Safetensors
English
bert
celeberty-small / trainer_state.json
davidmezzetti's picture
Add model
1800697
Raw
History Blame Contribute Delete
182 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 15.0,
"eval_steps": 500,
"global_step": 489135,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.015333190223557914,
"grad_norm": 0.5902125239372253,
"learning_rate": 0.0001996,
"loss": 7.87609912109375,
"step": 500
},
{
"epoch": 0.030666380447115827,
"grad_norm": 1.1883018016815186,
"learning_rate": 0.0003996,
"loss": 6.7239404296875,
"step": 1000
},
{
"epoch": 0.04599957067067374,
"grad_norm": 0.9555564522743225,
"learning_rate": 0.0005996,
"loss": 6.5268662109375,
"step": 1500
},
{
"epoch": 0.061332760894231654,
"grad_norm": 1.0314162969589233,
"learning_rate": 0.0007996,
"loss": 6.4513515625,
"step": 2000
},
{
"epoch": 0.07666595111778957,
"grad_norm": 0.6203042268753052,
"learning_rate": 0.0009996,
"loss": 6.4246337890625,
"step": 2500
},
{
"epoch": 0.09199914134134748,
"grad_norm": 1.0112814903259277,
"learning_rate": 0.0009989745908124158,
"loss": 6.40785888671875,
"step": 3000
},
{
"epoch": 0.10733233156490539,
"grad_norm": 0.8085325360298157,
"learning_rate": 0.0009979471266966002,
"loss": 6.38309375,
"step": 3500
},
{
"epoch": 0.12266552178846331,
"grad_norm": 1.1642123460769653,
"learning_rate": 0.0009969196625807843,
"loss": 6.35940771484375,
"step": 4000
},
{
"epoch": 0.1379987120120212,
"grad_norm": 0.9080906510353088,
"learning_rate": 0.0009958921984649686,
"loss": 6.311767578125,
"step": 4500
},
{
"epoch": 0.15333190223557913,
"grad_norm": 0.9173153638839722,
"learning_rate": 0.000994864734349153,
"loss": 6.1276005859375,
"step": 5000
},
{
"epoch": 0.16866509245913705,
"grad_norm": 1.2167878150939941,
"learning_rate": 0.0009938372702333372,
"loss": 5.86752001953125,
"step": 5500
},
{
"epoch": 0.18399828268269497,
"grad_norm": 1.0259867906570435,
"learning_rate": 0.0009928098061175214,
"loss": 5.35286865234375,
"step": 6000
},
{
"epoch": 0.1993314729062529,
"grad_norm": 1.083961844444275,
"learning_rate": 0.0009917823420017057,
"loss": 4.63032470703125,
"step": 6500
},
{
"epoch": 0.21466466312981078,
"grad_norm": 1.1050889492034912,
"learning_rate": 0.0009907548778858898,
"loss": 4.084828857421875,
"step": 7000
},
{
"epoch": 0.2299978533533687,
"grad_norm": 1.0633660554885864,
"learning_rate": 0.0009897274137700741,
"loss": 3.82325732421875,
"step": 7500
},
{
"epoch": 0.24533104357692662,
"grad_norm": 1.0864254236221313,
"learning_rate": 0.0009886999496542584,
"loss": 3.660906005859375,
"step": 8000
},
{
"epoch": 0.26066423380048453,
"grad_norm": 0.9320037364959717,
"learning_rate": 0.0009876724855384426,
"loss": 3.547222900390625,
"step": 8500
},
{
"epoch": 0.2759974240240424,
"grad_norm": 1.0103647708892822,
"learning_rate": 0.0009866450214226269,
"loss": 3.4493798828125,
"step": 9000
},
{
"epoch": 0.2913306142476004,
"grad_norm": 0.9672038555145264,
"learning_rate": 0.0009856175573068112,
"loss": 3.37892041015625,
"step": 9500
},
{
"epoch": 0.30666380447115826,
"grad_norm": 0.9260510206222534,
"learning_rate": 0.0009845900931909953,
"loss": 3.328186767578125,
"step": 10000
},
{
"epoch": 0.3219969946947162,
"grad_norm": 0.9813107252120972,
"learning_rate": 0.0009835626290751796,
"loss": 3.277389892578125,
"step": 10500
},
{
"epoch": 0.3373301849182741,
"grad_norm": 0.8876714706420898,
"learning_rate": 0.0009825351649593638,
"loss": 3.23009228515625,
"step": 11000
},
{
"epoch": 0.352663375141832,
"grad_norm": 0.919625997543335,
"learning_rate": 0.000981507700843548,
"loss": 3.189221435546875,
"step": 11500
},
{
"epoch": 0.36799656536538994,
"grad_norm": 0.8726804852485657,
"learning_rate": 0.0009804802367277324,
"loss": 3.165483642578125,
"step": 12000
},
{
"epoch": 0.38332975558894783,
"grad_norm": 0.9291759133338928,
"learning_rate": 0.0009794527726119165,
"loss": 3.137114013671875,
"step": 12500
},
{
"epoch": 0.3986629458125058,
"grad_norm": 0.9254608750343323,
"learning_rate": 0.0009784253084961008,
"loss": 3.10589501953125,
"step": 13000
},
{
"epoch": 0.41399613603606367,
"grad_norm": 0.9639304876327515,
"learning_rate": 0.000977397844380285,
"loss": 3.076835205078125,
"step": 13500
},
{
"epoch": 0.42932932625962156,
"grad_norm": 1.0396133661270142,
"learning_rate": 0.0009763703802644694,
"loss": 3.059826416015625,
"step": 14000
},
{
"epoch": 0.4446625164831795,
"grad_norm": 0.8993285298347473,
"learning_rate": 0.0009753429161486536,
"loss": 3.04440234375,
"step": 14500
},
{
"epoch": 0.4599957067067374,
"grad_norm": 0.9317373037338257,
"learning_rate": 0.0009743154520328378,
"loss": 3.02583447265625,
"step": 15000
},
{
"epoch": 0.47532889693029534,
"grad_norm": 0.9149473905563354,
"learning_rate": 0.000973287987917022,
"loss": 3.011250244140625,
"step": 15500
},
{
"epoch": 0.49066208715385323,
"grad_norm": 0.9477624297142029,
"learning_rate": 0.0009722605238012063,
"loss": 2.99285009765625,
"step": 16000
},
{
"epoch": 0.5059952773774111,
"grad_norm": 0.9451053142547607,
"learning_rate": 0.0009712330596853906,
"loss": 2.979430908203125,
"step": 16500
},
{
"epoch": 0.5213284676009691,
"grad_norm": 1.0122050046920776,
"learning_rate": 0.0009702055955695748,
"loss": 2.967337890625,
"step": 17000
},
{
"epoch": 0.536661657824527,
"grad_norm": 0.9523458480834961,
"learning_rate": 0.000969178131453759,
"loss": 2.94799755859375,
"step": 17500
},
{
"epoch": 0.5519948480480849,
"grad_norm": 0.9038415551185608,
"learning_rate": 0.0009681506673379432,
"loss": 2.92466845703125,
"step": 18000
},
{
"epoch": 0.5673280382716428,
"grad_norm": 0.8989875912666321,
"learning_rate": 0.0009671232032221275,
"loss": 2.9121552734375,
"step": 18500
},
{
"epoch": 0.5826612284952007,
"grad_norm": 0.9237350821495056,
"learning_rate": 0.0009660957391063118,
"loss": 2.903147216796875,
"step": 19000
},
{
"epoch": 0.5979944187187586,
"grad_norm": 0.8596885204315186,
"learning_rate": 0.000965068274990496,
"loss": 2.87883447265625,
"step": 19500
},
{
"epoch": 0.6133276089423165,
"grad_norm": 0.9314042329788208,
"learning_rate": 0.0009640408108746802,
"loss": 2.916532470703125,
"step": 20000
},
{
"epoch": 0.6286607991658745,
"grad_norm": 0.8612852692604065,
"learning_rate": 0.0009630133467588644,
"loss": 2.8758203125,
"step": 20500
},
{
"epoch": 0.6439939893894324,
"grad_norm": 0.9163721799850464,
"learning_rate": 0.0009619858826430488,
"loss": 2.8613203125,
"step": 21000
},
{
"epoch": 0.6593271796129903,
"grad_norm": 0.9759580492973328,
"learning_rate": 0.000960958418527233,
"loss": 2.850343505859375,
"step": 21500
},
{
"epoch": 0.6746603698365482,
"grad_norm": 0.9352011680603027,
"learning_rate": 0.0009599309544114172,
"loss": 2.846108642578125,
"step": 22000
},
{
"epoch": 0.6899935600601061,
"grad_norm": 0.8893311619758606,
"learning_rate": 0.0009589034902956014,
"loss": 2.835816650390625,
"step": 22500
},
{
"epoch": 0.705326750283664,
"grad_norm": 0.8881946206092834,
"learning_rate": 0.0009578760261797856,
"loss": 2.82673779296875,
"step": 23000
},
{
"epoch": 0.7206599405072219,
"grad_norm": 0.9216518402099609,
"learning_rate": 0.00095684856206397,
"loss": 2.80927490234375,
"step": 23500
},
{
"epoch": 0.7359931307307799,
"grad_norm": 0.9925842881202698,
"learning_rate": 0.0009558210979481542,
"loss": 2.80426953125,
"step": 24000
},
{
"epoch": 0.7513263209543377,
"grad_norm": 1.0029386281967163,
"learning_rate": 0.0009547936338323384,
"loss": 2.797843505859375,
"step": 24500
},
{
"epoch": 0.7666595111778957,
"grad_norm": 0.9868532419204712,
"learning_rate": 0.0009537661697165226,
"loss": 2.79123291015625,
"step": 25000
},
{
"epoch": 0.7819927014014536,
"grad_norm": 3.1048872470855713,
"learning_rate": 0.0009527387056007069,
"loss": 2.786656982421875,
"step": 25500
},
{
"epoch": 0.7973258916250116,
"grad_norm": 1.071280837059021,
"learning_rate": 0.0009517112414848912,
"loss": 2.825472412109375,
"step": 26000
},
{
"epoch": 0.8126590818485694,
"grad_norm": 1.0296210050582886,
"learning_rate": 0.0009506837773690754,
"loss": 2.791162353515625,
"step": 26500
},
{
"epoch": 0.8279922720721273,
"grad_norm": 0.9268739223480225,
"learning_rate": 0.0009496563132532596,
"loss": 2.768993896484375,
"step": 27000
},
{
"epoch": 0.8433254622956853,
"grad_norm": 0.9772447943687439,
"learning_rate": 0.0009486288491374438,
"loss": 2.7601396484375,
"step": 27500
},
{
"epoch": 0.8586586525192431,
"grad_norm": 0.9141217470169067,
"learning_rate": 0.0009476013850216281,
"loss": 2.753976318359375,
"step": 28000
},
{
"epoch": 0.8739918427428011,
"grad_norm": 0.9566735029220581,
"learning_rate": 0.0009465739209058124,
"loss": 2.769777099609375,
"step": 28500
},
{
"epoch": 0.889325032966359,
"grad_norm": 1.049375057220459,
"learning_rate": 0.0009455464567899966,
"loss": 2.7432177734375,
"step": 29000
},
{
"epoch": 0.9046582231899168,
"grad_norm": 0.8959460258483887,
"learning_rate": 0.0009445189926741808,
"loss": 2.738292724609375,
"step": 29500
},
{
"epoch": 0.9199914134134748,
"grad_norm": 0.9436110258102417,
"learning_rate": 0.000943491528558365,
"loss": 2.735663330078125,
"step": 30000
},
{
"epoch": 0.9353246036370327,
"grad_norm": 0.9505304098129272,
"learning_rate": 0.0009424640644425494,
"loss": 2.730620361328125,
"step": 30500
},
{
"epoch": 0.9506577938605907,
"grad_norm": 1.0711055994033813,
"learning_rate": 0.0009414366003267337,
"loss": 2.733042236328125,
"step": 31000
},
{
"epoch": 0.9659909840841485,
"grad_norm": 1.0136299133300781,
"learning_rate": 0.0009404091362109179,
"loss": 2.717471923828125,
"step": 31500
},
{
"epoch": 0.9813241743077065,
"grad_norm": 1.0794495344161987,
"learning_rate": 0.0009393816720951021,
"loss": 2.711216796875,
"step": 32000
},
{
"epoch": 0.9966573645312644,
"grad_norm": 0.9262813329696655,
"learning_rate": 0.0009383542079792864,
"loss": 2.7045615234375,
"step": 32500
},
{
"epoch": 1.0119905547548222,
"grad_norm": 1.5266720056533813,
"learning_rate": 0.0009373267438634706,
"loss": 2.70400634765625,
"step": 33000
},
{
"epoch": 1.0273237449783803,
"grad_norm": 1.0034760236740112,
"learning_rate": 0.0009362992797476549,
"loss": 2.70302392578125,
"step": 33500
},
{
"epoch": 1.0426569352019381,
"grad_norm": 0.8884544968605042,
"learning_rate": 0.0009352718156318391,
"loss": 2.685266357421875,
"step": 34000
},
{
"epoch": 1.057990125425496,
"grad_norm": 1.046505331993103,
"learning_rate": 0.0009342443515160233,
"loss": 2.676602294921875,
"step": 34500
},
{
"epoch": 1.073323315649054,
"grad_norm": 0.9898195862770081,
"learning_rate": 0.0009332168874002076,
"loss": 2.67285205078125,
"step": 35000
},
{
"epoch": 1.0886565058726119,
"grad_norm": 0.9700272083282471,
"learning_rate": 0.0009321894232843918,
"loss": 2.666143310546875,
"step": 35500
},
{
"epoch": 1.1039896960961697,
"grad_norm": 1.059637427330017,
"learning_rate": 0.0009311619591685761,
"loss": 2.66486669921875,
"step": 36000
},
{
"epoch": 1.1193228863197278,
"grad_norm": 0.9528005123138428,
"learning_rate": 0.0009301344950527603,
"loss": 2.660676513671875,
"step": 36500
},
{
"epoch": 1.1346560765432856,
"grad_norm": 0.9249218106269836,
"learning_rate": 0.0009291070309369446,
"loss": 2.65544287109375,
"step": 37000
},
{
"epoch": 1.1499892667668434,
"grad_norm": 1.242437720298767,
"learning_rate": 0.0009280795668211288,
"loss": 2.71104638671875,
"step": 37500
},
{
"epoch": 1.1653224569904015,
"grad_norm": 1.292757272720337,
"learning_rate": 0.000927052102705313,
"loss": 2.6573349609375,
"step": 38000
},
{
"epoch": 1.1806556472139593,
"grad_norm": 1.1269389390945435,
"learning_rate": 0.0009260246385894973,
"loss": 2.6435595703125,
"step": 38500
},
{
"epoch": 1.1959888374375172,
"grad_norm": 1.0906015634536743,
"learning_rate": 0.0009249971744736815,
"loss": 2.640878662109375,
"step": 39000
},
{
"epoch": 1.2113220276610752,
"grad_norm": 0.9825807809829712,
"learning_rate": 0.0009239697103578658,
"loss": 2.648980712890625,
"step": 39500
},
{
"epoch": 1.226655217884633,
"grad_norm": 0.9842659831047058,
"learning_rate": 0.00092294224624205,
"loss": 2.6298466796875,
"step": 40000
},
{
"epoch": 1.2419884081081909,
"grad_norm": 0.963578462600708,
"learning_rate": 0.0009219147821262342,
"loss": 2.630078125,
"step": 40500
},
{
"epoch": 1.257321598331749,
"grad_norm": 1.0084607601165771,
"learning_rate": 0.0009208873180104185,
"loss": 2.62867041015625,
"step": 41000
},
{
"epoch": 1.2726547885553068,
"grad_norm": 0.9019485712051392,
"learning_rate": 0.0009198598538946028,
"loss": 2.624818115234375,
"step": 41500
},
{
"epoch": 1.2879879787788648,
"grad_norm": 1.0349541902542114,
"learning_rate": 0.000918832389778787,
"loss": 2.61492529296875,
"step": 42000
},
{
"epoch": 1.3033211690024227,
"grad_norm": 1.0910366773605347,
"learning_rate": 0.0009178049256629712,
"loss": 2.611399169921875,
"step": 42500
},
{
"epoch": 1.3186543592259805,
"grad_norm": 0.955689013004303,
"learning_rate": 0.0009167774615471554,
"loss": 2.603817626953125,
"step": 43000
},
{
"epoch": 1.3339875494495383,
"grad_norm": 1.065265417098999,
"learning_rate": 0.0009157499974313397,
"loss": 2.600803466796875,
"step": 43500
},
{
"epoch": 1.3493207396730964,
"grad_norm": 1.3007278442382812,
"learning_rate": 0.000914722533315524,
"loss": 2.605083251953125,
"step": 44000
},
{
"epoch": 1.3646539298966542,
"grad_norm": 1.0439703464508057,
"learning_rate": 0.0009136950691997082,
"loss": 2.608570556640625,
"step": 44500
},
{
"epoch": 1.3799871201202123,
"grad_norm": 1.08986234664917,
"learning_rate": 0.0009126676050838924,
"loss": 2.60299169921875,
"step": 45000
},
{
"epoch": 1.3953203103437701,
"grad_norm": 0.9815884232521057,
"learning_rate": 0.0009116401409680766,
"loss": 2.59692333984375,
"step": 45500
},
{
"epoch": 1.410653500567328,
"grad_norm": 0.990138053894043,
"learning_rate": 0.0009106126768522609,
"loss": 2.5896708984375,
"step": 46000
},
{
"epoch": 1.425986690790886,
"grad_norm": 1.1070818901062012,
"learning_rate": 0.0009095852127364452,
"loss": 2.604625732421875,
"step": 46500
},
{
"epoch": 1.4413198810144439,
"grad_norm": 0.9439337849617004,
"learning_rate": 0.0009085577486206294,
"loss": 2.583058837890625,
"step": 47000
},
{
"epoch": 1.4566530712380017,
"grad_norm": 1.026479721069336,
"learning_rate": 0.0009075302845048136,
"loss": 2.5759638671875,
"step": 47500
},
{
"epoch": 1.4719862614615598,
"grad_norm": 1.0490233898162842,
"learning_rate": 0.000906502820388998,
"loss": 2.578414794921875,
"step": 48000
},
{
"epoch": 1.4873194516851176,
"grad_norm": 0.9949004650115967,
"learning_rate": 0.0009054753562731823,
"loss": 2.57360595703125,
"step": 48500
},
{
"epoch": 1.5026526419086754,
"grad_norm": 1.125211238861084,
"learning_rate": 0.0009044478921573665,
"loss": 2.566666748046875,
"step": 49000
},
{
"epoch": 1.5179858321322335,
"grad_norm": 0.9821215271949768,
"learning_rate": 0.0009034204280415507,
"loss": 2.567974609375,
"step": 49500
},
{
"epoch": 1.5333190223557913,
"grad_norm": 1.1235358715057373,
"learning_rate": 0.0009023929639257349,
"loss": 2.572336181640625,
"step": 50000
},
{
"epoch": 1.5486522125793494,
"grad_norm": 1.2190747261047363,
"learning_rate": 0.0009013654998099191,
"loss": 2.556693115234375,
"step": 50500
},
{
"epoch": 1.5639854028029072,
"grad_norm": 0.9801696538925171,
"learning_rate": 0.0009003380356941035,
"loss": 2.55803955078125,
"step": 51000
},
{
"epoch": 1.579318593026465,
"grad_norm": 1.704946517944336,
"learning_rate": 0.0008993105715782877,
"loss": 2.568549072265625,
"step": 51500
},
{
"epoch": 1.5946517832500229,
"grad_norm": 1.1942822933197021,
"learning_rate": 0.0008982831074624719,
"loss": 2.56319580078125,
"step": 52000
},
{
"epoch": 1.609984973473581,
"grad_norm": 1.0617578029632568,
"learning_rate": 0.0008972556433466561,
"loss": 2.559867431640625,
"step": 52500
},
{
"epoch": 1.6253181636971388,
"grad_norm": 0.9605861902236938,
"learning_rate": 0.0008962281792308405,
"loss": 2.548841064453125,
"step": 53000
},
{
"epoch": 1.6406513539206968,
"grad_norm": 1.1213620901107788,
"learning_rate": 0.0008952007151150247,
"loss": 2.550928955078125,
"step": 53500
},
{
"epoch": 1.6559845441442547,
"grad_norm": 1.9478049278259277,
"learning_rate": 0.0008941732509992089,
"loss": 2.543664794921875,
"step": 54000
},
{
"epoch": 1.6713177343678125,
"grad_norm": 1.0788325071334839,
"learning_rate": 0.0008931457868833931,
"loss": 2.5419150390625,
"step": 54500
},
{
"epoch": 1.6866509245913703,
"grad_norm": 1.0504772663116455,
"learning_rate": 0.0008921183227675773,
"loss": 2.550513427734375,
"step": 55000
},
{
"epoch": 1.7019841148149284,
"grad_norm": 0.9761285185813904,
"learning_rate": 0.0008910908586517617,
"loss": 2.555272705078125,
"step": 55500
},
{
"epoch": 1.7173173050384865,
"grad_norm": 1.0929040908813477,
"learning_rate": 0.0008900633945359459,
"loss": 2.55716748046875,
"step": 56000
},
{
"epoch": 1.7326504952620443,
"grad_norm": 1.0190736055374146,
"learning_rate": 0.0008890359304201301,
"loss": 2.54168701171875,
"step": 56500
},
{
"epoch": 1.7479836854856021,
"grad_norm": 1.105051875114441,
"learning_rate": 0.0008880084663043143,
"loss": 2.534852294921875,
"step": 57000
},
{
"epoch": 1.76331687570916,
"grad_norm": 1.1458134651184082,
"learning_rate": 0.0008869810021884985,
"loss": 2.533857666015625,
"step": 57500
},
{
"epoch": 1.778650065932718,
"grad_norm": 0.9536879658699036,
"learning_rate": 0.0008859535380726829,
"loss": 2.530094970703125,
"step": 58000
},
{
"epoch": 1.7939832561562759,
"grad_norm": 1.0406311750411987,
"learning_rate": 0.0008849260739568671,
"loss": 2.531776123046875,
"step": 58500
},
{
"epoch": 1.809316446379834,
"grad_norm": 1.0615878105163574,
"learning_rate": 0.0008838986098410513,
"loss": 2.527286865234375,
"step": 59000
},
{
"epoch": 1.8246496366033917,
"grad_norm": 1.0918337106704712,
"learning_rate": 0.0008828711457252355,
"loss": 2.529194091796875,
"step": 59500
},
{
"epoch": 1.8399828268269496,
"grad_norm": 1.0283782482147217,
"learning_rate": 0.0008818436816094198,
"loss": 2.521483154296875,
"step": 60000
},
{
"epoch": 1.8553160170505074,
"grad_norm": 0.9793681502342224,
"learning_rate": 0.000880816217493604,
"loss": 2.523812255859375,
"step": 60500
},
{
"epoch": 1.8706492072740655,
"grad_norm": 0.971653163433075,
"learning_rate": 0.0008797887533777883,
"loss": 2.518411865234375,
"step": 61000
},
{
"epoch": 1.8859823974976233,
"grad_norm": 1.2206299304962158,
"learning_rate": 0.0008787612892619725,
"loss": 2.510052001953125,
"step": 61500
},
{
"epoch": 1.9013155877211814,
"grad_norm": 1.032957673072815,
"learning_rate": 0.0008777338251461567,
"loss": 2.508514892578125,
"step": 62000
},
{
"epoch": 1.9166487779447392,
"grad_norm": 1.1681855916976929,
"learning_rate": 0.000876706361030341,
"loss": 2.5099169921875,
"step": 62500
},
{
"epoch": 1.931981968168297,
"grad_norm": 1.1712242364883423,
"learning_rate": 0.0008756788969145252,
"loss": 2.49520556640625,
"step": 63000
},
{
"epoch": 1.9473151583918549,
"grad_norm": 1.71385657787323,
"learning_rate": 0.0008746514327987095,
"loss": 2.51007666015625,
"step": 63500
},
{
"epoch": 1.962648348615413,
"grad_norm": 1.0384063720703125,
"learning_rate": 0.0008736239686828937,
"loss": 2.5085556640625,
"step": 64000
},
{
"epoch": 1.977981538838971,
"grad_norm": 1.014934778213501,
"learning_rate": 0.000872596504567078,
"loss": 2.50568408203125,
"step": 64500
},
{
"epoch": 1.9933147290625288,
"grad_norm": 1.1073269844055176,
"learning_rate": 0.0008715690404512623,
"loss": 2.507480712890625,
"step": 65000
},
{
"epoch": 2.0086479192860867,
"grad_norm": 1.0436677932739258,
"learning_rate": 0.0008705415763354466,
"loss": 2.498531494140625,
"step": 65500
},
{
"epoch": 2.0239811095096445,
"grad_norm": 1.062697410583496,
"learning_rate": 0.0008695141122196308,
"loss": 2.49360302734375,
"step": 66000
},
{
"epoch": 2.0393142997332023,
"grad_norm": 1.0030971765518188,
"learning_rate": 0.000868486648103815,
"loss": 2.488805908203125,
"step": 66500
},
{
"epoch": 2.0546474899567606,
"grad_norm": 1.0124140977859497,
"learning_rate": 0.0008674591839879993,
"loss": 2.495010498046875,
"step": 67000
},
{
"epoch": 2.0699806801803184,
"grad_norm": 1.070816159248352,
"learning_rate": 0.0008664317198721835,
"loss": 2.49730908203125,
"step": 67500
},
{
"epoch": 2.0853138704038763,
"grad_norm": 1.5643810033798218,
"learning_rate": 0.0008654042557563678,
"loss": 2.49456689453125,
"step": 68000
},
{
"epoch": 2.100647060627434,
"grad_norm": 1.0367401838302612,
"learning_rate": 0.000864376791640552,
"loss": 2.488412353515625,
"step": 68500
},
{
"epoch": 2.115980250850992,
"grad_norm": 1.1012048721313477,
"learning_rate": 0.0008633493275247362,
"loss": 2.493357666015625,
"step": 69000
},
{
"epoch": 2.13131344107455,
"grad_norm": 0.9959151744842529,
"learning_rate": 0.0008623218634089205,
"loss": 2.483527587890625,
"step": 69500
},
{
"epoch": 2.146646631298108,
"grad_norm": 1.040303349494934,
"learning_rate": 0.0008612943992931047,
"loss": 2.47618896484375,
"step": 70000
},
{
"epoch": 2.161979821521666,
"grad_norm": 1.0976628065109253,
"learning_rate": 0.000860266935177289,
"loss": 2.473078857421875,
"step": 70500
},
{
"epoch": 2.1773130117452237,
"grad_norm": 1.0886081457138062,
"learning_rate": 0.0008592394710614732,
"loss": 2.473018310546875,
"step": 71000
},
{
"epoch": 2.1926462019687816,
"grad_norm": 0.9382045269012451,
"learning_rate": 0.0008582120069456575,
"loss": 2.4686962890625,
"step": 71500
},
{
"epoch": 2.2079793921923394,
"grad_norm": 1.0487176179885864,
"learning_rate": 0.0008571845428298417,
"loss": 2.466439453125,
"step": 72000
},
{
"epoch": 2.2233125824158977,
"grad_norm": 1.019620418548584,
"learning_rate": 0.0008561570787140259,
"loss": 2.46345703125,
"step": 72500
},
{
"epoch": 2.2386457726394555,
"grad_norm": 0.9873740077018738,
"learning_rate": 0.0008551296145982102,
"loss": 2.46350244140625,
"step": 73000
},
{
"epoch": 2.2539789628630134,
"grad_norm": 1.1678472757339478,
"learning_rate": 0.0008541021504823944,
"loss": 2.46201611328125,
"step": 73500
},
{
"epoch": 2.269312153086571,
"grad_norm": 1.0411083698272705,
"learning_rate": 0.0008530746863665787,
"loss": 2.462214111328125,
"step": 74000
},
{
"epoch": 2.284645343310129,
"grad_norm": 1.1319035291671753,
"learning_rate": 0.0008520472222507629,
"loss": 2.46221533203125,
"step": 74500
},
{
"epoch": 2.299978533533687,
"grad_norm": 1.2074601650238037,
"learning_rate": 0.0008510197581349471,
"loss": 2.4605771484375,
"step": 75000
},
{
"epoch": 2.3153117237572447,
"grad_norm": 1.0101802349090576,
"learning_rate": 0.0008499922940191314,
"loss": 2.452893798828125,
"step": 75500
},
{
"epoch": 2.330644913980803,
"grad_norm": 1.0208394527435303,
"learning_rate": 0.0008489648299033157,
"loss": 2.459704833984375,
"step": 76000
},
{
"epoch": 2.345978104204361,
"grad_norm": 1.0759966373443604,
"learning_rate": 0.0008479373657874999,
"loss": 2.4663515625,
"step": 76500
},
{
"epoch": 2.3613112944279187,
"grad_norm": 1.3950154781341553,
"learning_rate": 0.0008469099016716841,
"loss": 2.47861962890625,
"step": 77000
},
{
"epoch": 2.3766444846514765,
"grad_norm": 1.06509530544281,
"learning_rate": 0.0008458824375558683,
"loss": 2.452545654296875,
"step": 77500
},
{
"epoch": 2.3919776748750343,
"grad_norm": 0.9840331673622131,
"learning_rate": 0.0008448549734400525,
"loss": 2.44589794921875,
"step": 78000
},
{
"epoch": 2.4073108650985926,
"grad_norm": 1.4551725387573242,
"learning_rate": 0.0008438275093242369,
"loss": 2.4442275390625,
"step": 78500
},
{
"epoch": 2.4226440553221504,
"grad_norm": 0.9275457859039307,
"learning_rate": 0.0008428000452084211,
"loss": 2.43882080078125,
"step": 79000
},
{
"epoch": 2.4379772455457083,
"grad_norm": 1.0269012451171875,
"learning_rate": 0.0008417725810926053,
"loss": 2.44102734375,
"step": 79500
},
{
"epoch": 2.453310435769266,
"grad_norm": 0.9632725119590759,
"learning_rate": 0.0008407451169767895,
"loss": 2.451830078125,
"step": 80000
},
{
"epoch": 2.468643625992824,
"grad_norm": 1.1963940858840942,
"learning_rate": 0.0008397176528609739,
"loss": 2.457018798828125,
"step": 80500
},
{
"epoch": 2.4839768162163818,
"grad_norm": 1.0788894891738892,
"learning_rate": 0.0008386901887451581,
"loss": 2.45520166015625,
"step": 81000
},
{
"epoch": 2.49931000643994,
"grad_norm": 0.9518547058105469,
"learning_rate": 0.0008376627246293423,
"loss": 2.453682373046875,
"step": 81500
},
{
"epoch": 2.514643196663498,
"grad_norm": 1.0291144847869873,
"learning_rate": 0.0008366352605135265,
"loss": 2.455847900390625,
"step": 82000
},
{
"epoch": 2.5299763868870557,
"grad_norm": 1.141697883605957,
"learning_rate": 0.0008356077963977108,
"loss": 2.438781005859375,
"step": 82500
},
{
"epoch": 2.5453095771106136,
"grad_norm": 1.0307334661483765,
"learning_rate": 0.0008345803322818952,
"loss": 2.429234130859375,
"step": 83000
},
{
"epoch": 2.5606427673341714,
"grad_norm": 1.1941356658935547,
"learning_rate": 0.0008335528681660794,
"loss": 2.429663330078125,
"step": 83500
},
{
"epoch": 2.5759759575577297,
"grad_norm": 1.006144642829895,
"learning_rate": 0.0008325254040502636,
"loss": 2.425216796875,
"step": 84000
},
{
"epoch": 2.5913091477812875,
"grad_norm": 1.0533301830291748,
"learning_rate": 0.0008314979399344478,
"loss": 2.4267919921875,
"step": 84500
},
{
"epoch": 2.6066423380048453,
"grad_norm": 1.1085325479507446,
"learning_rate": 0.000830470475818632,
"loss": 2.425106689453125,
"step": 85000
},
{
"epoch": 2.621975528228403,
"grad_norm": 0.966414749622345,
"learning_rate": 0.0008294430117028164,
"loss": 2.424360595703125,
"step": 85500
},
{
"epoch": 2.637308718451961,
"grad_norm": 1.9543871879577637,
"learning_rate": 0.0008284155475870006,
"loss": 2.4353818359375,
"step": 86000
},
{
"epoch": 2.652641908675519,
"grad_norm": 0.8990039825439453,
"learning_rate": 0.0008273880834711848,
"loss": 2.4239921875,
"step": 86500
},
{
"epoch": 2.6679750988990767,
"grad_norm": 0.9914394617080688,
"learning_rate": 0.000826360619355369,
"loss": 2.4215283203125,
"step": 87000
},
{
"epoch": 2.683308289122635,
"grad_norm": 1.1236220598220825,
"learning_rate": 0.0008253331552395533,
"loss": 2.419091064453125,
"step": 87500
},
{
"epoch": 2.698641479346193,
"grad_norm": 1.092773675918579,
"learning_rate": 0.0008243056911237376,
"loss": 2.41980224609375,
"step": 88000
},
{
"epoch": 2.7139746695697506,
"grad_norm": 1.0908790826797485,
"learning_rate": 0.0008232782270079218,
"loss": 2.417427734375,
"step": 88500
},
{
"epoch": 2.7293078597933085,
"grad_norm": 1.0773223638534546,
"learning_rate": 0.000822250762892106,
"loss": 2.41533544921875,
"step": 89000
},
{
"epoch": 2.7446410500168668,
"grad_norm": 1.0966140031814575,
"learning_rate": 0.0008212232987762902,
"loss": 2.415212890625,
"step": 89500
},
{
"epoch": 2.7599742402404246,
"grad_norm": 1.0292749404907227,
"learning_rate": 0.0008201958346604745,
"loss": 2.41604541015625,
"step": 90000
},
{
"epoch": 2.7753074304639824,
"grad_norm": 1.078901767730713,
"learning_rate": 0.0008191683705446588,
"loss": 2.4170439453125,
"step": 90500
},
{
"epoch": 2.7906406206875403,
"grad_norm": 1.0637688636779785,
"learning_rate": 0.000818140906428843,
"loss": 2.408758544921875,
"step": 91000
},
{
"epoch": 2.805973810911098,
"grad_norm": 1.0911078453063965,
"learning_rate": 0.0008171134423130272,
"loss": 2.410234130859375,
"step": 91500
},
{
"epoch": 2.821307001134656,
"grad_norm": 1.0348657369613647,
"learning_rate": 0.0008160859781972115,
"loss": 2.41000732421875,
"step": 92000
},
{
"epoch": 2.8366401913582138,
"grad_norm": 1.107820749282837,
"learning_rate": 0.0008150585140813957,
"loss": 2.40731787109375,
"step": 92500
},
{
"epoch": 2.851973381581772,
"grad_norm": 1.281421184539795,
"learning_rate": 0.00081403104996558,
"loss": 2.403383544921875,
"step": 93000
},
{
"epoch": 2.86730657180533,
"grad_norm": 1.0394543409347534,
"learning_rate": 0.0008130035858497642,
"loss": 2.418556640625,
"step": 93500
},
{
"epoch": 2.8826397620288877,
"grad_norm": 1.0160179138183594,
"learning_rate": 0.0008119761217339484,
"loss": 2.408310546875,
"step": 94000
},
{
"epoch": 2.8979729522524456,
"grad_norm": 1.0455384254455566,
"learning_rate": 0.0008109486576181327,
"loss": 2.40493359375,
"step": 94500
},
{
"epoch": 2.9133061424760034,
"grad_norm": 1.0490031242370605,
"learning_rate": 0.0008099211935023169,
"loss": 2.400986083984375,
"step": 95000
},
{
"epoch": 2.9286393326995617,
"grad_norm": 0.9743694067001343,
"learning_rate": 0.0008088937293865012,
"loss": 2.39481494140625,
"step": 95500
},
{
"epoch": 2.9439725229231195,
"grad_norm": 1.0077728033065796,
"learning_rate": 0.0008078662652706854,
"loss": 2.402294677734375,
"step": 96000
},
{
"epoch": 2.9593057131466773,
"grad_norm": 1.0075600147247314,
"learning_rate": 0.0008068388011548696,
"loss": 2.399803955078125,
"step": 96500
},
{
"epoch": 2.974638903370235,
"grad_norm": 126.66270446777344,
"learning_rate": 0.0008058113370390539,
"loss": 2.41233056640625,
"step": 97000
},
{
"epoch": 2.989972093593793,
"grad_norm": 1.0031794309616089,
"learning_rate": 0.0008047838729232381,
"loss": 2.399853759765625,
"step": 97500
},
{
"epoch": 3.005305283817351,
"grad_norm": 1.04261314868927,
"learning_rate": 0.0008037564088074224,
"loss": 2.40633349609375,
"step": 98000
},
{
"epoch": 3.020638474040909,
"grad_norm": 7.837483882904053,
"learning_rate": 0.0008027289446916066,
"loss": 2.4231552734375,
"step": 98500
},
{
"epoch": 3.035971664264467,
"grad_norm": 1.0534236431121826,
"learning_rate": 0.0008017014805757909,
"loss": 2.424142333984375,
"step": 99000
},
{
"epoch": 3.051304854488025,
"grad_norm": 1.0783402919769287,
"learning_rate": 0.0008006740164599752,
"loss": 2.401991943359375,
"step": 99500
},
{
"epoch": 3.0666380447115826,
"grad_norm": 11.349167823791504,
"learning_rate": 0.0007996465523441594,
"loss": 2.401349609375,
"step": 100000
},
{
"epoch": 3.0819712349351405,
"grad_norm": 1.2633163928985596,
"learning_rate": 0.0007986190882283437,
"loss": 2.39587548828125,
"step": 100500
},
{
"epoch": 3.0973044251586983,
"grad_norm": 0.8636135458946228,
"learning_rate": 0.0007975916241125279,
"loss": 2.394575439453125,
"step": 101000
},
{
"epoch": 3.1126376153822566,
"grad_norm": 1.0022070407867432,
"learning_rate": 0.0007965641599967122,
"loss": 2.39681689453125,
"step": 101500
},
{
"epoch": 3.1279708056058144,
"grad_norm": 1.1021260023117065,
"learning_rate": 0.0007955366958808964,
"loss": 2.402017822265625,
"step": 102000
},
{
"epoch": 3.1433039958293723,
"grad_norm": 1.0681781768798828,
"learning_rate": 0.0007945092317650806,
"loss": 2.415124755859375,
"step": 102500
},
{
"epoch": 3.15863718605293,
"grad_norm": 1.0853893756866455,
"learning_rate": 0.0007934817676492649,
"loss": 2.397299560546875,
"step": 103000
},
{
"epoch": 3.173970376276488,
"grad_norm": 1.0460395812988281,
"learning_rate": 0.0007924543035334492,
"loss": 2.386372802734375,
"step": 103500
},
{
"epoch": 3.189303566500046,
"grad_norm": 1.057936191558838,
"learning_rate": 0.0007914268394176334,
"loss": 2.387793212890625,
"step": 104000
},
{
"epoch": 3.204636756723604,
"grad_norm": 1.0139129161834717,
"learning_rate": 0.0007903993753018176,
"loss": 2.385367431640625,
"step": 104500
},
{
"epoch": 3.219969946947162,
"grad_norm": 0.972687840461731,
"learning_rate": 0.0007893719111860018,
"loss": 2.38626904296875,
"step": 105000
},
{
"epoch": 3.2353031371707197,
"grad_norm": 1.1815131902694702,
"learning_rate": 0.0007883444470701861,
"loss": 2.3891572265625,
"step": 105500
},
{
"epoch": 3.2506363273942775,
"grad_norm": 1.0027263164520264,
"learning_rate": 0.0007873169829543704,
"loss": 2.378604736328125,
"step": 106000
},
{
"epoch": 3.2659695176178354,
"grad_norm": 1.0525060892105103,
"learning_rate": 0.0007862895188385546,
"loss": 2.378078369140625,
"step": 106500
},
{
"epoch": 3.2813027078413937,
"grad_norm": 1.3268951177597046,
"learning_rate": 0.0007852620547227388,
"loss": 2.3823564453125,
"step": 107000
},
{
"epoch": 3.2966358980649515,
"grad_norm": 1.099571943283081,
"learning_rate": 0.000784234590606923,
"loss": 2.37035791015625,
"step": 107500
},
{
"epoch": 3.3119690882885093,
"grad_norm": 0.9928365349769592,
"learning_rate": 0.0007832071264911074,
"loss": 2.372281982421875,
"step": 108000
},
{
"epoch": 3.327302278512067,
"grad_norm": 0.9659942388534546,
"learning_rate": 0.0007821796623752916,
"loss": 2.373524169921875,
"step": 108500
},
{
"epoch": 3.342635468735625,
"grad_norm": 1.035586953163147,
"learning_rate": 0.0007811521982594758,
"loss": 2.370705322265625,
"step": 109000
},
{
"epoch": 3.357968658959183,
"grad_norm": 0.9759525060653687,
"learning_rate": 0.00078012473414366,
"loss": 2.372330810546875,
"step": 109500
},
{
"epoch": 3.373301849182741,
"grad_norm": 1.155532956123352,
"learning_rate": 0.0007790972700278442,
"loss": 2.36844091796875,
"step": 110000
},
{
"epoch": 3.388635039406299,
"grad_norm": 1.109609842300415,
"learning_rate": 0.0007780698059120286,
"loss": 2.36522314453125,
"step": 110500
},
{
"epoch": 3.403968229629857,
"grad_norm": 1.1023199558258057,
"learning_rate": 0.0007770423417962128,
"loss": 2.36473388671875,
"step": 111000
},
{
"epoch": 3.4193014198534146,
"grad_norm": 1.1913602352142334,
"learning_rate": 0.000776014877680397,
"loss": 2.3751591796875,
"step": 111500
},
{
"epoch": 3.4346346100769725,
"grad_norm": 1.0687745809555054,
"learning_rate": 0.0007749874135645812,
"loss": 2.366497802734375,
"step": 112000
},
{
"epoch": 3.4499678003005307,
"grad_norm": 0.9889634251594543,
"learning_rate": 0.0007739599494487654,
"loss": 2.365602294921875,
"step": 112500
},
{
"epoch": 3.4653009905240886,
"grad_norm": 0.9517545104026794,
"learning_rate": 0.0007729324853329498,
"loss": 2.353970947265625,
"step": 113000
},
{
"epoch": 3.4806341807476464,
"grad_norm": 1.074897289276123,
"learning_rate": 0.000771905021217134,
"loss": 2.355974609375,
"step": 113500
},
{
"epoch": 3.4959673709712042,
"grad_norm": 0.9137424230575562,
"learning_rate": 0.0007708775571013182,
"loss": 2.357901123046875,
"step": 114000
},
{
"epoch": 3.511300561194762,
"grad_norm": 1.1870537996292114,
"learning_rate": 0.0007698500929855024,
"loss": 2.36006494140625,
"step": 114500
},
{
"epoch": 3.52663375141832,
"grad_norm": 1.0832056999206543,
"learning_rate": 0.0007688226288696867,
"loss": 2.358577392578125,
"step": 115000
},
{
"epoch": 3.5419669416418778,
"grad_norm": 1.0135303735733032,
"learning_rate": 0.000767795164753871,
"loss": 2.351806884765625,
"step": 115500
},
{
"epoch": 3.557300131865436,
"grad_norm": 1.0838236808776855,
"learning_rate": 0.0007667677006380552,
"loss": 2.3514912109375,
"step": 116000
},
{
"epoch": 3.572633322088994,
"grad_norm": 0.9750739336013794,
"learning_rate": 0.0007657402365222394,
"loss": 2.350390869140625,
"step": 116500
},
{
"epoch": 3.5879665123125517,
"grad_norm": 0.8799856901168823,
"learning_rate": 0.0007647127724064237,
"loss": 2.34557568359375,
"step": 117000
},
{
"epoch": 3.6032997025361095,
"grad_norm": 1.1979016065597534,
"learning_rate": 0.000763685308290608,
"loss": 2.394179931640625,
"step": 117500
},
{
"epoch": 3.618632892759668,
"grad_norm": 1.188851237297058,
"learning_rate": 0.0007626578441747923,
"loss": 2.39966845703125,
"step": 118000
},
{
"epoch": 3.6339660829832257,
"grad_norm": 1.0725599527359009,
"learning_rate": 0.0007616303800589765,
"loss": 2.38725537109375,
"step": 118500
},
{
"epoch": 3.6492992732067835,
"grad_norm": 1.0177992582321167,
"learning_rate": 0.0007606029159431607,
"loss": 2.387532470703125,
"step": 119000
},
{
"epoch": 3.6646324634303413,
"grad_norm": 1.0341455936431885,
"learning_rate": 0.000759575451827345,
"loss": 2.39099169921875,
"step": 119500
},
{
"epoch": 3.679965653653899,
"grad_norm": 0.991851270198822,
"learning_rate": 0.0007585479877115293,
"loss": 2.389613037109375,
"step": 120000
},
{
"epoch": 3.695298843877457,
"grad_norm": 0.9739353656768799,
"learning_rate": 0.0007575205235957135,
"loss": 2.3843544921875,
"step": 120500
},
{
"epoch": 3.710632034101015,
"grad_norm": 0.9625571966171265,
"learning_rate": 0.0007564930594798977,
"loss": 2.384333740234375,
"step": 121000
},
{
"epoch": 3.725965224324573,
"grad_norm": 0.9422615766525269,
"learning_rate": 0.0007554655953640819,
"loss": 2.379560791015625,
"step": 121500
},
{
"epoch": 3.741298414548131,
"grad_norm": 0.9271089434623718,
"learning_rate": 0.0007544381312482662,
"loss": 2.385011474609375,
"step": 122000
},
{
"epoch": 3.756631604771689,
"grad_norm": 1.225429892539978,
"learning_rate": 0.0007534106671324504,
"loss": 2.36654638671875,
"step": 122500
},
{
"epoch": 3.7719647949952466,
"grad_norm": 0.9945045709609985,
"learning_rate": 0.0007523832030166347,
"loss": 2.341140380859375,
"step": 123000
},
{
"epoch": 3.787297985218805,
"grad_norm": 1.0410875082015991,
"learning_rate": 0.0007513557389008189,
"loss": 2.339957275390625,
"step": 123500
},
{
"epoch": 3.8026311754423627,
"grad_norm": 0.9847766160964966,
"learning_rate": 0.0007503282747850031,
"loss": 2.337027099609375,
"step": 124000
},
{
"epoch": 3.8179643656659206,
"grad_norm": 1.0451420545578003,
"learning_rate": 0.0007493008106691874,
"loss": 2.341636474609375,
"step": 124500
},
{
"epoch": 3.8332975558894784,
"grad_norm": 0.8977034687995911,
"learning_rate": 0.0007482733465533716,
"loss": 2.33426025390625,
"step": 125000
},
{
"epoch": 3.8486307461130362,
"grad_norm": 1.1214112043380737,
"learning_rate": 0.0007472458824375559,
"loss": 2.3320234375,
"step": 125500
},
{
"epoch": 3.863963936336594,
"grad_norm": 1.2016003131866455,
"learning_rate": 0.0007462184183217401,
"loss": 2.332626220703125,
"step": 126000
},
{
"epoch": 3.879297126560152,
"grad_norm": 0.9598099589347839,
"learning_rate": 0.0007451909542059244,
"loss": 2.33662158203125,
"step": 126500
},
{
"epoch": 3.8946303167837097,
"grad_norm": 1.105217695236206,
"learning_rate": 0.0007441634900901086,
"loss": 2.335081298828125,
"step": 127000
},
{
"epoch": 3.909963507007268,
"grad_norm": 1.055809736251831,
"learning_rate": 0.0007431360259742928,
"loss": 2.3350810546875,
"step": 127500
},
{
"epoch": 3.925296697230826,
"grad_norm": 0.9677720069885254,
"learning_rate": 0.0007421085618584771,
"loss": 2.32679296875,
"step": 128000
},
{
"epoch": 3.9406298874543837,
"grad_norm": 1.048683762550354,
"learning_rate": 0.0007410810977426613,
"loss": 2.329231201171875,
"step": 128500
},
{
"epoch": 3.9559630776779415,
"grad_norm": 1.0008125305175781,
"learning_rate": 0.0007400536336268456,
"loss": 2.33388330078125,
"step": 129000
},
{
"epoch": 3.9712962679015,
"grad_norm": 1.010535717010498,
"learning_rate": 0.0007390261695110298,
"loss": 2.329573974609375,
"step": 129500
},
{
"epoch": 3.9866294581250576,
"grad_norm": 1.0981321334838867,
"learning_rate": 0.000737998705395214,
"loss": 2.3347802734375,
"step": 130000
},
{
"epoch": 4.0019626483486155,
"grad_norm": 0.9873194694519043,
"learning_rate": 0.0007369712412793983,
"loss": 2.327262451171875,
"step": 130500
},
{
"epoch": 4.017295838572173,
"grad_norm": 1.0334597826004028,
"learning_rate": 0.0007359437771635826,
"loss": 2.323281982421875,
"step": 131000
},
{
"epoch": 4.032629028795731,
"grad_norm": 1.054046392440796,
"learning_rate": 0.0007349163130477668,
"loss": 2.32666259765625,
"step": 131500
},
{
"epoch": 4.047962219019289,
"grad_norm": 1.0472097396850586,
"learning_rate": 0.000733888848931951,
"loss": 2.320403076171875,
"step": 132000
},
{
"epoch": 4.063295409242847,
"grad_norm": 1.0481075048446655,
"learning_rate": 0.0007328613848161352,
"loss": 2.32290673828125,
"step": 132500
},
{
"epoch": 4.078628599466405,
"grad_norm": 1.0906621217727661,
"learning_rate": 0.0007318339207003195,
"loss": 2.31926025390625,
"step": 133000
},
{
"epoch": 4.0939617896899625,
"grad_norm": 0.986464262008667,
"learning_rate": 0.0007308064565845038,
"loss": 2.327080810546875,
"step": 133500
},
{
"epoch": 4.109294979913521,
"grad_norm": 1.0808234214782715,
"learning_rate": 0.0007297789924686881,
"loss": 2.324663330078125,
"step": 134000
},
{
"epoch": 4.124628170137079,
"grad_norm": 1.0530495643615723,
"learning_rate": 0.0007287515283528723,
"loss": 2.32747021484375,
"step": 134500
},
{
"epoch": 4.139961360360637,
"grad_norm": 1.5676161050796509,
"learning_rate": 0.0007277240642370566,
"loss": 2.326822998046875,
"step": 135000
},
{
"epoch": 4.155294550584195,
"grad_norm": 1.0009845495224,
"learning_rate": 0.0007266966001212409,
"loss": 2.32525244140625,
"step": 135500
},
{
"epoch": 4.170627740807753,
"grad_norm": 1.1091865301132202,
"learning_rate": 0.0007256691360054251,
"loss": 2.32842041015625,
"step": 136000
},
{
"epoch": 4.18596093103131,
"grad_norm": 0.9990963935852051,
"learning_rate": 0.0007246416718896093,
"loss": 2.326690185546875,
"step": 136500
},
{
"epoch": 4.201294121254868,
"grad_norm": 1.0287275314331055,
"learning_rate": 0.0007236142077737935,
"loss": 2.324251708984375,
"step": 137000
},
{
"epoch": 4.216627311478426,
"grad_norm": 0.9484901428222656,
"learning_rate": 0.0007225867436579778,
"loss": 2.32472998046875,
"step": 137500
},
{
"epoch": 4.231960501701984,
"grad_norm": 1.0180178880691528,
"learning_rate": 0.0007215592795421621,
"loss": 2.3212021484375,
"step": 138000
},
{
"epoch": 4.247293691925542,
"grad_norm": 1.0087257623672485,
"learning_rate": 0.0007205318154263463,
"loss": 2.325920654296875,
"step": 138500
},
{
"epoch": 4.2626268821491,
"grad_norm": 1.049051284790039,
"learning_rate": 0.0007195043513105305,
"loss": 2.31802099609375,
"step": 139000
},
{
"epoch": 4.277960072372657,
"grad_norm": 1.04050612449646,
"learning_rate": 0.0007184768871947147,
"loss": 2.319311767578125,
"step": 139500
},
{
"epoch": 4.293293262596216,
"grad_norm": 1.041749119758606,
"learning_rate": 0.000717449423078899,
"loss": 2.315538818359375,
"step": 140000
},
{
"epoch": 4.308626452819774,
"grad_norm": 1.0806001424789429,
"learning_rate": 0.0007164219589630833,
"loss": 2.318419921875,
"step": 140500
},
{
"epoch": 4.323959643043332,
"grad_norm": 0.9845504760742188,
"learning_rate": 0.0007153944948472675,
"loss": 2.319677978515625,
"step": 141000
},
{
"epoch": 4.33929283326689,
"grad_norm": 1.5578675270080566,
"learning_rate": 0.0007143670307314517,
"loss": 2.3137734375,
"step": 141500
},
{
"epoch": 4.3546260234904475,
"grad_norm": 0.9561862945556641,
"learning_rate": 0.0007133395666156359,
"loss": 2.311796875,
"step": 142000
},
{
"epoch": 4.369959213714005,
"grad_norm": 0.9891361594200134,
"learning_rate": 0.0007123121024998203,
"loss": 2.31162841796875,
"step": 142500
},
{
"epoch": 4.385292403937563,
"grad_norm": 0.9529083967208862,
"learning_rate": 0.0007112846383840045,
"loss": 2.31976171875,
"step": 143000
},
{
"epoch": 4.400625594161121,
"grad_norm": 1.074535846710205,
"learning_rate": 0.0007102571742681887,
"loss": 2.31236083984375,
"step": 143500
},
{
"epoch": 4.415958784384679,
"grad_norm": 0.9345454573631287,
"learning_rate": 0.0007092297101523729,
"loss": 2.3100966796875,
"step": 144000
},
{
"epoch": 4.431291974608237,
"grad_norm": 0.9117188453674316,
"learning_rate": 0.0007082022460365571,
"loss": 2.310595703125,
"step": 144500
},
{
"epoch": 4.446625164831795,
"grad_norm": 1.0388634204864502,
"learning_rate": 0.0007071747819207415,
"loss": 2.3145859375,
"step": 145000
},
{
"epoch": 4.461958355055353,
"grad_norm": 0.9499608874320984,
"learning_rate": 0.0007061473178049257,
"loss": 2.313781982421875,
"step": 145500
},
{
"epoch": 4.477291545278911,
"grad_norm": 0.9371971487998962,
"learning_rate": 0.0007051198536891099,
"loss": 2.307994384765625,
"step": 146000
},
{
"epoch": 4.492624735502469,
"grad_norm": 0.9107025861740112,
"learning_rate": 0.0007040923895732941,
"loss": 2.308412353515625,
"step": 146500
},
{
"epoch": 4.507957925726027,
"grad_norm": 0.9707034826278687,
"learning_rate": 0.0007030649254574784,
"loss": 2.30725,
"step": 147000
},
{
"epoch": 4.5232911159495846,
"grad_norm": 1.0450971126556396,
"learning_rate": 0.0007020374613416627,
"loss": 2.30588037109375,
"step": 147500
},
{
"epoch": 4.538624306173142,
"grad_norm": 0.9793196320533752,
"learning_rate": 0.0007010099972258469,
"loss": 2.305541748046875,
"step": 148000
},
{
"epoch": 4.5539574963967,
"grad_norm": 0.9398958086967468,
"learning_rate": 0.0006999825331100311,
"loss": 2.29575341796875,
"step": 148500
},
{
"epoch": 4.569290686620258,
"grad_norm": 1.0998700857162476,
"learning_rate": 0.0006989550689942153,
"loss": 2.30571923828125,
"step": 149000
},
{
"epoch": 4.584623876843816,
"grad_norm": 0.9057356119155884,
"learning_rate": 0.0006979276048783996,
"loss": 2.300291259765625,
"step": 149500
},
{
"epoch": 4.599957067067374,
"grad_norm": 1.0196291208267212,
"learning_rate": 0.0006969001407625839,
"loss": 2.3067138671875,
"step": 150000
},
{
"epoch": 4.615290257290932,
"grad_norm": 1.1155024766921997,
"learning_rate": 0.0006958726766467681,
"loss": 2.29611669921875,
"step": 150500
},
{
"epoch": 4.630623447514489,
"grad_norm": 1.0437560081481934,
"learning_rate": 0.0006948452125309523,
"loss": 2.2975673828125,
"step": 151000
},
{
"epoch": 4.645956637738048,
"grad_norm": 0.919367253780365,
"learning_rate": 0.0006938177484151366,
"loss": 2.303860595703125,
"step": 151500
},
{
"epoch": 4.661289827961606,
"grad_norm": 0.9895167946815491,
"learning_rate": 0.0006927902842993209,
"loss": 2.29558349609375,
"step": 152000
},
{
"epoch": 4.676623018185164,
"grad_norm": 1.0639727115631104,
"learning_rate": 0.0006917628201835052,
"loss": 2.3023212890625,
"step": 152500
},
{
"epoch": 4.691956208408722,
"grad_norm": 0.9420351982116699,
"learning_rate": 0.0006907353560676894,
"loss": 2.294746337890625,
"step": 153000
},
{
"epoch": 4.7072893986322795,
"grad_norm": 0.939965546131134,
"learning_rate": 0.0006897078919518736,
"loss": 2.29506201171875,
"step": 153500
},
{
"epoch": 4.722622588855837,
"grad_norm": 0.911388635635376,
"learning_rate": 0.0006886804278360579,
"loss": 2.297764892578125,
"step": 154000
},
{
"epoch": 4.737955779079395,
"grad_norm": 1.1255003213882446,
"learning_rate": 0.0006876529637202421,
"loss": 2.291568603515625,
"step": 154500
},
{
"epoch": 4.753288969302953,
"grad_norm": 1.0341004133224487,
"learning_rate": 0.0006866254996044264,
"loss": 2.29397412109375,
"step": 155000
},
{
"epoch": 4.768622159526511,
"grad_norm": 0.993778645992279,
"learning_rate": 0.0006855980354886106,
"loss": 2.29723876953125,
"step": 155500
},
{
"epoch": 4.783955349750069,
"grad_norm": 0.9582630395889282,
"learning_rate": 0.0006845705713727948,
"loss": 2.291623291015625,
"step": 156000
},
{
"epoch": 4.799288539973627,
"grad_norm": 1.0447920560836792,
"learning_rate": 0.0006835431072569791,
"loss": 2.297012451171875,
"step": 156500
},
{
"epoch": 4.814621730197185,
"grad_norm": 0.9290124773979187,
"learning_rate": 0.0006825156431411633,
"loss": 2.291688720703125,
"step": 157000
},
{
"epoch": 4.829954920420743,
"grad_norm": 0.99012291431427,
"learning_rate": 0.0006814881790253476,
"loss": 2.29289453125,
"step": 157500
},
{
"epoch": 4.845288110644301,
"grad_norm": 0.9243401288986206,
"learning_rate": 0.0006804607149095318,
"loss": 2.291481689453125,
"step": 158000
},
{
"epoch": 4.860621300867859,
"grad_norm": 0.9846622943878174,
"learning_rate": 0.0006794332507937161,
"loss": 2.36663818359375,
"step": 158500
},
{
"epoch": 4.8759544910914165,
"grad_norm": 1.0044904947280884,
"learning_rate": 0.0006784057866779003,
"loss": 2.283203857421875,
"step": 159000
},
{
"epoch": 4.891287681314974,
"grad_norm": 1.0759766101837158,
"learning_rate": 0.0006773783225620845,
"loss": 2.279999755859375,
"step": 159500
},
{
"epoch": 4.906620871538532,
"grad_norm": 0.9901242852210999,
"learning_rate": 0.0006763508584462688,
"loss": 2.341744384765625,
"step": 160000
},
{
"epoch": 4.92195406176209,
"grad_norm": 0.9589545726776123,
"learning_rate": 0.000675323394330453,
"loss": 2.276427734375,
"step": 160500
},
{
"epoch": 4.937287251985648,
"grad_norm": 1.1527578830718994,
"learning_rate": 0.0006742959302146373,
"loss": 2.281647216796875,
"step": 161000
},
{
"epoch": 4.952620442209206,
"grad_norm": 1.1275092363357544,
"learning_rate": 0.0006732684660988215,
"loss": 2.28138232421875,
"step": 161500
},
{
"epoch": 4.9679536324327636,
"grad_norm": 0.9733052849769592,
"learning_rate": 0.0006722410019830057,
"loss": 2.276751708984375,
"step": 162000
},
{
"epoch": 4.983286822656321,
"grad_norm": 0.9592075943946838,
"learning_rate": 0.00067121353786719,
"loss": 2.278658935546875,
"step": 162500
},
{
"epoch": 4.99862001287988,
"grad_norm": 0.9945816397666931,
"learning_rate": 0.0006701860737513742,
"loss": 2.285181884765625,
"step": 163000
},
{
"epoch": 5.013953203103438,
"grad_norm": 0.9680554866790771,
"learning_rate": 0.0006691586096355585,
"loss": 2.27959912109375,
"step": 163500
},
{
"epoch": 5.029286393326996,
"grad_norm": 0.9936657547950745,
"learning_rate": 0.0006681311455197427,
"loss": 2.2861240234375,
"step": 164000
},
{
"epoch": 5.044619583550554,
"grad_norm": 1.047868013381958,
"learning_rate": 0.0006671036814039269,
"loss": 2.278624267578125,
"step": 164500
},
{
"epoch": 5.0599527737741115,
"grad_norm": 0.9442141652107239,
"learning_rate": 0.0006660762172881112,
"loss": 2.272057373046875,
"step": 165000
},
{
"epoch": 5.075285963997669,
"grad_norm": 0.9988092184066772,
"learning_rate": 0.0006650487531722955,
"loss": 2.27026025390625,
"step": 165500
},
{
"epoch": 5.090619154221227,
"grad_norm": 1.1299848556518555,
"learning_rate": 0.0006640212890564797,
"loss": 2.267636962890625,
"step": 166000
},
{
"epoch": 5.105952344444785,
"grad_norm": 1.000954031944275,
"learning_rate": 0.0006629938249406639,
"loss": 2.267901123046875,
"step": 166500
},
{
"epoch": 5.121285534668343,
"grad_norm": 0.9552673101425171,
"learning_rate": 0.0006619663608248481,
"loss": 2.270672119140625,
"step": 167000
},
{
"epoch": 5.136618724891901,
"grad_norm": 1.0032132863998413,
"learning_rate": 0.0006609388967090324,
"loss": 2.269614013671875,
"step": 167500
},
{
"epoch": 5.151951915115459,
"grad_norm": 1.0271780490875244,
"learning_rate": 0.0006599114325932167,
"loss": 2.26814599609375,
"step": 168000
},
{
"epoch": 5.167285105339017,
"grad_norm": 0.9501484632492065,
"learning_rate": 0.000658883968477401,
"loss": 2.2711953125,
"step": 168500
},
{
"epoch": 5.182618295562575,
"grad_norm": 0.9047709703445435,
"learning_rate": 0.0006578565043615852,
"loss": 2.2666357421875,
"step": 169000
},
{
"epoch": 5.197951485786133,
"grad_norm": 1.0150599479675293,
"learning_rate": 0.0006568290402457694,
"loss": 2.26838525390625,
"step": 169500
},
{
"epoch": 5.213284676009691,
"grad_norm": 0.8929696083068848,
"learning_rate": 0.0006558015761299538,
"loss": 2.26212890625,
"step": 170000
},
{
"epoch": 5.2286178662332485,
"grad_norm": 1.0004863739013672,
"learning_rate": 0.000654774112014138,
"loss": 2.26541357421875,
"step": 170500
},
{
"epoch": 5.243951056456806,
"grad_norm": 1.0438615083694458,
"learning_rate": 0.0006537466478983222,
"loss": 2.261427978515625,
"step": 171000
},
{
"epoch": 5.259284246680364,
"grad_norm": 1.1246472597122192,
"learning_rate": 0.0006527191837825064,
"loss": 2.264583251953125,
"step": 171500
},
{
"epoch": 5.274617436903922,
"grad_norm": 0.9809620976448059,
"learning_rate": 0.0006516917196666906,
"loss": 2.263846923828125,
"step": 172000
},
{
"epoch": 5.28995062712748,
"grad_norm": 1.0542471408843994,
"learning_rate": 0.000650664255550875,
"loss": 2.264419921875,
"step": 172500
},
{
"epoch": 5.305283817351038,
"grad_norm": 0.9916055798530579,
"learning_rate": 0.0006496367914350592,
"loss": 2.263090576171875,
"step": 173000
},
{
"epoch": 5.3206170075745955,
"grad_norm": 0.9650833010673523,
"learning_rate": 0.0006486093273192434,
"loss": 2.269659912109375,
"step": 173500
},
{
"epoch": 5.335950197798154,
"grad_norm": 0.9465742707252502,
"learning_rate": 0.0006475818632034276,
"loss": 2.266365234375,
"step": 174000
},
{
"epoch": 5.351283388021712,
"grad_norm": 1.0650707483291626,
"learning_rate": 0.000646554399087612,
"loss": 2.27085498046875,
"step": 174500
},
{
"epoch": 5.36661657824527,
"grad_norm": 0.9889022707939148,
"learning_rate": 0.0006455269349717962,
"loss": 2.2664892578125,
"step": 175000
},
{
"epoch": 5.381949768468828,
"grad_norm": 0.9560773968696594,
"learning_rate": 0.0006444994708559804,
"loss": 2.271109375,
"step": 175500
},
{
"epoch": 5.397282958692386,
"grad_norm": 1.0954188108444214,
"learning_rate": 0.0006434720067401646,
"loss": 2.27003857421875,
"step": 176000
},
{
"epoch": 5.4126161489159434,
"grad_norm": 1.0397051572799683,
"learning_rate": 0.0006424445426243488,
"loss": 2.26712548828125,
"step": 176500
},
{
"epoch": 5.427949339139501,
"grad_norm": 0.895498514175415,
"learning_rate": 0.0006414170785085331,
"loss": 2.2700888671875,
"step": 177000
},
{
"epoch": 5.443282529363059,
"grad_norm": 0.925999104976654,
"learning_rate": 0.0006403896143927174,
"loss": 2.26104443359375,
"step": 177500
},
{
"epoch": 5.458615719586617,
"grad_norm": 1.0165995359420776,
"learning_rate": 0.0006393621502769016,
"loss": 2.25827978515625,
"step": 178000
},
{
"epoch": 5.473948909810175,
"grad_norm": 0.9287647008895874,
"learning_rate": 0.0006383346861610858,
"loss": 2.251656982421875,
"step": 178500
},
{
"epoch": 5.489282100033733,
"grad_norm": 0.9343289136886597,
"learning_rate": 0.00063730722204527,
"loss": 2.253738037109375,
"step": 179000
},
{
"epoch": 5.504615290257291,
"grad_norm": 0.9714416265487671,
"learning_rate": 0.0006362797579294543,
"loss": 2.25658935546875,
"step": 179500
},
{
"epoch": 5.519948480480849,
"grad_norm": 0.9686933755874634,
"learning_rate": 0.0006352522938136386,
"loss": 2.255177978515625,
"step": 180000
},
{
"epoch": 5.535281670704407,
"grad_norm": 0.9441134929656982,
"learning_rate": 0.0006342248296978228,
"loss": 2.2570390625,
"step": 180500
},
{
"epoch": 5.550614860927965,
"grad_norm": 0.97795569896698,
"learning_rate": 0.000633197365582007,
"loss": 2.25539013671875,
"step": 181000
},
{
"epoch": 5.565948051151523,
"grad_norm": 1.0169578790664673,
"learning_rate": 0.0006321699014661913,
"loss": 2.25522607421875,
"step": 181500
},
{
"epoch": 5.5812812413750805,
"grad_norm": 1.2037407159805298,
"learning_rate": 0.0006311424373503755,
"loss": 2.26430029296875,
"step": 182000
},
{
"epoch": 5.596614431598638,
"grad_norm": 1.067467451095581,
"learning_rate": 0.0006301149732345598,
"loss": 2.257678955078125,
"step": 182500
},
{
"epoch": 5.611947621822196,
"grad_norm": 0.9488410949707031,
"learning_rate": 0.000629087509118744,
"loss": 2.258622314453125,
"step": 183000
},
{
"epoch": 5.627280812045754,
"grad_norm": 1.0531996488571167,
"learning_rate": 0.0006280600450029282,
"loss": 2.255596435546875,
"step": 183500
},
{
"epoch": 5.642614002269312,
"grad_norm": 0.9988917112350464,
"learning_rate": 0.0006270325808871125,
"loss": 2.256611572265625,
"step": 184000
},
{
"epoch": 5.65794719249287,
"grad_norm": 1.0092180967330933,
"learning_rate": 0.0006260051167712967,
"loss": 2.255166259765625,
"step": 184500
},
{
"epoch": 5.6732803827164275,
"grad_norm": 1.084371566772461,
"learning_rate": 0.000624977652655481,
"loss": 2.25481640625,
"step": 185000
},
{
"epoch": 5.688613572939986,
"grad_norm": 1.0845147371292114,
"learning_rate": 0.0006239501885396652,
"loss": 2.25538818359375,
"step": 185500
},
{
"epoch": 5.703946763163544,
"grad_norm": 1.027005910873413,
"learning_rate": 0.0006229227244238496,
"loss": 2.246244384765625,
"step": 186000
},
{
"epoch": 5.719279953387102,
"grad_norm": 0.9159988164901733,
"learning_rate": 0.0006218952603080338,
"loss": 2.24983544921875,
"step": 186500
},
{
"epoch": 5.73461314361066,
"grad_norm": 1.1544363498687744,
"learning_rate": 0.000620867796192218,
"loss": 2.26296240234375,
"step": 187000
},
{
"epoch": 5.749946333834218,
"grad_norm": 0.9719838500022888,
"learning_rate": 0.0006198403320764023,
"loss": 2.25274951171875,
"step": 187500
},
{
"epoch": 5.765279524057775,
"grad_norm": 1.0159788131713867,
"learning_rate": 0.0006188128679605865,
"loss": 2.2626767578125,
"step": 188000
},
{
"epoch": 5.780612714281333,
"grad_norm": 0.9425241351127625,
"learning_rate": 0.0006177854038447708,
"loss": 2.249486083984375,
"step": 188500
},
{
"epoch": 5.795945904504891,
"grad_norm": 0.9763333797454834,
"learning_rate": 0.000616757939728955,
"loss": 2.251626708984375,
"step": 189000
},
{
"epoch": 5.811279094728449,
"grad_norm": 0.9173125624656677,
"learning_rate": 0.0006157304756131392,
"loss": 2.249015380859375,
"step": 189500
},
{
"epoch": 5.826612284952007,
"grad_norm": 1.048012137413025,
"learning_rate": 0.0006147030114973235,
"loss": 2.24950537109375,
"step": 190000
},
{
"epoch": 5.8419454751755655,
"grad_norm": 0.9468924403190613,
"learning_rate": 0.0006136755473815077,
"loss": 2.246572998046875,
"step": 190500
},
{
"epoch": 5.857278665399123,
"grad_norm": 0.8998527526855469,
"learning_rate": 0.000612648083265692,
"loss": 2.244881591796875,
"step": 191000
},
{
"epoch": 5.872611855622681,
"grad_norm": 0.9944680333137512,
"learning_rate": 0.0006116206191498762,
"loss": 2.24555859375,
"step": 191500
},
{
"epoch": 5.887945045846239,
"grad_norm": 0.9639532566070557,
"learning_rate": 0.0006105931550340604,
"loss": 2.242562744140625,
"step": 192000
},
{
"epoch": 5.903278236069797,
"grad_norm": 0.9046755433082581,
"learning_rate": 0.0006095656909182447,
"loss": 2.24507861328125,
"step": 192500
},
{
"epoch": 5.918611426293355,
"grad_norm": 0.9771453738212585,
"learning_rate": 0.000608538226802429,
"loss": 2.239072509765625,
"step": 193000
},
{
"epoch": 5.9339446165169125,
"grad_norm": 0.9813857674598694,
"learning_rate": 0.0006075107626866132,
"loss": 2.250779296875,
"step": 193500
},
{
"epoch": 5.94927780674047,
"grad_norm": 1.0077128410339355,
"learning_rate": 0.0006064832985707974,
"loss": 2.237977294921875,
"step": 194000
},
{
"epoch": 5.964610996964028,
"grad_norm": 0.963520348072052,
"learning_rate": 0.0006054558344549816,
"loss": 2.24643505859375,
"step": 194500
},
{
"epoch": 5.979944187187586,
"grad_norm": 0.9587693810462952,
"learning_rate": 0.0006044283703391659,
"loss": 2.249884765625,
"step": 195000
},
{
"epoch": 5.995277377411144,
"grad_norm": 0.9101521372795105,
"learning_rate": 0.0006034009062233502,
"loss": 2.246553466796875,
"step": 195500
},
{
"epoch": 6.010610567634702,
"grad_norm": 0.9489941596984863,
"learning_rate": 0.0006023734421075344,
"loss": 2.24535009765625,
"step": 196000
},
{
"epoch": 6.02594375785826,
"grad_norm": 0.9361016750335693,
"learning_rate": 0.0006013459779917186,
"loss": 2.2426767578125,
"step": 196500
},
{
"epoch": 6.041276948081818,
"grad_norm": 0.9649178981781006,
"learning_rate": 0.0006003185138759028,
"loss": 2.24096826171875,
"step": 197000
},
{
"epoch": 6.056610138305376,
"grad_norm": 0.8999959230422974,
"learning_rate": 0.0005992910497600872,
"loss": 2.247010009765625,
"step": 197500
},
{
"epoch": 6.071943328528934,
"grad_norm": 0.9264272451400757,
"learning_rate": 0.0005982635856442714,
"loss": 2.239556640625,
"step": 198000
},
{
"epoch": 6.087276518752492,
"grad_norm": 0.997776448726654,
"learning_rate": 0.0005972361215284556,
"loss": 2.23791552734375,
"step": 198500
},
{
"epoch": 6.10260970897605,
"grad_norm": 1.013379454612732,
"learning_rate": 0.0005962086574126398,
"loss": 2.2441943359375,
"step": 199000
},
{
"epoch": 6.117942899199607,
"grad_norm": 0.9030807018280029,
"learning_rate": 0.000595181193296824,
"loss": 2.237796142578125,
"step": 199500
},
{
"epoch": 6.133276089423165,
"grad_norm": 0.8709002733230591,
"learning_rate": 0.0005941537291810084,
"loss": 2.235516845703125,
"step": 200000
},
{
"epoch": 6.148609279646723,
"grad_norm": 0.9175723195075989,
"learning_rate": 0.0005931262650651926,
"loss": 2.240197021484375,
"step": 200500
},
{
"epoch": 6.163942469870281,
"grad_norm": 0.9393293857574463,
"learning_rate": 0.0005920988009493768,
"loss": 2.24956982421875,
"step": 201000
},
{
"epoch": 6.179275660093839,
"grad_norm": 1.037772536277771,
"learning_rate": 0.000591071336833561,
"loss": 2.237399169921875,
"step": 201500
},
{
"epoch": 6.194608850317397,
"grad_norm": 0.9020262360572815,
"learning_rate": 0.0005900438727177453,
"loss": 2.2604599609375,
"step": 202000
},
{
"epoch": 6.209942040540955,
"grad_norm": 0.9170573353767395,
"learning_rate": 0.0005890164086019296,
"loss": 2.24441845703125,
"step": 202500
},
{
"epoch": 6.225275230764513,
"grad_norm": 1.0186824798583984,
"learning_rate": 0.0005879889444861139,
"loss": 2.24244921875,
"step": 203000
},
{
"epoch": 6.240608420988071,
"grad_norm": 0.9553197026252747,
"learning_rate": 0.0005869614803702981,
"loss": 2.241346923828125,
"step": 203500
},
{
"epoch": 6.255941611211629,
"grad_norm": 0.9047817587852478,
"learning_rate": 0.0005859340162544823,
"loss": 2.23383544921875,
"step": 204000
},
{
"epoch": 6.271274801435187,
"grad_norm": 0.9347020387649536,
"learning_rate": 0.0005849065521386667,
"loss": 2.22915478515625,
"step": 204500
},
{
"epoch": 6.2866079916587445,
"grad_norm": 0.9690987467765808,
"learning_rate": 0.0005838790880228509,
"loss": 2.22935498046875,
"step": 205000
},
{
"epoch": 6.301941181882302,
"grad_norm": 0.9864622354507446,
"learning_rate": 0.0005828516239070351,
"loss": 2.231296875,
"step": 205500
},
{
"epoch": 6.31727437210586,
"grad_norm": 0.9501722455024719,
"learning_rate": 0.0005818241597912193,
"loss": 2.234732666015625,
"step": 206000
},
{
"epoch": 6.332607562329418,
"grad_norm": 0.9334090352058411,
"learning_rate": 0.0005807966956754035,
"loss": 2.228567138671875,
"step": 206500
},
{
"epoch": 6.347940752552976,
"grad_norm": 0.9500818252563477,
"learning_rate": 0.0005797692315595879,
"loss": 2.23051416015625,
"step": 207000
},
{
"epoch": 6.363273942776534,
"grad_norm": 0.9463828206062317,
"learning_rate": 0.0005787417674437721,
"loss": 2.227259033203125,
"step": 207500
},
{
"epoch": 6.378607133000092,
"grad_norm": 1.0424871444702148,
"learning_rate": 0.0005777143033279563,
"loss": 2.222300048828125,
"step": 208000
},
{
"epoch": 6.39394032322365,
"grad_norm": 0.9323693513870239,
"learning_rate": 0.0005766868392121405,
"loss": 2.223302490234375,
"step": 208500
},
{
"epoch": 6.409273513447208,
"grad_norm": 0.9248519539833069,
"learning_rate": 0.0005756593750963248,
"loss": 2.228900390625,
"step": 209000
},
{
"epoch": 6.424606703670766,
"grad_norm": 0.9006691575050354,
"learning_rate": 0.000574631910980509,
"loss": 2.22398388671875,
"step": 209500
},
{
"epoch": 6.439939893894324,
"grad_norm": 0.9456732273101807,
"learning_rate": 0.0005736044468646933,
"loss": 2.224505615234375,
"step": 210000
},
{
"epoch": 6.455273084117882,
"grad_norm": 0.9350338578224182,
"learning_rate": 0.0005725769827488775,
"loss": 2.220249267578125,
"step": 210500
},
{
"epoch": 6.470606274341439,
"grad_norm": 0.8714827299118042,
"learning_rate": 0.0005715495186330617,
"loss": 2.22928466796875,
"step": 211000
},
{
"epoch": 6.485939464564997,
"grad_norm": 0.9264363050460815,
"learning_rate": 0.000570522054517246,
"loss": 2.258007568359375,
"step": 211500
},
{
"epoch": 6.501272654788555,
"grad_norm": 0.9224889278411865,
"learning_rate": 0.0005694945904014303,
"loss": 2.252694580078125,
"step": 212000
},
{
"epoch": 6.516605845012113,
"grad_norm": 0.9216419458389282,
"learning_rate": 0.0005684671262856145,
"loss": 2.2246591796875,
"step": 212500
},
{
"epoch": 6.531939035235671,
"grad_norm": 0.9206384420394897,
"learning_rate": 0.0005674396621697987,
"loss": 2.220703857421875,
"step": 213000
},
{
"epoch": 6.5472722254592295,
"grad_norm": 0.9439525008201599,
"learning_rate": 0.000566412198053983,
"loss": 2.21642138671875,
"step": 213500
},
{
"epoch": 6.562605415682787,
"grad_norm": 0.9205273985862732,
"learning_rate": 0.0005653847339381672,
"loss": 2.2178076171875,
"step": 214000
},
{
"epoch": 6.577938605906345,
"grad_norm": 0.9845165014266968,
"learning_rate": 0.0005643572698223515,
"loss": 2.221718994140625,
"step": 214500
},
{
"epoch": 6.593271796129903,
"grad_norm": 1.006942629814148,
"learning_rate": 0.0005633298057065357,
"loss": 2.221414306640625,
"step": 215000
},
{
"epoch": 6.608604986353461,
"grad_norm": 0.9668125510215759,
"learning_rate": 0.0005623023415907199,
"loss": 2.218563720703125,
"step": 215500
},
{
"epoch": 6.623938176577019,
"grad_norm": 1.0040669441223145,
"learning_rate": 0.0005612748774749042,
"loss": 2.2148193359375,
"step": 216000
},
{
"epoch": 6.6392713668005765,
"grad_norm": 0.9871698617935181,
"learning_rate": 0.0005602474133590884,
"loss": 2.212528076171875,
"step": 216500
},
{
"epoch": 6.654604557024134,
"grad_norm": 0.9603434801101685,
"learning_rate": 0.0005592199492432726,
"loss": 2.207598876953125,
"step": 217000
},
{
"epoch": 6.669937747247692,
"grad_norm": 0.9901469349861145,
"learning_rate": 0.0005581924851274569,
"loss": 2.22165625,
"step": 217500
},
{
"epoch": 6.68527093747125,
"grad_norm": 0.9955246448516846,
"learning_rate": 0.0005571650210116411,
"loss": 2.214988037109375,
"step": 218000
},
{
"epoch": 6.700604127694808,
"grad_norm": 0.9757549166679382,
"learning_rate": 0.0005561375568958254,
"loss": 2.21699609375,
"step": 218500
},
{
"epoch": 6.715937317918366,
"grad_norm": 0.9231974482536316,
"learning_rate": 0.0005551100927800096,
"loss": 2.21194140625,
"step": 219000
},
{
"epoch": 6.7312705081419235,
"grad_norm": 0.950542151927948,
"learning_rate": 0.0005540826286641938,
"loss": 2.212541015625,
"step": 219500
},
{
"epoch": 6.746603698365482,
"grad_norm": 1.0565048456192017,
"learning_rate": 0.0005530551645483781,
"loss": 2.21347412109375,
"step": 220000
},
{
"epoch": 6.76193688858904,
"grad_norm": 1.0073930025100708,
"learning_rate": 0.0005520277004325625,
"loss": 2.208374755859375,
"step": 220500
},
{
"epoch": 6.777270078812598,
"grad_norm": 1.0372698307037354,
"learning_rate": 0.0005510002363167467,
"loss": 2.209357666015625,
"step": 221000
},
{
"epoch": 6.792603269036156,
"grad_norm": 0.9625072479248047,
"learning_rate": 0.0005499727722009309,
"loss": 2.206805419921875,
"step": 221500
},
{
"epoch": 6.807936459259714,
"grad_norm": 0.9764012694358826,
"learning_rate": 0.0005489453080851152,
"loss": 2.2055869140625,
"step": 222000
},
{
"epoch": 6.823269649483271,
"grad_norm": 0.9529429078102112,
"learning_rate": 0.0005479178439692994,
"loss": 2.208606689453125,
"step": 222500
},
{
"epoch": 6.838602839706829,
"grad_norm": 0.9485899806022644,
"learning_rate": 0.0005468903798534837,
"loss": 2.20398095703125,
"step": 223000
},
{
"epoch": 6.853936029930387,
"grad_norm": 0.9776471853256226,
"learning_rate": 0.0005458629157376679,
"loss": 2.21119189453125,
"step": 223500
},
{
"epoch": 6.869269220153945,
"grad_norm": 0.9755150079727173,
"learning_rate": 0.0005448354516218521,
"loss": 2.20704736328125,
"step": 224000
},
{
"epoch": 6.884602410377503,
"grad_norm": 1.0282044410705566,
"learning_rate": 0.0005438079875060364,
"loss": 2.20789208984375,
"step": 224500
},
{
"epoch": 6.8999356006010615,
"grad_norm": 0.9649658203125,
"learning_rate": 0.0005427805233902207,
"loss": 2.20995751953125,
"step": 225000
},
{
"epoch": 6.915268790824619,
"grad_norm": 1.0128750801086426,
"learning_rate": 0.0005417530592744049,
"loss": 2.2080166015625,
"step": 225500
},
{
"epoch": 6.930601981048177,
"grad_norm": 0.9281025528907776,
"learning_rate": 0.0005407255951585891,
"loss": 2.204848388671875,
"step": 226000
},
{
"epoch": 6.945935171271735,
"grad_norm": 0.986391544342041,
"learning_rate": 0.0005396981310427733,
"loss": 2.209612548828125,
"step": 226500
},
{
"epoch": 6.961268361495293,
"grad_norm": 0.9156123995780945,
"learning_rate": 0.0005386706669269576,
"loss": 2.206895263671875,
"step": 227000
},
{
"epoch": 6.976601551718851,
"grad_norm": 1.0013313293457031,
"learning_rate": 0.0005376432028111419,
"loss": 2.21271484375,
"step": 227500
},
{
"epoch": 6.9919347419424085,
"grad_norm": 1.0261515378952026,
"learning_rate": 0.0005366157386953261,
"loss": 2.2027060546875,
"step": 228000
},
{
"epoch": 7.007267932165966,
"grad_norm": 0.9307826161384583,
"learning_rate": 0.0005355882745795103,
"loss": 2.20712548828125,
"step": 228500
},
{
"epoch": 7.022601122389524,
"grad_norm": 0.9147453904151917,
"learning_rate": 0.0005345608104636945,
"loss": 2.201402587890625,
"step": 229000
},
{
"epoch": 7.037934312613082,
"grad_norm": 1.0509899854660034,
"learning_rate": 0.0005335333463478788,
"loss": 2.202632080078125,
"step": 229500
},
{
"epoch": 7.05326750283664,
"grad_norm": 0.9868361949920654,
"learning_rate": 0.0005325058822320631,
"loss": 2.205843017578125,
"step": 230000
},
{
"epoch": 7.068600693060198,
"grad_norm": 1.004503846168518,
"learning_rate": 0.0005314784181162473,
"loss": 2.20215283203125,
"step": 230500
},
{
"epoch": 7.083933883283756,
"grad_norm": 1.0334880352020264,
"learning_rate": 0.0005304509540004315,
"loss": 2.198786865234375,
"step": 231000
},
{
"epoch": 7.099267073507314,
"grad_norm": 1.1348960399627686,
"learning_rate": 0.0005294234898846157,
"loss": 2.30681884765625,
"step": 231500
},
{
"epoch": 7.114600263730872,
"grad_norm": 0.995057225227356,
"learning_rate": 0.0005283960257688001,
"loss": 2.20882763671875,
"step": 232000
},
{
"epoch": 7.12993345395443,
"grad_norm": 1.0143413543701172,
"learning_rate": 0.0005273685616529843,
"loss": 2.206244384765625,
"step": 232500
},
{
"epoch": 7.145266644177988,
"grad_norm": 0.9252045154571533,
"learning_rate": 0.0005263410975371685,
"loss": 2.2012646484375,
"step": 233000
},
{
"epoch": 7.160599834401546,
"grad_norm": 0.9729524254798889,
"learning_rate": 0.0005253136334213527,
"loss": 2.209430908203125,
"step": 233500
},
{
"epoch": 7.175933024625103,
"grad_norm": 0.8740706443786621,
"learning_rate": 0.0005242861693055369,
"loss": 2.20389599609375,
"step": 234000
},
{
"epoch": 7.191266214848661,
"grad_norm": 0.9020913243293762,
"learning_rate": 0.0005232587051897213,
"loss": 2.201166259765625,
"step": 234500
},
{
"epoch": 7.206599405072219,
"grad_norm": 0.9470095634460449,
"learning_rate": 0.0005222312410739055,
"loss": 2.1985302734375,
"step": 235000
},
{
"epoch": 7.221932595295777,
"grad_norm": 1.0765324831008911,
"learning_rate": 0.0005212037769580897,
"loss": 2.19733642578125,
"step": 235500
},
{
"epoch": 7.237265785519335,
"grad_norm": 0.9833400249481201,
"learning_rate": 0.0005201763128422739,
"loss": 2.23849853515625,
"step": 236000
},
{
"epoch": 7.2525989757428935,
"grad_norm": 1.003823161125183,
"learning_rate": 0.0005191488487264582,
"loss": 2.20395166015625,
"step": 236500
},
{
"epoch": 7.267932165966451,
"grad_norm": 0.9548781514167786,
"learning_rate": 0.0005181213846106425,
"loss": 2.194429443359375,
"step": 237000
},
{
"epoch": 7.283265356190009,
"grad_norm": 0.9909761548042297,
"learning_rate": 0.0005170939204948268,
"loss": 2.1935244140625,
"step": 237500
},
{
"epoch": 7.298598546413567,
"grad_norm": 1.0807112455368042,
"learning_rate": 0.000516066456379011,
"loss": 2.191677978515625,
"step": 238000
},
{
"epoch": 7.313931736637125,
"grad_norm": 0.9172926545143127,
"learning_rate": 0.0005150389922631952,
"loss": 2.1903837890625,
"step": 238500
},
{
"epoch": 7.329264926860683,
"grad_norm": 0.9298128485679626,
"learning_rate": 0.0005140115281473795,
"loss": 2.1929833984375,
"step": 239000
},
{
"epoch": 7.3445981170842405,
"grad_norm": 0.9095985889434814,
"learning_rate": 0.0005129840640315638,
"loss": 2.188815673828125,
"step": 239500
},
{
"epoch": 7.359931307307798,
"grad_norm": 0.909279465675354,
"learning_rate": 0.000511956599915748,
"loss": 2.188396728515625,
"step": 240000
},
{
"epoch": 7.375264497531356,
"grad_norm": 0.9723482131958008,
"learning_rate": 0.0005109291357999322,
"loss": 2.188496337890625,
"step": 240500
},
{
"epoch": 7.390597687754914,
"grad_norm": 0.9592594504356384,
"learning_rate": 0.0005099016716841165,
"loss": 2.191003173828125,
"step": 241000
},
{
"epoch": 7.405930877978472,
"grad_norm": 0.9337832927703857,
"learning_rate": 0.0005088742075683007,
"loss": 2.186651611328125,
"step": 241500
},
{
"epoch": 7.42126406820203,
"grad_norm": 0.8740912079811096,
"learning_rate": 0.000507846743452485,
"loss": 2.190474853515625,
"step": 242000
},
{
"epoch": 7.436597258425588,
"grad_norm": 1.0178817510604858,
"learning_rate": 0.0005068192793366692,
"loss": 2.185767578125,
"step": 242500
},
{
"epoch": 7.451930448649146,
"grad_norm": 0.9444199800491333,
"learning_rate": 0.0005057918152208534,
"loss": 2.184332763671875,
"step": 243000
},
{
"epoch": 7.467263638872704,
"grad_norm": 0.9648879766464233,
"learning_rate": 0.0005047643511050377,
"loss": 2.188568603515625,
"step": 243500
},
{
"epoch": 7.482596829096262,
"grad_norm": 0.9435598254203796,
"learning_rate": 0.0005037368869892219,
"loss": 2.186260498046875,
"step": 244000
},
{
"epoch": 7.49793001931982,
"grad_norm": 0.9051665663719177,
"learning_rate": 0.0005027094228734062,
"loss": 2.179791259765625,
"step": 244500
},
{
"epoch": 7.513263209543378,
"grad_norm": 0.9058200716972351,
"learning_rate": 0.0005016819587575904,
"loss": 2.183138427734375,
"step": 245000
},
{
"epoch": 7.528596399766935,
"grad_norm": 0.9183915853500366,
"learning_rate": 0.0005006544946417746,
"loss": 2.183941162109375,
"step": 245500
},
{
"epoch": 7.543929589990493,
"grad_norm": 0.9438408017158508,
"learning_rate": 0.0004996270305259589,
"loss": 2.18027099609375,
"step": 246000
},
{
"epoch": 7.559262780214051,
"grad_norm": 0.9633219242095947,
"learning_rate": 0.0004985995664101431,
"loss": 2.1832861328125,
"step": 246500
},
{
"epoch": 7.574595970437609,
"grad_norm": 1.0026354789733887,
"learning_rate": 0.0004975721022943274,
"loss": 2.180103271484375,
"step": 247000
},
{
"epoch": 7.589929160661168,
"grad_norm": 0.9692184329032898,
"learning_rate": 0.0004965446381785117,
"loss": 2.18084765625,
"step": 247500
},
{
"epoch": 7.6052623508847255,
"grad_norm": 0.9162648320198059,
"learning_rate": 0.0004955171740626959,
"loss": 2.185460205078125,
"step": 248000
},
{
"epoch": 7.620595541108283,
"grad_norm": 0.9270127415657043,
"learning_rate": 0.0004944897099468801,
"loss": 2.18500927734375,
"step": 248500
},
{
"epoch": 7.635928731331841,
"grad_norm": 1.0042341947555542,
"learning_rate": 0.0004934622458310643,
"loss": 2.178227783203125,
"step": 249000
},
{
"epoch": 7.651261921555399,
"grad_norm": 0.9266281723976135,
"learning_rate": 0.0004924347817152486,
"loss": 2.181103515625,
"step": 249500
},
{
"epoch": 7.666595111778957,
"grad_norm": 1.0151432752609253,
"learning_rate": 0.0004914073175994329,
"loss": 2.177479736328125,
"step": 250000
},
{
"epoch": 7.681928302002515,
"grad_norm": 1.0316983461380005,
"learning_rate": 0.0004903798534836171,
"loss": 2.17549169921875,
"step": 250500
},
{
"epoch": 7.6972614922260725,
"grad_norm": 1.0134527683258057,
"learning_rate": 0.0004893523893678013,
"loss": 2.18060888671875,
"step": 251000
},
{
"epoch": 7.71259468244963,
"grad_norm": 0.9627741575241089,
"learning_rate": 0.0004883249252519855,
"loss": 2.17560595703125,
"step": 251500
},
{
"epoch": 7.727927872673188,
"grad_norm": 0.9785253405570984,
"learning_rate": 0.00048729746113616986,
"loss": 2.18788623046875,
"step": 252000
},
{
"epoch": 7.743261062896746,
"grad_norm": 0.9112822413444519,
"learning_rate": 0.0004862699970203541,
"loss": 2.18787353515625,
"step": 252500
},
{
"epoch": 7.758594253120304,
"grad_norm": 0.9322875738143921,
"learning_rate": 0.0004852425329045383,
"loss": 2.182326904296875,
"step": 253000
},
{
"epoch": 7.773927443343862,
"grad_norm": 0.9323176741600037,
"learning_rate": 0.0004842150687887226,
"loss": 2.18578662109375,
"step": 253500
},
{
"epoch": 7.78926063356742,
"grad_norm": 0.9617714285850525,
"learning_rate": 0.0004831876046729068,
"loss": 2.18440576171875,
"step": 254000
},
{
"epoch": 7.804593823790978,
"grad_norm": 0.938381016254425,
"learning_rate": 0.00048216014055709106,
"loss": 2.180271728515625,
"step": 254500
},
{
"epoch": 7.819927014014536,
"grad_norm": 0.9258219003677368,
"learning_rate": 0.0004811326764412753,
"loss": 2.18286572265625,
"step": 255000
},
{
"epoch": 7.835260204238094,
"grad_norm": 0.9577145576477051,
"learning_rate": 0.00048010521232545955,
"loss": 2.182063720703125,
"step": 255500
},
{
"epoch": 7.850593394461652,
"grad_norm": 1.116279125213623,
"learning_rate": 0.00047907774820964377,
"loss": 2.22189208984375,
"step": 256000
},
{
"epoch": 7.8659265846852096,
"grad_norm": 0.8886823654174805,
"learning_rate": 0.00047805028409382804,
"loss": 2.1770849609375,
"step": 256500
},
{
"epoch": 7.881259774908767,
"grad_norm": 1.0181430578231812,
"learning_rate": 0.00047702281997801226,
"loss": 2.172475830078125,
"step": 257000
},
{
"epoch": 7.896592965132325,
"grad_norm": 0.9926044940948486,
"learning_rate": 0.0004759953558621965,
"loss": 2.172684326171875,
"step": 257500
},
{
"epoch": 7.911926155355883,
"grad_norm": 0.9689538478851318,
"learning_rate": 0.00047496789174638075,
"loss": 2.177263671875,
"step": 258000
},
{
"epoch": 7.927259345579441,
"grad_norm": 0.9273969531059265,
"learning_rate": 0.00047394042763056497,
"loss": 2.17976025390625,
"step": 258500
},
{
"epoch": 7.942592535803,
"grad_norm": 0.9241510629653931,
"learning_rate": 0.0004729129635147493,
"loss": 2.187328125,
"step": 259000
},
{
"epoch": 7.9579257260265575,
"grad_norm": 0.9800302386283875,
"learning_rate": 0.0004718854993989335,
"loss": 2.171550048828125,
"step": 259500
},
{
"epoch": 7.973258916250115,
"grad_norm": 0.9281045198440552,
"learning_rate": 0.0004708580352831178,
"loss": 2.18274755859375,
"step": 260000
},
{
"epoch": 7.988592106473673,
"grad_norm": 0.9374285340309143,
"learning_rate": 0.000469830571167302,
"loss": 2.17093701171875,
"step": 260500
},
{
"epoch": 8.003925296697231,
"grad_norm": 0.9211281538009644,
"learning_rate": 0.0004688031070514862,
"loss": 2.1726689453125,
"step": 261000
},
{
"epoch": 8.019258486920789,
"grad_norm": 0.9847639799118042,
"learning_rate": 0.0004677756429356705,
"loss": 2.1776376953125,
"step": 261500
},
{
"epoch": 8.034591677144347,
"grad_norm": 0.9983856081962585,
"learning_rate": 0.0004667481788198547,
"loss": 2.18103759765625,
"step": 262000
},
{
"epoch": 8.049924867367904,
"grad_norm": 0.8992999792098999,
"learning_rate": 0.000465720714704039,
"loss": 2.1789794921875,
"step": 262500
},
{
"epoch": 8.065258057591462,
"grad_norm": 0.8943489193916321,
"learning_rate": 0.0004646932505882232,
"loss": 2.173632568359375,
"step": 263000
},
{
"epoch": 8.08059124781502,
"grad_norm": 0.9369834065437317,
"learning_rate": 0.0004636657864724075,
"loss": 2.170263916015625,
"step": 263500
},
{
"epoch": 8.095924438038578,
"grad_norm": 0.9863760471343994,
"learning_rate": 0.0004626383223565917,
"loss": 2.171566162109375,
"step": 264000
},
{
"epoch": 8.111257628262136,
"grad_norm": 0.9283600449562073,
"learning_rate": 0.00046161085824077597,
"loss": 2.171884765625,
"step": 264500
},
{
"epoch": 8.126590818485694,
"grad_norm": 0.9468367695808411,
"learning_rate": 0.0004605833941249602,
"loss": 2.168829345703125,
"step": 265000
},
{
"epoch": 8.141924008709251,
"grad_norm": 1.0426998138427734,
"learning_rate": 0.0004595559300091444,
"loss": 2.170163330078125,
"step": 265500
},
{
"epoch": 8.15725719893281,
"grad_norm": 0.9358944892883301,
"learning_rate": 0.0004585284658933287,
"loss": 2.17309228515625,
"step": 266000
},
{
"epoch": 8.172590389156367,
"grad_norm": 0.8933295011520386,
"learning_rate": 0.0004575010017775129,
"loss": 2.17467041015625,
"step": 266500
},
{
"epoch": 8.187923579379925,
"grad_norm": 0.9726329445838928,
"learning_rate": 0.00045647353766169717,
"loss": 2.170869140625,
"step": 267000
},
{
"epoch": 8.203256769603485,
"grad_norm": 0.9581248164176941,
"learning_rate": 0.00045544607354588144,
"loss": 2.16552099609375,
"step": 267500
},
{
"epoch": 8.218589959827042,
"grad_norm": 0.9545334577560425,
"learning_rate": 0.0004544186094300657,
"loss": 2.163555908203125,
"step": 268000
},
{
"epoch": 8.2339231500506,
"grad_norm": 0.9880333542823792,
"learning_rate": 0.00045339114531424993,
"loss": 2.162514892578125,
"step": 268500
},
{
"epoch": 8.249256340274158,
"grad_norm": 0.9676527380943298,
"learning_rate": 0.00045236368119843415,
"loss": 2.1637099609375,
"step": 269000
},
{
"epoch": 8.264589530497716,
"grad_norm": 1.0039492845535278,
"learning_rate": 0.0004513362170826184,
"loss": 2.16224853515625,
"step": 269500
},
{
"epoch": 8.279922720721274,
"grad_norm": 0.9783391952514648,
"learning_rate": 0.00045030875296680264,
"loss": 2.16271435546875,
"step": 270000
},
{
"epoch": 8.295255910944832,
"grad_norm": 0.9354684352874756,
"learning_rate": 0.0004492812888509869,
"loss": 2.162572998046875,
"step": 270500
},
{
"epoch": 8.31058910116839,
"grad_norm": 1.0112619400024414,
"learning_rate": 0.0004482538247351711,
"loss": 2.166249755859375,
"step": 271000
},
{
"epoch": 8.325922291391947,
"grad_norm": 0.9312315583229065,
"learning_rate": 0.0004472263606193554,
"loss": 2.170287841796875,
"step": 271500
},
{
"epoch": 8.341255481615505,
"grad_norm": 0.910420298576355,
"learning_rate": 0.0004461988965035396,
"loss": 2.1701572265625,
"step": 272000
},
{
"epoch": 8.356588671839063,
"grad_norm": 0.9733853936195374,
"learning_rate": 0.00044517143238772383,
"loss": 2.1632666015625,
"step": 272500
},
{
"epoch": 8.37192186206262,
"grad_norm": 0.9422973394393921,
"learning_rate": 0.0004441439682719081,
"loss": 2.168470947265625,
"step": 273000
},
{
"epoch": 8.387255052286179,
"grad_norm": 0.9360983967781067,
"learning_rate": 0.0004431165041560923,
"loss": 2.167650634765625,
"step": 273500
},
{
"epoch": 8.402588242509736,
"grad_norm": 0.8891287446022034,
"learning_rate": 0.0004420890400402766,
"loss": 2.16269970703125,
"step": 274000
},
{
"epoch": 8.417921432733294,
"grad_norm": 0.977506697177887,
"learning_rate": 0.0004410615759244608,
"loss": 2.1626337890625,
"step": 274500
},
{
"epoch": 8.433254622956852,
"grad_norm": 0.938169002532959,
"learning_rate": 0.0004400341118086451,
"loss": 2.1620302734375,
"step": 275000
},
{
"epoch": 8.44858781318041,
"grad_norm": 1.0033444166183472,
"learning_rate": 0.0004390066476928293,
"loss": 2.162121337890625,
"step": 275500
},
{
"epoch": 8.463921003403968,
"grad_norm": 0.9575499296188354,
"learning_rate": 0.00043797918357701363,
"loss": 2.16051953125,
"step": 276000
},
{
"epoch": 8.479254193627526,
"grad_norm": 0.9749202728271484,
"learning_rate": 0.00043695171946119785,
"loss": 2.15986572265625,
"step": 276500
},
{
"epoch": 8.494587383851083,
"grad_norm": 0.9810478687286377,
"learning_rate": 0.00043592425534538207,
"loss": 2.153389404296875,
"step": 277000
},
{
"epoch": 8.509920574074641,
"grad_norm": 12.6534423828125,
"learning_rate": 0.00043489679122956634,
"loss": 2.644532470703125,
"step": 277500
},
{
"epoch": 8.5252537642982,
"grad_norm": 1.1211611032485962,
"learning_rate": 0.00043386932711375056,
"loss": 3.080705810546875,
"step": 278000
},
{
"epoch": 8.540586954521757,
"grad_norm": 0.9509645104408264,
"learning_rate": 0.00043284186299793483,
"loss": 2.183162353515625,
"step": 278500
},
{
"epoch": 8.555920144745315,
"grad_norm": 1.0484975576400757,
"learning_rate": 0.00043181439888211905,
"loss": 2.1676123046875,
"step": 279000
},
{
"epoch": 8.571253334968874,
"grad_norm": 0.9419655203819275,
"learning_rate": 0.0004307869347663033,
"loss": 2.167958740234375,
"step": 279500
},
{
"epoch": 8.586586525192432,
"grad_norm": 0.9268501400947571,
"learning_rate": 0.00042975947065048754,
"loss": 2.158396728515625,
"step": 280000
},
{
"epoch": 8.60191971541599,
"grad_norm": 0.9125949144363403,
"learning_rate": 0.00042873200653467176,
"loss": 2.15858154296875,
"step": 280500
},
{
"epoch": 8.617252905639548,
"grad_norm": 1.851557970046997,
"learning_rate": 0.00042770454241885603,
"loss": 2.154983642578125,
"step": 281000
},
{
"epoch": 8.632586095863106,
"grad_norm": 0.9740294814109802,
"learning_rate": 0.00042667707830304025,
"loss": 2.184305908203125,
"step": 281500
},
{
"epoch": 8.647919286086664,
"grad_norm": 1.0088683366775513,
"learning_rate": 0.0004256496141872245,
"loss": 2.154712890625,
"step": 282000
},
{
"epoch": 8.663252476310221,
"grad_norm": 0.9518082141876221,
"learning_rate": 0.00042462215007140874,
"loss": 2.1606416015625,
"step": 282500
},
{
"epoch": 8.67858566653378,
"grad_norm": 0.9166815280914307,
"learning_rate": 0.000423594685955593,
"loss": 2.16301708984375,
"step": 283000
},
{
"epoch": 8.693918856757337,
"grad_norm": 0.9738340973854065,
"learning_rate": 0.00042256722183977723,
"loss": 2.161161376953125,
"step": 283500
},
{
"epoch": 8.709252046980895,
"grad_norm": 0.9328674077987671,
"learning_rate": 0.0004215397577239615,
"loss": 2.1595078125,
"step": 284000
},
{
"epoch": 8.724585237204453,
"grad_norm": 4.924217224121094,
"learning_rate": 0.0004205122936081458,
"loss": 2.155772216796875,
"step": 284500
},
{
"epoch": 8.73991842742801,
"grad_norm": 3.7930402755737305,
"learning_rate": 0.00041948482949233,
"loss": 2.1722607421875,
"step": 285000
},
{
"epoch": 8.755251617651568,
"grad_norm": 0.9515505433082581,
"learning_rate": 0.00041845736537651426,
"loss": 2.16401220703125,
"step": 285500
},
{
"epoch": 8.770584807875126,
"grad_norm": 0.9334007501602173,
"learning_rate": 0.0004174299012606985,
"loss": 2.15022705078125,
"step": 286000
},
{
"epoch": 8.785917998098684,
"grad_norm": 0.9569264650344849,
"learning_rate": 0.00041640243714488275,
"loss": 2.16190478515625,
"step": 286500
},
{
"epoch": 8.801251188322242,
"grad_norm": 0.9780637621879578,
"learning_rate": 0.00041537497302906697,
"loss": 2.154711181640625,
"step": 287000
},
{
"epoch": 8.8165843785458,
"grad_norm": 0.9233477711677551,
"learning_rate": 0.00041434750891325124,
"loss": 2.157458251953125,
"step": 287500
},
{
"epoch": 8.831917568769358,
"grad_norm": 0.9175972938537598,
"learning_rate": 0.00041332004479743546,
"loss": 2.157989990234375,
"step": 288000
},
{
"epoch": 8.847250758992915,
"grad_norm": 0.9064671993255615,
"learning_rate": 0.0004122925806816197,
"loss": 2.15585986328125,
"step": 288500
},
{
"epoch": 8.862583949216473,
"grad_norm": 0.9852584004402161,
"learning_rate": 0.00041126511656580395,
"loss": 2.1588828125,
"step": 289000
},
{
"epoch": 8.877917139440031,
"grad_norm": 0.9182608127593994,
"learning_rate": 0.00041023765244998817,
"loss": 2.1511474609375,
"step": 289500
},
{
"epoch": 8.89325032966359,
"grad_norm": 1.0061551332473755,
"learning_rate": 0.00040921018833417244,
"loss": 2.153777099609375,
"step": 290000
},
{
"epoch": 8.908583519887149,
"grad_norm": 0.9124777317047119,
"learning_rate": 0.00040818272421835666,
"loss": 2.144818115234375,
"step": 290500
},
{
"epoch": 8.923916710110706,
"grad_norm": 0.910224199295044,
"learning_rate": 0.00040715526010254093,
"loss": 2.1789833984375,
"step": 291000
},
{
"epoch": 8.939249900334264,
"grad_norm": 0.9544818997383118,
"learning_rate": 0.00040612779598672515,
"loss": 2.15167138671875,
"step": 291500
},
{
"epoch": 8.954583090557822,
"grad_norm": 0.9717366099357605,
"learning_rate": 0.0004051003318709094,
"loss": 2.151651611328125,
"step": 292000
},
{
"epoch": 8.96991628078138,
"grad_norm": 0.9445834159851074,
"learning_rate": 0.00040407286775509364,
"loss": 2.154056884765625,
"step": 292500
},
{
"epoch": 8.985249471004938,
"grad_norm": 0.8973835110664368,
"learning_rate": 0.00040304540363927786,
"loss": 2.148269775390625,
"step": 293000
},
{
"epoch": 9.000582661228496,
"grad_norm": 0.9345681667327881,
"learning_rate": 0.0004020179395234622,
"loss": 2.154309326171875,
"step": 293500
},
{
"epoch": 9.015915851452053,
"grad_norm": 0.9269891977310181,
"learning_rate": 0.0004009904754076464,
"loss": 2.1445849609375,
"step": 294000
},
{
"epoch": 9.031249041675611,
"grad_norm": 0.9665656089782715,
"learning_rate": 0.0003999630112918307,
"loss": 2.142473388671875,
"step": 294500
},
{
"epoch": 9.046582231899169,
"grad_norm": 0.8849844336509705,
"learning_rate": 0.0003989355471760149,
"loss": 2.144214111328125,
"step": 295000
},
{
"epoch": 9.061915422122727,
"grad_norm": 11.574191093444824,
"learning_rate": 0.00039790808306019917,
"loss": 3.21774755859375,
"step": 295500
},
{
"epoch": 9.077248612346285,
"grad_norm": 2.6705760955810547,
"learning_rate": 0.0003968806189443834,
"loss": 3.997998046875,
"step": 296000
},
{
"epoch": 9.092581802569843,
"grad_norm": 1.4139961004257202,
"learning_rate": 0.0003958531548285676,
"loss": 3.25447314453125,
"step": 296500
},
{
"epoch": 9.1079149927934,
"grad_norm": 1.9016342163085938,
"learning_rate": 0.0003948256907127519,
"loss": 3.186572998046875,
"step": 297000
},
{
"epoch": 9.123248183016958,
"grad_norm": 1.194314956665039,
"learning_rate": 0.0003937982265969361,
"loss": 2.78134716796875,
"step": 297500
},
{
"epoch": 9.138581373240516,
"grad_norm": 1.0150920152664185,
"learning_rate": 0.00039277076248112037,
"loss": 2.463370361328125,
"step": 298000
},
{
"epoch": 9.153914563464074,
"grad_norm": 0.9405332207679749,
"learning_rate": 0.0003917432983653046,
"loss": 2.188419189453125,
"step": 298500
},
{
"epoch": 9.169247753687632,
"grad_norm": 0.9128625392913818,
"learning_rate": 0.00039071583424948886,
"loss": 2.168464111328125,
"step": 299000
},
{
"epoch": 9.18458094391119,
"grad_norm": 1.00675630569458,
"learning_rate": 0.0003896883701336731,
"loss": 2.159985107421875,
"step": 299500
},
{
"epoch": 9.199914134134747,
"grad_norm": 0.8720121383666992,
"learning_rate": 0.0003886609060178573,
"loss": 2.157592041015625,
"step": 300000
},
{
"epoch": 9.215247324358305,
"grad_norm": 0.9381896257400513,
"learning_rate": 0.00038763344190204156,
"loss": 2.156147705078125,
"step": 300500
},
{
"epoch": 9.230580514581863,
"grad_norm": 0.9941139817237854,
"learning_rate": 0.0003866059777862258,
"loss": 2.149791748046875,
"step": 301000
},
{
"epoch": 9.245913704805421,
"grad_norm": 0.9148195385932922,
"learning_rate": 0.00038557851367041005,
"loss": 2.1458505859375,
"step": 301500
},
{
"epoch": 9.26124689502898,
"grad_norm": 0.9031514525413513,
"learning_rate": 0.0003845510495545943,
"loss": 2.143084716796875,
"step": 302000
},
{
"epoch": 9.276580085252538,
"grad_norm": 1.0594336986541748,
"learning_rate": 0.0003835235854387786,
"loss": 2.139436767578125,
"step": 302500
},
{
"epoch": 9.291913275476096,
"grad_norm": 0.879939079284668,
"learning_rate": 0.0003824961213229628,
"loss": 2.14239794921875,
"step": 303000
},
{
"epoch": 9.307246465699654,
"grad_norm": 0.9320951700210571,
"learning_rate": 0.0003814686572071471,
"loss": 2.144962158203125,
"step": 303500
},
{
"epoch": 9.322579655923212,
"grad_norm": 0.8990065455436707,
"learning_rate": 0.0003804411930913313,
"loss": 2.14126318359375,
"step": 304000
},
{
"epoch": 9.33791284614677,
"grad_norm": 0.90969318151474,
"learning_rate": 0.0003794137289755155,
"loss": 2.14310986328125,
"step": 304500
},
{
"epoch": 9.353246036370328,
"grad_norm": 0.9072972536087036,
"learning_rate": 0.0003783862648596998,
"loss": 2.140880859375,
"step": 305000
},
{
"epoch": 9.368579226593885,
"grad_norm": 0.9800154566764832,
"learning_rate": 0.000377358800743884,
"loss": 2.13557958984375,
"step": 305500
},
{
"epoch": 9.383912416817443,
"grad_norm": 0.9199201464653015,
"learning_rate": 0.0003763313366280683,
"loss": 2.150316650390625,
"step": 306000
},
{
"epoch": 9.399245607041001,
"grad_norm": 0.8888324499130249,
"learning_rate": 0.0003753038725122525,
"loss": 2.13763818359375,
"step": 306500
},
{
"epoch": 9.414578797264559,
"grad_norm": 0.908806324005127,
"learning_rate": 0.0003742764083964368,
"loss": 2.140972900390625,
"step": 307000
},
{
"epoch": 9.429911987488117,
"grad_norm": 1.0923656225204468,
"learning_rate": 0.000373248944280621,
"loss": 2.130876220703125,
"step": 307500
},
{
"epoch": 9.445245177711675,
"grad_norm": 0.9099012017250061,
"learning_rate": 0.0003722214801648052,
"loss": 2.132488525390625,
"step": 308000
},
{
"epoch": 9.460578367935232,
"grad_norm": 0.9411154985427856,
"learning_rate": 0.0003711940160489895,
"loss": 2.13144873046875,
"step": 308500
},
{
"epoch": 9.47591155815879,
"grad_norm": 0.8911005258560181,
"learning_rate": 0.0003701665519331737,
"loss": 2.133062255859375,
"step": 309000
},
{
"epoch": 9.491244748382348,
"grad_norm": 0.943794310092926,
"learning_rate": 0.000369139087817358,
"loss": 2.133711669921875,
"step": 309500
},
{
"epoch": 9.506577938605906,
"grad_norm": 0.914375364780426,
"learning_rate": 0.0003681116237015422,
"loss": 2.131699462890625,
"step": 310000
},
{
"epoch": 9.521911128829464,
"grad_norm": 0.9958600401878357,
"learning_rate": 0.0003670841595857265,
"loss": 2.127068603515625,
"step": 310500
},
{
"epoch": 9.537244319053022,
"grad_norm": 0.9124085307121277,
"learning_rate": 0.00036605669546991074,
"loss": 2.135460693359375,
"step": 311000
},
{
"epoch": 9.55257750927658,
"grad_norm": 0.9349147081375122,
"learning_rate": 0.000365029231354095,
"loss": 2.130185791015625,
"step": 311500
},
{
"epoch": 9.567910699500137,
"grad_norm": 0.8774728178977966,
"learning_rate": 0.00036400176723827923,
"loss": 2.131975830078125,
"step": 312000
},
{
"epoch": 9.583243889723695,
"grad_norm": 0.8731340169906616,
"learning_rate": 0.00036297430312246345,
"loss": 2.152654296875,
"step": 312500
},
{
"epoch": 9.598577079947255,
"grad_norm": 0.9262880086898804,
"learning_rate": 0.0003619468390066477,
"loss": 2.1312138671875,
"step": 313000
},
{
"epoch": 9.613910270170813,
"grad_norm": 0.9379519820213318,
"learning_rate": 0.00036091937489083194,
"loss": 2.1309052734375,
"step": 313500
},
{
"epoch": 9.62924346039437,
"grad_norm": 0.8927226662635803,
"learning_rate": 0.0003598919107750162,
"loss": 2.12682373046875,
"step": 314000
},
{
"epoch": 9.644576650617928,
"grad_norm": 0.9276019334793091,
"learning_rate": 0.00035886444665920043,
"loss": 2.125686767578125,
"step": 314500
},
{
"epoch": 9.659909840841486,
"grad_norm": 1.0309480428695679,
"learning_rate": 0.0003578369825433847,
"loss": 2.127134033203125,
"step": 315000
},
{
"epoch": 9.675243031065044,
"grad_norm": 0.9474682807922363,
"learning_rate": 0.0003568095184275689,
"loss": 2.12454931640625,
"step": 315500
},
{
"epoch": 9.690576221288602,
"grad_norm": 0.9093489646911621,
"learning_rate": 0.00035578205431175314,
"loss": 2.123470458984375,
"step": 316000
},
{
"epoch": 9.70590941151216,
"grad_norm": 0.8616217970848083,
"learning_rate": 0.0003547545901959374,
"loss": 2.126158447265625,
"step": 316500
},
{
"epoch": 9.721242601735717,
"grad_norm": 0.928504228591919,
"learning_rate": 0.0003537271260801216,
"loss": 2.274539794921875,
"step": 317000
},
{
"epoch": 9.736575791959275,
"grad_norm": 0.9357199668884277,
"learning_rate": 0.0003526996619643059,
"loss": 2.12382861328125,
"step": 317500
},
{
"epoch": 9.751908982182833,
"grad_norm": 0.8808205723762512,
"learning_rate": 0.0003516721978484901,
"loss": 2.11723876953125,
"step": 318000
},
{
"epoch": 9.767242172406391,
"grad_norm": 0.8950467109680176,
"learning_rate": 0.0003506447337326744,
"loss": 2.123248046875,
"step": 318500
},
{
"epoch": 9.782575362629949,
"grad_norm": 1.010391116142273,
"learning_rate": 0.00034961726961685866,
"loss": 2.129399658203125,
"step": 319000
},
{
"epoch": 9.797908552853507,
"grad_norm": 0.8765254616737366,
"learning_rate": 0.0003485898055010429,
"loss": 2.126660888671875,
"step": 319500
},
{
"epoch": 9.813241743077064,
"grad_norm": 0.977301299571991,
"learning_rate": 0.00034756234138522715,
"loss": 2.12415966796875,
"step": 320000
},
{
"epoch": 9.828574933300622,
"grad_norm": 0.9465354681015015,
"learning_rate": 0.00034653487726941137,
"loss": 2.11918408203125,
"step": 320500
},
{
"epoch": 9.84390812352418,
"grad_norm": 0.9051035046577454,
"learning_rate": 0.00034550741315359564,
"loss": 2.1199619140625,
"step": 321000
},
{
"epoch": 9.859241313747738,
"grad_norm": 0.8718971014022827,
"learning_rate": 0.00034447994903777986,
"loss": 2.13548828125,
"step": 321500
},
{
"epoch": 9.874574503971296,
"grad_norm": 0.9003779888153076,
"learning_rate": 0.00034345248492196413,
"loss": 2.116677001953125,
"step": 322000
},
{
"epoch": 9.889907694194854,
"grad_norm": 0.9519785642623901,
"learning_rate": 0.00034242502080614835,
"loss": 2.2166533203125,
"step": 322500
},
{
"epoch": 9.905240884418411,
"grad_norm": 0.9673523902893066,
"learning_rate": 0.0003413975566903326,
"loss": 2.12233984375,
"step": 323000
},
{
"epoch": 9.92057407464197,
"grad_norm": 0.9066923260688782,
"learning_rate": 0.00034037009257451684,
"loss": 2.1174404296875,
"step": 323500
},
{
"epoch": 9.935907264865527,
"grad_norm": 0.8994842767715454,
"learning_rate": 0.00033934262845870106,
"loss": 2.11081494140625,
"step": 324000
},
{
"epoch": 9.951240455089085,
"grad_norm": 0.9179783463478088,
"learning_rate": 0.00033831516434288533,
"loss": 2.11522021484375,
"step": 324500
},
{
"epoch": 9.966573645312645,
"grad_norm": 0.9965870380401611,
"learning_rate": 0.00033728770022706955,
"loss": 2.115026611328125,
"step": 325000
},
{
"epoch": 9.981906835536202,
"grad_norm": 0.9637007713317871,
"learning_rate": 0.0003362602361112538,
"loss": 2.119259033203125,
"step": 325500
},
{
"epoch": 9.99724002575976,
"grad_norm": 0.8663878440856934,
"learning_rate": 0.00033523277199543804,
"loss": 2.2038720703125,
"step": 326000
},
{
"epoch": 10.012573215983318,
"grad_norm": 0.9896288514137268,
"learning_rate": 0.0003342053078796223,
"loss": 2.11087353515625,
"step": 326500
},
{
"epoch": 10.027906406206876,
"grad_norm": 0.9026309251785278,
"learning_rate": 0.00033317784376380653,
"loss": 2.109000732421875,
"step": 327000
},
{
"epoch": 10.043239596430434,
"grad_norm": 0.9134894013404846,
"learning_rate": 0.00033215037964799075,
"loss": 2.110432373046875,
"step": 327500
},
{
"epoch": 10.058572786653992,
"grad_norm": 0.8728117346763611,
"learning_rate": 0.0003311229155321751,
"loss": 2.1112392578125,
"step": 328000
},
{
"epoch": 10.07390597687755,
"grad_norm": 1.249224305152893,
"learning_rate": 0.0003300954514163593,
"loss": 2.116403076171875,
"step": 328500
},
{
"epoch": 10.089239167101107,
"grad_norm": 0.948402464389801,
"learning_rate": 0.00032906798730054357,
"loss": 2.11359228515625,
"step": 329000
},
{
"epoch": 10.104572357324665,
"grad_norm": 1.0774781703948975,
"learning_rate": 0.0003280405231847278,
"loss": 2.114695556640625,
"step": 329500
},
{
"epoch": 10.119905547548223,
"grad_norm": 0.932476282119751,
"learning_rate": 0.00032701305906891206,
"loss": 2.1198720703125,
"step": 330000
},
{
"epoch": 10.13523873777178,
"grad_norm": 0.9639845490455627,
"learning_rate": 0.0003259855949530963,
"loss": 2.115079345703125,
"step": 330500
},
{
"epoch": 10.150571927995339,
"grad_norm": 0.910460889339447,
"learning_rate": 0.00032495813083728055,
"loss": 2.117462890625,
"step": 331000
},
{
"epoch": 10.165905118218896,
"grad_norm": 0.8573005199432373,
"learning_rate": 0.00032393066672146476,
"loss": 2.10928564453125,
"step": 331500
},
{
"epoch": 10.181238308442454,
"grad_norm": 1.0561102628707886,
"learning_rate": 0.000322903202605649,
"loss": 2.11848291015625,
"step": 332000
},
{
"epoch": 10.196571498666012,
"grad_norm": 0.9995536208152771,
"learning_rate": 0.00032187573848983325,
"loss": 2.11069775390625,
"step": 332500
},
{
"epoch": 10.21190468888957,
"grad_norm": 0.9712611436843872,
"learning_rate": 0.00032084827437401747,
"loss": 2.11180810546875,
"step": 333000
},
{
"epoch": 10.227237879113128,
"grad_norm": 0.8606237173080444,
"learning_rate": 0.00031982081025820174,
"loss": 2.112678955078125,
"step": 333500
},
{
"epoch": 10.242571069336686,
"grad_norm": 0.8895093202590942,
"learning_rate": 0.00031879334614238596,
"loss": 2.104147705078125,
"step": 334000
},
{
"epoch": 10.257904259560243,
"grad_norm": 1.012330412864685,
"learning_rate": 0.00031776588202657023,
"loss": 2.11276416015625,
"step": 334500
},
{
"epoch": 10.273237449783801,
"grad_norm": 0.8904862999916077,
"learning_rate": 0.00031673841791075445,
"loss": 2.109542724609375,
"step": 335000
},
{
"epoch": 10.288570640007359,
"grad_norm": 0.908838152885437,
"learning_rate": 0.00031571095379493867,
"loss": 2.106905029296875,
"step": 335500
},
{
"epoch": 10.303903830230919,
"grad_norm": 0.8937394618988037,
"learning_rate": 0.00031468348967912294,
"loss": 2.110737060546875,
"step": 336000
},
{
"epoch": 10.319237020454477,
"grad_norm": 0.9888873100280762,
"learning_rate": 0.0003136560255633072,
"loss": 2.117593017578125,
"step": 336500
},
{
"epoch": 10.334570210678034,
"grad_norm": 0.9693204760551453,
"learning_rate": 0.0003126285614474915,
"loss": 2.10942724609375,
"step": 337000
},
{
"epoch": 10.349903400901592,
"grad_norm": 0.9552232027053833,
"learning_rate": 0.0003116010973316757,
"loss": 2.105618408203125,
"step": 337500
},
{
"epoch": 10.36523659112515,
"grad_norm": 1.0517728328704834,
"learning_rate": 0.00031057363321586,
"loss": 2.108389892578125,
"step": 338000
},
{
"epoch": 10.380569781348708,
"grad_norm": 0.8790879249572754,
"learning_rate": 0.0003095461691000442,
"loss": 2.105360107421875,
"step": 338500
},
{
"epoch": 10.395902971572266,
"grad_norm": 0.8989303112030029,
"learning_rate": 0.00030851870498422847,
"loss": 2.137923095703125,
"step": 339000
},
{
"epoch": 10.411236161795824,
"grad_norm": 0.9769546389579773,
"learning_rate": 0.0003074912408684127,
"loss": 2.110882080078125,
"step": 339500
},
{
"epoch": 10.426569352019381,
"grad_norm": 0.8892316222190857,
"learning_rate": 0.0003064637767525969,
"loss": 2.11351318359375,
"step": 340000
},
{
"epoch": 10.44190254224294,
"grad_norm": 0.9441830515861511,
"learning_rate": 0.0003054363126367812,
"loss": 2.104545654296875,
"step": 340500
},
{
"epoch": 10.457235732466497,
"grad_norm": 1.9213511943817139,
"learning_rate": 0.0003044088485209654,
"loss": 2.103302490234375,
"step": 341000
},
{
"epoch": 10.472568922690055,
"grad_norm": 0.9644396305084229,
"learning_rate": 0.00030338138440514967,
"loss": 2.10061181640625,
"step": 341500
},
{
"epoch": 10.487902112913613,
"grad_norm": 1.0395784378051758,
"learning_rate": 0.0003023539202893339,
"loss": 2.101993408203125,
"step": 342000
},
{
"epoch": 10.50323530313717,
"grad_norm": 0.9795703887939453,
"learning_rate": 0.00030132645617351816,
"loss": 2.10506640625,
"step": 342500
},
{
"epoch": 10.518568493360728,
"grad_norm": 0.9227430820465088,
"learning_rate": 0.0003002989920577024,
"loss": 2.10047705078125,
"step": 343000
},
{
"epoch": 10.533901683584286,
"grad_norm": 0.8801776170730591,
"learning_rate": 0.0002992715279418866,
"loss": 2.101152587890625,
"step": 343500
},
{
"epoch": 10.549234873807844,
"grad_norm": 0.8818261027336121,
"learning_rate": 0.00029824406382607087,
"loss": 2.100804931640625,
"step": 344000
},
{
"epoch": 10.564568064031402,
"grad_norm": 0.8552242517471313,
"learning_rate": 0.0002972165997102551,
"loss": 2.09761376953125,
"step": 344500
},
{
"epoch": 10.57990125425496,
"grad_norm": 0.9127684831619263,
"learning_rate": 0.0002961891355944394,
"loss": 2.0985595703125,
"step": 345000
},
{
"epoch": 10.595234444478518,
"grad_norm": 0.8820174932479858,
"learning_rate": 0.00029516167147862363,
"loss": 2.098114501953125,
"step": 345500
},
{
"epoch": 10.610567634702075,
"grad_norm": 0.9346983432769775,
"learning_rate": 0.0002941342073628079,
"loss": 2.098978759765625,
"step": 346000
},
{
"epoch": 10.625900824925633,
"grad_norm": 0.91888427734375,
"learning_rate": 0.0002931067432469921,
"loss": 2.0967607421875,
"step": 346500
},
{
"epoch": 10.641234015149191,
"grad_norm": 0.9148829579353333,
"learning_rate": 0.00029207927913117634,
"loss": 2.093981689453125,
"step": 347000
},
{
"epoch": 10.656567205372749,
"grad_norm": 0.8980809450149536,
"learning_rate": 0.0002910518150153606,
"loss": 2.0962724609375,
"step": 347500
},
{
"epoch": 10.671900395596309,
"grad_norm": 0.9673241376876831,
"learning_rate": 0.0002900243508995448,
"loss": 2.09750048828125,
"step": 348000
},
{
"epoch": 10.687233585819866,
"grad_norm": 0.8674719929695129,
"learning_rate": 0.0002889968867837291,
"loss": 2.094869873046875,
"step": 348500
},
{
"epoch": 10.702566776043424,
"grad_norm": 0.9313533902168274,
"learning_rate": 0.0002879694226679133,
"loss": 2.09409765625,
"step": 349000
},
{
"epoch": 10.717899966266982,
"grad_norm": 0.9745606780052185,
"learning_rate": 0.0002869419585520976,
"loss": 2.095867431640625,
"step": 349500
},
{
"epoch": 10.73323315649054,
"grad_norm": 0.8968507051467896,
"learning_rate": 0.0002859144944362818,
"loss": 2.0949326171875,
"step": 350000
},
{
"epoch": 10.748566346714098,
"grad_norm": 1.0235860347747803,
"learning_rate": 0.0002848870303204661,
"loss": 2.097340576171875,
"step": 350500
},
{
"epoch": 10.763899536937656,
"grad_norm": 0.9557115435600281,
"learning_rate": 0.0002838595662046503,
"loss": 2.092142822265625,
"step": 351000
},
{
"epoch": 10.779232727161213,
"grad_norm": 0.9046032428741455,
"learning_rate": 0.0002828321020888345,
"loss": 2.09010400390625,
"step": 351500
},
{
"epoch": 10.794565917384771,
"grad_norm": 2.994778633117676,
"learning_rate": 0.0002818046379730188,
"loss": 2.0924736328125,
"step": 352000
},
{
"epoch": 10.809899107608329,
"grad_norm": 0.9183769822120667,
"learning_rate": 0.000280777173857203,
"loss": 2.09469580078125,
"step": 352500
},
{
"epoch": 10.825232297831887,
"grad_norm": 0.9352908134460449,
"learning_rate": 0.0002797497097413873,
"loss": 2.092965576171875,
"step": 353000
},
{
"epoch": 10.840565488055445,
"grad_norm": 0.9282081723213196,
"learning_rate": 0.00027872224562557155,
"loss": 2.0887353515625,
"step": 353500
},
{
"epoch": 10.855898678279003,
"grad_norm": 0.9089400768280029,
"learning_rate": 0.0002776947815097558,
"loss": 2.091353759765625,
"step": 354000
},
{
"epoch": 10.87123186850256,
"grad_norm": 0.9882640838623047,
"learning_rate": 0.00027666731739394004,
"loss": 2.094681396484375,
"step": 354500
},
{
"epoch": 10.886565058726118,
"grad_norm": 0.9600873589515686,
"learning_rate": 0.00027563985327812426,
"loss": 2.0882900390625,
"step": 355000
},
{
"epoch": 10.901898248949676,
"grad_norm": 0.8674167394638062,
"learning_rate": 0.00027461238916230853,
"loss": 2.089852294921875,
"step": 355500
},
{
"epoch": 10.917231439173234,
"grad_norm": 0.8718811869621277,
"learning_rate": 0.00027358492504649275,
"loss": 2.088258056640625,
"step": 356000
},
{
"epoch": 10.932564629396792,
"grad_norm": 0.9575982093811035,
"learning_rate": 0.000272557460930677,
"loss": 2.082998046875,
"step": 356500
},
{
"epoch": 10.94789781962035,
"grad_norm": 0.9020726084709167,
"learning_rate": 0.00027152999681486124,
"loss": 2.0901455078125,
"step": 357000
},
{
"epoch": 10.963231009843907,
"grad_norm": 0.8743335604667664,
"learning_rate": 0.0002705025326990455,
"loss": 2.091321044921875,
"step": 357500
},
{
"epoch": 10.978564200067465,
"grad_norm": 0.8623536825180054,
"learning_rate": 0.00026947506858322973,
"loss": 2.088880615234375,
"step": 358000
},
{
"epoch": 10.993897390291025,
"grad_norm": 0.9811893701553345,
"learning_rate": 0.000268447604467414,
"loss": 2.105343017578125,
"step": 358500
},
{
"epoch": 11.009230580514583,
"grad_norm": 0.9964583516120911,
"learning_rate": 0.0002674201403515982,
"loss": 2.08487841796875,
"step": 359000
},
{
"epoch": 11.02456377073814,
"grad_norm": 0.9179002046585083,
"learning_rate": 0.00026639267623578244,
"loss": 2.088212158203125,
"step": 359500
},
{
"epoch": 11.039896960961698,
"grad_norm": 0.8934545516967773,
"learning_rate": 0.0002653652121199667,
"loss": 2.0840595703125,
"step": 360000
},
{
"epoch": 11.055230151185256,
"grad_norm": 0.9651426672935486,
"learning_rate": 0.00026433774800415093,
"loss": 2.086373046875,
"step": 360500
},
{
"epoch": 11.070563341408814,
"grad_norm": 21.365129470825195,
"learning_rate": 0.0002633102838883352,
"loss": 2.1591787109375,
"step": 361000
},
{
"epoch": 11.085896531632372,
"grad_norm": 0.9601545333862305,
"learning_rate": 0.0002622828197725194,
"loss": 2.11279541015625,
"step": 361500
},
{
"epoch": 11.10122972185593,
"grad_norm": 0.9134415984153748,
"learning_rate": 0.00026125535565670375,
"loss": 2.08732421875,
"step": 362000
},
{
"epoch": 11.116562912079488,
"grad_norm": 0.8562743067741394,
"learning_rate": 0.00026022789154088796,
"loss": 2.08519189453125,
"step": 362500
},
{
"epoch": 11.131896102303045,
"grad_norm": 0.9580739736557007,
"learning_rate": 0.0002592004274250722,
"loss": 2.087134033203125,
"step": 363000
},
{
"epoch": 11.147229292526603,
"grad_norm": 0.9790233373641968,
"learning_rate": 0.00025817296330925645,
"loss": 2.0866435546875,
"step": 363500
},
{
"epoch": 11.162562482750161,
"grad_norm": 0.8779180645942688,
"learning_rate": 0.00025714549919344067,
"loss": 2.08897607421875,
"step": 364000
},
{
"epoch": 11.177895672973719,
"grad_norm": 0.9718131422996521,
"learning_rate": 0.00025611803507762494,
"loss": 2.0864853515625,
"step": 364500
},
{
"epoch": 11.193228863197277,
"grad_norm": 0.9318755269050598,
"learning_rate": 0.00025509057096180916,
"loss": 2.077686767578125,
"step": 365000
},
{
"epoch": 11.208562053420835,
"grad_norm": 0.9992250204086304,
"learning_rate": 0.00025406310684599343,
"loss": 2.07693115234375,
"step": 365500
},
{
"epoch": 11.223895243644392,
"grad_norm": 0.9219459891319275,
"learning_rate": 0.00025303564273017765,
"loss": 2.072782958984375,
"step": 366000
},
{
"epoch": 11.23922843386795,
"grad_norm": 0.8979462385177612,
"learning_rate": 0.00025200817861436187,
"loss": 2.074909912109375,
"step": 366500
},
{
"epoch": 11.254561624091508,
"grad_norm": 0.8906501531600952,
"learning_rate": 0.00025098071449854614,
"loss": 2.0798525390625,
"step": 367000
},
{
"epoch": 11.269894814315066,
"grad_norm": 0.8725651502609253,
"learning_rate": 0.0002499532503827304,
"loss": 2.082548095703125,
"step": 367500
},
{
"epoch": 11.285228004538624,
"grad_norm": 0.9534032940864563,
"learning_rate": 0.00024892578626691463,
"loss": 2.079510986328125,
"step": 368000
},
{
"epoch": 11.300561194762182,
"grad_norm": 0.9377962350845337,
"learning_rate": 0.00024789832215109885,
"loss": 2.079551025390625,
"step": 368500
},
{
"epoch": 11.31589438498574,
"grad_norm": 0.8636206984519958,
"learning_rate": 0.0002468708580352831,
"loss": 2.073847900390625,
"step": 369000
},
{
"epoch": 11.331227575209297,
"grad_norm": 0.923529326915741,
"learning_rate": 0.00024584339391946734,
"loss": 2.0712060546875,
"step": 369500
},
{
"epoch": 11.346560765432855,
"grad_norm": 0.8881893754005432,
"learning_rate": 0.0002448159298036516,
"loss": 2.08044287109375,
"step": 370000
},
{
"epoch": 11.361893955656415,
"grad_norm": 0.946375846862793,
"learning_rate": 0.00024378846568783586,
"loss": 2.076953857421875,
"step": 370500
},
{
"epoch": 11.377227145879973,
"grad_norm": 0.9080908894538879,
"learning_rate": 0.0002427610015720201,
"loss": 2.074939453125,
"step": 371000
},
{
"epoch": 11.39256033610353,
"grad_norm": 0.9097892045974731,
"learning_rate": 0.00024173353745620435,
"loss": 2.07150537109375,
"step": 371500
},
{
"epoch": 11.407893526327088,
"grad_norm": 1.0230767726898193,
"learning_rate": 0.0002407060733403886,
"loss": 2.073179443359375,
"step": 372000
},
{
"epoch": 11.423226716550646,
"grad_norm": 0.9200460314750671,
"learning_rate": 0.0002396786092245728,
"loss": 2.0725185546875,
"step": 372500
},
{
"epoch": 11.438559906774204,
"grad_norm": 0.900130569934845,
"learning_rate": 0.00023865114510875708,
"loss": 2.07635546875,
"step": 373000
},
{
"epoch": 11.453893096997762,
"grad_norm": 0.937246561050415,
"learning_rate": 0.00023762368099294133,
"loss": 2.079601318359375,
"step": 373500
},
{
"epoch": 11.46922628722132,
"grad_norm": 0.8678611516952515,
"learning_rate": 0.00023659621687712558,
"loss": 2.07959912109375,
"step": 374000
},
{
"epoch": 11.484559477444877,
"grad_norm": 0.9533618688583374,
"learning_rate": 0.00023556875276130982,
"loss": 2.066445068359375,
"step": 374500
},
{
"epoch": 11.499892667668435,
"grad_norm": 0.9741430878639221,
"learning_rate": 0.00023454128864549407,
"loss": 2.07080810546875,
"step": 375000
},
{
"epoch": 11.515225857891993,
"grad_norm": 0.9919676184654236,
"learning_rate": 0.0002335138245296783,
"loss": 2.066888427734375,
"step": 375500
},
{
"epoch": 11.53055904811555,
"grad_norm": 0.9268702864646912,
"learning_rate": 0.00023248636041386256,
"loss": 2.061945556640625,
"step": 376000
},
{
"epoch": 11.545892238339109,
"grad_norm": 0.878388524055481,
"learning_rate": 0.00023145889629804677,
"loss": 2.064855712890625,
"step": 376500
},
{
"epoch": 11.561225428562667,
"grad_norm": 0.9411481022834778,
"learning_rate": 0.00023043143218223102,
"loss": 2.059548095703125,
"step": 377000
},
{
"epoch": 11.576558618786224,
"grad_norm": 0.9315140843391418,
"learning_rate": 0.0002294039680664153,
"loss": 2.068278076171875,
"step": 377500
},
{
"epoch": 11.591891809009782,
"grad_norm": 0.8674419522285461,
"learning_rate": 0.00022837650395059954,
"loss": 2.06922998046875,
"step": 378000
},
{
"epoch": 11.60722499923334,
"grad_norm": 0.9791417121887207,
"learning_rate": 0.00022734903983478378,
"loss": 2.06313720703125,
"step": 378500
},
{
"epoch": 11.622558189456898,
"grad_norm": 0.9076842069625854,
"learning_rate": 0.00022632157571896803,
"loss": 2.060241943359375,
"step": 379000
},
{
"epoch": 11.637891379680456,
"grad_norm": 0.939439058303833,
"learning_rate": 0.00022529411160315227,
"loss": 2.0638212890625,
"step": 379500
},
{
"epoch": 11.653224569904014,
"grad_norm": 0.905220091342926,
"learning_rate": 0.00022426664748733652,
"loss": 2.057696044921875,
"step": 380000
},
{
"epoch": 11.668557760127571,
"grad_norm": 0.9110855460166931,
"learning_rate": 0.00022323918337152074,
"loss": 2.0625986328125,
"step": 380500
},
{
"epoch": 11.68389095035113,
"grad_norm": 0.9120020866394043,
"learning_rate": 0.00022221171925570498,
"loss": 2.06264013671875,
"step": 381000
},
{
"epoch": 11.699224140574689,
"grad_norm": 1.0115516185760498,
"learning_rate": 0.00022118425513988925,
"loss": 2.0600927734375,
"step": 381500
},
{
"epoch": 11.714557330798247,
"grad_norm": 0.923702597618103,
"learning_rate": 0.0002201567910240735,
"loss": 2.063225830078125,
"step": 382000
},
{
"epoch": 11.729890521021805,
"grad_norm": 0.9295987486839294,
"learning_rate": 0.00021912932690825774,
"loss": 2.0610546875,
"step": 382500
},
{
"epoch": 11.745223711245362,
"grad_norm": 0.9020257592201233,
"learning_rate": 0.000218101862792442,
"loss": 2.063927978515625,
"step": 383000
},
{
"epoch": 11.76055690146892,
"grad_norm": 0.9038189649581909,
"learning_rate": 0.00021707439867662623,
"loss": 2.06617919921875,
"step": 383500
},
{
"epoch": 11.775890091692478,
"grad_norm": 0.9462250471115112,
"learning_rate": 0.00021604693456081048,
"loss": 2.06784716796875,
"step": 384000
},
{
"epoch": 11.791223281916036,
"grad_norm": 0.8941705226898193,
"learning_rate": 0.0002150194704449947,
"loss": 2.058805908203125,
"step": 384500
},
{
"epoch": 11.806556472139594,
"grad_norm": 0.8866847157478333,
"learning_rate": 0.00021399200632917894,
"loss": 2.06271875,
"step": 385000
},
{
"epoch": 11.821889662363152,
"grad_norm": 0.9080286622047424,
"learning_rate": 0.0002129645422133632,
"loss": 2.061570556640625,
"step": 385500
},
{
"epoch": 11.83722285258671,
"grad_norm": 0.9109781980514526,
"learning_rate": 0.00021193707809754746,
"loss": 2.050642333984375,
"step": 386000
},
{
"epoch": 11.852556042810267,
"grad_norm": 0.9107651710510254,
"learning_rate": 0.0002109096139817317,
"loss": 2.06415869140625,
"step": 386500
},
{
"epoch": 11.867889233033825,
"grad_norm": 0.8821139335632324,
"learning_rate": 0.00020988214986591595,
"loss": 2.06503369140625,
"step": 387000
},
{
"epoch": 11.883222423257383,
"grad_norm": 0.9155604243278503,
"learning_rate": 0.0002088546857501002,
"loss": 2.0608984375,
"step": 387500
},
{
"epoch": 11.89855561348094,
"grad_norm": 0.9363220930099487,
"learning_rate": 0.0002078272216342844,
"loss": 2.059036376953125,
"step": 388000
},
{
"epoch": 11.913888803704499,
"grad_norm": 0.8350512385368347,
"learning_rate": 0.00020679975751846866,
"loss": 2.057044921875,
"step": 388500
},
{
"epoch": 11.929221993928056,
"grad_norm": 0.9073978662490845,
"learning_rate": 0.0002057722934026529,
"loss": 2.06073779296875,
"step": 389000
},
{
"epoch": 11.944555184151614,
"grad_norm": 0.9082510471343994,
"learning_rate": 0.00020474482928683715,
"loss": 2.059148193359375,
"step": 389500
},
{
"epoch": 11.959888374375172,
"grad_norm": 0.9777054786682129,
"learning_rate": 0.00020371736517102142,
"loss": 2.055695068359375,
"step": 390000
},
{
"epoch": 11.97522156459873,
"grad_norm": 0.960303783416748,
"learning_rate": 0.00020268990105520567,
"loss": 2.0562041015625,
"step": 390500
},
{
"epoch": 11.990554754822288,
"grad_norm": 0.8981908559799194,
"learning_rate": 0.0002016624369393899,
"loss": 2.058037353515625,
"step": 391000
},
{
"epoch": 12.005887945045846,
"grad_norm": 0.9634485244750977,
"learning_rate": 0.00020063497282357416,
"loss": 2.05224609375,
"step": 391500
},
{
"epoch": 12.021221135269403,
"grad_norm": 0.9679660201072693,
"learning_rate": 0.00019960750870775837,
"loss": 2.052638916015625,
"step": 392000
},
{
"epoch": 12.036554325492961,
"grad_norm": 0.9284456968307495,
"learning_rate": 0.00019858004459194262,
"loss": 2.048908203125,
"step": 392500
},
{
"epoch": 12.05188751571652,
"grad_norm": 0.8738199472427368,
"learning_rate": 0.00019755258047612686,
"loss": 2.051658935546875,
"step": 393000
},
{
"epoch": 12.067220705940079,
"grad_norm": 0.9431732296943665,
"learning_rate": 0.0001965251163603111,
"loss": 2.04869482421875,
"step": 393500
},
{
"epoch": 12.082553896163637,
"grad_norm": 0.9679067730903625,
"learning_rate": 0.00019549765224449535,
"loss": 2.053001708984375,
"step": 394000
},
{
"epoch": 12.097887086387194,
"grad_norm": 0.8935375213623047,
"learning_rate": 0.00019447018812867963,
"loss": 2.047596435546875,
"step": 394500
},
{
"epoch": 12.113220276610752,
"grad_norm": 0.91364586353302,
"learning_rate": 0.00019344272401286387,
"loss": 2.051766845703125,
"step": 395000
},
{
"epoch": 12.12855346683431,
"grad_norm": 0.9564911723136902,
"learning_rate": 0.00019241525989704812,
"loss": 2.0480810546875,
"step": 395500
},
{
"epoch": 12.143886657057868,
"grad_norm": 0.8961449861526489,
"learning_rate": 0.00019138779578123233,
"loss": 2.0425924072265627,
"step": 396000
},
{
"epoch": 12.159219847281426,
"grad_norm": 0.9604061841964722,
"learning_rate": 0.00019036033166541658,
"loss": 2.04914453125,
"step": 396500
},
{
"epoch": 12.174553037504984,
"grad_norm": 1.0155346393585205,
"learning_rate": 0.00018933286754960083,
"loss": 2.047794921875,
"step": 397000
},
{
"epoch": 12.189886227728541,
"grad_norm": 0.9058077931404114,
"learning_rate": 0.00018830540343378507,
"loss": 2.0435338134765626,
"step": 397500
},
{
"epoch": 12.2052194179521,
"grad_norm": 0.9513629078865051,
"learning_rate": 0.00018727793931796932,
"loss": 2.0454945068359374,
"step": 398000
},
{
"epoch": 12.220552608175657,
"grad_norm": 0.9357339143753052,
"learning_rate": 0.00018625047520215356,
"loss": 2.0360347900390625,
"step": 398500
},
{
"epoch": 12.235885798399215,
"grad_norm": 0.8954430222511292,
"learning_rate": 0.00018522301108633783,
"loss": 2.050967529296875,
"step": 399000
},
{
"epoch": 12.251218988622773,
"grad_norm": 0.8790498971939087,
"learning_rate": 0.00018419554697052208,
"loss": 2.0475169677734373,
"step": 399500
},
{
"epoch": 12.26655217884633,
"grad_norm": 0.9526025056838989,
"learning_rate": 0.0001831680828547063,
"loss": 2.04253076171875,
"step": 400000
},
{
"epoch": 12.281885369069888,
"grad_norm": 0.9807749390602112,
"learning_rate": 0.00018214061873889054,
"loss": 2.039591064453125,
"step": 400500
},
{
"epoch": 12.297218559293446,
"grad_norm": 0.94532710313797,
"learning_rate": 0.0001811131546230748,
"loss": 2.0433421630859376,
"step": 401000
},
{
"epoch": 12.312551749517004,
"grad_norm": 0.9574890732765198,
"learning_rate": 0.00018008569050725903,
"loss": 2.0430377197265623,
"step": 401500
},
{
"epoch": 12.327884939740562,
"grad_norm": 1.125672459602356,
"learning_rate": 0.00017905822639144328,
"loss": 2.04189306640625,
"step": 402000
},
{
"epoch": 12.34321812996412,
"grad_norm": 0.8750975728034973,
"learning_rate": 0.00017803076227562752,
"loss": 2.0426529541015626,
"step": 402500
},
{
"epoch": 12.358551320187678,
"grad_norm": 0.9902825355529785,
"learning_rate": 0.0001770032981598118,
"loss": 2.03882373046875,
"step": 403000
},
{
"epoch": 12.373884510411235,
"grad_norm": 0.8706808686256409,
"learning_rate": 0.00017597583404399604,
"loss": 2.040285888671875,
"step": 403500
},
{
"epoch": 12.389217700634793,
"grad_norm": 0.9436490535736084,
"learning_rate": 0.00017494836992818026,
"loss": 2.0391044921875,
"step": 404000
},
{
"epoch": 12.404550890858353,
"grad_norm": 0.9359686374664307,
"learning_rate": 0.0001739209058123645,
"loss": 2.040439697265625,
"step": 404500
},
{
"epoch": 12.41988408108191,
"grad_norm": 8.439556121826172,
"learning_rate": 0.00017289344169654875,
"loss": 2.052749755859375,
"step": 405000
},
{
"epoch": 12.435217271305469,
"grad_norm": 0.8805736899375916,
"learning_rate": 0.000171865977580733,
"loss": 2.049886962890625,
"step": 405500
},
{
"epoch": 12.450550461529026,
"grad_norm": 0.9033536314964294,
"learning_rate": 0.00017083851346491724,
"loss": 2.0480576171875,
"step": 406000
},
{
"epoch": 12.465883651752584,
"grad_norm": 0.9548377394676208,
"learning_rate": 0.00016981104934910148,
"loss": 2.032974853515625,
"step": 406500
},
{
"epoch": 12.481216841976142,
"grad_norm": 0.8679157495498657,
"learning_rate": 0.00016878358523328573,
"loss": 2.039317626953125,
"step": 407000
},
{
"epoch": 12.4965500321997,
"grad_norm": 0.9515289068222046,
"learning_rate": 0.00016775612111747,
"loss": 2.0402255859375,
"step": 407500
},
{
"epoch": 12.511883222423258,
"grad_norm": 0.8877235054969788,
"learning_rate": 0.00016672865700165422,
"loss": 2.0382567138671877,
"step": 408000
},
{
"epoch": 12.527216412646816,
"grad_norm": 0.93044513463974,
"learning_rate": 0.00016570119288583846,
"loss": 2.042698486328125,
"step": 408500
},
{
"epoch": 12.542549602870373,
"grad_norm": 0.9317615628242493,
"learning_rate": 0.0001646737287700227,
"loss": 2.0413780517578126,
"step": 409000
},
{
"epoch": 12.557882793093931,
"grad_norm": 0.9409142732620239,
"learning_rate": 0.00016364626465420695,
"loss": 2.0424705810546877,
"step": 409500
},
{
"epoch": 12.573215983317489,
"grad_norm": 0.9483290910720825,
"learning_rate": 0.0001626188005383912,
"loss": 2.04281787109375,
"step": 410000
},
{
"epoch": 12.588549173541047,
"grad_norm": 0.9657753705978394,
"learning_rate": 0.00016159133642257544,
"loss": 2.0340877685546874,
"step": 410500
},
{
"epoch": 12.603882363764605,
"grad_norm": 1.6388766765594482,
"learning_rate": 0.0001605638723067597,
"loss": 2.0358363037109375,
"step": 411000
},
{
"epoch": 12.619215553988163,
"grad_norm": 0.9225006699562073,
"learning_rate": 0.0001595364081909439,
"loss": 2.0344635009765626,
"step": 411500
},
{
"epoch": 12.63454874421172,
"grad_norm": 0.9115180969238281,
"learning_rate": 0.00015850894407512818,
"loss": 2.0352620849609373,
"step": 412000
},
{
"epoch": 12.649881934435278,
"grad_norm": 0.9649976491928101,
"learning_rate": 0.00015748147995931243,
"loss": 2.0334176025390627,
"step": 412500
},
{
"epoch": 12.665215124658836,
"grad_norm": 0.950308084487915,
"learning_rate": 0.00015645401584349667,
"loss": 2.034465087890625,
"step": 413000
},
{
"epoch": 12.680548314882394,
"grad_norm": 0.9390722513198853,
"learning_rate": 0.00015542655172768092,
"loss": 2.03224560546875,
"step": 413500
},
{
"epoch": 12.695881505105952,
"grad_norm": 0.8929451704025269,
"learning_rate": 0.00015439908761186516,
"loss": 2.0324617919921875,
"step": 414000
},
{
"epoch": 12.71121469532951,
"grad_norm": 0.9831867218017578,
"learning_rate": 0.0001533716234960494,
"loss": 2.0264443359375,
"step": 414500
},
{
"epoch": 12.726547885553067,
"grad_norm": 0.9382458925247192,
"learning_rate": 0.00015234415938023365,
"loss": 2.0276663818359375,
"step": 415000
},
{
"epoch": 12.741881075776625,
"grad_norm": 0.9780635237693787,
"learning_rate": 0.00015131669526441787,
"loss": 2.0322786865234375,
"step": 415500
},
{
"epoch": 12.757214266000185,
"grad_norm": 0.9014304280281067,
"learning_rate": 0.00015028923114860214,
"loss": 2.0272650146484374,
"step": 416000
},
{
"epoch": 12.772547456223743,
"grad_norm": 0.9504732489585876,
"learning_rate": 0.0001492617670327864,
"loss": 2.0319107666015626,
"step": 416500
},
{
"epoch": 12.7878806464473,
"grad_norm": 0.9370633959770203,
"learning_rate": 0.00014823430291697063,
"loss": 2.0276944580078125,
"step": 417000
},
{
"epoch": 12.803213836670858,
"grad_norm": 0.9153697490692139,
"learning_rate": 0.00014720683880115488,
"loss": 2.0308023681640623,
"step": 417500
},
{
"epoch": 12.818547026894416,
"grad_norm": 0.9047188758850098,
"learning_rate": 0.00014617937468533912,
"loss": 2.0271561279296875,
"step": 418000
},
{
"epoch": 12.833880217117974,
"grad_norm": 0.9106605052947998,
"learning_rate": 0.00014515191056952337,
"loss": 2.0228441162109374,
"step": 418500
},
{
"epoch": 12.849213407341532,
"grad_norm": 0.9332911372184753,
"learning_rate": 0.0001441244464537076,
"loss": 2.028383544921875,
"step": 419000
},
{
"epoch": 12.86454659756509,
"grad_norm": 0.9788764715194702,
"learning_rate": 0.00014309698233789183,
"loss": 2.0307564697265623,
"step": 419500
},
{
"epoch": 12.879879787788648,
"grad_norm": 0.9490921497344971,
"learning_rate": 0.00014206951822207608,
"loss": 2.032196533203125,
"step": 420000
},
{
"epoch": 12.895212978012205,
"grad_norm": 0.9331722259521484,
"learning_rate": 0.00014104205410626035,
"loss": 2.02581201171875,
"step": 420500
},
{
"epoch": 12.910546168235763,
"grad_norm": 0.9405240416526794,
"learning_rate": 0.0001400145899904446,
"loss": 2.0220235595703127,
"step": 421000
},
{
"epoch": 12.925879358459321,
"grad_norm": 0.9027151465415955,
"learning_rate": 0.00013898712587462884,
"loss": 2.023098388671875,
"step": 421500
},
{
"epoch": 12.941212548682879,
"grad_norm": 0.9208540320396423,
"learning_rate": 0.00013795966175881308,
"loss": 2.0254830322265627,
"step": 422000
},
{
"epoch": 12.956545738906437,
"grad_norm": 0.9127742648124695,
"learning_rate": 0.00013693219764299733,
"loss": 2.021534912109375,
"step": 422500
},
{
"epoch": 12.971878929129995,
"grad_norm": 0.9001808762550354,
"learning_rate": 0.00013590473352718157,
"loss": 2.0235911865234373,
"step": 423000
},
{
"epoch": 12.987212119353552,
"grad_norm": 1.0127555131912231,
"learning_rate": 0.0001348772694113658,
"loss": 2.02138330078125,
"step": 423500
},
{
"epoch": 13.00254530957711,
"grad_norm": 0.9362192749977112,
"learning_rate": 0.00013384980529555004,
"loss": 2.0196815185546875,
"step": 424000
},
{
"epoch": 13.017878499800668,
"grad_norm": 0.9487162828445435,
"learning_rate": 0.0001328223411797343,
"loss": 2.018721435546875,
"step": 424500
},
{
"epoch": 13.033211690024226,
"grad_norm": 0.948177695274353,
"learning_rate": 0.00013179487706391855,
"loss": 2.019599365234375,
"step": 425000
},
{
"epoch": 13.048544880247784,
"grad_norm": 0.8995773792266846,
"learning_rate": 0.0001307674129481028,
"loss": 2.0233519287109374,
"step": 425500
},
{
"epoch": 13.063878070471342,
"grad_norm": 0.9270117878913879,
"learning_rate": 0.00012973994883228704,
"loss": 2.017077880859375,
"step": 426000
},
{
"epoch": 13.0792112606949,
"grad_norm": 0.8888282179832458,
"learning_rate": 0.0001287124847164713,
"loss": 2.0177393798828125,
"step": 426500
},
{
"epoch": 13.094544450918459,
"grad_norm": 1.00867760181427,
"learning_rate": 0.00012768502060065553,
"loss": 2.01686376953125,
"step": 427000
},
{
"epoch": 13.109877641142017,
"grad_norm": 0.9551187753677368,
"learning_rate": 0.00012665755648483975,
"loss": 2.015956298828125,
"step": 427500
},
{
"epoch": 13.125210831365575,
"grad_norm": 1.001699686050415,
"learning_rate": 0.000125630092369024,
"loss": 2.018053955078125,
"step": 428000
},
{
"epoch": 13.140544021589132,
"grad_norm": 0.8624700903892517,
"learning_rate": 0.00012460262825320824,
"loss": 2.0144296875,
"step": 428500
},
{
"epoch": 13.15587721181269,
"grad_norm": 0.9099005460739136,
"learning_rate": 0.00012357516413739252,
"loss": 2.014934326171875,
"step": 429000
},
{
"epoch": 13.171210402036248,
"grad_norm": 0.9618480801582336,
"learning_rate": 0.00012254770002157676,
"loss": 2.0125889892578126,
"step": 429500
},
{
"epoch": 13.186543592259806,
"grad_norm": 0.9395309090614319,
"learning_rate": 0.000121520235905761,
"loss": 2.01322705078125,
"step": 430000
},
{
"epoch": 13.201876782483364,
"grad_norm": 0.9544713497161865,
"learning_rate": 0.00012049277178994524,
"loss": 2.0199425048828124,
"step": 430500
},
{
"epoch": 13.217209972706922,
"grad_norm": 0.9401258826255798,
"learning_rate": 0.00011946530767412948,
"loss": 2.01855322265625,
"step": 431000
},
{
"epoch": 13.23254316293048,
"grad_norm": 0.953959047794342,
"learning_rate": 0.00011843784355831373,
"loss": 2.01529150390625,
"step": 431500
},
{
"epoch": 13.247876353154037,
"grad_norm": 0.9421747326850891,
"learning_rate": 0.00011741037944249797,
"loss": 2.0098516845703127,
"step": 432000
},
{
"epoch": 13.263209543377595,
"grad_norm": 0.88308185338974,
"learning_rate": 0.00011638291532668222,
"loss": 2.0104246826171877,
"step": 432500
},
{
"epoch": 13.278542733601153,
"grad_norm": 0.9535665512084961,
"learning_rate": 0.00011535545121086646,
"loss": 2.0113984375,
"step": 433000
},
{
"epoch": 13.29387592382471,
"grad_norm": 0.8829141855239868,
"learning_rate": 0.00011432798709505071,
"loss": 2.0163486328125,
"step": 433500
},
{
"epoch": 13.309209114048269,
"grad_norm": 0.9442151784896851,
"learning_rate": 0.00011330052297923494,
"loss": 2.0098338623046876,
"step": 434000
},
{
"epoch": 13.324542304271827,
"grad_norm": 0.8852311968803406,
"learning_rate": 0.0001122730588634192,
"loss": 2.01128857421875,
"step": 434500
},
{
"epoch": 13.339875494495384,
"grad_norm": 0.9344000220298767,
"learning_rate": 0.00011124559474760344,
"loss": 2.00576123046875,
"step": 435000
},
{
"epoch": 13.355208684718942,
"grad_norm": 0.9251855611801147,
"learning_rate": 0.00011021813063178769,
"loss": 2.01498876953125,
"step": 435500
},
{
"epoch": 13.3705418749425,
"grad_norm": 0.9849985241889954,
"learning_rate": 0.00010919066651597192,
"loss": 2.0109813232421874,
"step": 436000
},
{
"epoch": 13.385875065166058,
"grad_norm": 0.9185920357704163,
"learning_rate": 0.00010816320240015618,
"loss": 2.013811279296875,
"step": 436500
},
{
"epoch": 13.401208255389616,
"grad_norm": 0.9730367660522461,
"learning_rate": 0.00010713573828434042,
"loss": 2.00818798828125,
"step": 437000
},
{
"epoch": 13.416541445613174,
"grad_norm": 4.857619762420654,
"learning_rate": 0.00010610827416852467,
"loss": 2.00988037109375,
"step": 437500
},
{
"epoch": 13.431874635836731,
"grad_norm": 0.9473047852516174,
"learning_rate": 0.0001050808100527089,
"loss": 2.005110595703125,
"step": 438000
},
{
"epoch": 13.44720782606029,
"grad_norm": 0.9315809607505798,
"learning_rate": 0.00010405334593689316,
"loss": 2.0071646728515624,
"step": 438500
},
{
"epoch": 13.462541016283849,
"grad_norm": 0.9317981004714966,
"learning_rate": 0.0001030258818210774,
"loss": 2.0065947265625,
"step": 439000
},
{
"epoch": 13.477874206507407,
"grad_norm": 0.9032931923866272,
"learning_rate": 0.00010199841770526165,
"loss": 2.001392333984375,
"step": 439500
},
{
"epoch": 13.493207396730964,
"grad_norm": 0.9399937391281128,
"learning_rate": 0.00010097095358944588,
"loss": 2.0057232666015623,
"step": 440000
},
{
"epoch": 13.508540586954522,
"grad_norm": 0.9294759631156921,
"learning_rate": 9.994348947363013e-05,
"loss": 2.0005076904296875,
"step": 440500
},
{
"epoch": 13.52387377717808,
"grad_norm": 0.9257121682167053,
"learning_rate": 9.891602535781439e-05,
"loss": 2.00176904296875,
"step": 441000
},
{
"epoch": 13.539206967401638,
"grad_norm": 0.9932507276535034,
"learning_rate": 9.788856124199863e-05,
"loss": 2.0089398193359376,
"step": 441500
},
{
"epoch": 13.554540157625196,
"grad_norm": 0.9416031241416931,
"learning_rate": 9.686109712618286e-05,
"loss": 2.0063912353515625,
"step": 442000
},
{
"epoch": 13.569873347848754,
"grad_norm": 0.9554703235626221,
"learning_rate": 9.583363301036711e-05,
"loss": 2.0038231201171874,
"step": 442500
},
{
"epoch": 13.585206538072311,
"grad_norm": 0.9067648649215698,
"learning_rate": 9.480616889455137e-05,
"loss": 2.0017965087890626,
"step": 443000
},
{
"epoch": 13.60053972829587,
"grad_norm": 0.9855777025222778,
"learning_rate": 9.377870477873561e-05,
"loss": 2.006759765625,
"step": 443500
},
{
"epoch": 13.615872918519427,
"grad_norm": 0.9550586938858032,
"learning_rate": 9.275124066291984e-05,
"loss": 2.00286181640625,
"step": 444000
},
{
"epoch": 13.631206108742985,
"grad_norm": 0.9628958702087402,
"learning_rate": 9.172377654710409e-05,
"loss": 2.0019564208984373,
"step": 444500
},
{
"epoch": 13.646539298966543,
"grad_norm": 0.9518256783485413,
"learning_rate": 9.069631243128835e-05,
"loss": 1.9976925048828125,
"step": 445000
},
{
"epoch": 13.6618724891901,
"grad_norm": 0.8920614719390869,
"learning_rate": 8.966884831547259e-05,
"loss": 2.0026600341796876,
"step": 445500
},
{
"epoch": 13.677205679413659,
"grad_norm": 0.9062846899032593,
"learning_rate": 8.864138419965682e-05,
"loss": 2.0046416015625,
"step": 446000
},
{
"epoch": 13.692538869637216,
"grad_norm": 0.8731572031974792,
"learning_rate": 8.761392008384107e-05,
"loss": 1.9965362548828125,
"step": 446500
},
{
"epoch": 13.707872059860774,
"grad_norm": 1.01910400390625,
"learning_rate": 8.658645596802533e-05,
"loss": 1.999089599609375,
"step": 447000
},
{
"epoch": 13.723205250084332,
"grad_norm": 0.9068704843521118,
"learning_rate": 8.555899185220957e-05,
"loss": 2.00007763671875,
"step": 447500
},
{
"epoch": 13.73853844030789,
"grad_norm": 0.9264721870422363,
"learning_rate": 8.45315277363938e-05,
"loss": 1.997175537109375,
"step": 448000
},
{
"epoch": 13.753871630531448,
"grad_norm": 0.9075028300285339,
"learning_rate": 8.350406362057805e-05,
"loss": 1.9973629150390626,
"step": 448500
},
{
"epoch": 13.769204820755006,
"grad_norm": 0.9317379593849182,
"learning_rate": 8.24765995047623e-05,
"loss": 1.996767333984375,
"step": 449000
},
{
"epoch": 13.784538010978565,
"grad_norm": 0.9457451105117798,
"learning_rate": 8.144913538894655e-05,
"loss": 1.998953857421875,
"step": 449500
},
{
"epoch": 13.799871201202123,
"grad_norm": 0.9694564938545227,
"learning_rate": 8.042167127313079e-05,
"loss": 1.994564697265625,
"step": 450000
},
{
"epoch": 13.81520439142568,
"grad_norm": 0.9401931762695312,
"learning_rate": 7.939420715731503e-05,
"loss": 1.99537109375,
"step": 450500
},
{
"epoch": 13.830537581649239,
"grad_norm": 0.9121079444885254,
"learning_rate": 7.836674304149928e-05,
"loss": 1.9967298583984374,
"step": 451000
},
{
"epoch": 13.845870771872796,
"grad_norm": 0.9305955767631531,
"learning_rate": 7.733927892568353e-05,
"loss": 1.9936163330078125,
"step": 451500
},
{
"epoch": 13.861203962096354,
"grad_norm": 0.9616714119911194,
"learning_rate": 7.631181480986777e-05,
"loss": 1.9941680908203125,
"step": 452000
},
{
"epoch": 13.876537152319912,
"grad_norm": 0.887218177318573,
"learning_rate": 7.528435069405201e-05,
"loss": 1.9960093994140624,
"step": 452500
},
{
"epoch": 13.89187034254347,
"grad_norm": 0.9147011041641235,
"learning_rate": 7.425688657823626e-05,
"loss": 1.9913450927734375,
"step": 453000
},
{
"epoch": 13.907203532767028,
"grad_norm": 0.9264606237411499,
"learning_rate": 7.32294224624205e-05,
"loss": 1.9949853515625,
"step": 453500
},
{
"epoch": 13.922536722990586,
"grad_norm": 0.930642306804657,
"learning_rate": 7.220195834660475e-05,
"loss": 1.9899346923828125,
"step": 454000
},
{
"epoch": 13.937869913214143,
"grad_norm": 0.854496955871582,
"learning_rate": 7.117449423078899e-05,
"loss": 1.998737548828125,
"step": 454500
},
{
"epoch": 13.953203103437701,
"grad_norm": 0.9076685309410095,
"learning_rate": 7.014703011497324e-05,
"loss": 1.9913065185546874,
"step": 455000
},
{
"epoch": 13.96853629366126,
"grad_norm": 0.895515501499176,
"learning_rate": 6.911956599915747e-05,
"loss": 1.9948375244140626,
"step": 455500
},
{
"epoch": 13.983869483884817,
"grad_norm": 0.9749481081962585,
"learning_rate": 6.809210188334173e-05,
"loss": 1.9882486572265625,
"step": 456000
},
{
"epoch": 13.999202674108375,
"grad_norm": 0.9416682124137878,
"learning_rate": 6.706463776752597e-05,
"loss": 1.98831982421875,
"step": 456500
},
{
"epoch": 14.014535864331933,
"grad_norm": 0.9078933000564575,
"learning_rate": 6.603717365171022e-05,
"loss": 1.987869873046875,
"step": 457000
},
{
"epoch": 14.02986905455549,
"grad_norm": 0.9235718250274658,
"learning_rate": 6.500970953589445e-05,
"loss": 1.99305859375,
"step": 457500
},
{
"epoch": 14.045202244779048,
"grad_norm": 0.9646036028862,
"learning_rate": 6.398224542007871e-05,
"loss": 1.9931959228515626,
"step": 458000
},
{
"epoch": 14.060535435002606,
"grad_norm": 1.0133482217788696,
"learning_rate": 6.295478130426295e-05,
"loss": 1.9905670166015625,
"step": 458500
},
{
"epoch": 14.075868625226164,
"grad_norm": 0.9169586300849915,
"learning_rate": 6.19273171884472e-05,
"loss": 1.992046142578125,
"step": 459000
},
{
"epoch": 14.091201815449722,
"grad_norm": 1.002814531326294,
"learning_rate": 6.089985307263144e-05,
"loss": 1.98700048828125,
"step": 459500
},
{
"epoch": 14.10653500567328,
"grad_norm": 0.9563496112823486,
"learning_rate": 5.987238895681568e-05,
"loss": 1.98731005859375,
"step": 460000
},
{
"epoch": 14.121868195896838,
"grad_norm": 0.9733923673629761,
"learning_rate": 5.884492484099993e-05,
"loss": 1.98325830078125,
"step": 460500
},
{
"epoch": 14.137201386120395,
"grad_norm": 0.9921107888221741,
"learning_rate": 5.781746072518417e-05,
"loss": 1.9850029296875,
"step": 461000
},
{
"epoch": 14.152534576343955,
"grad_norm": 0.9491721391677856,
"learning_rate": 5.6789996609368424e-05,
"loss": 1.9857259521484374,
"step": 461500
},
{
"epoch": 14.167867766567513,
"grad_norm": 0.8804633021354675,
"learning_rate": 5.576253249355266e-05,
"loss": 1.98178076171875,
"step": 462000
},
{
"epoch": 14.18320095679107,
"grad_norm": 0.9189617037773132,
"learning_rate": 5.473506837773691e-05,
"loss": 1.9854525146484374,
"step": 462500
},
{
"epoch": 14.198534147014628,
"grad_norm": 0.9965745210647583,
"learning_rate": 5.370760426192115e-05,
"loss": 1.98847705078125,
"step": 463000
},
{
"epoch": 14.213867337238186,
"grad_norm": 0.9616757035255432,
"learning_rate": 5.26801401461054e-05,
"loss": 1.985622802734375,
"step": 463500
},
{
"epoch": 14.229200527461744,
"grad_norm": 1.0184119939804077,
"learning_rate": 5.165267603028964e-05,
"loss": 1.989092041015625,
"step": 464000
},
{
"epoch": 14.244533717685302,
"grad_norm": 0.9619824290275574,
"learning_rate": 5.062521191447389e-05,
"loss": 1.98259423828125,
"step": 464500
},
{
"epoch": 14.25986690790886,
"grad_norm": 0.9330005645751953,
"learning_rate": 4.959774779865813e-05,
"loss": 1.9782156982421875,
"step": 465000
},
{
"epoch": 14.275200098132418,
"grad_norm": 0.9333367943763733,
"learning_rate": 4.857028368284238e-05,
"loss": 1.985651611328125,
"step": 465500
},
{
"epoch": 14.290533288355975,
"grad_norm": 0.9643346071243286,
"learning_rate": 4.754281956702662e-05,
"loss": 1.987324951171875,
"step": 466000
},
{
"epoch": 14.305866478579533,
"grad_norm": 0.9358896613121033,
"learning_rate": 4.651535545121087e-05,
"loss": 1.98485693359375,
"step": 466500
},
{
"epoch": 14.321199668803091,
"grad_norm": 0.9450936913490295,
"learning_rate": 4.5487891335395114e-05,
"loss": 1.9812447509765625,
"step": 467000
},
{
"epoch": 14.336532859026649,
"grad_norm": 0.8825874924659729,
"learning_rate": 4.446042721957936e-05,
"loss": 1.9822239990234376,
"step": 467500
},
{
"epoch": 14.351866049250207,
"grad_norm": 0.8992685675621033,
"learning_rate": 4.3432963103763604e-05,
"loss": 1.981752685546875,
"step": 468000
},
{
"epoch": 14.367199239473765,
"grad_norm": 0.9332273602485657,
"learning_rate": 4.240549898794785e-05,
"loss": 1.9774483642578125,
"step": 468500
},
{
"epoch": 14.382532429697322,
"grad_norm": 0.9260332584381104,
"learning_rate": 4.137803487213209e-05,
"loss": 1.983718505859375,
"step": 469000
},
{
"epoch": 14.39786561992088,
"grad_norm": 0.9751363396644592,
"learning_rate": 4.035057075631634e-05,
"loss": 1.9811126708984375,
"step": 469500
},
{
"epoch": 14.413198810144438,
"grad_norm": 0.95609450340271,
"learning_rate": 3.932310664050058e-05,
"loss": 1.981505126953125,
"step": 470000
},
{
"epoch": 14.428532000367996,
"grad_norm": 0.9047972559928894,
"learning_rate": 3.829564252468483e-05,
"loss": 1.9790108642578126,
"step": 470500
},
{
"epoch": 14.443865190591554,
"grad_norm": 0.951911985874176,
"learning_rate": 3.726817840886907e-05,
"loss": 1.9770003662109374,
"step": 471000
},
{
"epoch": 14.459198380815112,
"grad_norm": 0.98056560754776,
"learning_rate": 3.624071429305332e-05,
"loss": 1.98037890625,
"step": 471500
},
{
"epoch": 14.47453157103867,
"grad_norm": 0.9287095665931702,
"learning_rate": 3.521325017723756e-05,
"loss": 1.975437744140625,
"step": 472000
},
{
"epoch": 14.489864761262229,
"grad_norm": 0.9857012629508972,
"learning_rate": 3.418578606142181e-05,
"loss": 1.9744571533203126,
"step": 472500
},
{
"epoch": 14.505197951485787,
"grad_norm": 0.9061793088912964,
"learning_rate": 3.315832194560605e-05,
"loss": 1.97807421875,
"step": 473000
},
{
"epoch": 14.520531141709345,
"grad_norm": 0.9368249773979187,
"learning_rate": 3.21308578297903e-05,
"loss": 1.9766177978515624,
"step": 473500
},
{
"epoch": 14.535864331932903,
"grad_norm": 0.9416869282722473,
"learning_rate": 3.110339371397454e-05,
"loss": 1.97445361328125,
"step": 474000
},
{
"epoch": 14.55119752215646,
"grad_norm": 0.9494808912277222,
"learning_rate": 3.0075929598158784e-05,
"loss": 1.980681884765625,
"step": 474500
},
{
"epoch": 14.566530712380018,
"grad_norm": 0.9582274556159973,
"learning_rate": 2.904846548234303e-05,
"loss": 1.9713182373046876,
"step": 475000
},
{
"epoch": 14.581863902603576,
"grad_norm": 0.8838064670562744,
"learning_rate": 2.8021001366527275e-05,
"loss": 1.975603759765625,
"step": 475500
},
{
"epoch": 14.597197092827134,
"grad_norm": 0.9534491896629333,
"learning_rate": 2.699353725071152e-05,
"loss": 1.972388427734375,
"step": 476000
},
{
"epoch": 14.612530283050692,
"grad_norm": 0.96030193567276,
"learning_rate": 2.5966073134895765e-05,
"loss": 1.9766041259765625,
"step": 476500
},
{
"epoch": 14.62786347327425,
"grad_norm": 0.9311888217926025,
"learning_rate": 2.493860901908001e-05,
"loss": 1.9739630126953125,
"step": 477000
},
{
"epoch": 14.643196663497807,
"grad_norm": 0.9285484552383423,
"learning_rate": 2.3911144903264255e-05,
"loss": 1.97049609375,
"step": 477500
},
{
"epoch": 14.658529853721365,
"grad_norm": 0.9047533869743347,
"learning_rate": 2.28836807874485e-05,
"loss": 1.972680419921875,
"step": 478000
},
{
"epoch": 14.673863043944923,
"grad_norm": 0.9683652520179749,
"learning_rate": 2.1856216671632746e-05,
"loss": 1.9756221923828126,
"step": 478500
},
{
"epoch": 14.689196234168481,
"grad_norm": 0.8978590965270996,
"learning_rate": 2.082875255581699e-05,
"loss": 1.973049560546875,
"step": 479000
},
{
"epoch": 14.704529424392039,
"grad_norm": 0.9055408835411072,
"learning_rate": 1.9801288440001232e-05,
"loss": 1.973763671875,
"step": 479500
},
{
"epoch": 14.719862614615597,
"grad_norm": 0.8995787501335144,
"learning_rate": 1.8773824324185478e-05,
"loss": 1.9763970947265626,
"step": 480000
},
{
"epoch": 14.735195804839154,
"grad_norm": 0.9438687562942505,
"learning_rate": 1.7746360208369723e-05,
"loss": 1.971931396484375,
"step": 480500
},
{
"epoch": 14.750528995062712,
"grad_norm": 0.9335545301437378,
"learning_rate": 1.6718896092553968e-05,
"loss": 1.966485107421875,
"step": 481000
},
{
"epoch": 14.76586218528627,
"grad_norm": 0.9144617915153503,
"learning_rate": 1.5691431976738213e-05,
"loss": 1.9701805419921874,
"step": 481500
},
{
"epoch": 14.781195375509828,
"grad_norm": 1.01540207862854,
"learning_rate": 1.4663967860922457e-05,
"loss": 1.966320068359375,
"step": 482000
},
{
"epoch": 14.796528565733386,
"grad_norm": 0.9367926120758057,
"learning_rate": 1.3636503745106702e-05,
"loss": 1.9681883544921874,
"step": 482500
},
{
"epoch": 14.811861755956944,
"grad_norm": 0.9213653802871704,
"learning_rate": 1.2609039629290947e-05,
"loss": 1.97171142578125,
"step": 483000
},
{
"epoch": 14.827194946180501,
"grad_norm": 0.9544906616210938,
"learning_rate": 1.1581575513475192e-05,
"loss": 1.9717967529296876,
"step": 483500
},
{
"epoch": 14.84252813640406,
"grad_norm": 0.9205822944641113,
"learning_rate": 1.0554111397659437e-05,
"loss": 1.966518798828125,
"step": 484000
},
{
"epoch": 14.857861326627619,
"grad_norm": 0.9487109780311584,
"learning_rate": 9.52664728184368e-06,
"loss": 1.9655050048828124,
"step": 484500
},
{
"epoch": 14.873194516851177,
"grad_norm": 0.9579548239707947,
"learning_rate": 8.499183166027926e-06,
"loss": 1.9715491943359376,
"step": 485000
},
{
"epoch": 14.888527707074735,
"grad_norm": 0.9350531697273254,
"learning_rate": 7.471719050212171e-06,
"loss": 1.9708812255859376,
"step": 485500
},
{
"epoch": 14.903860897298292,
"grad_norm": 0.9493932723999023,
"learning_rate": 6.444254934396416e-06,
"loss": 1.968130615234375,
"step": 486000
},
{
"epoch": 14.91919408752185,
"grad_norm": 0.9266709089279175,
"learning_rate": 5.416790818580661e-06,
"loss": 1.9653431396484375,
"step": 486500
},
{
"epoch": 14.934527277745408,
"grad_norm": 0.9095956683158875,
"learning_rate": 4.3893267027649065e-06,
"loss": 1.9693433837890626,
"step": 487000
},
{
"epoch": 14.949860467968966,
"grad_norm": 0.9118430614471436,
"learning_rate": 3.361862586949151e-06,
"loss": 1.9688089599609375,
"step": 487500
},
{
"epoch": 14.965193658192524,
"grad_norm": 0.884293794631958,
"learning_rate": 2.334398471133396e-06,
"loss": 1.962416748046875,
"step": 488000
},
{
"epoch": 14.980526848416082,
"grad_norm": 0.9306288957595825,
"learning_rate": 1.3069343553176407e-06,
"loss": 1.9678748779296875,
"step": 488500
},
{
"epoch": 14.99586003863964,
"grad_norm": 0.8706272840499878,
"learning_rate": 2.7947023950188536e-07,
"loss": 1.9680279541015624,
"step": 489000
},
{
"epoch": 15.0,
"step": 489135,
"total_flos": 1.0411608145803264e+18,
"train_loss": 2.3089358288195254,
"train_runtime": 82476.5142,
"train_samples_per_second": 379.552,
"train_steps_per_second": 5.931
}
],
"logging_steps": 500,
"max_steps": 489135,
"num_input_tokens_seen": 0,
"num_train_epochs": 15,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.0411608145803264e+18,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}