ASD9987's picture
Add files using upload-large-folder tool
e861bd3 verified
Raw
History Blame Contribute Delete
55.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1227,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002449479485609308,
"grad_norm": 6.991664886474609,
"learning_rate": 8.064516129032259e-08,
"loss": 1.1740529537200928,
"step": 1,
"token_acc": 0.6749786507258754
},
{
"epoch": 0.01224739742804654,
"grad_norm": 7.713718891143799,
"learning_rate": 4.032258064516129e-07,
"loss": 1.169565200805664,
"step": 5,
"token_acc": 0.6757037037037037
},
{
"epoch": 0.02449479485609308,
"grad_norm": 9.855337142944336,
"learning_rate": 8.064516129032258e-07,
"loss": 1.1896038055419922,
"step": 10,
"token_acc": 0.6696705041683207
},
{
"epoch": 0.03674219228413962,
"grad_norm": 7.122698783874512,
"learning_rate": 1.2096774193548388e-06,
"loss": 1.1736026763916017,
"step": 15,
"token_acc": 0.6726278392080156
},
{
"epoch": 0.04898958971218616,
"grad_norm": 6.2926530838012695,
"learning_rate": 1.6129032258064516e-06,
"loss": 1.120145034790039,
"step": 20,
"token_acc": 0.6823338735818476
},
{
"epoch": 0.0612369871402327,
"grad_norm": 5.297961711883545,
"learning_rate": 2.0161290322580646e-06,
"loss": 1.0726344108581543,
"step": 25,
"token_acc": 0.6913205133330668
},
{
"epoch": 0.07348438456827924,
"grad_norm": 6.803680419921875,
"learning_rate": 2.4193548387096776e-06,
"loss": 1.0197155952453614,
"step": 30,
"token_acc": 0.7046268889235083
},
{
"epoch": 0.08573178199632578,
"grad_norm": 6.257656574249268,
"learning_rate": 2.822580645161291e-06,
"loss": 0.9886772155761718,
"step": 35,
"token_acc": 0.7097298395287427
},
{
"epoch": 0.09797917942437231,
"grad_norm": 5.260560035705566,
"learning_rate": 3.225806451612903e-06,
"loss": 0.9901582717895507,
"step": 40,
"token_acc": 0.7109607795763873
},
{
"epoch": 0.11022657685241886,
"grad_norm": 5.54268217086792,
"learning_rate": 3.6290322580645166e-06,
"loss": 0.9580869674682617,
"step": 45,
"token_acc": 0.7128370317346695
},
{
"epoch": 0.1224739742804654,
"grad_norm": 7.221277236938477,
"learning_rate": 4.032258064516129e-06,
"loss": 0.9542367935180665,
"step": 50,
"token_acc": 0.7172453703703704
},
{
"epoch": 0.13472137170851195,
"grad_norm": 4.5115251541137695,
"learning_rate": 4.435483870967742e-06,
"loss": 0.9257698059082031,
"step": 55,
"token_acc": 0.7236826311367183
},
{
"epoch": 0.14696876913655849,
"grad_norm": 4.466303825378418,
"learning_rate": 4.838709677419355e-06,
"loss": 0.9245384216308594,
"step": 60,
"token_acc": 0.7216376582278481
},
{
"epoch": 0.15921616656460502,
"grad_norm": 7.5447163581848145,
"learning_rate": 4.999918191571812e-06,
"loss": 0.9373905181884765,
"step": 65,
"token_acc": 0.7212690577604395
},
{
"epoch": 0.17146356399265156,
"grad_norm": 4.495761394500732,
"learning_rate": 4.999418270566514e-06,
"loss": 0.9201144218444824,
"step": 70,
"token_acc": 0.7242019451955285
},
{
"epoch": 0.1837109614206981,
"grad_norm": 4.361832141876221,
"learning_rate": 4.99846396845553e-06,
"loss": 0.8885965347290039,
"step": 75,
"token_acc": 0.7290909090909091
},
{
"epoch": 0.19595835884874463,
"grad_norm": 4.68657922744751,
"learning_rate": 4.9970554587259724e-06,
"loss": 0.913387680053711,
"step": 80,
"token_acc": 0.7288353101302741
},
{
"epoch": 0.2082057562767912,
"grad_norm": 4.364779472351074,
"learning_rate": 4.995192997437504e-06,
"loss": 0.9117264747619629,
"step": 85,
"token_acc": 0.7253237353781151
},
{
"epoch": 0.22045315370483773,
"grad_norm": 4.398984909057617,
"learning_rate": 4.992876923175802e-06,
"loss": 0.9025796890258789,
"step": 90,
"token_acc": 0.7308422301304863
},
{
"epoch": 0.23270055113288426,
"grad_norm": 6.909200191497803,
"learning_rate": 4.990107656990993e-06,
"loss": 0.930146598815918,
"step": 95,
"token_acc": 0.7192499903149576
},
{
"epoch": 0.2449479485609308,
"grad_norm": 7.305383682250977,
"learning_rate": 4.986885702321115e-06,
"loss": 0.9064167022705079,
"step": 100,
"token_acc": 0.7296472453428459
},
{
"epoch": 0.25719534598897736,
"grad_norm": 4.521721363067627,
"learning_rate": 4.983211644900594e-06,
"loss": 0.8999384880065918,
"step": 105,
"token_acc": 0.7293990920981765
},
{
"epoch": 0.2694427434170239,
"grad_norm": 4.209871292114258,
"learning_rate": 4.9790861526537596e-06,
"loss": 0.8929675102233887,
"step": 110,
"token_acc": 0.73212523194757
},
{
"epoch": 0.28169014084507044,
"grad_norm": 4.252216815948486,
"learning_rate": 4.974509975573422e-06,
"loss": 0.857850456237793,
"step": 115,
"token_acc": 0.7368546182192812
},
{
"epoch": 0.29393753827311697,
"grad_norm": 4.586759567260742,
"learning_rate": 4.9694839455845215e-06,
"loss": 0.9065601348876953,
"step": 120,
"token_acc": 0.7309617100515964
},
{
"epoch": 0.3061849357011635,
"grad_norm": 4.332535743713379,
"learning_rate": 4.964008976392897e-06,
"loss": 0.9019945144653321,
"step": 125,
"token_acc": 0.7280488760084594
},
{
"epoch": 0.31843233312921004,
"grad_norm": 4.497773170471191,
"learning_rate": 4.958086063319172e-06,
"loss": 0.9245548248291016,
"step": 130,
"token_acc": 0.7241136141038198
},
{
"epoch": 0.3306797305572566,
"grad_norm": 4.905919551849365,
"learning_rate": 4.951716283117817e-06,
"loss": 0.8866440773010253,
"step": 135,
"token_acc": 0.7321524217614201
},
{
"epoch": 0.3429271279853031,
"grad_norm": 7.825119972229004,
"learning_rate": 4.944900793781394e-06,
"loss": 0.8727359771728516,
"step": 140,
"token_acc": 0.7362950081024465
},
{
"epoch": 0.35517452541334965,
"grad_norm": 6.268953323364258,
"learning_rate": 4.937640834330046e-06,
"loss": 0.891855525970459,
"step": 145,
"token_acc": 0.7312945818788891
},
{
"epoch": 0.3674219228413962,
"grad_norm": 4.484751224517822,
"learning_rate": 4.929937724586244e-06,
"loss": 0.8802884101867676,
"step": 150,
"token_acc": 0.7316405180171368
},
{
"epoch": 0.3796693202694427,
"grad_norm": 4.88800048828125,
"learning_rate": 4.921792864934858e-06,
"loss": 0.8927088737487793,
"step": 155,
"token_acc": 0.7282748915210507
},
{
"epoch": 0.39191671769748926,
"grad_norm": 4.61529541015625,
"learning_rate": 4.913207736068565e-06,
"loss": 0.9064929962158204,
"step": 160,
"token_acc": 0.726866373902133
},
{
"epoch": 0.40416411512553585,
"grad_norm": 4.808589935302734,
"learning_rate": 4.904183898718672e-06,
"loss": 0.8911245346069336,
"step": 165,
"token_acc": 0.7311148608876505
},
{
"epoch": 0.4164115125535824,
"grad_norm": 5.910472869873047,
"learning_rate": 4.894722993371386e-06,
"loss": 0.8989328384399414,
"step": 170,
"token_acc": 0.7308344092963202
},
{
"epoch": 0.4286589099816289,
"grad_norm": 5.084550380706787,
"learning_rate": 4.884826739969575e-06,
"loss": 0.8640045166015625,
"step": 175,
"token_acc": 0.7379365396128518
},
{
"epoch": 0.44090630740967546,
"grad_norm": 8.873269081115723,
"learning_rate": 4.8744969376000965e-06,
"loss": 0.875187873840332,
"step": 180,
"token_acc": 0.7372251099560176
},
{
"epoch": 0.453153704837722,
"grad_norm": 4.072157859802246,
"learning_rate": 4.863735464166733e-06,
"loss": 0.8567923545837403,
"step": 185,
"token_acc": 0.7411574323803652
},
{
"epoch": 0.46540110226576853,
"grad_norm": 4.482424259185791,
"learning_rate": 4.8525442760487965e-06,
"loss": 0.8537416458129883,
"step": 190,
"token_acc": 0.7406744141741284
},
{
"epoch": 0.47764849969381507,
"grad_norm": 5.709517002105713,
"learning_rate": 4.840925407745471e-06,
"loss": 0.8654661178588867,
"step": 195,
"token_acc": 0.7388517279821628
},
{
"epoch": 0.4898958971218616,
"grad_norm": 6.7080817222595215,
"learning_rate": 4.828880971505951e-06,
"loss": 0.868289566040039,
"step": 200,
"token_acc": 0.7351872827403039
},
{
"epoch": 0.5021432945499081,
"grad_norm": 4.263020038604736,
"learning_rate": 4.816413156945441e-06,
"loss": 0.8504949569702148,
"step": 205,
"token_acc": 0.7372483488501631
},
{
"epoch": 0.5143906919779547,
"grad_norm": 4.941431522369385,
"learning_rate": 4.803524230647105e-06,
"loss": 0.87947998046875,
"step": 210,
"token_acc": 0.7340966921119593
},
{
"epoch": 0.5266380894060012,
"grad_norm": 3.8537113666534424,
"learning_rate": 4.790216535750006e-06,
"loss": 0.8551071166992188,
"step": 215,
"token_acc": 0.7394460138104206
},
{
"epoch": 0.5388854868340478,
"grad_norm": 6.776486873626709,
"learning_rate": 4.776492491523135e-06,
"loss": 0.8690934181213379,
"step": 220,
"token_acc": 0.735881841876629
},
{
"epoch": 0.5511328842620943,
"grad_norm": 5.204588890075684,
"learning_rate": 4.7623545929256045e-06,
"loss": 0.8763906478881835,
"step": 225,
"token_acc": 0.7309547163642308
},
{
"epoch": 0.5633802816901409,
"grad_norm": 3.7976465225219727,
"learning_rate": 4.747805410153077e-06,
"loss": 0.8602134704589843,
"step": 230,
"token_acc": 0.7396070560080287
},
{
"epoch": 0.5756276791181874,
"grad_norm": 4.394361972808838,
"learning_rate": 4.7328475881705154e-06,
"loss": 0.883229923248291,
"step": 235,
"token_acc": 0.7299810844892812
},
{
"epoch": 0.5878750765462339,
"grad_norm": 4.36138916015625,
"learning_rate": 4.717483846231345e-06,
"loss": 0.8553913116455079,
"step": 240,
"token_acc": 0.7385073441842
},
{
"epoch": 0.6001224739742804,
"grad_norm": 5.331536769866943,
"learning_rate": 4.701716977383105e-06,
"loss": 0.876121711730957,
"step": 245,
"token_acc": 0.7354679610130348
},
{
"epoch": 0.612369871402327,
"grad_norm": 3.9598731994628906,
"learning_rate": 4.685549847959687e-06,
"loss": 0.8674356460571289,
"step": 250,
"token_acc": 0.7374639427826293
},
{
"epoch": 0.6246172688303735,
"grad_norm": 4.912125110626221,
"learning_rate": 4.668985397060254e-06,
"loss": 0.8524209976196289,
"step": 255,
"token_acc": 0.7395983357337174
},
{
"epoch": 0.6368646662584201,
"grad_norm": 4.637007236480713,
"learning_rate": 4.6520266360149215e-06,
"loss": 0.8363597869873047,
"step": 260,
"token_acc": 0.7452462593516209
},
{
"epoch": 0.6491120636864667,
"grad_norm": 5.870323657989502,
"learning_rate": 4.63467664783732e-06,
"loss": 0.8575145721435546,
"step": 265,
"token_acc": 0.7388337231146724
},
{
"epoch": 0.6613594611145132,
"grad_norm": 3.967334032058716,
"learning_rate": 4.6169385866641185e-06,
"loss": 0.8751319885253906,
"step": 270,
"token_acc": 0.7333723425004877
},
{
"epoch": 0.6736068585425597,
"grad_norm": 3.4224841594696045,
"learning_rate": 4.59881567718161e-06,
"loss": 0.8715348243713379,
"step": 275,
"token_acc": 0.7351949172484117
},
{
"epoch": 0.6858542559706062,
"grad_norm": 5.422490119934082,
"learning_rate": 4.580311214039495e-06,
"loss": 0.8651253700256347,
"step": 280,
"token_acc": 0.7377377377377378
},
{
"epoch": 0.6981016533986528,
"grad_norm": 3.816480875015259,
"learning_rate": 4.561428561251919e-06,
"loss": 0.8501185417175293,
"step": 285,
"token_acc": 0.7394655305964368
},
{
"epoch": 0.7103490508266993,
"grad_norm": 6.062622547149658,
"learning_rate": 4.542171151585921e-06,
"loss": 0.8345209121704101,
"step": 290,
"token_acc": 0.7421023534986574
},
{
"epoch": 0.7225964482547459,
"grad_norm": 4.085614204406738,
"learning_rate": 4.522542485937369e-06,
"loss": 0.852170467376709,
"step": 295,
"token_acc": 0.7420508587823763
},
{
"epoch": 0.7348438456827924,
"grad_norm": 5.013440132141113,
"learning_rate": 4.502546132694518e-06,
"loss": 0.8618827819824219,
"step": 300,
"token_acc": 0.7351885098743267
},
{
"epoch": 0.747091243110839,
"grad_norm": 4.537975311279297,
"learning_rate": 4.482185727089293e-06,
"loss": 0.8637663841247558,
"step": 305,
"token_acc": 0.7376827179671008
},
{
"epoch": 0.7593386405388854,
"grad_norm": 11.192097663879395,
"learning_rate": 4.4614649705364295e-06,
"loss": 0.8398235321044922,
"step": 310,
"token_acc": 0.7414437965011059
},
{
"epoch": 0.771586037966932,
"grad_norm": 6.570138931274414,
"learning_rate": 4.440387629960568e-06,
"loss": 0.8514525413513183,
"step": 315,
"token_acc": 0.740827387218778
},
{
"epoch": 0.7838334353949785,
"grad_norm": 15.270551681518555,
"learning_rate": 4.418957537111453e-06,
"loss": 0.8647954940795899,
"step": 320,
"token_acc": 0.7368656837989737
},
{
"epoch": 0.7960808328230251,
"grad_norm": 7.683502674102783,
"learning_rate": 4.397178587867335e-06,
"loss": 0.849925422668457,
"step": 325,
"token_acc": 0.7428682320978071
},
{
"epoch": 0.8083282302510717,
"grad_norm": 5.864404201507568,
"learning_rate": 4.375054741526726e-06,
"loss": 0.8434003829956055,
"step": 330,
"token_acc": 0.7431786771964461
},
{
"epoch": 0.8205756276791182,
"grad_norm": 5.582365989685059,
"learning_rate": 4.352590020088617e-06,
"loss": 0.847816276550293,
"step": 335,
"token_acc": 0.7405130213432544
},
{
"epoch": 0.8328230251071648,
"grad_norm": 10.045018196105957,
"learning_rate": 4.3297885075213e-06,
"loss": 0.8608020782470703,
"step": 340,
"token_acc": 0.737972873758646
},
{
"epoch": 0.8450704225352113,
"grad_norm": 8.338502883911133,
"learning_rate": 4.306654349019925e-06,
"loss": 0.8354406356811523,
"step": 345,
"token_acc": 0.7470999492247002
},
{
"epoch": 0.8573178199632578,
"grad_norm": 5.34677791595459,
"learning_rate": 4.283191750252923e-06,
"loss": 0.8256412506103515,
"step": 350,
"token_acc": 0.7468063671166633
},
{
"epoch": 0.8695652173913043,
"grad_norm": 5.408663272857666,
"learning_rate": 4.259404976597442e-06,
"loss": 0.827094554901123,
"step": 355,
"token_acc": 0.7478493978313928
},
{
"epoch": 0.8818126148193509,
"grad_norm": 5.741966247558594,
"learning_rate": 4.2352983523639166e-06,
"loss": 0.8458335876464844,
"step": 360,
"token_acc": 0.7402354091149758
},
{
"epoch": 0.8940600122473974,
"grad_norm": 9.3356294631958,
"learning_rate": 4.21087626000994e-06,
"loss": 0.8354385375976563,
"step": 365,
"token_acc": 0.7407901761879337
},
{
"epoch": 0.906307409675444,
"grad_norm": 5.660871982574463,
"learning_rate": 4.186143139343549e-06,
"loss": 0.8446069717407226,
"step": 370,
"token_acc": 0.7382692384489438
},
{
"epoch": 0.9185548071034905,
"grad_norm": 4.74036979675293,
"learning_rate": 4.161103486716095e-06,
"loss": 0.8566476821899414,
"step": 375,
"token_acc": 0.7369387675354617
},
{
"epoch": 0.9308022045315371,
"grad_norm": 4.942912578582764,
"learning_rate": 4.135761854204825e-06,
"loss": 0.8482555389404297,
"step": 380,
"token_acc": 0.7435846508302216
},
{
"epoch": 0.9430496019595835,
"grad_norm": 4.156243324279785,
"learning_rate": 4.1101228487853475e-06,
"loss": 0.833610725402832,
"step": 385,
"token_acc": 0.7458942144854476
},
{
"epoch": 0.9552969993876301,
"grad_norm": 5.1740241050720215,
"learning_rate": 4.084191131494097e-06,
"loss": 0.8583192825317383,
"step": 390,
"token_acc": 0.7402297936627837
},
{
"epoch": 0.9675443968156767,
"grad_norm": 4.792290687561035,
"learning_rate": 4.057971416580992e-06,
"loss": 0.8479979515075684,
"step": 395,
"token_acc": 0.7428756971502789
},
{
"epoch": 0.9797917942437232,
"grad_norm": 5.586516380310059,
"learning_rate": 4.0314684706524056e-06,
"loss": 0.8402324676513672,
"step": 400,
"token_acc": 0.744393517760681
},
{
"epoch": 0.9920391916717698,
"grad_norm": 5.168452262878418,
"learning_rate": 4.004687111804616e-06,
"loss": 0.8356870651245117,
"step": 405,
"token_acc": 0.7430509531792133
},
{
"epoch": 1.0024494794856094,
"grad_norm": 5.489582538604736,
"learning_rate": 3.977632208747911e-06,
"loss": 0.8233718872070312,
"step": 410,
"token_acc": 0.7534750035339018
},
{
"epoch": 1.014696876913656,
"grad_norm": 5.425813674926758,
"learning_rate": 3.950308679921478e-06,
"loss": 0.6470079898834229,
"step": 415,
"token_acc": 0.7951952920311177
},
{
"epoch": 1.0269442743417023,
"grad_norm": 5.050281524658203,
"learning_rate": 3.922721492599255e-06,
"loss": 0.660604190826416,
"step": 420,
"token_acc": 0.7930899096976836
},
{
"epoch": 1.039191671769749,
"grad_norm": 15.731215476989746,
"learning_rate": 3.894875661986912e-06,
"loss": 0.6408588409423828,
"step": 425,
"token_acc": 0.7979481556895607
},
{
"epoch": 1.0514390691977955,
"grad_norm": 7.563592910766602,
"learning_rate": 3.866776250310107e-06,
"loss": 0.6218419075012207,
"step": 430,
"token_acc": 0.80126932211969
},
{
"epoch": 1.063686466625842,
"grad_norm": 10.406630516052246,
"learning_rate": 3.838428365894205e-06,
"loss": 0.6513051986694336,
"step": 435,
"token_acc": 0.7884600236313509
},
{
"epoch": 1.0759338640538885,
"grad_norm": 5.691953659057617,
"learning_rate": 3.8098371622356067e-06,
"loss": 0.6410561084747315,
"step": 440,
"token_acc": 0.7947625808487143
},
{
"epoch": 1.088181261481935,
"grad_norm": 10.795446395874023,
"learning_rate": 3.781007837064875e-06,
"loss": 0.6436257839202881,
"step": 445,
"token_acc": 0.7951966938282194
},
{
"epoch": 1.1004286589099816,
"grad_norm": 3.864281177520752,
"learning_rate": 3.7519456314018105e-06,
"loss": 0.6299556732177735,
"step": 450,
"token_acc": 0.7965261529920292
},
{
"epoch": 1.1126760563380282,
"grad_norm": 6.612410068511963,
"learning_rate": 3.7226558286026633e-06,
"loss": 0.6525215625762939,
"step": 455,
"token_acc": 0.7905346534653466
},
{
"epoch": 1.1249234537660746,
"grad_norm": 4.40219783782959,
"learning_rate": 3.6931437533996484e-06,
"loss": 0.6335238456726074,
"step": 460,
"token_acc": 0.7977228682170543
},
{
"epoch": 1.1371708511941212,
"grad_norm": 5.150554180145264,
"learning_rate": 3.663414770932933e-06,
"loss": 0.664974594116211,
"step": 465,
"token_acc": 0.7850033178500332
},
{
"epoch": 1.1494182486221678,
"grad_norm": 5.686324596405029,
"learning_rate": 3.6334742857752882e-06,
"loss": 0.6400838851928711,
"step": 470,
"token_acc": 0.7946844109388823
},
{
"epoch": 1.1616656460502144,
"grad_norm": 5.249683380126953,
"learning_rate": 3.603327740949559e-06,
"loss": 0.6418388366699219,
"step": 475,
"token_acc": 0.7934316249317419
},
{
"epoch": 1.1739130434782608,
"grad_norm": 5.586535930633545,
"learning_rate": 3.572980616939159e-06,
"loss": 0.6471261024475098,
"step": 480,
"token_acc": 0.7916852024103207
},
{
"epoch": 1.1861604409063073,
"grad_norm": 6.5408244132995605,
"learning_rate": 3.5424384306917392e-06,
"loss": 0.6485323429107666,
"step": 485,
"token_acc": 0.7901527623569049
},
{
"epoch": 1.198407838334354,
"grad_norm": 4.752018451690674,
"learning_rate": 3.511706734616244e-06,
"loss": 0.630111837387085,
"step": 490,
"token_acc": 0.7973464412740279
},
{
"epoch": 1.2106552357624005,
"grad_norm": 4.9688568115234375,
"learning_rate": 3.480791115573506e-06,
"loss": 0.651498556137085,
"step": 495,
"token_acc": 0.7929363712809835
},
{
"epoch": 1.2229026331904471,
"grad_norm": 4.710268974304199,
"learning_rate": 3.4496971938605877e-06,
"loss": 0.653368330001831,
"step": 500,
"token_acc": 0.7921513944223108
},
{
"epoch": 1.2351500306184935,
"grad_norm": 5.565318584442139,
"learning_rate": 3.41843062218904e-06,
"loss": 0.660006332397461,
"step": 505,
"token_acc": 0.7896928682977615
},
{
"epoch": 1.24739742804654,
"grad_norm": 5.7004313468933105,
"learning_rate": 3.3869970846572687e-06,
"loss": 0.6148884296417236,
"step": 510,
"token_acc": 0.8024188813652198
},
{
"epoch": 1.2596448254745867,
"grad_norm": 5.074853897094727,
"learning_rate": 3.3554022957171956e-06,
"loss": 0.645864200592041,
"step": 515,
"token_acc": 0.7956607946312819
},
{
"epoch": 1.2718922229026333,
"grad_norm": 8.077428817749023,
"learning_rate": 3.3236519991354e-06,
"loss": 0.6333592891693115,
"step": 520,
"token_acc": 0.7961572867644199
},
{
"epoch": 1.2841396203306799,
"grad_norm": 4.3480658531188965,
"learning_rate": 3.2917519669489344e-06,
"loss": 0.6460052490234375,
"step": 525,
"token_acc": 0.7965608260228243
},
{
"epoch": 1.2963870177587262,
"grad_norm": 4.82958984375,
"learning_rate": 3.259707998415998e-06,
"loss": 0.6225022792816162,
"step": 530,
"token_acc": 0.8001334013418605
},
{
"epoch": 1.3086344151867728,
"grad_norm": 5.863722324371338,
"learning_rate": 3.227525918961659e-06,
"loss": 0.631247615814209,
"step": 535,
"token_acc": 0.7949552800878708
},
{
"epoch": 1.3208818126148194,
"grad_norm": 5.4803853034973145,
"learning_rate": 3.195211579118828e-06,
"loss": 0.6522356510162354,
"step": 540,
"token_acc": 0.7866347496509076
},
{
"epoch": 1.3331292100428658,
"grad_norm": 4.040228366851807,
"learning_rate": 3.1627708534646552e-06,
"loss": 0.6348727226257325,
"step": 545,
"token_acc": 0.7953030659452678
},
{
"epoch": 1.3453766074709124,
"grad_norm": 4.5015363693237305,
"learning_rate": 3.1302096395525726e-06,
"loss": 0.654164981842041,
"step": 550,
"token_acc": 0.7893705416634005
},
{
"epoch": 1.357624004898959,
"grad_norm": 6.569249153137207,
"learning_rate": 3.0975338568401385e-06,
"loss": 0.6519898414611817,
"step": 555,
"token_acc": 0.7915543219666931
},
{
"epoch": 1.3698714023270055,
"grad_norm": 4.45184850692749,
"learning_rate": 3.0647494456129217e-06,
"loss": 0.6468002319335937,
"step": 560,
"token_acc": 0.7931159712694698
},
{
"epoch": 1.3821187997550521,
"grad_norm": 4.584598541259766,
"learning_rate": 3.031862365904581e-06,
"loss": 0.6463231086730957,
"step": 565,
"token_acc": 0.7936663658125025
},
{
"epoch": 1.3943661971830985,
"grad_norm": 6.708564281463623,
"learning_rate": 2.9988785964133665e-06,
"loss": 0.6645674705505371,
"step": 570,
"token_acc": 0.7846387300105054
},
{
"epoch": 1.406613594611145,
"grad_norm": 5.134814262390137,
"learning_rate": 2.9658041334152233e-06,
"loss": 0.6450769424438476,
"step": 575,
"token_acc": 0.7883453748667535
},
{
"epoch": 1.4188609920391917,
"grad_norm": 4.8022780418396,
"learning_rate": 2.9326449896736975e-06,
"loss": 0.645802116394043,
"step": 580,
"token_acc": 0.7931034482758621
},
{
"epoch": 1.4311083894672383,
"grad_norm": 11.037680625915527,
"learning_rate": 2.8994071933468527e-06,
"loss": 0.6355294227600098,
"step": 585,
"token_acc": 0.795571780017459
},
{
"epoch": 1.4433557868952849,
"grad_norm": 5.600420951843262,
"learning_rate": 2.8660967868913747e-06,
"loss": 0.6302547454833984,
"step": 590,
"token_acc": 0.7944136788067424
},
{
"epoch": 1.4556031843233312,
"grad_norm": 5.480545997619629,
"learning_rate": 2.832719825964093e-06,
"loss": 0.6494641304016113,
"step": 595,
"token_acc": 0.7888609058255233
},
{
"epoch": 1.4678505817513778,
"grad_norm": 5.16481351852417,
"learning_rate": 2.7992823783210893e-06,
"loss": 0.6297293663024902,
"step": 600,
"token_acc": 0.7950167588741267
},
{
"epoch": 1.4800979791794244,
"grad_norm": 4.243412971496582,
"learning_rate": 2.76579052271461e-06,
"loss": 0.6522404670715332,
"step": 605,
"token_acc": 0.7899070385126162
},
{
"epoch": 1.4923453766074708,
"grad_norm": 4.794040679931641,
"learning_rate": 2.732250347787986e-06,
"loss": 0.6418954372406006,
"step": 610,
"token_acc": 0.7959231766230705
},
{
"epoch": 1.5045927740355176,
"grad_norm": 7.080262660980225,
"learning_rate": 2.6986679509687457e-06,
"loss": 0.6371477127075196,
"step": 615,
"token_acc": 0.7943812344139651
},
{
"epoch": 1.516840171463564,
"grad_norm": 5.95667839050293,
"learning_rate": 2.6650494373601355e-06,
"loss": 0.640924596786499,
"step": 620,
"token_acc": 0.7974332292750607
},
{
"epoch": 1.5290875688916106,
"grad_norm": 4.725330829620361,
"learning_rate": 2.6314009186312437e-06,
"loss": 0.6558629512786865,
"step": 625,
"token_acc": 0.7909005555120883
},
{
"epoch": 1.5413349663196572,
"grad_norm": 4.347110271453857,
"learning_rate": 2.5977285119059347e-06,
"loss": 0.6575963973999024,
"step": 630,
"token_acc": 0.789889599070308
},
{
"epoch": 1.5535823637477035,
"grad_norm": 3.9792187213897705,
"learning_rate": 2.5640383386507812e-06,
"loss": 0.6560555458068847,
"step": 635,
"token_acc": 0.7933890786322774
},
{
"epoch": 1.5658297611757501,
"grad_norm": 6.149561405181885,
"learning_rate": 2.53033652356222e-06,
"loss": 0.6462264060974121,
"step": 640,
"token_acc": 0.7909087306305235
},
{
"epoch": 1.5780771586037967,
"grad_norm": 6.400949954986572,
"learning_rate": 2.4966291934531112e-06,
"loss": 0.64112229347229,
"step": 645,
"token_acc": 0.7910521891953848
},
{
"epoch": 1.590324556031843,
"grad_norm": 3.783074378967285,
"learning_rate": 2.462922476138918e-06,
"loss": 0.652773094177246,
"step": 650,
"token_acc": 0.7924428490459097
},
{
"epoch": 1.60257195345989,
"grad_norm": 5.220808506011963,
"learning_rate": 2.429222499323698e-06,
"loss": 0.6376447200775146,
"step": 655,
"token_acc": 0.7944650165659716
},
{
"epoch": 1.6148193508879363,
"grad_norm": 5.370139122009277,
"learning_rate": 2.3955353894861234e-06,
"loss": 0.6519615650177002,
"step": 660,
"token_acc": 0.7908662415336476
},
{
"epoch": 1.6270667483159829,
"grad_norm": 4.1884355545043945,
"learning_rate": 2.3618672707657176e-06,
"loss": 0.6450218677520752,
"step": 665,
"token_acc": 0.7953612078069386
},
{
"epoch": 1.6393141457440294,
"grad_norm": 5.350536823272705,
"learning_rate": 2.328224263849518e-06,
"loss": 0.6367081642150879,
"step": 670,
"token_acc": 0.7957014830676501
},
{
"epoch": 1.6515615431720758,
"grad_norm": 3.792351007461548,
"learning_rate": 2.2946124848593695e-06,
"loss": 0.6231406688690185,
"step": 675,
"token_acc": 0.8002352405524359
},
{
"epoch": 1.6638089406001226,
"grad_norm": 3.6872265338897705,
"learning_rate": 2.261038044240042e-06,
"loss": 0.6401480197906494,
"step": 680,
"token_acc": 0.7936702844014893
},
{
"epoch": 1.676056338028169,
"grad_norm": 3.924558162689209,
"learning_rate": 2.2275070456483862e-06,
"loss": 0.6520843029022216,
"step": 685,
"token_acc": 0.7895960927203659
},
{
"epoch": 1.6883037354562156,
"grad_norm": 3.8300726413726807,
"learning_rate": 2.1940255848437216e-06,
"loss": 0.6518651962280273,
"step": 690,
"token_acc": 0.7888732063043989
},
{
"epoch": 1.7005511328842622,
"grad_norm": 4.321022033691406,
"learning_rate": 2.160599748579658e-06,
"loss": 0.6466531753540039,
"step": 695,
"token_acc": 0.7931918008784773
},
{
"epoch": 1.7127985303123086,
"grad_norm": 6.859404563903809,
"learning_rate": 2.127235613497556e-06,
"loss": 0.636614990234375,
"step": 700,
"token_acc": 0.7948687207768853
},
{
"epoch": 1.7250459277403551,
"grad_norm": 3.922469139099121,
"learning_rate": 2.093939245021826e-06,
"loss": 0.630801773071289,
"step": 705,
"token_acc": 0.7957414509427932
},
{
"epoch": 1.7372933251684017,
"grad_norm": 4.203063488006592,
"learning_rate": 2.060716696257262e-06,
"loss": 0.6456500053405761,
"step": 710,
"token_acc": 0.7914672556959954
},
{
"epoch": 1.749540722596448,
"grad_norm": 4.85980749130249,
"learning_rate": 2.0275740068886293e-06,
"loss": 0.6230958938598633,
"step": 715,
"token_acc": 0.7996174841614535
},
{
"epoch": 1.761788120024495,
"grad_norm": 5.844249248504639,
"learning_rate": 1.9945172020826696e-06,
"loss": 0.6419669151306152,
"step": 720,
"token_acc": 0.7938420348058902
},
{
"epoch": 1.7740355174525413,
"grad_norm": 4.3660359382629395,
"learning_rate": 1.961552291392763e-06,
"loss": 0.6448030471801758,
"step": 725,
"token_acc": 0.7937423939072744
},
{
"epoch": 1.7862829148805879,
"grad_norm": 11.928579330444336,
"learning_rate": 1.9286852676664234e-06,
"loss": 0.6479958534240723,
"step": 730,
"token_acc": 0.7932216905901116
},
{
"epoch": 1.7985303123086345,
"grad_norm": 3.94968581199646,
"learning_rate": 1.8959221059558286e-06,
"loss": 0.6348490715026855,
"step": 735,
"token_acc": 0.795221027479092
},
{
"epoch": 1.8107777097366808,
"grad_norm": 4.603184700012207,
"learning_rate": 1.8632687624315885e-06,
"loss": 0.6441236972808838,
"step": 740,
"token_acc": 0.7924422561146471
},
{
"epoch": 1.8230251071647277,
"grad_norm": 10.49752426147461,
"learning_rate": 1.8307311732999433e-06,
"loss": 0.6095977783203125,
"step": 745,
"token_acc": 0.8042950884867767
},
{
"epoch": 1.835272504592774,
"grad_norm": 3.515354633331299,
"learning_rate": 1.798315253723593e-06,
"loss": 0.6375907897949219,
"step": 750,
"token_acc": 0.7981048989541558
},
{
"epoch": 1.8475199020208206,
"grad_norm": 4.343272686004639,
"learning_rate": 1.7660268967463515e-06,
"loss": 0.6393054008483887,
"step": 755,
"token_acc": 0.7919190309212624
},
{
"epoch": 1.8597672994488672,
"grad_norm": 4.648293495178223,
"learning_rate": 1.733871972221824e-06,
"loss": 0.6467394351959228,
"step": 760,
"token_acc": 0.7936601693246702
},
{
"epoch": 1.8720146968769136,
"grad_norm": 3.9130358695983887,
"learning_rate": 1.7018563257462956e-06,
"loss": 0.6282668113708496,
"step": 765,
"token_acc": 0.7994340427181456
},
{
"epoch": 1.8842620943049602,
"grad_norm": 4.328963756561279,
"learning_rate": 1.6699857775960343e-06,
"loss": 0.6343886375427246,
"step": 770,
"token_acc": 0.7952363738603044
},
{
"epoch": 1.8965094917330068,
"grad_norm": 5.430701732635498,
"learning_rate": 1.6382661216691968e-06,
"loss": 0.63662748336792,
"step": 775,
"token_acc": 0.7960699548044803
},
{
"epoch": 1.9087568891610531,
"grad_norm": 6.405421257019043,
"learning_rate": 1.606703124432522e-06,
"loss": 0.6205535888671875,
"step": 780,
"token_acc": 0.7973083356023183
},
{
"epoch": 1.9210042865891,
"grad_norm": 5.377999305725098,
"learning_rate": 1.5753025238730294e-06,
"loss": 0.6296541690826416,
"step": 785,
"token_acc": 0.7959222130097233
},
{
"epoch": 1.9332516840171463,
"grad_norm": 3.7075588703155518,
"learning_rate": 1.5440700284548708e-06,
"loss": 0.6614448070526123,
"step": 790,
"token_acc": 0.7879409034653465
},
{
"epoch": 1.945499081445193,
"grad_norm": 6.081875324249268,
"learning_rate": 1.513011316081567e-06,
"loss": 0.6293749809265137,
"step": 795,
"token_acc": 0.7978190505548706
},
{
"epoch": 1.9577464788732395,
"grad_norm": 4.014639854431152,
"learning_rate": 1.4821320330637962e-06,
"loss": 0.6189370632171631,
"step": 800,
"token_acc": 0.8002206635668689
},
{
"epoch": 1.9699938763012859,
"grad_norm": 4.775810718536377,
"learning_rate": 1.4514377930929235e-06,
"loss": 0.6305100440979003,
"step": 805,
"token_acc": 0.7961676553231346
},
{
"epoch": 1.9822412737293327,
"grad_norm": 11.449211120605469,
"learning_rate": 1.4209341762204587e-06,
"loss": 0.6374061107635498,
"step": 810,
"token_acc": 0.7956634175452556
},
{
"epoch": 1.994488671157379,
"grad_norm": 6.428242206573486,
"learning_rate": 1.3906267278436359e-06,
"loss": 0.6145306587219238,
"step": 815,
"token_acc": 0.8039574191451324
},
{
"epoch": 2.0048989589712187,
"grad_norm": 3.371020793914795,
"learning_rate": 1.3605209576972843e-06,
"loss": 0.5897501468658447,
"step": 820,
"token_acc": 0.8192170148691887
},
{
"epoch": 2.017146356399265,
"grad_norm": 3.5891268253326416,
"learning_rate": 1.3306223388521939e-06,
"loss": 0.4845391273498535,
"step": 825,
"token_acc": 0.8451054706940776
},
{
"epoch": 2.029393753827312,
"grad_norm": 4.491250514984131,
"learning_rate": 1.3009363067201371e-06,
"loss": 0.4741975784301758,
"step": 830,
"token_acc": 0.8496166699429919
},
{
"epoch": 2.0416411512553583,
"grad_norm": 3.665771722793579,
"learning_rate": 1.2714682580657379e-06,
"loss": 0.4662147045135498,
"step": 835,
"token_acc": 0.8462080300540032
},
{
"epoch": 2.0538885486834046,
"grad_norm": 4.234457015991211,
"learning_rate": 1.2422235500253687e-06,
"loss": 0.4687329292297363,
"step": 840,
"token_acc": 0.8483358695228062
},
{
"epoch": 2.0661359461114515,
"grad_norm": 13.02021312713623,
"learning_rate": 1.2132074991332542e-06,
"loss": 0.4781540870666504,
"step": 845,
"token_acc": 0.8460030227523385
},
{
"epoch": 2.078383343539498,
"grad_norm": 4.561012268066406,
"learning_rate": 1.18442538035495e-06,
"loss": 0.4608255386352539,
"step": 850,
"token_acc": 0.8514021748017333
},
{
"epoch": 2.090630740967544,
"grad_norm": 5.0508270263671875,
"learning_rate": 1.1558824261283829e-06,
"loss": 0.4783477783203125,
"step": 855,
"token_acc": 0.8410647004010857
},
{
"epoch": 2.102878138395591,
"grad_norm": 10.500925064086914,
"learning_rate": 1.1275838254126214e-06,
"loss": 0.47881574630737306,
"step": 860,
"token_acc": 0.8417210313762038
},
{
"epoch": 2.1151255358236374,
"grad_norm": 4.909471035003662,
"learning_rate": 1.0995347227445502e-06,
"loss": 0.4551267623901367,
"step": 865,
"token_acc": 0.8498848017796139
},
{
"epoch": 2.127372933251684,
"grad_norm": 5.86276388168335,
"learning_rate": 1.0717402173036187e-06,
"loss": 0.46577911376953124,
"step": 870,
"token_acc": 0.8471839306813707
},
{
"epoch": 2.1396203306797306,
"grad_norm": 3.647233486175537,
"learning_rate": 1.044205361984842e-06,
"loss": 0.4693206787109375,
"step": 875,
"token_acc": 0.8472078387636421
},
{
"epoch": 2.151867728107777,
"grad_norm": 6.777562141418457,
"learning_rate": 1.0169351624802035e-06,
"loss": 0.4774106502532959,
"step": 880,
"token_acc": 0.844955431095685
},
{
"epoch": 2.1641151255358237,
"grad_norm": 4.635525226593018,
"learning_rate": 9.899345763686538e-07,
"loss": 0.47169036865234376,
"step": 885,
"token_acc": 0.8453473635700296
},
{
"epoch": 2.17636252296387,
"grad_norm": 5.107778072357178,
"learning_rate": 9.632085122148424e-07,
"loss": 0.4596005916595459,
"step": 890,
"token_acc": 0.8491451821133759
},
{
"epoch": 2.188609920391917,
"grad_norm": 5.593906879425049,
"learning_rate": 9.367618286767711e-07,
"loss": 0.4577906131744385,
"step": 895,
"token_acc": 0.852890604103779
},
{
"epoch": 2.2008573178199633,
"grad_norm": 8.50526237487793,
"learning_rate": 9.105993336225161e-07,
"loss": 0.4655435562133789,
"step": 900,
"token_acc": 0.8478589420654912
},
{
"epoch": 2.2131047152480097,
"grad_norm": 4.241094589233398,
"learning_rate": 8.847257832561781e-07,
"loss": 0.47768077850341795,
"step": 905,
"token_acc": 0.8461073744411777
},
{
"epoch": 2.2253521126760565,
"grad_norm": 4.346312999725342,
"learning_rate": 8.591458812532336e-07,
"loss": 0.4508483409881592,
"step": 910,
"token_acc": 0.8519886145088084
},
{
"epoch": 2.237599510104103,
"grad_norm": 4.39501953125,
"learning_rate": 8.338642779054288e-07,
"loss": 0.4618703365325928,
"step": 915,
"token_acc": 0.8477535805375711
},
{
"epoch": 2.249846907532149,
"grad_norm": 5.004550933837891,
"learning_rate": 8.088855692753828e-07,
"loss": 0.47423648834228516,
"step": 920,
"token_acc": 0.8433487623264239
},
{
"epoch": 2.262094304960196,
"grad_norm": 8.192081451416016,
"learning_rate": 7.842142963610467e-07,
"loss": 0.46451425552368164,
"step": 925,
"token_acc": 0.8506801161546691
},
{
"epoch": 2.2743417023882424,
"grad_norm": 5.936365127563477,
"learning_rate": 7.598549442701747e-07,
"loss": 0.46504778861999513,
"step": 930,
"token_acc": 0.8484742468415938
},
{
"epoch": 2.286589099816289,
"grad_norm": 6.64162015914917,
"learning_rate": 7.358119414049542e-07,
"loss": 0.48408074378967286,
"step": 935,
"token_acc": 0.8455108359133127
},
{
"epoch": 2.2988364972443356,
"grad_norm": 5.138242721557617,
"learning_rate": 7.120896586569453e-07,
"loss": 0.4564502716064453,
"step": 940,
"token_acc": 0.851229078482389
},
{
"epoch": 2.311083894672382,
"grad_norm": 5.581968784332275,
"learning_rate": 6.886924086124786e-07,
"loss": 0.4458012580871582,
"step": 945,
"token_acc": 0.8538615446178471
},
{
"epoch": 2.3233312921004288,
"grad_norm": 4.3089070320129395,
"learning_rate": 6.656244447686458e-07,
"loss": 0.4544158935546875,
"step": 950,
"token_acc": 0.849496071740338
},
{
"epoch": 2.335578689528475,
"grad_norm": 8.022623062133789,
"learning_rate": 6.428899607600364e-07,
"loss": 0.46418256759643556,
"step": 955,
"token_acc": 0.8481595706332359
},
{
"epoch": 2.3478260869565215,
"grad_norm": 5.561613082885742,
"learning_rate": 6.204930895963607e-07,
"loss": 0.4830313682556152,
"step": 960,
"token_acc": 0.8458221652488055
},
{
"epoch": 2.3600734843845683,
"grad_norm": 4.618040084838867,
"learning_rate": 5.984379029110888e-07,
"loss": 0.4596832275390625,
"step": 965,
"token_acc": 0.8473798694878386
},
{
"epoch": 2.3723208818126147,
"grad_norm": 4.180395126342773,
"learning_rate": 5.767284102212464e-07,
"loss": 0.47627959251403806,
"step": 970,
"token_acc": 0.8443898443898444
},
{
"epoch": 2.3845682792406615,
"grad_norm": 7.8003411293029785,
"learning_rate": 5.553685581985085e-07,
"loss": 0.4734807014465332,
"step": 975,
"token_acc": 0.8454314525440689
},
{
"epoch": 2.396815676668708,
"grad_norm": 6.798686504364014,
"learning_rate": 5.343622299517143e-07,
"loss": 0.4706712245941162,
"step": 980,
"token_acc": 0.8457135996157849
},
{
"epoch": 2.4090630740967542,
"grad_norm": 6.040155410766602,
"learning_rate": 5.137132443209384e-07,
"loss": 0.47454032897949217,
"step": 985,
"token_acc": 0.8463217554474782
},
{
"epoch": 2.421310471524801,
"grad_norm": 4.184665679931641,
"learning_rate": 4.934253551832463e-07,
"loss": 0.47580909729003906,
"step": 990,
"token_acc": 0.8450045575238775
},
{
"epoch": 2.4335578689528474,
"grad_norm": 6.029951572418213,
"learning_rate": 4.7350225077025805e-07,
"loss": 0.4867083549499512,
"step": 995,
"token_acc": 0.840796420942324
},
{
"epoch": 2.4458052663808942,
"grad_norm": 4.318844318389893,
"learning_rate": 4.5394755299764867e-07,
"loss": 0.46222858428955077,
"step": 1000,
"token_acc": 0.8506385341476957
},
{
"epoch": 2.4580526638089406,
"grad_norm": 5.949728965759277,
"learning_rate": 4.3476481680670175e-07,
"loss": 0.4552288055419922,
"step": 1005,
"token_acc": 0.8509888388486391
},
{
"epoch": 2.470300061236987,
"grad_norm": 4.026670932769775,
"learning_rate": 4.159575295180396e-07,
"loss": 0.47368178367614744,
"step": 1010,
"token_acc": 0.8461627058778363
},
{
"epoch": 2.482547458665034,
"grad_norm": 4.262962341308594,
"learning_rate": 3.9752911019764946e-07,
"loss": 0.45987939834594727,
"step": 1015,
"token_acc": 0.8497596153846154
},
{
"epoch": 2.49479485609308,
"grad_norm": 6.769244194030762,
"learning_rate": 3.7948290903531166e-07,
"loss": 0.49054980278015137,
"step": 1020,
"token_acc": 0.8425439826874525
},
{
"epoch": 2.507042253521127,
"grad_norm": 6.347579479217529,
"learning_rate": 3.6182220673555256e-07,
"loss": 0.46785907745361327,
"step": 1025,
"token_acc": 0.8451467812537649
},
{
"epoch": 2.5192896509491733,
"grad_norm": 4.114010810852051,
"learning_rate": 3.445502139212334e-07,
"loss": 0.4741077423095703,
"step": 1030,
"token_acc": 0.845554834523037
},
{
"epoch": 2.5315370483772197,
"grad_norm": 5.250743865966797,
"learning_rate": 3.276700705498745e-07,
"loss": 0.4721240043640137,
"step": 1035,
"token_acc": 0.8438785808555236
},
{
"epoch": 2.5437844458052665,
"grad_norm": 6.302319526672363,
"learning_rate": 3.1118484534282484e-07,
"loss": 0.456193208694458,
"step": 1040,
"token_acc": 0.8506296827674159
},
{
"epoch": 2.556031843233313,
"grad_norm": 5.0382981300354,
"learning_rate": 2.95097535227388e-07,
"loss": 0.47198972702026365,
"step": 1045,
"token_acc": 0.8453859456060367
},
{
"epoch": 2.5682792406613597,
"grad_norm": 10.161928176879883,
"learning_rate": 2.7941106479199426e-07,
"loss": 0.4501551628112793,
"step": 1050,
"token_acc": 0.8524753614296123
},
{
"epoch": 2.580526638089406,
"grad_norm": 4.283596515655518,
"learning_rate": 2.641282857545274e-07,
"loss": 0.4893691062927246,
"step": 1055,
"token_acc": 0.8429578004135713
},
{
"epoch": 2.5927740355174524,
"grad_norm": 7.293058395385742,
"learning_rate": 2.492519764438966e-07,
"loss": 0.4803295612335205,
"step": 1060,
"token_acc": 0.8443132501857137
},
{
"epoch": 2.605021432945499,
"grad_norm": 6.88516902923584,
"learning_rate": 2.3478484129495077e-07,
"loss": 0.46808528900146484,
"step": 1065,
"token_acc": 0.845524172306196
},
{
"epoch": 2.6172688303735456,
"grad_norm": 5.870273590087891,
"learning_rate": 2.2072951035682688e-07,
"loss": 0.46234817504882814,
"step": 1070,
"token_acc": 0.8491432540907687
},
{
"epoch": 2.629516227801592,
"grad_norm": 5.054083824157715,
"learning_rate": 2.0708853881481973e-07,
"loss": 0.4716303825378418,
"step": 1075,
"token_acc": 0.8471902074556872
},
{
"epoch": 2.641763625229639,
"grad_norm": 5.593235969543457,
"learning_rate": 1.9386440652586546e-07,
"loss": 0.46905932426452634,
"step": 1080,
"token_acc": 0.8457278789533206
},
{
"epoch": 2.654011022657685,
"grad_norm": 4.2795867919921875,
"learning_rate": 1.8105951756771322e-07,
"loss": 0.4618710517883301,
"step": 1085,
"token_acc": 0.8501392320665176
},
{
"epoch": 2.6662584200857316,
"grad_norm": 4.194334983825684,
"learning_rate": 1.686761998018785e-07,
"loss": 0.4832133769989014,
"step": 1090,
"token_acc": 0.8424084816963393
},
{
"epoch": 2.6785058175137784,
"grad_norm": 7.414880752563477,
"learning_rate": 1.5671670445044918e-07,
"loss": 0.4764682769775391,
"step": 1095,
"token_acc": 0.8440724158068912
},
{
"epoch": 2.6907532149418247,
"grad_norm": 4.725894451141357,
"learning_rate": 1.4518320568682602e-07,
"loss": 0.4642933368682861,
"step": 1100,
"token_acc": 0.848500968955428
},
{
"epoch": 2.7030006123698715,
"grad_norm": 12.555808067321777,
"learning_rate": 1.340778002404694e-07,
"loss": 0.4585245609283447,
"step": 1105,
"token_acc": 0.8510247293356238
},
{
"epoch": 2.715248009797918,
"grad_norm": 4.7355732917785645,
"learning_rate": 1.234025070157238e-07,
"loss": 0.47963943481445315,
"step": 1110,
"token_acc": 0.8441808271882013
},
{
"epoch": 2.7274954072259643,
"grad_norm": 6.5449700355529785,
"learning_rate": 1.1315926672479243e-07,
"loss": 0.4840541839599609,
"step": 1115,
"token_acc": 0.843370607028754
},
{
"epoch": 2.739742804654011,
"grad_norm": 6.476162910461426,
"learning_rate": 1.033499415349265e-07,
"loss": 0.47246541976928713,
"step": 1120,
"token_acc": 0.8454825462012321
},
{
"epoch": 2.7519902020820575,
"grad_norm": 6.02608585357666,
"learning_rate": 9.397631472989138e-08,
"loss": 0.44602155685424805,
"step": 1125,
"token_acc": 0.8545568529112139
},
{
"epoch": 2.7642375995101043,
"grad_norm": 4.15132999420166,
"learning_rate": 8.504009038577621e-08,
"loss": 0.4540902614593506,
"step": 1130,
"token_acc": 0.8520612768687841
},
{
"epoch": 2.7764849969381507,
"grad_norm": 4.35493803024292,
"learning_rate": 7.654289306120194e-08,
"loss": 0.4731804847717285,
"step": 1135,
"token_acc": 0.8463782973857498
},
{
"epoch": 2.788732394366197,
"grad_norm": 5.5318074226379395,
"learning_rate": 6.848626750198428e-08,
"loss": 0.45829362869262696,
"step": 1140,
"token_acc": 0.8508665545166465
},
{
"epoch": 2.800979791794244,
"grad_norm": 4.269839286804199,
"learning_rate": 6.08716783603075e-08,
"loss": 0.4502554416656494,
"step": 1145,
"token_acc": 0.8539763779527559
},
{
"epoch": 2.81322718922229,
"grad_norm": 3.8356523513793945,
"learning_rate": 5.3700509928458765e-08,
"loss": 0.46181259155273435,
"step": 1150,
"token_acc": 0.8495603645561489
},
{
"epoch": 2.825474586650337,
"grad_norm": 3.848315954208374,
"learning_rate": 4.697406588717007e-08,
"loss": 0.4596513271331787,
"step": 1155,
"token_acc": 0.8490898179635927
},
{
"epoch": 2.8377219840783834,
"grad_norm": 9.545782089233398,
"learning_rate": 4.069356906861721e-08,
"loss": 0.48660125732421877,
"step": 1160,
"token_acc": 0.8441344075456868
},
{
"epoch": 2.8499693815064298,
"grad_norm": 4.807938098907471,
"learning_rate": 3.486016123411345e-08,
"loss": 0.4720099449157715,
"step": 1165,
"token_acc": 0.8459315341718565
},
{
"epoch": 2.8622167789344766,
"grad_norm": 4.377501964569092,
"learning_rate": 2.947490286654503e-08,
"loss": 0.45624866485595705,
"step": 1170,
"token_acc": 0.8516121486217786
},
{
"epoch": 2.874464176362523,
"grad_norm": 4.4703874588012695,
"learning_rate": 2.453877297758117e-08,
"loss": 0.45848474502563474,
"step": 1175,
"token_acc": 0.8506992867673427
},
{
"epoch": 2.8867115737905698,
"grad_norm": 4.576780796051025,
"learning_rate": 2.0052668929692875e-08,
"loss": 0.4643035888671875,
"step": 1180,
"token_acc": 0.8491360817504583
},
{
"epoch": 2.898958971218616,
"grad_norm": 7.080748081207275,
"learning_rate": 1.601740627301951e-08,
"loss": 0.45264711380004885,
"step": 1185,
"token_acc": 0.8515248796147673
},
{
"epoch": 2.9112063686466625,
"grad_norm": 6.231855869293213,
"learning_rate": 1.2433718597103795e-08,
"loss": 0.47052750587463377,
"step": 1190,
"token_acc": 0.848204354044176
},
{
"epoch": 2.923453766074709,
"grad_norm": 5.396893501281738,
"learning_rate": 9.302257397531267e-09,
"loss": 0.47264981269836426,
"step": 1195,
"token_acc": 0.843019680196802
},
{
"epoch": 2.9357011635027557,
"grad_norm": 4.027212142944336,
"learning_rate": 6.623591957490016e-09,
"loss": 0.46438822746276853,
"step": 1200,
"token_acc": 0.8480934093642112
},
{
"epoch": 2.947948560930802,
"grad_norm": 7.147536277770996,
"learning_rate": 4.398209244279305e-09,
"loss": 0.47541131973266604,
"step": 1205,
"token_acc": 0.845031355523396
},
{
"epoch": 2.960195958358849,
"grad_norm": 15.318401336669922,
"learning_rate": 2.6265138207806628e-09,
"loss": 0.45806264877319336,
"step": 1210,
"token_acc": 0.8491041543610947
},
{
"epoch": 2.9724433557868952,
"grad_norm": 4.848905086517334,
"learning_rate": 1.308827771910881e-09,
"loss": 0.47438344955444334,
"step": 1215,
"token_acc": 0.8430241870895403
},
{
"epoch": 2.9846907532149416,
"grad_norm": 6.5829548835754395,
"learning_rate": 4.4539064606829996e-10,
"loss": 0.46582889556884766,
"step": 1220,
"token_acc": 0.8473073768891318
},
{
"epoch": 2.9969381506429884,
"grad_norm": 5.647032737731934,
"learning_rate": 3.6359411584308e-11,
"loss": 0.46216769218444825,
"step": 1225,
"token_acc": 0.8522442377678933
}
],
"logging_steps": 5,
"max_steps": 1227,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.3867353301859697e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}