Xinyuan-QS-72B / trainer_state.json
jztdd's picture
Upload 10 files
c962e57 verified
raw
history blame
64.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3694,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0027070925825663237,
"grad_norm": 2.7909258086527373,
"learning_rate": 4.999977397559719e-05,
"loss": 1.4602,
"step": 10
},
{
"epoch": 0.005414185165132647,
"grad_norm": 2.9246419242894404,
"learning_rate": 4.9999095906475714e-05,
"loss": 1.4843,
"step": 20
},
{
"epoch": 0.008121277747698972,
"grad_norm": 2.60296538082041,
"learning_rate": 4.999796580489639e-05,
"loss": 1.3953,
"step": 30
},
{
"epoch": 0.010828370330265295,
"grad_norm": 2.427882849139683,
"learning_rate": 4.9996383691293656e-05,
"loss": 1.3886,
"step": 40
},
{
"epoch": 0.01353546291283162,
"grad_norm": 2.362478801666027,
"learning_rate": 4.999434959427522e-05,
"loss": 1.4087,
"step": 50
},
{
"epoch": 0.016242555495397944,
"grad_norm": 2.0562878551243253,
"learning_rate": 4.999186355062154e-05,
"loss": 1.3925,
"step": 60
},
{
"epoch": 0.018949648077964266,
"grad_norm": 1.4997595683614129,
"learning_rate": 4.998892560528511e-05,
"loss": 1.4251,
"step": 70
},
{
"epoch": 0.02165674066053059,
"grad_norm": 1.8396372795029057,
"learning_rate": 4.9985535811389726e-05,
"loss": 1.4299,
"step": 80
},
{
"epoch": 0.024363833243096916,
"grad_norm": 3.016445354360887,
"learning_rate": 4.99816942302295e-05,
"loss": 1.4179,
"step": 90
},
{
"epoch": 0.02707092582566324,
"grad_norm": 2.285685249444506,
"learning_rate": 4.997740093126769e-05,
"loss": 1.4453,
"step": 100
},
{
"epoch": 0.02977801840822956,
"grad_norm": 2.3733951587674054,
"learning_rate": 4.997265599213554e-05,
"loss": 1.3905,
"step": 110
},
{
"epoch": 0.03248511099079589,
"grad_norm": 3.8476837651749336,
"learning_rate": 4.996745949863082e-05,
"loss": 1.4418,
"step": 120
},
{
"epoch": 0.03519220357336221,
"grad_norm": 1.6916145193283187,
"learning_rate": 4.996181154471625e-05,
"loss": 1.4637,
"step": 130
},
{
"epoch": 0.03789929615592853,
"grad_norm": 1.6551273153572528,
"learning_rate": 4.9955712232517896e-05,
"loss": 1.422,
"step": 140
},
{
"epoch": 0.040606388738494856,
"grad_norm": 1.842180664471275,
"learning_rate": 4.99491616723232e-05,
"loss": 1.4768,
"step": 150
},
{
"epoch": 0.04331348132106118,
"grad_norm": 2.8401045455632503,
"learning_rate": 4.994215998257909e-05,
"loss": 1.4626,
"step": 160
},
{
"epoch": 0.0460205739036275,
"grad_norm": 2.137479556495621,
"learning_rate": 4.993470728988979e-05,
"loss": 1.4147,
"step": 170
},
{
"epoch": 0.04872766648619383,
"grad_norm": 1.9073264544958397,
"learning_rate": 4.992680372901454e-05,
"loss": 1.4119,
"step": 180
},
{
"epoch": 0.051434759068760154,
"grad_norm": 1.5685293387125376,
"learning_rate": 4.9918449442865126e-05,
"loss": 1.428,
"step": 190
},
{
"epoch": 0.05414185165132648,
"grad_norm": 2.128149762022408,
"learning_rate": 4.9909644582503366e-05,
"loss": 1.4497,
"step": 200
},
{
"epoch": 0.0568489442338928,
"grad_norm": 1.8550357912320867,
"learning_rate": 4.990038930713833e-05,
"loss": 1.4661,
"step": 210
},
{
"epoch": 0.05955603681645912,
"grad_norm": 1.811618523801653,
"learning_rate": 4.989068378412346e-05,
"loss": 1.418,
"step": 220
},
{
"epoch": 0.062263129399025445,
"grad_norm": 1.691670012971929,
"learning_rate": 4.9880528188953556e-05,
"loss": 1.4446,
"step": 230
},
{
"epoch": 0.06497022198159177,
"grad_norm": 1.7432757441956248,
"learning_rate": 4.986992270526161e-05,
"loss": 1.3911,
"step": 240
},
{
"epoch": 0.0676773145641581,
"grad_norm": 1.709102114001383,
"learning_rate": 4.9858867524815456e-05,
"loss": 1.4912,
"step": 250
},
{
"epoch": 0.07038440714672442,
"grad_norm": 1.7869823709380956,
"learning_rate": 4.984736284751436e-05,
"loss": 1.454,
"step": 260
},
{
"epoch": 0.07309149972929074,
"grad_norm": 1.6458598808862888,
"learning_rate": 4.983540888138534e-05,
"loss": 1.4278,
"step": 270
},
{
"epoch": 0.07579859231185707,
"grad_norm": 2.4415366747183054,
"learning_rate": 4.982300584257943e-05,
"loss": 1.4511,
"step": 280
},
{
"epoch": 0.07850568489442339,
"grad_norm": 1.7712176478354567,
"learning_rate": 4.98101539553678e-05,
"loss": 1.4221,
"step": 290
},
{
"epoch": 0.08121277747698971,
"grad_norm": 1.408329894177912,
"learning_rate": 4.979685345213766e-05,
"loss": 1.4153,
"step": 300
},
{
"epoch": 0.08391987005955603,
"grad_norm": 2.8322848403176106,
"learning_rate": 4.978310457338806e-05,
"loss": 1.4544,
"step": 310
},
{
"epoch": 0.08662696264212236,
"grad_norm": 1.349216831725713,
"learning_rate": 4.976890756772558e-05,
"loss": 1.4232,
"step": 320
},
{
"epoch": 0.08933405522468868,
"grad_norm": 3.302218333832659,
"learning_rate": 4.9754262691859786e-05,
"loss": 1.4716,
"step": 330
},
{
"epoch": 0.092041147807255,
"grad_norm": 2.049533306517282,
"learning_rate": 4.9739170210598646e-05,
"loss": 1.469,
"step": 340
},
{
"epoch": 0.09474824038982133,
"grad_norm": 2.329244123317743,
"learning_rate": 4.9723630396843665e-05,
"loss": 1.4268,
"step": 350
},
{
"epoch": 0.09745533297238766,
"grad_norm": 1.6230256143926636,
"learning_rate": 4.9707643531585026e-05,
"loss": 1.44,
"step": 360
},
{
"epoch": 0.10016242555495398,
"grad_norm": 2.048460766648452,
"learning_rate": 4.969120990389645e-05,
"loss": 1.4171,
"step": 370
},
{
"epoch": 0.10286951813752031,
"grad_norm": 1.8949361232151039,
"learning_rate": 4.967432981093002e-05,
"loss": 1.4555,
"step": 380
},
{
"epoch": 0.10557661072008663,
"grad_norm": 1.454874136363247,
"learning_rate": 4.965700355791075e-05,
"loss": 1.4637,
"step": 390
},
{
"epoch": 0.10828370330265295,
"grad_norm": 1.6866618414734464,
"learning_rate": 4.9639231458131155e-05,
"loss": 1.4882,
"step": 400
},
{
"epoch": 0.11099079588521928,
"grad_norm": 1.767137745890727,
"learning_rate": 4.962101383294546e-05,
"loss": 1.4643,
"step": 410
},
{
"epoch": 0.1136978884677856,
"grad_norm": 1.4955900585918869,
"learning_rate": 4.9602351011763916e-05,
"loss": 1.4916,
"step": 420
},
{
"epoch": 0.11640498105035192,
"grad_norm": 1.4089383124912003,
"learning_rate": 4.958324333204675e-05,
"loss": 1.4558,
"step": 430
},
{
"epoch": 0.11911207363291824,
"grad_norm": 1.794803201497381,
"learning_rate": 4.9563691139298115e-05,
"loss": 1.4685,
"step": 440
},
{
"epoch": 0.12181916621548457,
"grad_norm": 4.657348417565155,
"learning_rate": 4.954369478705984e-05,
"loss": 1.4517,
"step": 450
},
{
"epoch": 0.12452625879805089,
"grad_norm": 1.727020331207215,
"learning_rate": 4.952325463690499e-05,
"loss": 1.4905,
"step": 460
},
{
"epoch": 0.12723335138061723,
"grad_norm": 1.677461362340949,
"learning_rate": 4.95023710584314e-05,
"loss": 1.4786,
"step": 470
},
{
"epoch": 0.12994044396318355,
"grad_norm": 1.3718142212853295,
"learning_rate": 4.948104442925493e-05,
"loss": 1.5028,
"step": 480
},
{
"epoch": 0.13264753654574987,
"grad_norm": 1.762804612117205,
"learning_rate": 4.945927513500267e-05,
"loss": 1.53,
"step": 490
},
{
"epoch": 0.1353546291283162,
"grad_norm": 2.616132608044088,
"learning_rate": 4.943706356930596e-05,
"loss": 1.4513,
"step": 500
},
{
"epoch": 0.13806172171088252,
"grad_norm": 1.5396425075226725,
"learning_rate": 4.941441013379328e-05,
"loss": 1.4783,
"step": 510
},
{
"epoch": 0.14076881429344884,
"grad_norm": 1.5753718924201408,
"learning_rate": 4.939131523808295e-05,
"loss": 1.469,
"step": 520
},
{
"epoch": 0.14347590687601516,
"grad_norm": 1.990561438026612,
"learning_rate": 4.936777929977578e-05,
"loss": 1.5288,
"step": 530
},
{
"epoch": 0.1461829994585815,
"grad_norm": 1.535027221106918,
"learning_rate": 4.9343802744447485e-05,
"loss": 1.4864,
"step": 540
},
{
"epoch": 0.1488900920411478,
"grad_norm": 1.5120704229625987,
"learning_rate": 4.9319386005641e-05,
"loss": 1.5288,
"step": 550
},
{
"epoch": 0.15159718462371413,
"grad_norm": 1.3272029647679429,
"learning_rate": 4.9294529524858605e-05,
"loss": 1.4571,
"step": 560
},
{
"epoch": 0.15430427720628045,
"grad_norm": 1.3970253346744186,
"learning_rate": 4.926923375155402e-05,
"loss": 1.4252,
"step": 570
},
{
"epoch": 0.15701136978884678,
"grad_norm": 1.8691453221841268,
"learning_rate": 4.924349914312421e-05,
"loss": 1.4326,
"step": 580
},
{
"epoch": 0.1597184623714131,
"grad_norm": 1.949334637283444,
"learning_rate": 4.921732616490111e-05,
"loss": 1.4056,
"step": 590
},
{
"epoch": 0.16242555495397942,
"grad_norm": 1.5924101773217003,
"learning_rate": 4.919071529014329e-05,
"loss": 1.4681,
"step": 600
},
{
"epoch": 0.16513264753654575,
"grad_norm": 1.6221588411171692,
"learning_rate": 4.91636670000273e-05,
"loss": 1.509,
"step": 610
},
{
"epoch": 0.16783974011911207,
"grad_norm": 1.5149301503130714,
"learning_rate": 4.913618178363904e-05,
"loss": 1.5005,
"step": 620
},
{
"epoch": 0.1705468327016784,
"grad_norm": 1.6644176831544628,
"learning_rate": 4.9108260137964865e-05,
"loss": 1.4638,
"step": 630
},
{
"epoch": 0.17325392528424471,
"grad_norm": 1.569403594423818,
"learning_rate": 4.9079902567882665e-05,
"loss": 1.4826,
"step": 640
},
{
"epoch": 0.17596101786681104,
"grad_norm": 1.3674634000662147,
"learning_rate": 4.9051109586152634e-05,
"loss": 1.4344,
"step": 650
},
{
"epoch": 0.17866811044937736,
"grad_norm": 1.1930764967763101,
"learning_rate": 4.902188171340811e-05,
"loss": 1.4396,
"step": 660
},
{
"epoch": 0.18137520303194368,
"grad_norm": 1.89545649756294,
"learning_rate": 4.8992219478146094e-05,
"loss": 1.4858,
"step": 670
},
{
"epoch": 0.18408229561451,
"grad_norm": 4.121840228113138,
"learning_rate": 4.8962123416717706e-05,
"loss": 1.4912,
"step": 680
},
{
"epoch": 0.18678938819707633,
"grad_norm": 1.292830871259127,
"learning_rate": 4.893159407331849e-05,
"loss": 1.5105,
"step": 690
},
{
"epoch": 0.18949648077964265,
"grad_norm": 2.0378465705988797,
"learning_rate": 4.8900631999978565e-05,
"loss": 1.4478,
"step": 700
},
{
"epoch": 0.19220357336220897,
"grad_norm": 1.4086096037731015,
"learning_rate": 4.886923775655267e-05,
"loss": 1.4748,
"step": 710
},
{
"epoch": 0.19491066594477532,
"grad_norm": 1.6471666031866425,
"learning_rate": 4.883741191071001e-05,
"loss": 1.4616,
"step": 720
},
{
"epoch": 0.19761775852734165,
"grad_norm": 1.3016146758942742,
"learning_rate": 4.880515503792402e-05,
"loss": 1.4502,
"step": 730
},
{
"epoch": 0.20032485110990797,
"grad_norm": 1.7755038959773146,
"learning_rate": 4.8772467721461916e-05,
"loss": 1.4929,
"step": 740
},
{
"epoch": 0.2030319436924743,
"grad_norm": 1.8173068291163943,
"learning_rate": 4.8739350552374206e-05,
"loss": 1.4514,
"step": 750
},
{
"epoch": 0.20573903627504062,
"grad_norm": 1.3957806556643733,
"learning_rate": 4.870580412948397e-05,
"loss": 1.4901,
"step": 760
},
{
"epoch": 0.20844612885760694,
"grad_norm": 1.7574648378127529,
"learning_rate": 4.8671829059375996e-05,
"loss": 1.4239,
"step": 770
},
{
"epoch": 0.21115322144017326,
"grad_norm": 1.988134441146335,
"learning_rate": 4.863742595638589e-05,
"loss": 1.4929,
"step": 780
},
{
"epoch": 0.21386031402273958,
"grad_norm": 1.6248880174612887,
"learning_rate": 4.8602595442588926e-05,
"loss": 1.5049,
"step": 790
},
{
"epoch": 0.2165674066053059,
"grad_norm": 1.4759465960672522,
"learning_rate": 4.856733814778878e-05,
"loss": 1.4617,
"step": 800
},
{
"epoch": 0.21927449918787223,
"grad_norm": 1.6391142091975623,
"learning_rate": 4.853165470950617e-05,
"loss": 1.496,
"step": 810
},
{
"epoch": 0.22198159177043855,
"grad_norm": 1.5875776432159403,
"learning_rate": 4.8495545772967336e-05,
"loss": 1.4696,
"step": 820
},
{
"epoch": 0.22468868435300487,
"grad_norm": 1.3488488116129105,
"learning_rate": 4.8459011991092335e-05,
"loss": 1.4865,
"step": 830
},
{
"epoch": 0.2273957769355712,
"grad_norm": 1.2491528956283275,
"learning_rate": 4.842205402448326e-05,
"loss": 1.4818,
"step": 840
},
{
"epoch": 0.23010286951813752,
"grad_norm": 1.4146633075540191,
"learning_rate": 4.8384672541412314e-05,
"loss": 1.4461,
"step": 850
},
{
"epoch": 0.23280996210070384,
"grad_norm": 2.8216763132669405,
"learning_rate": 4.834686821780966e-05,
"loss": 1.4905,
"step": 860
},
{
"epoch": 0.23551705468327017,
"grad_norm": 1.3681470128440785,
"learning_rate": 4.83086417372513e-05,
"loss": 1.4762,
"step": 870
},
{
"epoch": 0.2382241472658365,
"grad_norm": 1.6949909678615045,
"learning_rate": 4.826999379094662e-05,
"loss": 1.4548,
"step": 880
},
{
"epoch": 0.2409312398484028,
"grad_norm": 1.7298574479777804,
"learning_rate": 4.823092507772593e-05,
"loss": 1.4457,
"step": 890
},
{
"epoch": 0.24363833243096913,
"grad_norm": 1.3751587504284029,
"learning_rate": 4.8191436304027846e-05,
"loss": 1.4485,
"step": 900
},
{
"epoch": 0.24634542501353546,
"grad_norm": 1.98272939662037,
"learning_rate": 4.815152818388647e-05,
"loss": 1.4827,
"step": 910
},
{
"epoch": 0.24905251759610178,
"grad_norm": 1.48867070663095,
"learning_rate": 4.811120143891855e-05,
"loss": 1.4961,
"step": 920
},
{
"epoch": 0.2517596101786681,
"grad_norm": 1.6582119972239975,
"learning_rate": 4.807045679831035e-05,
"loss": 1.5125,
"step": 930
},
{
"epoch": 0.25446670276123445,
"grad_norm": 2.2411742740260143,
"learning_rate": 4.8029294998804515e-05,
"loss": 1.481,
"step": 940
},
{
"epoch": 0.25717379534380075,
"grad_norm": 3.1196395093384757,
"learning_rate": 4.798771678468673e-05,
"loss": 1.4766,
"step": 950
},
{
"epoch": 0.2598808879263671,
"grad_norm": 1.747388174474549,
"learning_rate": 4.794572290777228e-05,
"loss": 1.4664,
"step": 960
},
{
"epoch": 0.2625879805089334,
"grad_norm": 1.244225229661525,
"learning_rate": 4.790331412739246e-05,
"loss": 1.4881,
"step": 970
},
{
"epoch": 0.26529507309149974,
"grad_norm": 1.5361914036706255,
"learning_rate": 4.786049121038078e-05,
"loss": 1.4481,
"step": 980
},
{
"epoch": 0.26800216567406604,
"grad_norm": 2.004415613855818,
"learning_rate": 4.78172549310592e-05,
"loss": 1.4654,
"step": 990
},
{
"epoch": 0.2707092582566324,
"grad_norm": 1.6485133719660112,
"learning_rate": 4.777360607122405e-05,
"loss": 1.4569,
"step": 1000
},
{
"epoch": 0.2734163508391987,
"grad_norm": 1.3237479316055099,
"learning_rate": 4.772954542013193e-05,
"loss": 1.4616,
"step": 1010
},
{
"epoch": 0.27612344342176504,
"grad_norm": 1.4682831465817014,
"learning_rate": 4.7685073774485424e-05,
"loss": 1.4312,
"step": 1020
},
{
"epoch": 0.27883053600433133,
"grad_norm": 2.66972568412979,
"learning_rate": 4.764019193841871e-05,
"loss": 1.4638,
"step": 1030
},
{
"epoch": 0.2815376285868977,
"grad_norm": 1.8874806530406238,
"learning_rate": 4.7594900723482994e-05,
"loss": 1.4342,
"step": 1040
},
{
"epoch": 0.284244721169464,
"grad_norm": 1.7012798675260252,
"learning_rate": 4.7549200948631875e-05,
"loss": 1.4808,
"step": 1050
},
{
"epoch": 0.2869518137520303,
"grad_norm": 1.5041238189640276,
"learning_rate": 4.750309344020649e-05,
"loss": 1.477,
"step": 1060
},
{
"epoch": 0.2896589063345966,
"grad_norm": 1.445182392004961,
"learning_rate": 4.74565790319206e-05,
"loss": 1.4684,
"step": 1070
},
{
"epoch": 0.292365998917163,
"grad_norm": 1.5350175117888023,
"learning_rate": 4.7409658564845526e-05,
"loss": 1.4786,
"step": 1080
},
{
"epoch": 0.29507309149972927,
"grad_norm": 1.6518274612672268,
"learning_rate": 4.73623328873949e-05,
"loss": 1.4819,
"step": 1090
},
{
"epoch": 0.2977801840822956,
"grad_norm": 1.7095589003674432,
"learning_rate": 4.731460285530936e-05,
"loss": 1.4976,
"step": 1100
},
{
"epoch": 0.3004872766648619,
"grad_norm": 1.723367767756465,
"learning_rate": 4.726646933164108e-05,
"loss": 1.4504,
"step": 1110
},
{
"epoch": 0.30319436924742826,
"grad_norm": 1.9040390036164423,
"learning_rate": 4.7217933186738114e-05,
"loss": 1.4579,
"step": 1120
},
{
"epoch": 0.30590146182999456,
"grad_norm": 1.2294419736565405,
"learning_rate": 4.716899529822874e-05,
"loss": 1.5031,
"step": 1130
},
{
"epoch": 0.3086085544125609,
"grad_norm": 2.0118944901286997,
"learning_rate": 4.7119656551005505e-05,
"loss": 1.4532,
"step": 1140
},
{
"epoch": 0.31131564699512726,
"grad_norm": 1.6217590845134082,
"learning_rate": 4.706991783720928e-05,
"loss": 1.4871,
"step": 1150
},
{
"epoch": 0.31402273957769355,
"grad_norm": 1.509260025154433,
"learning_rate": 4.701978005621311e-05,
"loss": 1.4464,
"step": 1160
},
{
"epoch": 0.3167298321602599,
"grad_norm": 1.5171464183166403,
"learning_rate": 4.696924411460596e-05,
"loss": 1.4486,
"step": 1170
},
{
"epoch": 0.3194369247428262,
"grad_norm": 1.2572808716835593,
"learning_rate": 4.6918310926176306e-05,
"loss": 1.4458,
"step": 1180
},
{
"epoch": 0.32214401732539255,
"grad_norm": 1.7849046370327453,
"learning_rate": 4.686698141189564e-05,
"loss": 1.4404,
"step": 1190
},
{
"epoch": 0.32485110990795885,
"grad_norm": 1.914525739200728,
"learning_rate": 4.6815256499901775e-05,
"loss": 1.4898,
"step": 1200
},
{
"epoch": 0.3275582024905252,
"grad_norm": 1.611833069659245,
"learning_rate": 4.6763137125482104e-05,
"loss": 1.4847,
"step": 1210
},
{
"epoch": 0.3302652950730915,
"grad_norm": 1.753641811163595,
"learning_rate": 4.671062423105667e-05,
"loss": 1.433,
"step": 1220
},
{
"epoch": 0.33297238765565784,
"grad_norm": 1.1996803367158615,
"learning_rate": 4.66577187661611e-05,
"loss": 1.4533,
"step": 1230
},
{
"epoch": 0.33567948023822414,
"grad_norm": 1.554114829537543,
"learning_rate": 4.660442168742951e-05,
"loss": 1.4708,
"step": 1240
},
{
"epoch": 0.3383865728207905,
"grad_norm": 6.405248874344029,
"learning_rate": 4.6550733958577115e-05,
"loss": 1.4721,
"step": 1250
},
{
"epoch": 0.3410936654033568,
"grad_norm": 1.5464469028099914,
"learning_rate": 4.6496656550382864e-05,
"loss": 1.4524,
"step": 1260
},
{
"epoch": 0.34380075798592313,
"grad_norm": 1.5438246148321,
"learning_rate": 4.6442190440671886e-05,
"loss": 1.5217,
"step": 1270
},
{
"epoch": 0.34650785056848943,
"grad_norm": 1.6459917338592895,
"learning_rate": 4.638733661429775e-05,
"loss": 1.431,
"step": 1280
},
{
"epoch": 0.3492149431510558,
"grad_norm": 1.7332054776628096,
"learning_rate": 4.633209606312473e-05,
"loss": 1.4896,
"step": 1290
},
{
"epoch": 0.3519220357336221,
"grad_norm": 1.7861896691639296,
"learning_rate": 4.6276469786009846e-05,
"loss": 1.4587,
"step": 1300
},
{
"epoch": 0.3546291283161884,
"grad_norm": 1.380983484664312,
"learning_rate": 4.622045878878477e-05,
"loss": 1.4928,
"step": 1310
},
{
"epoch": 0.3573362208987547,
"grad_norm": 1.6722892346510752,
"learning_rate": 4.616406408423769e-05,
"loss": 1.3885,
"step": 1320
},
{
"epoch": 0.36004331348132107,
"grad_norm": 1.302543220749753,
"learning_rate": 4.610728669209494e-05,
"loss": 1.4281,
"step": 1330
},
{
"epoch": 0.36275040606388737,
"grad_norm": 1.3027170210142245,
"learning_rate": 4.605012763900264e-05,
"loss": 1.4695,
"step": 1340
},
{
"epoch": 0.3654574986464537,
"grad_norm": 1.604586924613264,
"learning_rate": 4.599258795850803e-05,
"loss": 1.449,
"step": 1350
},
{
"epoch": 0.36816459122902,
"grad_norm": 1.2848250120256988,
"learning_rate": 4.593466869104087e-05,
"loss": 1.41,
"step": 1360
},
{
"epoch": 0.37087168381158636,
"grad_norm": 1.5746776369025801,
"learning_rate": 4.5876370883894606e-05,
"loss": 1.5164,
"step": 1370
},
{
"epoch": 0.37357877639415266,
"grad_norm": 1.8869095435477306,
"learning_rate": 4.581769559120738e-05,
"loss": 1.4649,
"step": 1380
},
{
"epoch": 0.376285868976719,
"grad_norm": 1.717413158097039,
"learning_rate": 4.575864387394304e-05,
"loss": 1.4801,
"step": 1390
},
{
"epoch": 0.3789929615592853,
"grad_norm": 2.539587222307521,
"learning_rate": 4.569921679987192e-05,
"loss": 1.4385,
"step": 1400
},
{
"epoch": 0.38170005414185165,
"grad_norm": 1.2980114991302223,
"learning_rate": 4.563941544355152e-05,
"loss": 1.4934,
"step": 1410
},
{
"epoch": 0.38440714672441795,
"grad_norm": 1.6715775537496502,
"learning_rate": 4.5579240886307136e-05,
"loss": 1.4538,
"step": 1420
},
{
"epoch": 0.3871142393069843,
"grad_norm": 2.709253749500051,
"learning_rate": 4.551869421621221e-05,
"loss": 1.4554,
"step": 1430
},
{
"epoch": 0.38982133188955065,
"grad_norm": 1.5064962735011314,
"learning_rate": 4.5457776528068743e-05,
"loss": 1.5152,
"step": 1440
},
{
"epoch": 0.39252842447211694,
"grad_norm": 1.4866135650495016,
"learning_rate": 4.539648892338747e-05,
"loss": 1.4701,
"step": 1450
},
{
"epoch": 0.3952355170546833,
"grad_norm": 1.425267538841413,
"learning_rate": 4.533483251036792e-05,
"loss": 1.4794,
"step": 1460
},
{
"epoch": 0.3979426096372496,
"grad_norm": 1.4920137928311075,
"learning_rate": 4.527280840387842e-05,
"loss": 1.4884,
"step": 1470
},
{
"epoch": 0.40064970221981594,
"grad_norm": 1.4659663002416385,
"learning_rate": 4.5210417725435896e-05,
"loss": 1.4436,
"step": 1480
},
{
"epoch": 0.40335679480238223,
"grad_norm": 1.4406570198350586,
"learning_rate": 4.514766160318561e-05,
"loss": 1.473,
"step": 1490
},
{
"epoch": 0.4060638873849486,
"grad_norm": 1.4236232583730681,
"learning_rate": 4.508454117188077e-05,
"loss": 1.441,
"step": 1500
},
{
"epoch": 0.4087709799675149,
"grad_norm": 1.658008187554401,
"learning_rate": 4.5021057572862e-05,
"loss": 1.4656,
"step": 1510
},
{
"epoch": 0.41147807255008123,
"grad_norm": 1.5768571336837218,
"learning_rate": 4.495721195403671e-05,
"loss": 1.4681,
"step": 1520
},
{
"epoch": 0.4141851651326475,
"grad_norm": 1.770673573444082,
"learning_rate": 4.4893005469858316e-05,
"loss": 1.42,
"step": 1530
},
{
"epoch": 0.4168922577152139,
"grad_norm": 1.3645638511742517,
"learning_rate": 4.482843928130541e-05,
"loss": 1.4911,
"step": 1540
},
{
"epoch": 0.41959935029778017,
"grad_norm": 1.6871925830755095,
"learning_rate": 4.476351455586072e-05,
"loss": 1.5269,
"step": 1550
},
{
"epoch": 0.4223064428803465,
"grad_norm": 1.4062614507732203,
"learning_rate": 4.469823246749003e-05,
"loss": 1.445,
"step": 1560
},
{
"epoch": 0.4250135354629128,
"grad_norm": 1.732815716768682,
"learning_rate": 4.4632594196620945e-05,
"loss": 1.5041,
"step": 1570
},
{
"epoch": 0.42772062804547917,
"grad_norm": 1.2938163974151866,
"learning_rate": 4.456660093012155e-05,
"loss": 1.4565,
"step": 1580
},
{
"epoch": 0.43042772062804546,
"grad_norm": 1.4591593663238542,
"learning_rate": 4.4500253861278926e-05,
"loss": 1.457,
"step": 1590
},
{
"epoch": 0.4331348132106118,
"grad_norm": 1.619966018635216,
"learning_rate": 4.443355418977761e-05,
"loss": 1.4848,
"step": 1600
},
{
"epoch": 0.4358419057931781,
"grad_norm": 1.5713970862575257,
"learning_rate": 4.436650312167787e-05,
"loss": 1.4311,
"step": 1610
},
{
"epoch": 0.43854899837574446,
"grad_norm": 4.836628503687221,
"learning_rate": 4.429910186939392e-05,
"loss": 1.5039,
"step": 1620
},
{
"epoch": 0.44125609095831075,
"grad_norm": 1.5617479359446638,
"learning_rate": 4.4231351651671985e-05,
"loss": 1.4387,
"step": 1630
},
{
"epoch": 0.4439631835408771,
"grad_norm": 4.127569209273957,
"learning_rate": 4.4163253693568266e-05,
"loss": 1.4654,
"step": 1640
},
{
"epoch": 0.4466702761234434,
"grad_norm": 1.9045020579112555,
"learning_rate": 4.409480922642677e-05,
"loss": 1.403,
"step": 1650
},
{
"epoch": 0.44937736870600975,
"grad_norm": 1.444246465499322,
"learning_rate": 4.40260194878571e-05,
"loss": 1.4568,
"step": 1660
},
{
"epoch": 0.45208446128857604,
"grad_norm": 1.3435184719673208,
"learning_rate": 4.395688572171203e-05,
"loss": 1.5321,
"step": 1670
},
{
"epoch": 0.4547915538711424,
"grad_norm": 1.8192617831889093,
"learning_rate": 4.3887409178064995e-05,
"loss": 1.4693,
"step": 1680
},
{
"epoch": 0.4574986464537087,
"grad_norm": 1.3323020396326049,
"learning_rate": 4.3817591113187563e-05,
"loss": 1.4333,
"step": 1690
},
{
"epoch": 0.46020573903627504,
"grad_norm": 1.2996234776965605,
"learning_rate": 4.374743278952662e-05,
"loss": 1.4034,
"step": 1700
},
{
"epoch": 0.4629128316188414,
"grad_norm": 1.4381391819706373,
"learning_rate": 4.367693547568163e-05,
"loss": 1.4924,
"step": 1710
},
{
"epoch": 0.4656199242014077,
"grad_norm": 1.8146691731599425,
"learning_rate": 4.360610044638167e-05,
"loss": 1.4284,
"step": 1720
},
{
"epoch": 0.46832701678397404,
"grad_norm": 2.968939444897473,
"learning_rate": 4.353492898246234e-05,
"loss": 1.4586,
"step": 1730
},
{
"epoch": 0.47103410936654033,
"grad_norm": 2.510758529664001,
"learning_rate": 4.346342237084267e-05,
"loss": 1.4913,
"step": 1740
},
{
"epoch": 0.4737412019491067,
"grad_norm": 1.6561919074934996,
"learning_rate": 4.339158190450176e-05,
"loss": 1.474,
"step": 1750
},
{
"epoch": 0.476448294531673,
"grad_norm": 1.6475904490305402,
"learning_rate": 4.331940888245552e-05,
"loss": 1.4331,
"step": 1760
},
{
"epoch": 0.47915538711423933,
"grad_norm": 1.5072770205173374,
"learning_rate": 4.3246904609733074e-05,
"loss": 1.4949,
"step": 1770
},
{
"epoch": 0.4818624796968056,
"grad_norm": 1.3220720078913577,
"learning_rate": 4.317407039735323e-05,
"loss": 1.4037,
"step": 1780
},
{
"epoch": 0.484569572279372,
"grad_norm": 2.630264685866369,
"learning_rate": 4.310090756230072e-05,
"loss": 1.4366,
"step": 1790
},
{
"epoch": 0.48727666486193827,
"grad_norm": 1.8523936235855012,
"learning_rate": 4.3027417427502435e-05,
"loss": 1.4757,
"step": 1800
},
{
"epoch": 0.4899837574445046,
"grad_norm": 1.6259780791806493,
"learning_rate": 4.2953601321803494e-05,
"loss": 1.4554,
"step": 1810
},
{
"epoch": 0.4926908500270709,
"grad_norm": 1.3829948059505168,
"learning_rate": 4.287946057994318e-05,
"loss": 1.444,
"step": 1820
},
{
"epoch": 0.49539794260963727,
"grad_norm": 1.472490908240983,
"learning_rate": 4.2804996542530854e-05,
"loss": 1.4799,
"step": 1830
},
{
"epoch": 0.49810503519220356,
"grad_norm": 1.4418412770727913,
"learning_rate": 4.273021055602168e-05,
"loss": 1.4539,
"step": 1840
},
{
"epoch": 0.5008121277747699,
"grad_norm": 1.3579847745967637,
"learning_rate": 4.265510397269229e-05,
"loss": 1.4866,
"step": 1850
},
{
"epoch": 0.5035192203573362,
"grad_norm": 1.2861587741021074,
"learning_rate": 4.2579678150616344e-05,
"loss": 1.4509,
"step": 1860
},
{
"epoch": 0.5062263129399025,
"grad_norm": 1.69680502271894,
"learning_rate": 4.2503934453639945e-05,
"loss": 1.4417,
"step": 1870
},
{
"epoch": 0.5089334055224689,
"grad_norm": 1.374930502329713,
"learning_rate": 4.2427874251357e-05,
"loss": 1.4459,
"step": 1880
},
{
"epoch": 0.5116404981050352,
"grad_norm": 1.6333223217571196,
"learning_rate": 4.235149891908447e-05,
"loss": 1.4187,
"step": 1890
},
{
"epoch": 0.5143475906876015,
"grad_norm": 1.2550594285330077,
"learning_rate": 4.2274809837837444e-05,
"loss": 1.4777,
"step": 1900
},
{
"epoch": 0.5170546832701678,
"grad_norm": 1.7713884342868287,
"learning_rate": 4.219780839430424e-05,
"loss": 1.4742,
"step": 1910
},
{
"epoch": 0.5197617758527342,
"grad_norm": 1.8526614078728505,
"learning_rate": 4.212049598082126e-05,
"loss": 1.4639,
"step": 1920
},
{
"epoch": 0.5224688684353005,
"grad_norm": 1.7077160912171838,
"learning_rate": 4.204287399534791e-05,
"loss": 1.4678,
"step": 1930
},
{
"epoch": 0.5251759610178668,
"grad_norm": 1.7342679697841041,
"learning_rate": 4.196494384144119e-05,
"loss": 1.4587,
"step": 1940
},
{
"epoch": 0.5278830536004331,
"grad_norm": 1.3760999544704575,
"learning_rate": 4.188670692823042e-05,
"loss": 1.4364,
"step": 1950
},
{
"epoch": 0.5305901461829995,
"grad_norm": 1.4927106037863223,
"learning_rate": 4.1808164670391747e-05,
"loss": 1.4636,
"step": 1960
},
{
"epoch": 0.5332972387655658,
"grad_norm": 1.3148570909335389,
"learning_rate": 4.172931848812251e-05,
"loss": 1.4781,
"step": 1970
},
{
"epoch": 0.5360043313481321,
"grad_norm": 1.3941339635980279,
"learning_rate": 4.1650169807115623e-05,
"loss": 1.4214,
"step": 1980
},
{
"epoch": 0.5387114239306985,
"grad_norm": 1.1834123922199669,
"learning_rate": 4.157072005853373e-05,
"loss": 1.4226,
"step": 1990
},
{
"epoch": 0.5414185165132648,
"grad_norm": 1.564745160522187,
"learning_rate": 4.149097067898342e-05,
"loss": 1.4631,
"step": 2000
},
{
"epoch": 0.5441256090958311,
"grad_norm": 1.5351238663722413,
"learning_rate": 4.141092311048914e-05,
"loss": 1.4231,
"step": 2010
},
{
"epoch": 0.5468327016783974,
"grad_norm": 1.4673872230064178,
"learning_rate": 4.133057880046721e-05,
"loss": 1.4308,
"step": 2020
},
{
"epoch": 0.5495397942609638,
"grad_norm": 2.0555674903653576,
"learning_rate": 4.1249939201699605e-05,
"loss": 1.4054,
"step": 2030
},
{
"epoch": 0.5522468868435301,
"grad_norm": 1.4275260085968484,
"learning_rate": 4.116900577230769e-05,
"loss": 1.4171,
"step": 2040
},
{
"epoch": 0.5549539794260964,
"grad_norm": 1.2596642799013085,
"learning_rate": 4.108777997572588e-05,
"loss": 1.4064,
"step": 2050
},
{
"epoch": 0.5576610720086627,
"grad_norm": 1.475728319316803,
"learning_rate": 4.100626328067515e-05,
"loss": 1.451,
"step": 2060
},
{
"epoch": 0.5603681645912291,
"grad_norm": 1.4445720383429121,
"learning_rate": 4.0924457161136465e-05,
"loss": 1.4357,
"step": 2070
},
{
"epoch": 0.5630752571737954,
"grad_norm": 1.277001561017148,
"learning_rate": 4.0842363096324186e-05,
"loss": 1.4152,
"step": 2080
},
{
"epoch": 0.5657823497563617,
"grad_norm": 4.077914450657582,
"learning_rate": 4.075998257065927e-05,
"loss": 1.4219,
"step": 2090
},
{
"epoch": 0.568489442338928,
"grad_norm": 1.4881560211628038,
"learning_rate": 4.067731707374245e-05,
"loss": 1.4558,
"step": 2100
},
{
"epoch": 0.5711965349214944,
"grad_norm": 1.8102222242604538,
"learning_rate": 4.0594368100327276e-05,
"loss": 1.4418,
"step": 2110
},
{
"epoch": 0.5739036275040607,
"grad_norm": 1.5342688729092908,
"learning_rate": 4.051113715029314e-05,
"loss": 1.4186,
"step": 2120
},
{
"epoch": 0.576610720086627,
"grad_norm": 1.4776107653456874,
"learning_rate": 4.0427625728618094e-05,
"loss": 1.4241,
"step": 2130
},
{
"epoch": 0.5793178126691932,
"grad_norm": 1.8005823424782537,
"learning_rate": 4.034383534535169e-05,
"loss": 1.4604,
"step": 2140
},
{
"epoch": 0.5820249052517596,
"grad_norm": 1.4602346950120104,
"learning_rate": 4.0259767515587607e-05,
"loss": 1.4164,
"step": 2150
},
{
"epoch": 0.584731997834326,
"grad_norm": 2.2063404920756025,
"learning_rate": 4.017542375943635e-05,
"loss": 1.4546,
"step": 2160
},
{
"epoch": 0.5874390904168922,
"grad_norm": 1.3913645582438086,
"learning_rate": 4.009080560199767e-05,
"loss": 1.4499,
"step": 2170
},
{
"epoch": 0.5901461829994585,
"grad_norm": 1.4511449754518972,
"learning_rate": 4.0005914573333076e-05,
"loss": 1.3874,
"step": 2180
},
{
"epoch": 0.5928532755820249,
"grad_norm": 1.374243913815498,
"learning_rate": 3.992075220843806e-05,
"loss": 1.3881,
"step": 2190
},
{
"epoch": 0.5955603681645912,
"grad_norm": 1.4550895039236171,
"learning_rate": 3.983532004721446e-05,
"loss": 1.4356,
"step": 2200
},
{
"epoch": 0.5982674607471575,
"grad_norm": 2.8598201176092597,
"learning_rate": 3.974961963444252e-05,
"loss": 1.4194,
"step": 2210
},
{
"epoch": 0.6009745533297238,
"grad_norm": 1.2900018758134322,
"learning_rate": 3.9663652519753016e-05,
"loss": 1.4254,
"step": 2220
},
{
"epoch": 0.6036816459122902,
"grad_norm": 1.3550238391234248,
"learning_rate": 3.95774202575992e-05,
"loss": 1.4265,
"step": 2230
},
{
"epoch": 0.6063887384948565,
"grad_norm": 1.2357167108126608,
"learning_rate": 3.9490924407228725e-05,
"loss": 1.439,
"step": 2240
},
{
"epoch": 0.6090958310774228,
"grad_norm": 1.7838096381387492,
"learning_rate": 3.940416653265542e-05,
"loss": 1.3754,
"step": 2250
},
{
"epoch": 0.6118029236599891,
"grad_norm": 1.5918950572416704,
"learning_rate": 3.931714820263104e-05,
"loss": 1.4193,
"step": 2260
},
{
"epoch": 0.6145100162425555,
"grad_norm": 1.9008726278943935,
"learning_rate": 3.922987099061685e-05,
"loss": 1.4014,
"step": 2270
},
{
"epoch": 0.6172171088251218,
"grad_norm": 1.5355669123167321,
"learning_rate": 3.914233647475525e-05,
"loss": 1.4242,
"step": 2280
},
{
"epoch": 0.6199242014076881,
"grad_norm": 8.645566079065198,
"learning_rate": 3.9054546237841156e-05,
"loss": 1.3978,
"step": 2290
},
{
"epoch": 0.6226312939902545,
"grad_norm": 1.4299000700088789,
"learning_rate": 3.896650186729345e-05,
"loss": 1.4444,
"step": 2300
},
{
"epoch": 0.6253383865728208,
"grad_norm": 1.5568306708216317,
"learning_rate": 3.8878204955126224e-05,
"loss": 1.4535,
"step": 2310
},
{
"epoch": 0.6280454791553871,
"grad_norm": 1.4298727206313604,
"learning_rate": 3.878965709792003e-05,
"loss": 1.3964,
"step": 2320
},
{
"epoch": 0.6307525717379534,
"grad_norm": 1.9081562507364116,
"learning_rate": 3.8700859896793004e-05,
"loss": 1.4637,
"step": 2330
},
{
"epoch": 0.6334596643205198,
"grad_norm": 1.7135745696260376,
"learning_rate": 3.861181495737187e-05,
"loss": 1.4429,
"step": 2340
},
{
"epoch": 0.6361667569030861,
"grad_norm": 1.5360272794198329,
"learning_rate": 3.852252388976299e-05,
"loss": 1.4268,
"step": 2350
},
{
"epoch": 0.6388738494856524,
"grad_norm": 1.2103422788055682,
"learning_rate": 3.843298830852317e-05,
"loss": 1.4469,
"step": 2360
},
{
"epoch": 0.6415809420682187,
"grad_norm": 1.2503359557443374,
"learning_rate": 3.834320983263052e-05,
"loss": 1.4208,
"step": 2370
},
{
"epoch": 0.6442880346507851,
"grad_norm": 1.704839306648677,
"learning_rate": 3.8253190085455136e-05,
"loss": 1.4205,
"step": 2380
},
{
"epoch": 0.6469951272333514,
"grad_norm": 1.7016731029603525,
"learning_rate": 3.816293069472981e-05,
"loss": 1.4311,
"step": 2390
},
{
"epoch": 0.6497022198159177,
"grad_norm": 1.4362792263340989,
"learning_rate": 3.8072433292520526e-05,
"loss": 1.38,
"step": 2400
},
{
"epoch": 0.652409312398484,
"grad_norm": 1.5654681480694452,
"learning_rate": 3.7981699515196975e-05,
"loss": 1.4094,
"step": 2410
},
{
"epoch": 0.6551164049810504,
"grad_norm": 1.4659564972472934,
"learning_rate": 3.789073100340298e-05,
"loss": 1.409,
"step": 2420
},
{
"epoch": 0.6578234975636167,
"grad_norm": 7.417972847111027,
"learning_rate": 3.779952940202686e-05,
"loss": 1.4294,
"step": 2430
},
{
"epoch": 0.660530590146183,
"grad_norm": 1.3497894617366657,
"learning_rate": 3.770809636017158e-05,
"loss": 1.407,
"step": 2440
},
{
"epoch": 0.6632376827287493,
"grad_norm": 1.3081753915783962,
"learning_rate": 3.7616433531125035e-05,
"loss": 1.3756,
"step": 2450
},
{
"epoch": 0.6659447753113157,
"grad_norm": 1.527466315347287,
"learning_rate": 3.752454257233013e-05,
"loss": 1.4118,
"step": 2460
},
{
"epoch": 0.668651867893882,
"grad_norm": 1.6627023176259257,
"learning_rate": 3.7432425145354804e-05,
"loss": 1.4329,
"step": 2470
},
{
"epoch": 0.6713589604764483,
"grad_norm": 1.309294160787172,
"learning_rate": 3.734008291586195e-05,
"loss": 1.3788,
"step": 2480
},
{
"epoch": 0.6740660530590146,
"grad_norm": 1.4372591663740029,
"learning_rate": 3.724751755357937e-05,
"loss": 1.3044,
"step": 2490
},
{
"epoch": 0.676773145641581,
"grad_norm": 1.8886812778978754,
"learning_rate": 3.71547307322695e-05,
"loss": 1.3798,
"step": 2500
},
{
"epoch": 0.6794802382241473,
"grad_norm": 1.9753319009276904,
"learning_rate": 3.7061724129699233e-05,
"loss": 1.4362,
"step": 2510
},
{
"epoch": 0.6821873308067136,
"grad_norm": 1.5870592037312612,
"learning_rate": 3.6968499427609493e-05,
"loss": 1.434,
"step": 2520
},
{
"epoch": 0.6848944233892799,
"grad_norm": 1.2460580041038836,
"learning_rate": 3.6875058311684896e-05,
"loss": 1.3991,
"step": 2530
},
{
"epoch": 0.6876015159718463,
"grad_norm": 1.984255011434617,
"learning_rate": 3.6781402471523244e-05,
"loss": 1.426,
"step": 2540
},
{
"epoch": 0.6903086085544126,
"grad_norm": 1.7368812988388407,
"learning_rate": 3.6687533600604955e-05,
"loss": 1.3894,
"step": 2550
},
{
"epoch": 0.6930157011369789,
"grad_norm": 1.4619686462990809,
"learning_rate": 3.659345339626247e-05,
"loss": 1.4415,
"step": 2560
},
{
"epoch": 0.6957227937195453,
"grad_norm": 1.2916377265272263,
"learning_rate": 3.649916355964956e-05,
"loss": 1.4081,
"step": 2570
},
{
"epoch": 0.6984298863021116,
"grad_norm": 1.4083989609601546,
"learning_rate": 3.640466579571052e-05,
"loss": 1.3883,
"step": 2580
},
{
"epoch": 0.7011369788846779,
"grad_norm": 1.7091470425749693,
"learning_rate": 3.630996181314944e-05,
"loss": 1.4253,
"step": 2590
},
{
"epoch": 0.7038440714672441,
"grad_norm": 1.3001138215713586,
"learning_rate": 3.621505332439918e-05,
"loss": 1.4529,
"step": 2600
},
{
"epoch": 0.7065511640498106,
"grad_norm": 1.7216936795211795,
"learning_rate": 3.611994204559051e-05,
"loss": 1.4673,
"step": 2610
},
{
"epoch": 0.7092582566323768,
"grad_norm": 1.781463187900956,
"learning_rate": 3.602462969652102e-05,
"loss": 1.4071,
"step": 2620
},
{
"epoch": 0.7119653492149431,
"grad_norm": 1.3210728327580619,
"learning_rate": 3.5929118000624064e-05,
"loss": 1.394,
"step": 2630
},
{
"epoch": 0.7146724417975094,
"grad_norm": 1.7986279315081266,
"learning_rate": 3.5833408684937563e-05,
"loss": 1.4134,
"step": 2640
},
{
"epoch": 0.7173795343800758,
"grad_norm": 1.8124093789053668,
"learning_rate": 3.573750348007278e-05,
"loss": 1.395,
"step": 2650
},
{
"epoch": 0.7200866269626421,
"grad_norm": 1.3733544619692244,
"learning_rate": 3.564140412018306e-05,
"loss": 1.4952,
"step": 2660
},
{
"epoch": 0.7227937195452084,
"grad_norm": 1.1625577227023642,
"learning_rate": 3.554511234293243e-05,
"loss": 1.3995,
"step": 2670
},
{
"epoch": 0.7255008121277747,
"grad_norm": 1.1967978911566575,
"learning_rate": 3.54486298894642e-05,
"loss": 1.3751,
"step": 2680
},
{
"epoch": 0.7282079047103411,
"grad_norm": 1.5901060329868977,
"learning_rate": 3.5351958504369504e-05,
"loss": 1.3714,
"step": 2690
},
{
"epoch": 0.7309149972929074,
"grad_norm": 1.4331430962829053,
"learning_rate": 3.525509993565569e-05,
"loss": 1.4538,
"step": 2700
},
{
"epoch": 0.7336220898754737,
"grad_norm": 1.578052835087518,
"learning_rate": 3.5158055934714784e-05,
"loss": 1.4321,
"step": 2710
},
{
"epoch": 0.73632918245804,
"grad_norm": 1.4900021275863136,
"learning_rate": 3.506082825629176e-05,
"loss": 1.4224,
"step": 2720
},
{
"epoch": 0.7390362750406064,
"grad_norm": 1.803432159658249,
"learning_rate": 3.496341865845286e-05,
"loss": 1.4066,
"step": 2730
},
{
"epoch": 0.7417433676231727,
"grad_norm": 1.8906057321665828,
"learning_rate": 3.4865828902553786e-05,
"loss": 1.3412,
"step": 2740
},
{
"epoch": 0.744450460205739,
"grad_norm": 1.5022839535897015,
"learning_rate": 3.476806075320784e-05,
"loss": 1.4204,
"step": 2750
},
{
"epoch": 0.7471575527883053,
"grad_norm": 1.2487314478311355,
"learning_rate": 3.467011597825402e-05,
"loss": 1.3934,
"step": 2760
},
{
"epoch": 0.7498646453708717,
"grad_norm": 1.546439548647951,
"learning_rate": 3.4571996348725066e-05,
"loss": 1.3971,
"step": 2770
},
{
"epoch": 0.752571737953438,
"grad_norm": 1.2557741585940914,
"learning_rate": 3.447370363881543e-05,
"loss": 1.4221,
"step": 2780
},
{
"epoch": 0.7552788305360043,
"grad_norm": 1.3960693603390584,
"learning_rate": 3.437523962584923e-05,
"loss": 1.4134,
"step": 2790
},
{
"epoch": 0.7579859231185706,
"grad_norm": 1.5039396517235502,
"learning_rate": 3.427660609024799e-05,
"loss": 1.3927,
"step": 2800
},
{
"epoch": 0.760693015701137,
"grad_norm": 1.0815563013041047,
"learning_rate": 3.417780481549863e-05,
"loss": 1.3605,
"step": 2810
},
{
"epoch": 0.7634001082837033,
"grad_norm": 1.3586955443715294,
"learning_rate": 3.407883758812106e-05,
"loss": 1.4041,
"step": 2820
},
{
"epoch": 0.7661072008662696,
"grad_norm": 1.0920930772742923,
"learning_rate": 3.397970619763597e-05,
"loss": 1.428,
"step": 2830
},
{
"epoch": 0.7688142934488359,
"grad_norm": 1.3785331176013547,
"learning_rate": 3.388041243653242e-05,
"loss": 1.4384,
"step": 2840
},
{
"epoch": 0.7715213860314023,
"grad_norm": 1.3052185999792532,
"learning_rate": 3.3780958100235445e-05,
"loss": 1.3429,
"step": 2850
},
{
"epoch": 0.7742284786139686,
"grad_norm": 2.042603724777467,
"learning_rate": 3.3681344987073615e-05,
"loss": 1.3328,
"step": 2860
},
{
"epoch": 0.7769355711965349,
"grad_norm": 1.441878338125671,
"learning_rate": 3.358157489824648e-05,
"loss": 1.4065,
"step": 2870
},
{
"epoch": 0.7796426637791013,
"grad_norm": 1.35241437286716,
"learning_rate": 3.348164963779202e-05,
"loss": 1.4298,
"step": 2880
},
{
"epoch": 0.7823497563616676,
"grad_norm": 1.6143507738425422,
"learning_rate": 3.3381571012554024e-05,
"loss": 1.3649,
"step": 2890
},
{
"epoch": 0.7850568489442339,
"grad_norm": 1.4514578355155894,
"learning_rate": 3.328134083214941e-05,
"loss": 1.3838,
"step": 2900
},
{
"epoch": 0.7877639415268002,
"grad_norm": 1.515687318078472,
"learning_rate": 3.318096090893551e-05,
"loss": 1.4035,
"step": 2910
},
{
"epoch": 0.7904710341093666,
"grad_norm": 1.2884021166011788,
"learning_rate": 3.308043305797729e-05,
"loss": 1.3559,
"step": 2920
},
{
"epoch": 0.7931781266919329,
"grad_norm": 1.4337183263775037,
"learning_rate": 3.297975909701457e-05,
"loss": 1.4059,
"step": 2930
},
{
"epoch": 0.7958852192744992,
"grad_norm": 1.5371766509521272,
"learning_rate": 3.2878940846429085e-05,
"loss": 1.3774,
"step": 2940
},
{
"epoch": 0.7985923118570655,
"grad_norm": 1.3405567316289324,
"learning_rate": 3.277798012921163e-05,
"loss": 1.3904,
"step": 2950
},
{
"epoch": 0.8012994044396319,
"grad_norm": 1.4555652196907567,
"learning_rate": 3.2676878770929075e-05,
"loss": 1.4146,
"step": 2960
},
{
"epoch": 0.8040064970221982,
"grad_norm": 1.4152140878656931,
"learning_rate": 3.257563859969134e-05,
"loss": 1.4307,
"step": 2970
},
{
"epoch": 0.8067135896047645,
"grad_norm": 1.4286723786741538,
"learning_rate": 3.2474261446118384e-05,
"loss": 1.3802,
"step": 2980
},
{
"epoch": 0.8094206821873308,
"grad_norm": 1.7253495020201928,
"learning_rate": 3.237274914330704e-05,
"loss": 1.3832,
"step": 2990
},
{
"epoch": 0.8121277747698972,
"grad_norm": 1.42537117722243,
"learning_rate": 3.227110352679791e-05,
"loss": 1.4196,
"step": 3000
},
{
"epoch": 0.8148348673524635,
"grad_norm": 1.6839610738971422,
"learning_rate": 3.216932643454219e-05,
"loss": 1.3631,
"step": 3010
},
{
"epoch": 0.8175419599350298,
"grad_norm": 1.971059453029315,
"learning_rate": 3.2067419706868404e-05,
"loss": 1.3943,
"step": 3020
},
{
"epoch": 0.8202490525175961,
"grad_norm": 1.4912077914008548,
"learning_rate": 3.196538518644911e-05,
"loss": 1.3607,
"step": 3030
},
{
"epoch": 0.8229561451001625,
"grad_norm": 2.642611457217887,
"learning_rate": 3.186322471826765e-05,
"loss": 1.3885,
"step": 3040
},
{
"epoch": 0.8256632376827288,
"grad_norm": 2.0970669091185052,
"learning_rate": 3.176094014958473e-05,
"loss": 1.4174,
"step": 3050
},
{
"epoch": 0.828370330265295,
"grad_norm": 1.6411214953570172,
"learning_rate": 3.1658533329905034e-05,
"loss": 1.3908,
"step": 3060
},
{
"epoch": 0.8310774228478613,
"grad_norm": 1.4488420914686069,
"learning_rate": 3.155600611094377e-05,
"loss": 1.3525,
"step": 3070
},
{
"epoch": 0.8337845154304278,
"grad_norm": 1.3921616791086928,
"learning_rate": 3.145336034659322e-05,
"loss": 1.3815,
"step": 3080
},
{
"epoch": 0.836491608012994,
"grad_norm": 1.3485960002369286,
"learning_rate": 3.135059789288919e-05,
"loss": 1.3718,
"step": 3090
},
{
"epoch": 0.8391987005955603,
"grad_norm": 3.0257814626066333,
"learning_rate": 3.1247720607977465e-05,
"loss": 1.3773,
"step": 3100
},
{
"epoch": 0.8419057931781266,
"grad_norm": 1.607580731614463,
"learning_rate": 3.11447303520802e-05,
"loss": 1.3527,
"step": 3110
},
{
"epoch": 0.844612885760693,
"grad_norm": 1.370395503929159,
"learning_rate": 3.1041628987462256e-05,
"loss": 1.3978,
"step": 3120
},
{
"epoch": 0.8473199783432593,
"grad_norm": 1.6192384678646978,
"learning_rate": 3.0938418378397615e-05,
"loss": 1.3757,
"step": 3130
},
{
"epoch": 0.8500270709258256,
"grad_norm": 1.4273589649993463,
"learning_rate": 3.0835100391135555e-05,
"loss": 1.3831,
"step": 3140
},
{
"epoch": 0.852734163508392,
"grad_norm": 1.35446316177469,
"learning_rate": 3.0731676893866995e-05,
"loss": 1.3877,
"step": 3150
},
{
"epoch": 0.8554412560909583,
"grad_norm": 2.163498528349308,
"learning_rate": 3.062814975669067e-05,
"loss": 1.3685,
"step": 3160
},
{
"epoch": 0.8581483486735246,
"grad_norm": 1.3900238288203155,
"learning_rate": 3.052452085157933e-05,
"loss": 1.358,
"step": 3170
},
{
"epoch": 0.8608554412560909,
"grad_norm": 1.216239709612049,
"learning_rate": 3.042079205234589e-05,
"loss": 1.3548,
"step": 3180
},
{
"epoch": 0.8635625338386573,
"grad_norm": 1.349220629569809,
"learning_rate": 3.0316965234609535e-05,
"loss": 1.3874,
"step": 3190
},
{
"epoch": 0.8662696264212236,
"grad_norm": 1.1250681117090537,
"learning_rate": 3.021304227576182e-05,
"loss": 1.3876,
"step": 3200
},
{
"epoch": 0.8689767190037899,
"grad_norm": 1.3913714515749762,
"learning_rate": 3.0109025054932744e-05,
"loss": 1.347,
"step": 3210
},
{
"epoch": 0.8716838115863562,
"grad_norm": 1.0609773784043655,
"learning_rate": 3.0004915452956695e-05,
"loss": 1.3667,
"step": 3220
},
{
"epoch": 0.8743909041689226,
"grad_norm": 1.4538812767833513,
"learning_rate": 2.9900715352338538e-05,
"loss": 1.3696,
"step": 3230
},
{
"epoch": 0.8770979967514889,
"grad_norm": 1.7108604924584878,
"learning_rate": 2.979642663721951e-05,
"loss": 1.3795,
"step": 3240
},
{
"epoch": 0.8798050893340552,
"grad_norm": 2.5139594534456546,
"learning_rate": 2.9692051193343184e-05,
"loss": 1.3734,
"step": 3250
},
{
"epoch": 0.8825121819166215,
"grad_norm": 2.5623897245796736,
"learning_rate": 2.958759090802134e-05,
"loss": 1.372,
"step": 3260
},
{
"epoch": 0.8852192744991879,
"grad_norm": 1.7887583210364995,
"learning_rate": 2.948304767009986e-05,
"loss": 1.3814,
"step": 3270
},
{
"epoch": 0.8879263670817542,
"grad_norm": 1.0710727849835535,
"learning_rate": 2.93784233699246e-05,
"loss": 1.4349,
"step": 3280
},
{
"epoch": 0.8906334596643205,
"grad_norm": 2.563978550658464,
"learning_rate": 2.927371989930714e-05,
"loss": 1.3437,
"step": 3290
},
{
"epoch": 0.8933405522468868,
"grad_norm": 2.145609531812958,
"learning_rate": 2.9168939151490638e-05,
"loss": 1.4138,
"step": 3300
},
{
"epoch": 0.8960476448294532,
"grad_norm": 1.4751882863646832,
"learning_rate": 2.906408302111556e-05,
"loss": 1.3525,
"step": 3310
},
{
"epoch": 0.8987547374120195,
"grad_norm": 1.499040784897915,
"learning_rate": 2.8959153404185468e-05,
"loss": 1.3717,
"step": 3320
},
{
"epoch": 0.9014618299945858,
"grad_norm": 1.3438190150311098,
"learning_rate": 2.885415219803267e-05,
"loss": 1.3545,
"step": 3330
},
{
"epoch": 0.9041689225771521,
"grad_norm": 1.4506182307044833,
"learning_rate": 2.8749081301283955e-05,
"loss": 1.3841,
"step": 3340
},
{
"epoch": 0.9068760151597185,
"grad_norm": 1.5878698130305084,
"learning_rate": 2.8643942613826264e-05,
"loss": 1.3263,
"step": 3350
},
{
"epoch": 0.9095831077422848,
"grad_norm": 1.3742244372155115,
"learning_rate": 2.8538738036772318e-05,
"loss": 1.3403,
"step": 3360
},
{
"epoch": 0.9122902003248511,
"grad_norm": 1.2893381483305373,
"learning_rate": 2.8433469472426257e-05,
"loss": 1.3252,
"step": 3370
},
{
"epoch": 0.9149972929074174,
"grad_norm": 1.7256165310292995,
"learning_rate": 2.832813882424923e-05,
"loss": 1.3595,
"step": 3380
},
{
"epoch": 0.9177043854899838,
"grad_norm": 1.574410258409617,
"learning_rate": 2.8222747996824973e-05,
"loss": 1.3498,
"step": 3390
},
{
"epoch": 0.9204114780725501,
"grad_norm": 1.4304720741459473,
"learning_rate": 2.8117298895825417e-05,
"loss": 1.3613,
"step": 3400
},
{
"epoch": 0.9231185706551164,
"grad_norm": 1.8155878547283097,
"learning_rate": 2.8011793427976152e-05,
"loss": 1.3675,
"step": 3410
},
{
"epoch": 0.9258256632376828,
"grad_norm": 1.568315689782668,
"learning_rate": 2.7906233501022005e-05,
"loss": 1.3878,
"step": 3420
},
{
"epoch": 0.9285327558202491,
"grad_norm": 1.3449271497175002,
"learning_rate": 2.780062102369255e-05,
"loss": 1.3451,
"step": 3430
},
{
"epoch": 0.9312398484028154,
"grad_norm": 1.8796135561662028,
"learning_rate": 2.7694957905667533e-05,
"loss": 1.3541,
"step": 3440
},
{
"epoch": 0.9339469409853817,
"grad_norm": 1.5881147795174977,
"learning_rate": 2.7589246057542424e-05,
"loss": 1.3481,
"step": 3450
},
{
"epoch": 0.9366540335679481,
"grad_norm": 1.5669861875980806,
"learning_rate": 2.74834873907938e-05,
"loss": 1.3049,
"step": 3460
},
{
"epoch": 0.9393611261505144,
"grad_norm": 1.3159700318764556,
"learning_rate": 2.7377683817744825e-05,
"loss": 1.3724,
"step": 3470
},
{
"epoch": 0.9420682187330807,
"grad_norm": 1.1861068665908672,
"learning_rate": 2.727183725153066e-05,
"loss": 1.3996,
"step": 3480
},
{
"epoch": 0.944775311315647,
"grad_norm": 1.588796314194285,
"learning_rate": 2.7165949606063838e-05,
"loss": 1.3868,
"step": 3490
},
{
"epoch": 0.9474824038982134,
"grad_norm": 1.8405833406640584,
"learning_rate": 2.7060022795999722e-05,
"loss": 1.3248,
"step": 3500
},
{
"epoch": 0.9501894964807797,
"grad_norm": 1.4845932153084007,
"learning_rate": 2.6954058736701825e-05,
"loss": 1.3113,
"step": 3510
},
{
"epoch": 0.952896589063346,
"grad_norm": 2.2421879491898347,
"learning_rate": 2.684805934420721e-05,
"loss": 1.3569,
"step": 3520
},
{
"epoch": 0.9556036816459123,
"grad_norm": 1.8263161858131314,
"learning_rate": 2.674202653519183e-05,
"loss": 1.353,
"step": 3530
},
{
"epoch": 0.9583107742284787,
"grad_norm": 1.6830556154140768,
"learning_rate": 2.6635962226935844e-05,
"loss": 1.3581,
"step": 3540
},
{
"epoch": 0.961017866811045,
"grad_norm": 1.1962057897073894,
"learning_rate": 2.652986833728905e-05,
"loss": 1.3377,
"step": 3550
},
{
"epoch": 0.9637249593936112,
"grad_norm": 1.1605034457482457,
"learning_rate": 2.642374678463606e-05,
"loss": 1.3142,
"step": 3560
},
{
"epoch": 0.9664320519761775,
"grad_norm": 1.6116484287446706,
"learning_rate": 2.6317599487861732e-05,
"loss": 1.3704,
"step": 3570
},
{
"epoch": 0.969139144558744,
"grad_norm": 2.022789120877381,
"learning_rate": 2.6211428366316414e-05,
"loss": 1.3188,
"step": 3580
},
{
"epoch": 0.9718462371413102,
"grad_norm": 2.0186900985710245,
"learning_rate": 2.6105235339781254e-05,
"loss": 1.3168,
"step": 3590
},
{
"epoch": 0.9745533297238765,
"grad_norm": 2.040600986429446,
"learning_rate": 2.599902232843348e-05,
"loss": 1.361,
"step": 3600
},
{
"epoch": 0.9772604223064428,
"grad_norm": 1.2311848491778494,
"learning_rate": 2.5892791252811694e-05,
"loss": 1.3658,
"step": 3610
},
{
"epoch": 0.9799675148890092,
"grad_norm": 1.8540679978665582,
"learning_rate": 2.578654403378112e-05,
"loss": 1.373,
"step": 3620
},
{
"epoch": 0.9826746074715755,
"grad_norm": 1.3073823909407571,
"learning_rate": 2.5680282592498913e-05,
"loss": 1.3845,
"step": 3630
},
{
"epoch": 0.9853817000541418,
"grad_norm": 1.6720630501211762,
"learning_rate": 2.5574008850379366e-05,
"loss": 1.3438,
"step": 3640
},
{
"epoch": 0.9880887926367081,
"grad_norm": 1.4546581832324275,
"learning_rate": 2.546772472905921e-05,
"loss": 1.2815,
"step": 3650
},
{
"epoch": 0.9907958852192745,
"grad_norm": 1.4509964601293808,
"learning_rate": 2.5361432150362848e-05,
"loss": 1.3469,
"step": 3660
},
{
"epoch": 0.9935029778018408,
"grad_norm": 1.4316417505770653,
"learning_rate": 2.5255133036267608e-05,
"loss": 1.3315,
"step": 3670
},
{
"epoch": 0.9962100703844071,
"grad_norm": 1.7706024949129144,
"learning_rate": 2.5148829308869004e-05,
"loss": 1.3607,
"step": 3680
},
{
"epoch": 0.9989171629669734,
"grad_norm": 1.5222816405267459,
"learning_rate": 2.5042522890345936e-05,
"loss": 1.2815,
"step": 3690
}
],
"logging_steps": 10,
"max_steps": 7388,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"total_flos": 1754891515854848.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}