate": 0.001,
"loss": 1.2589,
"step": 273100
},
{
"epoch": 88.29993535875889,
"grad_norm": 1.4493634700775146,
"learning_rate": 0.001,
"loss": 1.2684,
"step": 273200
},
{
"epoch": 88.33225597931481,
"grad_norm": 1.6324255466461182,
"learning_rate": 0.001,
"loss": 1.2729,
"step": 273300
},
{
"epoch": 88.36457659987072,
"grad_norm": 1.597145676612854,
"learning_rate": 0.001,
"loss": 1.265,
"step": 273400
},
{
"epoch": 88.39689722042664,
"grad_norm": 1.5736491680145264,
"learning_rate": 0.001,
"loss": 1.2631,
"step": 273500
},
{
"epoch": 88.42921784098255,
"grad_norm": 1.4064427614212036,
"learning_rate": 0.001,
"loss": 1.283,
"step": 273600
},
{
"epoch": 88.46153846153847,
"grad_norm": 1.119673728942871,
"learning_rate": 0.001,
"loss": 1.2887,
"step": 273700
},
{
"epoch": 88.49385908209437,
"grad_norm": 1.693988561630249,
"learning_rate": 0.001,
"loss": 1.2808,
"step": 273800
},
{
"epoch": 88.5261797026503,
"grad_norm": 1.2078344821929932,
"learning_rate": 0.001,
"loss": 1.2897,
"step": 273900
},
{
"epoch": 88.5585003232062,
"grad_norm": 1.3367725610733032,
"learning_rate": 0.001,
"loss": 1.301,
"step": 274000
},
{
"epoch": 88.59082094376213,
"grad_norm": 1.3235373497009277,
"learning_rate": 0.001,
"loss": 1.3032,
"step": 274100
},
{
"epoch": 88.62314156431803,
"grad_norm": 1.6101962327957153,
"learning_rate": 0.001,
"loss": 1.3075,
"step": 274200
},
{
"epoch": 88.65546218487395,
"grad_norm": 1.2641617059707642,
"learning_rate": 0.001,
"loss": 1.3054,
"step": 274300
},
{
"epoch": 88.68778280542986,
"grad_norm": 1.3416248559951782,
"learning_rate": 0.001,
"loss": 1.3062,
"step": 274400
},
{
"epoch": 88.72010342598578,
"grad_norm": 1.562116026878357,
"learning_rate": 0.001,
"loss": 1.3059,
"step": 274500
},
{
"epoch": 88.75242404654169,
"grad_norm": 1.4060194492340088,
"learning_rate": 0.001,
"loss": 1.3267,
"step": 274600
},
{
"epoch": 88.78474466709761,
"grad_norm": 1.39121413230896,
"learning_rate": 0.001,
"loss": 1.3195,
"step": 274700
},
{
"epoch": 88.81706528765352,
"grad_norm": 1.2430040836334229,
"learning_rate": 0.001,
"loss": 1.3274,
"step": 274800
},
{
"epoch": 88.84938590820944,
"grad_norm": 1.948536992073059,
"learning_rate": 0.001,
"loss": 1.3204,
"step": 274900
},
{
"epoch": 88.88170652876535,
"grad_norm": 1.3875020742416382,
"learning_rate": 0.001,
"loss": 1.3251,
"step": 275000
},
{
"epoch": 88.91402714932127,
"grad_norm": 1.2047786712646484,
"learning_rate": 0.001,
"loss": 1.3333,
"step": 275100
},
{
"epoch": 88.94634776987718,
"grad_norm": 1.0818095207214355,
"learning_rate": 0.001,
"loss": 1.3397,
"step": 275200
},
{
"epoch": 88.9786683904331,
"grad_norm": 1.2731244564056396,
"learning_rate": 0.001,
"loss": 1.3399,
"step": 275300
},
{
"epoch": 89.01098901098901,
"grad_norm": 1.4436392784118652,
"learning_rate": 0.001,
"loss": 1.2948,
"step": 275400
},
{
"epoch": 89.04330963154493,
"grad_norm": 1.208919644355774,
"learning_rate": 0.001,
"loss": 1.2368,
"step": 275500
},
{
"epoch": 89.07563025210084,
"grad_norm": 1.2179805040359497,
"learning_rate": 0.001,
"loss": 1.239,
"step": 275600
},
{
"epoch": 89.10795087265676,
"grad_norm": 1.3445192575454712,
"learning_rate": 0.001,
"loss": 1.2441,
"step": 275700
},
{
"epoch": 89.14027149321267,
"grad_norm": 1.353281021118164,
"learning_rate": 0.001,
"loss": 1.2335,
"step": 275800
},
{
"epoch": 89.17259211376859,
"grad_norm": 1.3525140285491943,
"learning_rate": 0.001,
"loss": 1.2548,
"step": 275900
},
{
"epoch": 89.2049127343245,
"grad_norm": 1.5240520238876343,
"learning_rate": 0.001,
"loss": 1.2397,
"step": 276000
},
{
"epoch": 89.23723335488042,
"grad_norm": 1.335564136505127,
"learning_rate": 0.001,
"loss": 1.2559,
"step": 276100
},
{
"epoch": 89.26955397543632,
"grad_norm": 1.603509545326233,
"learning_rate": 0.001,
"loss": 1.2473,
"step": 276200
},
{
"epoch": 89.30187459599225,
"grad_norm": 1.552399754524231,
"learning_rate": 0.001,
"loss": 1.2749,
"step": 276300
},
{
"epoch": 89.33419521654815,
"grad_norm": 1.3112140893936157,
"learning_rate": 0.001,
"loss": 1.2611,
"step": 276400
},
{
"epoch": 89.36651583710407,
"grad_norm": 1.3453762531280518,
"learning_rate": 0.001,
"loss": 1.2655,
"step": 276500
},
{
"epoch": 89.39883645765998,
"grad_norm": 1.7586126327514648,
"learning_rate": 0.001,
"loss": 1.2817,
"step": 276600
},
{
"epoch": 89.4311570782159,
"grad_norm": 1.3448916673660278,
"learning_rate": 0.001,
"loss": 1.2659,
"step": 276700
},
{
"epoch": 89.46347769877181,
"grad_norm": 1.836043357849121,
"learning_rate": 0.001,
"loss": 1.2748,
"step": 276800
},
{
"epoch": 89.49579831932773,
"grad_norm": 1.819887638092041,
"learning_rate": 0.001,
"loss": 1.2764,
"step": 276900
},
{
"epoch": 89.52811893988364,
"grad_norm": 1.4810632467269897,
"learning_rate": 0.001,
"loss": 1.2805,
"step": 277000
},
{
"epoch": 89.56043956043956,
"grad_norm": 1.3329464197158813,
"learning_rate": 0.001,
"loss": 1.2896,
"step": 277100
},
{
"epoch": 89.59276018099547,
"grad_norm": 2.099147319793701,
"learning_rate": 0.001,
"loss": 1.3037,
"step": 277200
},
{
"epoch": 89.62508080155139,
"grad_norm": 1.7932400703430176,
"learning_rate": 0.001,
"loss": 1.3015,
"step": 277300
},
{
"epoch": 89.6574014221073,
"grad_norm": 2.0583438873291016,
"learning_rate": 0.001,
"loss": 1.2991,
"step": 277400
},
{
"epoch": 89.68972204266322,
"grad_norm": 1.4134740829467773,
"learning_rate": 0.001,
"loss": 1.2986,
"step": 277500
},
{
"epoch": 89.72204266321913,
"grad_norm": 2.4827754497528076,
"learning_rate": 0.001,
"loss": 1.3211,
"step": 277600
},
{
"epoch": 89.75436328377505,
"grad_norm": 1.5677183866500854,
"learning_rate": 0.001,
"loss": 1.3036,
"step": 277700
},
{
"epoch": 89.78668390433096,
"grad_norm": 1.4511871337890625,
"learning_rate": 0.001,
"loss": 1.2971,
"step": 277800
},
{
"epoch": 89.81900452488688,
"grad_norm": 1.4890848398208618,
"learning_rate": 0.001,
"loss": 1.3212,
"step": 277900
},
{
"epoch": 89.85132514544279,
"grad_norm": 1.3750056028366089,
"learning_rate": 0.001,
"loss": 1.3236,
"step": 278000
},
{
"epoch": 89.88364576599871,
"grad_norm": 1.318659782409668,
"learning_rate": 0.001,
"loss": 1.3317,
"step": 278100
},
{
"epoch": 89.91596638655462,
"grad_norm": 1.7443362474441528,
"learning_rate": 0.001,
"loss": 1.3196,
"step": 278200
},
{
"epoch": 89.94828700711054,
"grad_norm": 1.5030484199523926,
"learning_rate": 0.001,
"loss": 1.3295,
"step": 278300
},
{
"epoch": 89.98060762766644,
"grad_norm": 1.2083711624145508,
"learning_rate": 0.001,
"loss": 1.3228,
"step": 278400
},
{
"epoch": 90.01292824822237,
"grad_norm": 1.5732769966125488,
"learning_rate": 0.001,
"loss": 1.2681,
"step": 278500
},
{
"epoch": 90.04524886877829,
"grad_norm": 1.7650502920150757,
"learning_rate": 0.001,
"loss": 1.2236,
"step": 278600
},
{
"epoch": 90.0775694893342,
"grad_norm": 1.5268874168395996,
"learning_rate": 0.001,
"loss": 1.2362,
"step": 278700
},
{
"epoch": 90.10989010989012,
"grad_norm": 2.05912709236145,
"learning_rate": 0.001,
"loss": 1.2283,
"step": 278800
},
{
"epoch": 90.14221073044602,
"grad_norm": 1.7886790037155151,
"learning_rate": 0.001,
"loss": 1.2406,
"step": 278900
},
{
"epoch": 90.17453135100195,
"grad_norm": 2.1991751194000244,
"learning_rate": 0.001,
"loss": 1.2394,
"step": 279000
},
{
"epoch": 90.20685197155785,
"grad_norm": 2.0550949573516846,
"learning_rate": 0.001,
"loss": 1.2508,
"step": 279100
},
{
"epoch": 90.23917259211377,
"grad_norm": 1.589145302772522,
"learning_rate": 0.001,
"loss": 1.2411,
"step": 279200
},
{
"epoch": 90.27149321266968,
"grad_norm": 1.8807960748672485,
"learning_rate": 0.001,
"loss": 1.2499,
"step": 279300
},
{
"epoch": 90.3038138332256,
"grad_norm": 1.9496572017669678,
"learning_rate": 0.001,
"loss": 1.2388,
"step": 279400
},
{
"epoch": 90.33613445378151,
"grad_norm": 1.5636197328567505,
"learning_rate": 0.001,
"loss": 1.2575,
"step": 279500
},
{
"epoch": 90.36845507433743,
"grad_norm": 1.9738800525665283,
"learning_rate": 0.001,
"loss": 1.2586,
"step": 279600
},
{
"epoch": 90.40077569489334,
"grad_norm": 1.8361268043518066,
"learning_rate": 0.001,
"loss": 1.2811,
"step": 279700
},
{
"epoch": 90.43309631544926,
"grad_norm": 1.5236778259277344,
"learning_rate": 0.001,
"loss": 1.2711,
"step": 279800
},
{
"epoch": 90.46541693600517,
"grad_norm": 1.494449496269226,
"learning_rate": 0.001,
"loss": 1.2674,
"step": 279900
},
{
"epoch": 90.49773755656109,
"grad_norm": 1.6337898969650269,
"learning_rate": 0.001,
"loss": 1.284,
"step": 280000
},
{
"epoch": 90.530058177117,
"grad_norm": 1.714060664176941,
"learning_rate": 0.001,
"loss": 1.2715,
"step": 280100
},
{
"epoch": 90.56237879767292,
"grad_norm": 1.6631816625595093,
"learning_rate": 0.001,
"loss": 1.2965,
"step": 280200
},
{
"epoch": 90.59469941822883,
"grad_norm": 1.3032524585723877,
"learning_rate": 0.001,
"loss": 1.2745,
"step": 280300
},
{
"epoch": 90.62702003878475,
"grad_norm": 1.5754284858703613,
"learning_rate": 0.001,
"loss": 1.2896,
"step": 280400
},
{
"epoch": 90.65934065934066,
"grad_norm": 1.6375526189804077,
"learning_rate": 0.001,
"loss": 1.2907,
"step": 280500
},
{
"epoch": 90.69166127989658,
"grad_norm": 1.5295196771621704,
"learning_rate": 0.001,
"loss": 1.2915,
"step": 280600
},
{
"epoch": 90.72398190045249,
"grad_norm": 1.4669249057769775,
"learning_rate": 0.001,
"loss": 1.2999,
"step": 280700
},
{
"epoch": 90.75630252100841,
"grad_norm": 1.7439963817596436,
"learning_rate": 0.001,
"loss": 1.3056,
"step": 280800
},
{
"epoch": 90.78862314156432,
"grad_norm": 1.8686052560806274,
"learning_rate": 0.001,
"loss": 1.3203,
"step": 280900
},
{
"epoch": 90.82094376212024,
"grad_norm": 1.313255786895752,
"learning_rate": 0.001,
"loss": 1.3031,
"step": 281000
},
{
"epoch": 90.85326438267614,
"grad_norm": 1.739733338356018,
"learning_rate": 0.001,
"loss": 1.3256,
"step": 281100
},
{
"epoch": 90.88558500323207,
"grad_norm": 1.5114035606384277,
"learning_rate": 0.001,
"loss": 1.32,
"step": 281200
},
{
"epoch": 90.91790562378797,
"grad_norm": 1.4134900569915771,
"learning_rate": 0.001,
"loss": 1.3058,
"step": 281300
},
{
"epoch": 90.9502262443439,
"grad_norm": 1.872104525566101,
"learning_rate": 0.001,
"loss": 1.3184,
"step": 281400
},
{
"epoch": 90.9825468648998,
"grad_norm": 1.8291857242584229,
"learning_rate": 0.001,
"loss": 1.3336,
"step": 281500
},
{
"epoch": 91.01486748545572,
"grad_norm": 2.1532673835754395,
"learning_rate": 0.001,
"loss": 1.2458,
"step": 281600
},
{
"epoch": 91.04718810601163,
"grad_norm": 1.7323987483978271,
"learning_rate": 0.001,
"loss": 1.2135,
"step": 281700
},
{
"epoch": 91.07950872656755,
"grad_norm": 1.5196421146392822,
"learning_rate": 0.001,
"loss": 1.217,
"step": 281800
},
{
"epoch": 91.11182934712346,
"grad_norm": 1.6187310218811035,
"learning_rate": 0.001,
"loss": 1.2276,
"step": 281900
},
{
"epoch": 91.14414996767938,
"grad_norm": 1.5262999534606934,
"learning_rate": 0.001,
"loss": 1.2282,
"step": 282000
},
{
"epoch": 91.17647058823529,
"grad_norm": 1.8699413537979126,
"learning_rate": 0.001,
"loss": 1.2363,
"step": 282100
},
{
"epoch": 91.20879120879121,
"grad_norm": 1.6775938272476196,
"learning_rate": 0.001,
"loss": 1.2205,
"step": 282200
},
{
"epoch": 91.24111182934712,
"grad_norm": 1.3796979188919067,
"learning_rate": 0.001,
"loss": 1.2452,
"step": 282300
},
{
"epoch": 91.27343244990304,
"grad_norm": 1.5739753246307373,
"learning_rate": 0.001,
"loss": 1.2414,
"step": 282400
},
{
"epoch": 91.30575307045895,
"grad_norm": 1.6853272914886475,
"learning_rate": 0.001,
"loss": 1.2476,
"step": 282500
},
{
"epoch": 91.33807369101487,
"grad_norm": 1.5356518030166626,
"learning_rate": 0.001,
"loss": 1.2497,
"step": 282600
},
{
"epoch": 91.37039431157078,
"grad_norm": 2.0445168018341064,
"learning_rate": 0.001,
"loss": 1.2624,
"step": 282700
},
{
"epoch": 91.4027149321267,
"grad_norm": 1.7084234952926636,
"learning_rate": 0.001,
"loss": 1.2655,
"step": 282800
},
{
"epoch": 91.4350355526826,
"grad_norm": 1.8382023572921753,
"learning_rate": 0.001,
"loss": 1.2619,
"step": 282900
},
{
"epoch": 91.46735617323853,
"grad_norm": 1.7835328578948975,
"learning_rate": 0.001,
"loss": 1.2576,
"step": 283000
},
{
"epoch": 91.49967679379444,
"grad_norm": 1.803221583366394,
"learning_rate": 0.001,
"loss": 1.2713,
"step": 283100
},
{
"epoch": 91.53199741435036,
"grad_norm": 2.1018903255462646,
"learning_rate": 0.001,
"loss": 1.2506,
"step": 283200
},
{
"epoch": 91.56431803490626,
"grad_norm": 2.113609790802002,
"learning_rate": 0.001,
"loss": 1.2879,
"step": 283300
},
{
"epoch": 91.59663865546219,
"grad_norm": 2.2016968727111816,
"learning_rate": 0.001,
"loss": 1.2952,
"step": 283400
},
{
"epoch": 91.6289592760181,
"grad_norm": 2.4190940856933594,
"learning_rate": 0.001,
"loss": 1.2733,
"step": 283500
},
{
"epoch": 91.66127989657402,
"grad_norm": 1.8203952312469482,
"learning_rate": 0.001,
"loss": 1.3156,
"step": 283600
},
{
"epoch": 91.69360051712992,
"grad_norm": 1.5039267539978027,
"learning_rate": 0.001,
"loss": 1.3124,
"step": 283700
},
{
"epoch": 91.72592113768584,
"grad_norm": 1.5443657636642456,
"learning_rate": 0.001,
"loss": 1.2934,
"step": 283800
},
{
"epoch": 91.75824175824175,
"grad_norm": 1.5294469594955444,
"learning_rate": 0.001,
"loss": 1.3075,
"step": 283900
},
{
"epoch": 91.79056237879767,
"grad_norm": 1.8437509536743164,
"learning_rate": 0.001,
"loss": 1.3035,
"step": 284000
},
{
"epoch": 91.82288299935358,
"grad_norm": 1.7995463609695435,
"learning_rate": 0.001,
"loss": 1.2974,
"step": 284100
},
{
"epoch": 91.8552036199095,
"grad_norm": 1.6432673931121826,
"learning_rate": 0.001,
"loss": 1.2899,
"step": 284200
},
{
"epoch": 91.88752424046541,
"grad_norm": 2.490947723388672,
"learning_rate": 0.001,
"loss": 1.312,
"step": 284300
},
{
"epoch": 91.91984486102133,
"grad_norm": 2.030097007751465,
"learning_rate": 0.001,
"loss": 1.3157,
"step": 284400
},
{
"epoch": 91.95216548157724,
"grad_norm": 2.9488537311553955,
"learning_rate": 0.001,
"loss": 1.3004,
"step": 284500
},
{
"epoch": 91.98448610213316,
"grad_norm": 1.4872292280197144,
"learning_rate": 0.001,
"loss": 1.3234,
"step": 284600
},
{
"epoch": 92.01680672268908,
"grad_norm": 1.9539153575897217,
"learning_rate": 0.001,
"loss": 1.2429,
"step": 284700
},
{
"epoch": 92.04912734324499,
"grad_norm": 1.7706369161605835,
"learning_rate": 0.001,
"loss": 1.2026,
"step": 284800
},
{
"epoch": 92.08144796380091,
"grad_norm": 2.6587302684783936,
"learning_rate": 0.001,
"loss": 1.2096,
"step": 284900
},
{
"epoch": 92.11376858435682,
"grad_norm": 1.708607792854309,
"learning_rate": 0.001,
"loss": 1.2147,
"step": 285000
},
{
"epoch": 92.14608920491274,
"grad_norm": 2.2268874645233154,
"learning_rate": 0.001,
"loss": 1.226,
"step": 285100
},
{
"epoch": 92.17840982546865,
"grad_norm": 1.5876752138137817,
"learning_rate": 0.001,
"loss": 1.2316,
"step": 285200
},
{
"epoch": 92.21073044602457,
"grad_norm": 1.9572148323059082,
"learning_rate": 0.001,
"loss": 1.2429,
"step": 285300
},
{
"epoch": 92.24305106658048,
"grad_norm": 1.697906255722046,
"learning_rate": 0.001,
"loss": 1.2302,
"step": 285400
},
{
"epoch": 92.2753716871364,
"grad_norm": 2.0902249813079834,
"learning_rate": 0.001,
"loss": 1.2441,
"step": 285500
},
{
"epoch": 92.3076923076923,
"grad_norm": 3.7891249656677246,
"learning_rate": 0.001,
"loss": 1.2448,
"step": 285600
},
{
"epoch": 92.34001292824823,
"grad_norm": 3.3546688556671143,
"learning_rate": 0.001,
"loss": 1.2287,
"step": 285700
},
{
"epoch": 92.37233354880414,
"grad_norm": 2.449063301086426,
"learning_rate": 0.001,
"loss": 1.2333,
"step": 285800
},
{
"epoch": 92.40465416936006,
"grad_norm": 1.6470502614974976,
"learning_rate": 0.001,
"loss": 1.2556,
"step": 285900
},
{
"epoch": 92.43697478991596,
"grad_norm": 3.0614566802978516,
"learning_rate": 0.001,
"loss": 1.2638,
"step": 286000
},
{
"epoch": 92.46929541047189,
"grad_norm": 1.4969353675842285,
"learning_rate": 0.001,
"loss": 1.2848,
"step": 286100
},
{
"epoch": 92.5016160310278,
"grad_norm": 1.6264548301696777,
"learning_rate": 0.001,
"loss": 1.2691,
"step": 286200
},
{
"epoch": 92.53393665158372,
"grad_norm": 2.711341381072998,
"learning_rate": 0.001,
"loss": 1.273,
"step": 286300
},
{
"epoch": 92.56625727213962,
"grad_norm": 1.4267327785491943,
"learning_rate": 0.001,
"loss": 1.2576,
"step": 286400
},
{
"epoch": 92.59857789269554,
"grad_norm": 2.5398762226104736,
"learning_rate": 0.001,
"loss": 1.2652,
"step": 286500
},
{
"epoch": 92.63089851325145,
"grad_norm": 1.5481876134872437,
"learning_rate": 0.001,
"loss": 1.2863,
"step": 286600
},
{
"epoch": 92.66321913380737,
"grad_norm": 3.0440077781677246,
"learning_rate": 0.001,
"loss": 1.2803,
"step": 286700
},
{
"epoch": 92.69553975436328,
"grad_norm": 1.9892159700393677,
"learning_rate": 0.001,
"loss": 1.2965,
"step": 286800
},
{
"epoch": 92.7278603749192,
"grad_norm": 3.3127901554107666,
"learning_rate": 0.001,
"loss": 1.2769,
"step": 286900
},
{
"epoch": 92.76018099547511,
"grad_norm": 2.0899786949157715,
"learning_rate": 0.001,
"loss": 1.2826,
"step": 287000
},
{
"epoch": 92.79250161603103,
"grad_norm": 2.527277946472168,
"learning_rate": 0.001,
"loss": 1.2975,
"step": 287100
},
{
"epoch": 92.82482223658694,
"grad_norm": 2.2242958545684814,
"learning_rate": 0.001,
"loss": 1.302,
"step": 287200
},
{
"epoch": 92.85714285714286,
"grad_norm": 2.0108234882354736,
"learning_rate": 0.001,
"loss": 1.3099,
"step": 287300
},
{
"epoch": 92.88946347769877,
"grad_norm": 2.542722702026367,
"learning_rate": 0.001,
"loss": 1.3074,
"step": 287400
},
{
"epoch": 92.92178409825469,
"grad_norm": 3.474771738052368,
"learning_rate": 0.001,
"loss": 1.3182,
"step": 287500
},
{
"epoch": 92.9541047188106,
"grad_norm": 2.7254514694213867,
"learning_rate": 0.001,
"loss": 1.306,
"step": 287600
},
{
"epoch": 92.98642533936652,
"grad_norm": 1.9012902975082397,
"learning_rate": 0.001,
"loss": 1.3006,
"step": 287700
},
{
"epoch": 93.01874595992243,
"grad_norm": 1.2151107788085938,
"learning_rate": 0.001,
"loss": 1.2586,
"step": 287800
},
{
"epoch": 93.05106658047835,
"grad_norm": 1.3145076036453247,
"learning_rate": 0.001,
"loss": 1.1907,
"step": 287900
},
{
"epoch": 93.08338720103426,
"grad_norm": 1.1841275691986084,
"learning_rate": 0.001,
"loss": 1.2155,
"step": 288000
},
{
"epoch": 93.11570782159018,
"grad_norm": 1.4306896924972534,
"learning_rate": 0.001,
"loss": 1.212,
"step": 288100
},
{
"epoch": 93.14802844214609,
"grad_norm": 1.5732502937316895,
"learning_rate": 0.001,
"loss": 1.2304,
"step": 288200
},
{
"epoch": 93.180349062702,
"grad_norm": 1.4531052112579346,
"learning_rate": 0.001,
"loss": 1.2174,
"step": 288300
},
{
"epoch": 93.21266968325791,
"grad_norm": 1.8546981811523438,
"learning_rate": 0.001,
"loss": 1.2203,
"step": 288400
},
{
"epoch": 93.24499030381384,
"grad_norm": 1.5319015979766846,
"learning_rate": 0.001,
"loss": 1.2282,
"step": 288500
},
{
"epoch": 93.27731092436974,
"grad_norm": 1.888697624206543,
"learning_rate": 0.001,
"loss": 1.2493,
"step": 288600
},
{
"epoch": 93.30963154492567,
"grad_norm": 1.4183104038238525,
"learning_rate": 0.001,
"loss": 1.2342,
"step": 288700
},
{
"epoch": 93.34195216548157,
"grad_norm": 1.1921762228012085,
"learning_rate": 0.001,
"loss": 1.2513,
"step": 288800
},
{
"epoch": 93.3742727860375,
"grad_norm": 2.578413724899292,
"learning_rate": 0.001,
"loss": 1.2586,
"step": 288900
},
{
"epoch": 93.4065934065934,
"grad_norm": 1.7396384477615356,
"learning_rate": 0.001,
"loss": 1.2548,
"step": 289000
},
{
"epoch": 93.43891402714932,
"grad_norm": 1.2900177240371704,
"learning_rate": 0.001,
"loss": 1.2492,
"step": 289100
},
{
"epoch": 93.47123464770523,
"grad_norm": 1.5431060791015625,
"learning_rate": 0.001,
"loss": 1.2553,
"step": 289200
},
{
"epoch": 93.50355526826115,
"grad_norm": 1.6249630451202393,
"learning_rate": 0.001,
"loss": 1.2581,
"step": 289300
},
{
"epoch": 93.53587588881706,
"grad_norm": 1.2563765048980713,
"learning_rate": 0.001,
"loss": 1.2642,
"step": 289400
},
{
"epoch": 93.56819650937298,
"grad_norm": 1.4387561082839966,
"learning_rate": 0.001,
"loss": 1.2686,
"step": 289500
},
{
"epoch": 93.60051712992889,
"grad_norm": 1.8408147096633911,
"learning_rate": 0.001,
"loss": 1.2701,
"step": 289600
},
{
"epoch": 93.63283775048481,
"grad_norm": 1.6275147199630737,
"learning_rate": 0.001,
"loss": 1.2577,
"step": 289700
},
{
"epoch": 93.66515837104072,
"grad_norm": 1.4617375135421753,
"learning_rate": 0.001,
"loss": 1.2648,
"step": 289800
},
{
"epoch": 93.69747899159664,
"grad_norm": 1.2824053764343262,
"learning_rate": 0.001,
"loss": 1.2661,
"step": 289900
},
{
"epoch": 93.72979961215255,
"grad_norm": 1.387819766998291,
"learning_rate": 0.001,
"loss": 1.289,
"step": 290000
},
{
"epoch": 93.76212023270847,
"grad_norm": 1.6011693477630615,
"learning_rate": 0.001,
"loss": 1.2842,
"step": 290100
},
{
"epoch": 93.79444085326438,
"grad_norm": 1.2590481042861938,
"learning_rate": 0.001,
"loss": 1.2912,
"step": 290200
},
{
"epoch": 93.8267614738203,
"grad_norm": 1.555876612663269,
"learning_rate": 0.001,
"loss": 1.2938,
"step": 290300
},
{
"epoch": 93.8590820943762,
"grad_norm": 1.9848129749298096,
"learning_rate": 0.001,
"loss": 1.2919,
"step": 290400
},
{
"epoch": 93.89140271493213,
"grad_norm": 1.3681098222732544,
"learning_rate": 0.001,
"loss": 1.3025,
"step": 290500
},
{
"epoch": 93.92372333548803,
"grad_norm": 1.2209434509277344,
"learning_rate": 0.001,
"loss": 1.3117,
"step": 290600
},
{
"epoch": 93.95604395604396,
"grad_norm": 2.183659076690674,
"learning_rate": 0.001,
"loss": 1.3009,
"step": 290700
},
{
"epoch": 93.98836457659988,
"grad_norm": 1.590404987335205,
"learning_rate": 0.001,
"loss": 1.2985,
"step": 290800
},
{
"epoch": 94.02068519715579,
"grad_norm": 1.3527745008468628,
"learning_rate": 0.001,
"loss": 1.2485,
"step": 290900
},
{
"epoch": 94.0530058177117,
"grad_norm": 1.115366816520691,
"learning_rate": 0.001,
"loss": 1.1863,
"step": 291000
},
{
"epoch": 94.08532643826761,
"grad_norm": 1.7399864196777344,
"learning_rate": 0.001,
"loss": 1.2096,
"step": 291100
},
{
"epoch": 94.11764705882354,
"grad_norm": 1.2375653982162476,
"learning_rate": 0.001,
"loss": 1.2216,
"step": 291200
},
{
"epoch": 94.14996767937944,
"grad_norm": 1.97869074344635,
"learning_rate": 0.001,
"loss": 1.1938,
"step": 291300
},
{
"epoch": 94.18228829993537,
"grad_norm": 1.4840346574783325,
"learning_rate": 0.001,
"loss": 1.2258,
"step": 291400
},
{
"epoch": 94.21460892049127,
"grad_norm": 1.7346757650375366,
"learning_rate": 0.001,
"loss": 1.2156,
"step": 291500
},
{
"epoch": 94.2469295410472,
"grad_norm": 1.629726767539978,
"learning_rate": 0.001,
"loss": 1.2284,
"step": 291600
},
{
"epoch": 94.2792501616031,
"grad_norm": 1.2896777391433716,
"learning_rate": 0.001,
"loss": 1.2242,
"step": 291700
},
{
"epoch": 94.31157078215902,
"grad_norm": 1.416365623474121,
"learning_rate": 0.001,
"loss": 1.2422,
"step": 291800
},
{
"epoch": 94.34389140271493,
"grad_norm": 1.3081306219100952,
"learning_rate": 0.001,
"loss": 1.221,
"step": 291900
},
{
"epoch": 94.37621202327085,
"grad_norm": 1.4403235912322998,
"learning_rate": 0.001,
"loss": 1.2429,
"step": 292000
},
{
"epoch": 94.40853264382676,
"grad_norm": 1.3355509042739868,
"learning_rate": 0.001,
"loss": 1.2335,
"step": 292100
},
{
"epoch": 94.44085326438268,
"grad_norm": 1.8607823848724365,
"learning_rate": 0.001,
"loss": 1.2437,
"step": 292200
},
{
"epoch": 94.47317388493859,
"grad_norm": 1.584852933883667,
"learning_rate": 0.001,
"loss": 1.2451,
"step": 292300
},
{
"epoch": 94.50549450549451,
"grad_norm": 1.5097578763961792,
"learning_rate": 0.001,
"loss": 1.2562,
"step": 292400
},
{
"epoch": 94.53781512605042,
"grad_norm": 1.6666420698165894,
"learning_rate": 0.001,
"loss": 1.2592,
"step": 292500
},
{
"epoch": 94.57013574660634,
"grad_norm": 1.3373353481292725,
"learning_rate": 0.001,
"loss": 1.2517,
"step": 292600
},
{
"epoch": 94.60245636716225,
"grad_norm": 1.1197723150253296,
"learning_rate": 0.001,
"loss": 1.2504,
"step": 292700
},
{
"epoch": 94.63477698771817,
"grad_norm": 1.3049250841140747,
"learning_rate": 0.001,
"loss": 1.27,
"step": 292800
},
{
"epoch": 94.66709760827408,
"grad_norm": 1.1198012828826904,
"learning_rate": 0.001,
"loss": 1.267,
"step": 292900
},
{
"epoch": 94.69941822883,
"grad_norm": 2.1118505001068115,
"learning_rate": 0.001,
"loss": 1.2917,
"step": 293000
},
{
"epoch": 94.7317388493859,
"grad_norm": 1.9926177263259888,
"learning_rate": 0.001,
"loss": 1.2819,
"step": 293100
},
{
"epoch": 94.76405946994183,
"grad_norm": 1.731961727142334,
"learning_rate": 0.001,
"loss": 1.2782,
"step": 293200
},
{
"epoch": 94.79638009049773,
"grad_norm": 1.3176889419555664,
"learning_rate": 0.001,
"loss": 1.2801,
"step": 293300
},
{
"epoch": 94.82870071105366,
"grad_norm": 1.6108160018920898,
"learning_rate": 0.001,
"loss": 1.302,
"step": 293400
},
{
"epoch": 94.86102133160956,
"grad_norm": 1.3358465433120728,
"learning_rate": 0.001,
"loss": 1.2877,
"step": 293500
},
{
"epoch": 94.89334195216549,
"grad_norm": 1.6573939323425293,
"learning_rate": 0.001,
"loss": 1.2933,
"step": 293600
},
{
"epoch": 94.9256625727214,
"grad_norm": 2.2432045936584473,
"learning_rate": 0.001,
"loss": 1.2933,
"step": 293700
},
{
"epoch": 94.95798319327731,
"grad_norm": 1.3388210535049438,
"learning_rate": 0.001,
"loss": 1.2973,
"step": 293800
},
{
"epoch": 94.99030381383322,
"grad_norm": 1.348464012145996,
"learning_rate": 0.001,
"loss": 1.2986,
"step": 293900
},
{
"epoch": 95.02262443438914,
"grad_norm": 1.73881995677948,
"learning_rate": 0.001,
"loss": 1.2369,
"step": 294000
},
{
"epoch": 95.05494505494505,
"grad_norm": 1.3996957540512085,
"learning_rate": 0.001,
"loss": 1.2076,
"step": 294100
},
{
"epoch": 95.08726567550097,
"grad_norm": 1.9472845792770386,
"learning_rate": 0.001,
"loss": 1.2159,
"step": 294200
},
{
"epoch": 95.11958629605688,
"grad_norm": 1.5158746242523193,
"learning_rate": 0.001,
"loss": 1.1974,
"step": 294300
},
{
"epoch": 95.1519069166128,
"grad_norm": 1.6360524892807007,
"learning_rate": 0.001,
"loss": 1.1959,
"step": 294400
},
{
"epoch": 95.18422753716871,
"grad_norm": 2.1060550212860107,
"learning_rate": 0.001,
"loss": 1.2141,
"step": 294500
},
{
"epoch": 95.21654815772463,
"grad_norm": 2.401296377182007,
"learning_rate": 0.001,
"loss": 1.2202,
"step": 294600
},
{
"epoch": 95.24886877828054,
"grad_norm": 1.393446445465088,
"learning_rate": 0.001,
"loss": 1.2225,
"step": 294700
},
{
"epoch": 95.28118939883646,
"grad_norm": 1.2871545553207397,
"learning_rate": 0.001,
"loss": 1.2283,
"step": 294800
},
{
"epoch": 95.31351001939237,
"grad_norm": 1.6274781227111816,
"learning_rate": 0.001,
"loss": 1.215,
"step": 294900
},
{
"epoch": 95.34583063994829,
"grad_norm": 1.3049452304840088,
"learning_rate": 0.001,
"loss": 1.2231,
"step": 295000
},
{
"epoch": 95.3781512605042,
"grad_norm": 1.4392836093902588,
"learning_rate": 0.001,
"loss": 1.2347,
"step": 295100
},
{
"epoch": 95.41047188106012,
"grad_norm": 1.6895877122879028,
"learning_rate": 0.001,
"loss": 1.2284,
"step": 295200
},
{
"epoch": 95.44279250161603,
"grad_norm": 1.4769349098205566,
"learning_rate": 0.001,
"loss": 1.238,
"step": 295300
},
{
"epoch": 95.47511312217195,
"grad_norm": 1.87929105758667,
"learning_rate": 0.001,
"loss": 1.2448,
"step": 295400
},
{
"epoch": 95.50743374272786,
"grad_norm": 1.3902554512023926,
"learning_rate": 0.001,
"loss": 1.2439,
"step": 295500
},
{
"epoch": 95.53975436328378,
"grad_norm": 1.49274480342865,
"learning_rate": 0.001,
"loss": 1.2534,
"step": 295600
},
{
"epoch": 95.57207498383968,
"grad_norm": 1.6464723348617554,
"learning_rate": 0.001,
"loss": 1.2545,
"step": 295700
},
{
"epoch": 95.6043956043956,
"grad_norm": 1.642978310585022,
"learning_rate": 0.001,
"loss": 1.2554,
"step": 295800
},
{
"epoch": 95.63671622495151,
"grad_norm": 1.5612927675247192,
"learning_rate": 0.001,
"loss": 1.2677,
"step": 295900
},
{
"epoch": 95.66903684550743,
"grad_norm": 1.1381276845932007,
"learning_rate": 0.001,
"loss": 1.2708,
"step": 296000
},
{
"epoch": 95.70135746606334,
"grad_norm": 1.2459452152252197,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 296100
},
{
"epoch": 95.73367808661926,
"grad_norm": 1.2951840162277222,
"learning_rate": 0.001,
"loss": 1.271,
"step": 296200
},
{
"epoch": 95.76599870717517,
"grad_norm": 1.3488235473632812,
"learning_rate": 0.001,
"loss": 1.2678,
"step": 296300
},
{
"epoch": 95.7983193277311,
"grad_norm": 1.4154523611068726,
"learning_rate": 0.001,
"loss": 1.2818,
"step": 296400
},
{
"epoch": 95.830639948287,
"grad_norm": 1.1226747035980225,
"learning_rate": 0.001,
"loss": 1.2769,
"step": 296500
},
{
"epoch": 95.86296056884292,
"grad_norm": 1.3886140584945679,
"learning_rate": 0.001,
"loss": 1.295,
"step": 296600
},
{
"epoch": 95.89528118939883,
"grad_norm": 2.038051128387451,
"learning_rate": 0.001,
"loss": 1.2677,
"step": 296700
},
{
"epoch": 95.92760180995475,
"grad_norm": 1.6668192148208618,
"learning_rate": 0.001,
"loss": 1.2819,
"step": 296800
},
{
"epoch": 95.95992243051066,
"grad_norm": 1.619035243988037,
"learning_rate": 0.001,
"loss": 1.2846,
"step": 296900
},
{
"epoch": 95.99224305106658,
"grad_norm": 1.4912338256835938,
"learning_rate": 0.001,
"loss": 1.2981,
"step": 297000
},
{
"epoch": 96.0245636716225,
"grad_norm": 1.6322486400604248,
"learning_rate": 0.001,
"loss": 1.2208,
"step": 297100
},
{
"epoch": 96.05688429217841,
"grad_norm": 1.1914606094360352,
"learning_rate": 0.001,
"loss": 1.18,
"step": 297200
},
{
"epoch": 96.08920491273433,
"grad_norm": 1.3940045833587646,
"learning_rate": 0.001,
"loss": 1.1991,
"step": 297300
},
{
"epoch": 96.12152553329024,
"grad_norm": 1.2746566534042358,
"learning_rate": 0.001,
"loss": 1.2011,
"step": 297400
},
{
"epoch": 96.15384615384616,
"grad_norm": 1.8106060028076172,
"learning_rate": 0.001,
"loss": 1.2108,
"step": 297500
},
{
"epoch": 96.18616677440207,
"grad_norm": 1.2734863758087158,
"learning_rate": 0.001,
"loss": 1.2121,
"step": 297600
},
{
"epoch": 96.21848739495799,
"grad_norm": 1.7641494274139404,
"learning_rate": 0.001,
"loss": 1.2023,
"step": 297700
},
{
"epoch": 96.2508080155139,
"grad_norm": 1.2450460195541382,
"learning_rate": 0.001,
"loss": 1.225,
"step": 297800
},
{
"epoch": 96.28312863606982,
"grad_norm": 1.7588828802108765,
"learning_rate": 0.001,
"loss": 1.2199,
"step": 297900
},
{
"epoch": 96.31544925662573,
"grad_norm": 1.5999172925949097,
"learning_rate": 0.001,
"loss": 1.2359,
"step": 298000
},
{
"epoch": 96.34776987718165,
"grad_norm": 1.4087074995040894,
"learning_rate": 0.001,
"loss": 1.2187,
"step": 298100
},
{
"epoch": 96.38009049773756,
"grad_norm": 1.8337302207946777,
"learning_rate": 0.001,
"loss": 1.2249,
"step": 298200
},
{
"epoch": 96.41241111829348,
"grad_norm": 1.6353654861450195,
"learning_rate": 0.001,
"loss": 1.2268,
"step": 298300
},
{
"epoch": 96.44473173884938,
"grad_norm": 1.2645622491836548,
"learning_rate": 0.001,
"loss": 1.2402,
"step": 298400
},
{
"epoch": 96.4770523594053,
"grad_norm": 1.8545764684677124,
"learning_rate": 0.001,
"loss": 1.2355,
"step": 298500
},
{
"epoch": 96.50937297996121,
"grad_norm": 1.5485554933547974,
"learning_rate": 0.001,
"loss": 1.2414,
"step": 298600
},
{
"epoch": 96.54169360051714,
"grad_norm": 1.0770783424377441,
"learning_rate": 0.001,
"loss": 1.2577,
"step": 298700
},
{
"epoch": 96.57401422107304,
"grad_norm": 1.9056742191314697,
"learning_rate": 0.001,
"loss": 1.2308,
"step": 298800
},
{
"epoch": 96.60633484162896,
"grad_norm": 1.3413792848587036,
"learning_rate": 0.001,
"loss": 1.2474,
"step": 298900
},
{
"epoch": 96.63865546218487,
"grad_norm": 1.6371670961380005,
"learning_rate": 0.001,
"loss": 1.2583,
"step": 299000
},
{
"epoch": 96.6709760827408,
"grad_norm": 1.3188087940216064,
"learning_rate": 0.001,
"loss": 1.2615,
"step": 299100
},
{
"epoch": 96.7032967032967,
"grad_norm": 1.1330523490905762,
"learning_rate": 0.001,
"loss": 1.2591,
"step": 299200
},
{
"epoch": 96.73561732385262,
"grad_norm": 1.6628353595733643,
"learning_rate": 0.001,
"loss": 1.2597,
"step": 299300
},
{
"epoch": 96.76793794440853,
"grad_norm": 1.4252923727035522,
"learning_rate": 0.001,
"loss": 1.2862,
"step": 299400
},
{
"epoch": 96.80025856496445,
"grad_norm": 1.1442095041275024,
"learning_rate": 0.001,
"loss": 1.2666,
"step": 299500
},
{
"epoch": 96.83257918552036,
"grad_norm": 1.5159488916397095,
"learning_rate": 0.001,
"loss": 1.2795,
"step": 299600
},
{
"epoch": 96.86489980607628,
"grad_norm": 1.2380590438842773,
"learning_rate": 0.001,
"loss": 1.2688,
"step": 299700
},
{
"epoch": 96.89722042663219,
"grad_norm": 1.3822698593139648,
"learning_rate": 0.001,
"loss": 1.2589,
"step": 299800
},
{
"epoch": 96.92954104718811,
"grad_norm": 1.676533579826355,
"learning_rate": 0.001,
"loss": 1.2799,
"step": 299900
},
{
"epoch": 96.96186166774402,
"grad_norm": 1.413086175918579,
"learning_rate": 0.001,
"loss": 1.2872,
"step": 300000
},
{
"epoch": 96.99418228829994,
"grad_norm": 1.2737962007522583,
"learning_rate": 0.001,
"loss": 1.2724,
"step": 300100
},
{
"epoch": 97.02650290885585,
"grad_norm": 1.336970567703247,
"learning_rate": 0.001,
"loss": 1.21,
"step": 300200
},
{
"epoch": 97.05882352941177,
"grad_norm": 1.431416392326355,
"learning_rate": 0.001,
"loss": 1.1857,
"step": 300300
},
{
"epoch": 97.09114414996768,
"grad_norm": 1.4808313846588135,
"learning_rate": 0.001,
"loss": 1.1833,
"step": 300400
},
{
"epoch": 97.1234647705236,
"grad_norm": 1.8020944595336914,
"learning_rate": 0.001,
"loss": 1.1712,
"step": 300500
},
{
"epoch": 97.1557853910795,
"grad_norm": 2.30206036567688,
"learning_rate": 0.001,
"loss": 1.1915,
"step": 300600
},
{
"epoch": 97.18810601163543,
"grad_norm": 2.694793939590454,
"learning_rate": 0.001,
"loss": 1.2045,
"step": 300700
},
{
"epoch": 97.22042663219133,
"grad_norm": 1.1539241075515747,
"learning_rate": 0.001,
"loss": 1.2206,
"step": 300800
},
{
"epoch": 97.25274725274726,
"grad_norm": 1.5353020429611206,
"learning_rate": 0.001,
"loss": 1.2012,
"step": 300900
},
{
"epoch": 97.28506787330316,
"grad_norm": 2.7079131603240967,
"learning_rate": 0.001,
"loss": 1.2227,
"step": 301000
},
{
"epoch": 97.31738849385908,
"grad_norm": 1.341162085533142,
"learning_rate": 0.001,
"loss": 1.2148,
"step": 301100
},
{
"epoch": 97.34970911441499,
"grad_norm": 1.5532795190811157,
"learning_rate": 0.001,
"loss": 1.2105,
"step": 301200
},
{
"epoch": 97.38202973497091,
"grad_norm": 1.2931928634643555,
"learning_rate": 0.001,
"loss": 1.221,
"step": 301300
},
{
"epoch": 97.41435035552682,
"grad_norm": 1.5295546054840088,
"learning_rate": 0.001,
"loss": 1.2319,
"step": 301400
},
{
"epoch": 97.44667097608274,
"grad_norm": 1.7937957048416138,
"learning_rate": 0.001,
"loss": 1.2374,
"step": 301500
},
{
"epoch": 97.47899159663865,
"grad_norm": 1.9637655019760132,
"learning_rate": 0.001,
"loss": 1.2392,
"step": 301600
},
{
"epoch": 97.51131221719457,
"grad_norm": 2.6429688930511475,
"learning_rate": 0.001,
"loss": 1.2481,
"step": 301700
},
{
"epoch": 97.54363283775048,
"grad_norm": 1.6325021982192993,
"learning_rate": 0.001,
"loss": 1.2325,
"step": 301800
},
{
"epoch": 97.5759534583064,
"grad_norm": 1.498071551322937,
"learning_rate": 0.001,
"loss": 1.2497,
"step": 301900
},
{
"epoch": 97.60827407886231,
"grad_norm": 1.7910757064819336,
"learning_rate": 0.001,
"loss": 1.2404,
"step": 302000
},
{
"epoch": 97.64059469941823,
"grad_norm": 1.675830602645874,
"learning_rate": 0.001,
"loss": 1.2417,
"step": 302100
},
{
"epoch": 97.67291531997414,
"grad_norm": 1.691313624382019,
"learning_rate": 0.001,
"loss": 1.2463,
"step": 302200
},
{
"epoch": 97.70523594053006,
"grad_norm": 1.329566240310669,
"learning_rate": 0.001,
"loss": 1.272,
"step": 302300
},
{
"epoch": 97.73755656108597,
"grad_norm": 1.98543119430542,
"learning_rate": 0.001,
"loss": 1.2633,
"step": 302400
},
{
"epoch": 97.76987718164189,
"grad_norm": 1.4446206092834473,
"learning_rate": 0.001,
"loss": 1.2629,
"step": 302500
},
{
"epoch": 97.8021978021978,
"grad_norm": 1.3561028242111206,
"learning_rate": 0.001,
"loss": 1.2556,
"step": 302600
},
{
"epoch": 97.83451842275372,
"grad_norm": 1.306576132774353,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 302700
},
{
"epoch": 97.86683904330962,
"grad_norm": 2.1355342864990234,
"learning_rate": 0.001,
"loss": 1.2636,
"step": 302800
},
{
"epoch": 97.89915966386555,
"grad_norm": 1.790517807006836,
"learning_rate": 0.001,
"loss": 1.2809,
"step": 302900
},
{
"epoch": 97.93148028442145,
"grad_norm": 1.942603588104248,
"learning_rate": 0.001,
"loss": 1.2834,
"step": 303000
},
{
"epoch": 97.96380090497738,
"grad_norm": 1.5469906330108643,
"learning_rate": 0.001,
"loss": 1.2697,
"step": 303100
},
{
"epoch": 97.99612152553328,
"grad_norm": 1.294941782951355,
"learning_rate": 0.001,
"loss": 1.2827,
"step": 303200
},
{
"epoch": 98.0284421460892,
"grad_norm": 1.4144631624221802,
"learning_rate": 0.001,
"loss": 1.1949,
"step": 303300
},
{
"epoch": 98.06076276664513,
"grad_norm": 1.364519476890564,
"learning_rate": 0.001,
"loss": 1.1738,
"step": 303400
},
{
"epoch": 98.09308338720103,
"grad_norm": 1.5441681146621704,
"learning_rate": 0.001,
"loss": 1.1732,
"step": 303500
},
{
"epoch": 98.12540400775696,
"grad_norm": 1.8025354146957397,
"learning_rate": 0.001,
"loss": 1.1845,
"step": 303600
},
{
"epoch": 98.15772462831286,
"grad_norm": 1.320715069770813,
"learning_rate": 0.001,
"loss": 1.1865,
"step": 303700
},
{
"epoch": 98.19004524886878,
"grad_norm": 2.114084482192993,
"learning_rate": 0.001,
"loss": 1.1936,
"step": 303800
},
{
"epoch": 98.22236586942469,
"grad_norm": 1.6143896579742432,
"learning_rate": 0.001,
"loss": 1.1968,
"step": 303900
},
{
"epoch": 98.25468648998061,
"grad_norm": 1.443982481956482,
"learning_rate": 0.001,
"loss": 1.1783,
"step": 304000
},
{
"epoch": 98.28700711053652,
"grad_norm": 1.64999258518219,
"learning_rate": 0.001,
"loss": 1.2074,
"step": 304100
},
{
"epoch": 98.31932773109244,
"grad_norm": 1.5300345420837402,
"learning_rate": 0.001,
"loss": 1.2132,
"step": 304200
},
{
"epoch": 98.35164835164835,
"grad_norm": 1.6809816360473633,
"learning_rate": 0.001,
"loss": 1.2259,
"step": 304300
},
{
"epoch": 98.38396897220427,
"grad_norm": 1.9328725337982178,
"learning_rate": 0.001,
"loss": 1.2312,
"step": 304400
},
{
"epoch": 98.41628959276018,
"grad_norm": 1.5429282188415527,
"learning_rate": 0.001,
"loss": 1.2226,
"step": 304500
},
{
"epoch": 98.4486102133161,
"grad_norm": 1.4246468544006348,
"learning_rate": 0.001,
"loss": 1.2442,
"step": 304600
},
{
"epoch": 98.48093083387201,
"grad_norm": 1.5565303564071655,
"learning_rate": 0.001,
"loss": 1.2259,
"step": 304700
},
{
"epoch": 98.51325145442793,
"grad_norm": 2.1429333686828613,
"learning_rate": 0.001,
"loss": 1.2475,
"step": 304800
},
{
"epoch": 98.54557207498384,
"grad_norm": 2.1533963680267334,
"learning_rate": 0.001,
"loss": 1.2391,
"step": 304900
},
{
"epoch": 98.57789269553976,
"grad_norm": 1.9747862815856934,
"learning_rate": 0.001,
"loss": 1.2367,
"step": 305000
},
{
"epoch": 98.61021331609567,
"grad_norm": 1.6359381675720215,
"learning_rate": 0.001,
"loss": 1.2427,
"step": 305100
},
{
"epoch": 98.64253393665159,
"grad_norm": 1.8843120336532593,
"learning_rate": 0.001,
"loss": 1.2373,
"step": 305200
},
{
"epoch": 98.6748545572075,
"grad_norm": 1.4711463451385498,
"learning_rate": 0.001,
"loss": 1.2469,
"step": 305300
},
{
"epoch": 98.70717517776342,
"grad_norm": 1.9563316106796265,
"learning_rate": 0.001,
"loss": 1.2345,
"step": 305400
},
{
"epoch": 98.73949579831933,
"grad_norm": 1.6488176584243774,
"learning_rate": 0.001,
"loss": 1.2563,
"step": 305500
},
{
"epoch": 98.77181641887525,
"grad_norm": 1.4608503580093384,
"learning_rate": 0.001,
"loss": 1.2748,
"step": 305600
},
{
"epoch": 98.80413703943115,
"grad_norm": 1.7756547927856445,
"learning_rate": 0.001,
"loss": 1.2494,
"step": 305700
},
{
"epoch": 98.83645765998708,
"grad_norm": 2.4544506072998047,
"learning_rate": 0.001,
"loss": 1.2582,
"step": 305800
},
{
"epoch": 98.86877828054298,
"grad_norm": 2.120619058609009,
"learning_rate": 0.001,
"loss": 1.2558,
"step": 305900
},
{
"epoch": 98.9010989010989,
"grad_norm": 1.5788395404815674,
"learning_rate": 0.001,
"loss": 1.2499,
"step": 306000
},
{
"epoch": 98.93341952165481,
"grad_norm": 1.230873465538025,
"learning_rate": 0.001,
"loss": 1.2738,
"step": 306100
},
{
"epoch": 98.96574014221073,
"grad_norm": 1.6700103282928467,
"learning_rate": 0.001,
"loss": 1.2731,
"step": 306200
},
{
"epoch": 98.99806076276664,
"grad_norm": 1.2944440841674805,
"learning_rate": 0.001,
"loss": 1.2669,
"step": 306300
},
{
"epoch": 99.03038138332256,
"grad_norm": 1.3641657829284668,
"learning_rate": 0.001,
"loss": 1.1849,
"step": 306400
},
{
"epoch": 99.06270200387847,
"grad_norm": 2.0977532863616943,
"learning_rate": 0.001,
"loss": 1.168,
"step": 306500
},
{
"epoch": 99.09502262443439,
"grad_norm": 1.7131783962249756,
"learning_rate": 0.001,
"loss": 1.1785,
"step": 306600
},
{
"epoch": 99.1273432449903,
"grad_norm": 2.3469855785369873,
"learning_rate": 0.001,
"loss": 1.1632,
"step": 306700
},
{
"epoch": 99.15966386554622,
"grad_norm": 1.6222341060638428,
"learning_rate": 0.001,
"loss": 1.1836,
"step": 306800
},
{
"epoch": 99.19198448610213,
"grad_norm": 1.9087047576904297,
"learning_rate": 0.001,
"loss": 1.2028,
"step": 306900
},
{
"epoch": 99.22430510665805,
"grad_norm": 1.636622428894043,
"learning_rate": 0.001,
"loss": 1.1924,
"step": 307000
},
{
"epoch": 99.25662572721396,
"grad_norm": 1.309688925743103,
"learning_rate": 0.001,
"loss": 1.1942,
"step": 307100
},
{
"epoch": 99.28894634776988,
"grad_norm": 2.204533338546753,
"learning_rate": 0.001,
"loss": 1.2124,
"step": 307200
},
{
"epoch": 99.32126696832579,
"grad_norm": 1.620397686958313,
"learning_rate": 0.001,
"loss": 1.221,
"step": 307300
},
{
"epoch": 99.35358758888171,
"grad_norm": 1.5576107501983643,
"learning_rate": 0.001,
"loss": 1.2064,
"step": 307400
},
{
"epoch": 99.38590820943762,
"grad_norm": 1.6092556715011597,
"learning_rate": 0.001,
"loss": 1.2254,
"step": 307500
},
{
"epoch": 99.41822882999354,
"grad_norm": 1.7601780891418457,
"learning_rate": 0.001,
"loss": 1.2083,
"step": 307600
},
{
"epoch": 99.45054945054945,
"grad_norm": 1.8288476467132568,
"learning_rate": 0.001,
"loss": 1.218,
"step": 307700
},
{
"epoch": 99.48287007110537,
"grad_norm": 1.7833744287490845,
"learning_rate": 0.001,
"loss": 1.2525,
"step": 307800
},
{
"epoch": 99.51519069166127,
"grad_norm": 1.4205470085144043,
"learning_rate": 0.001,
"loss": 1.2277,
"step": 307900
},
{
"epoch": 99.5475113122172,
"grad_norm": 1.9912147521972656,
"learning_rate": 0.001,
"loss": 1.2277,
"step": 308000
},
{
"epoch": 99.5798319327731,
"grad_norm": 1.6823841333389282,
"learning_rate": 0.001,
"loss": 1.228,
"step": 308100
},
{
"epoch": 99.61215255332903,
"grad_norm": 1.8055593967437744,
"learning_rate": 0.001,
"loss": 1.221,
"step": 308200
},
{
"epoch": 99.64447317388493,
"grad_norm": 1.5161564350128174,
"learning_rate": 0.001,
"loss": 1.2135,
"step": 308300
},
{
"epoch": 99.67679379444085,
"grad_norm": 1.8130974769592285,
"learning_rate": 0.001,
"loss": 1.2592,
"step": 308400
},
{
"epoch": 99.70911441499676,
"grad_norm": 1.6617481708526611,
"learning_rate": 0.001,
"loss": 1.2478,
"step": 308500
},
{
"epoch": 99.74143503555268,
"grad_norm": 1.4916375875473022,
"learning_rate": 0.001,
"loss": 1.2515,
"step": 308600
},
{
"epoch": 99.77375565610859,
"grad_norm": 1.6607589721679688,
"learning_rate": 0.001,
"loss": 1.2388,
"step": 308700
},
{
"epoch": 99.80607627666451,
"grad_norm": 1.589477300643921,
"learning_rate": 0.001,
"loss": 1.2579,
"step": 308800
},
{
"epoch": 99.83839689722042,
"grad_norm": 1.507033348083496,
"learning_rate": 0.001,
"loss": 1.2585,
"step": 308900
},
{
"epoch": 99.87071751777634,
"grad_norm": 1.7399075031280518,
"learning_rate": 0.001,
"loss": 1.2578,
"step": 309000
},
{
"epoch": 99.90303813833225,
"grad_norm": 1.6721640825271606,
"learning_rate": 0.001,
"loss": 1.2621,
"step": 309100
},
{
"epoch": 99.93535875888817,
"grad_norm": 2.365220308303833,
"learning_rate": 0.001,
"loss": 1.252,
"step": 309200
},
{
"epoch": 99.96767937944408,
"grad_norm": 2.6448214054107666,
"learning_rate": 0.001,
"loss": 1.249,
"step": 309300
},
{
"epoch": 100.0,
"grad_norm": 2.781751871109009,
"learning_rate": 0.001,
"loss": 1.2299,
"step": 309400
}
],
"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": 6.973099854336e+17,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}