terry69's picture
Model save
423560a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2414,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00041425020712510354,
"grad_norm": 24.495001410392245,
"learning_rate": 4.132231404958678e-08,
"loss": 1.4066,
"step": 1
},
{
"epoch": 0.002071251035625518,
"grad_norm": 24.278159346480546,
"learning_rate": 2.066115702479339e-07,
"loss": 1.4104,
"step": 5
},
{
"epoch": 0.004142502071251036,
"grad_norm": 15.207256152484726,
"learning_rate": 4.132231404958678e-07,
"loss": 1.3724,
"step": 10
},
{
"epoch": 0.006213753106876553,
"grad_norm": 8.665623744138836,
"learning_rate": 6.198347107438018e-07,
"loss": 1.2508,
"step": 15
},
{
"epoch": 0.008285004142502071,
"grad_norm": 11.05134062750248,
"learning_rate": 8.264462809917356e-07,
"loss": 1.1327,
"step": 20
},
{
"epoch": 0.010356255178127589,
"grad_norm": 4.379402402170414,
"learning_rate": 1.0330578512396695e-06,
"loss": 1.0227,
"step": 25
},
{
"epoch": 0.012427506213753107,
"grad_norm": 3.3489843705809355,
"learning_rate": 1.2396694214876035e-06,
"loss": 0.9669,
"step": 30
},
{
"epoch": 0.014498757249378625,
"grad_norm": 3.3543061730911634,
"learning_rate": 1.4462809917355372e-06,
"loss": 0.9568,
"step": 35
},
{
"epoch": 0.016570008285004142,
"grad_norm": 3.0246929252712045,
"learning_rate": 1.6528925619834712e-06,
"loss": 0.9285,
"step": 40
},
{
"epoch": 0.018641259320629662,
"grad_norm": 3.0205344615509735,
"learning_rate": 1.859504132231405e-06,
"loss": 0.9156,
"step": 45
},
{
"epoch": 0.020712510356255178,
"grad_norm": 3.2104230970271694,
"learning_rate": 2.066115702479339e-06,
"loss": 0.9066,
"step": 50
},
{
"epoch": 0.022783761391880698,
"grad_norm": 3.0769403333765366,
"learning_rate": 2.2727272727272728e-06,
"loss": 0.9123,
"step": 55
},
{
"epoch": 0.024855012427506214,
"grad_norm": 3.1192139331466073,
"learning_rate": 2.479338842975207e-06,
"loss": 0.8824,
"step": 60
},
{
"epoch": 0.026926263463131733,
"grad_norm": 3.174458267006907,
"learning_rate": 2.6859504132231405e-06,
"loss": 0.8872,
"step": 65
},
{
"epoch": 0.02899751449875725,
"grad_norm": 3.3704437475031614,
"learning_rate": 2.8925619834710743e-06,
"loss": 0.876,
"step": 70
},
{
"epoch": 0.03106876553438277,
"grad_norm": 3.2746704156283384,
"learning_rate": 3.0991735537190086e-06,
"loss": 0.8687,
"step": 75
},
{
"epoch": 0.033140016570008285,
"grad_norm": 3.324353070402648,
"learning_rate": 3.3057851239669424e-06,
"loss": 0.8715,
"step": 80
},
{
"epoch": 0.035211267605633804,
"grad_norm": 2.9659428539001875,
"learning_rate": 3.5123966942148763e-06,
"loss": 0.8686,
"step": 85
},
{
"epoch": 0.037282518641259324,
"grad_norm": 3.1234867961603885,
"learning_rate": 3.71900826446281e-06,
"loss": 0.8637,
"step": 90
},
{
"epoch": 0.03935376967688484,
"grad_norm": 3.1123775472434705,
"learning_rate": 3.925619834710744e-06,
"loss": 0.8553,
"step": 95
},
{
"epoch": 0.041425020712510356,
"grad_norm": 3.115177689829066,
"learning_rate": 4.132231404958678e-06,
"loss": 0.8695,
"step": 100
},
{
"epoch": 0.043496271748135876,
"grad_norm": 3.0340671696442714,
"learning_rate": 4.338842975206612e-06,
"loss": 0.8712,
"step": 105
},
{
"epoch": 0.045567522783761395,
"grad_norm": 3.073424358865733,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.8353,
"step": 110
},
{
"epoch": 0.04763877381938691,
"grad_norm": 2.9777104137179227,
"learning_rate": 4.75206611570248e-06,
"loss": 0.8473,
"step": 115
},
{
"epoch": 0.04971002485501243,
"grad_norm": 3.034579402136387,
"learning_rate": 4.958677685950414e-06,
"loss": 0.8475,
"step": 120
},
{
"epoch": 0.05178127589063795,
"grad_norm": 3.0813316554860704,
"learning_rate": 5.165289256198347e-06,
"loss": 0.8507,
"step": 125
},
{
"epoch": 0.053852526926263466,
"grad_norm": 3.06671775904819,
"learning_rate": 5.371900826446281e-06,
"loss": 0.8463,
"step": 130
},
{
"epoch": 0.05592377796188898,
"grad_norm": 3.2475108548122957,
"learning_rate": 5.578512396694216e-06,
"loss": 0.8335,
"step": 135
},
{
"epoch": 0.0579950289975145,
"grad_norm": 3.05517684817537,
"learning_rate": 5.785123966942149e-06,
"loss": 0.8197,
"step": 140
},
{
"epoch": 0.06006628003314002,
"grad_norm": 3.209092202417741,
"learning_rate": 5.991735537190083e-06,
"loss": 0.838,
"step": 145
},
{
"epoch": 0.06213753106876554,
"grad_norm": 3.317353575296279,
"learning_rate": 6.198347107438017e-06,
"loss": 0.8424,
"step": 150
},
{
"epoch": 0.06420878210439106,
"grad_norm": 3.50463295880707,
"learning_rate": 6.404958677685951e-06,
"loss": 0.833,
"step": 155
},
{
"epoch": 0.06628003314001657,
"grad_norm": 2.966129268982567,
"learning_rate": 6.611570247933885e-06,
"loss": 0.8198,
"step": 160
},
{
"epoch": 0.06835128417564208,
"grad_norm": 3.060604077046953,
"learning_rate": 6.818181818181818e-06,
"loss": 0.8123,
"step": 165
},
{
"epoch": 0.07042253521126761,
"grad_norm": 3.106442872246759,
"learning_rate": 7.0247933884297525e-06,
"loss": 0.8256,
"step": 170
},
{
"epoch": 0.07249378624689312,
"grad_norm": 3.2061341459353603,
"learning_rate": 7.231404958677687e-06,
"loss": 0.8272,
"step": 175
},
{
"epoch": 0.07456503728251865,
"grad_norm": 3.261013201810741,
"learning_rate": 7.43801652892562e-06,
"loss": 0.8226,
"step": 180
},
{
"epoch": 0.07663628831814416,
"grad_norm": 3.0418461202096387,
"learning_rate": 7.644628099173555e-06,
"loss": 0.8134,
"step": 185
},
{
"epoch": 0.07870753935376967,
"grad_norm": 2.925954602214559,
"learning_rate": 7.851239669421489e-06,
"loss": 0.8106,
"step": 190
},
{
"epoch": 0.0807787903893952,
"grad_norm": 3.096505735396263,
"learning_rate": 8.057851239669421e-06,
"loss": 0.8219,
"step": 195
},
{
"epoch": 0.08285004142502071,
"grad_norm": 3.0178444567465883,
"learning_rate": 8.264462809917356e-06,
"loss": 0.8139,
"step": 200
},
{
"epoch": 0.08492129246064622,
"grad_norm": 3.085074860786418,
"learning_rate": 8.47107438016529e-06,
"loss": 0.8079,
"step": 205
},
{
"epoch": 0.08699254349627175,
"grad_norm": 2.9513280908493438,
"learning_rate": 8.677685950413224e-06,
"loss": 0.791,
"step": 210
},
{
"epoch": 0.08906379453189726,
"grad_norm": 3.0685126689734843,
"learning_rate": 8.884297520661158e-06,
"loss": 0.8065,
"step": 215
},
{
"epoch": 0.09113504556752279,
"grad_norm": 3.3564757893140067,
"learning_rate": 9.090909090909091e-06,
"loss": 0.817,
"step": 220
},
{
"epoch": 0.0932062966031483,
"grad_norm": 3.144009095079128,
"learning_rate": 9.297520661157025e-06,
"loss": 0.799,
"step": 225
},
{
"epoch": 0.09527754763877382,
"grad_norm": 2.9152439896159703,
"learning_rate": 9.50413223140496e-06,
"loss": 0.8123,
"step": 230
},
{
"epoch": 0.09734879867439934,
"grad_norm": 2.959984059385148,
"learning_rate": 9.710743801652894e-06,
"loss": 0.7906,
"step": 235
},
{
"epoch": 0.09942004971002485,
"grad_norm": 3.0028265600360067,
"learning_rate": 9.917355371900828e-06,
"loss": 0.8028,
"step": 240
},
{
"epoch": 0.10149130074565037,
"grad_norm": 2.8289351291922693,
"learning_rate": 9.999952928077044e-06,
"loss": 0.7937,
"step": 245
},
{
"epoch": 0.1035625517812759,
"grad_norm": 2.8859057720780013,
"learning_rate": 9.999665269535307e-06,
"loss": 0.7891,
"step": 250
},
{
"epoch": 0.1056338028169014,
"grad_norm": 2.8213578917004445,
"learning_rate": 9.99911611854702e-06,
"loss": 0.7941,
"step": 255
},
{
"epoch": 0.10770505385252693,
"grad_norm": 2.9934763093271486,
"learning_rate": 9.998305503833872e-06,
"loss": 0.8038,
"step": 260
},
{
"epoch": 0.10977630488815245,
"grad_norm": 3.1184221834210617,
"learning_rate": 9.997233467792626e-06,
"loss": 0.7883,
"step": 265
},
{
"epoch": 0.11184755592377796,
"grad_norm": 3.080255277284623,
"learning_rate": 9.995900066492902e-06,
"loss": 0.7908,
"step": 270
},
{
"epoch": 0.11391880695940348,
"grad_norm": 2.961023177669383,
"learning_rate": 9.994305369674242e-06,
"loss": 0.8018,
"step": 275
},
{
"epoch": 0.115990057995029,
"grad_norm": 3.2994763750087572,
"learning_rate": 9.992449460742464e-06,
"loss": 0.7929,
"step": 280
},
{
"epoch": 0.11806130903065451,
"grad_norm": 3.118171539850542,
"learning_rate": 9.9903324367653e-06,
"loss": 0.7756,
"step": 285
},
{
"epoch": 0.12013256006628004,
"grad_norm": 2.93509763542972,
"learning_rate": 9.98795440846732e-06,
"loss": 0.7893,
"step": 290
},
{
"epoch": 0.12220381110190555,
"grad_norm": 3.005099209957883,
"learning_rate": 9.985315500224135e-06,
"loss": 0.7837,
"step": 295
},
{
"epoch": 0.12427506213753108,
"grad_norm": 3.0065053233637262,
"learning_rate": 9.982415850055902e-06,
"loss": 0.775,
"step": 300
},
{
"epoch": 0.1263463131731566,
"grad_norm": 2.8875884081237726,
"learning_rate": 9.979255609620095e-06,
"loss": 0.773,
"step": 305
},
{
"epoch": 0.12841756420878211,
"grad_norm": 2.953711663180962,
"learning_rate": 9.975834944203581e-06,
"loss": 0.7833,
"step": 310
},
{
"epoch": 0.1304888152444076,
"grad_norm": 3.0971637074544875,
"learning_rate": 9.972154032713973e-06,
"loss": 0.778,
"step": 315
},
{
"epoch": 0.13256006628003314,
"grad_norm": 2.9856834222834334,
"learning_rate": 9.968213067670265e-06,
"loss": 0.7632,
"step": 320
},
{
"epoch": 0.13463131731565867,
"grad_norm": 3.026025328201696,
"learning_rate": 9.964012255192776e-06,
"loss": 0.76,
"step": 325
},
{
"epoch": 0.13670256835128416,
"grad_norm": 3.1660979738179336,
"learning_rate": 9.959551814992364e-06,
"loss": 0.7553,
"step": 330
},
{
"epoch": 0.1387738193869097,
"grad_norm": 3.082731646367652,
"learning_rate": 9.954831980358928e-06,
"loss": 0.7724,
"step": 335
},
{
"epoch": 0.14084507042253522,
"grad_norm": 2.808393207259557,
"learning_rate": 9.949852998149217e-06,
"loss": 0.7741,
"step": 340
},
{
"epoch": 0.14291632145816072,
"grad_norm": 3.1777761122613013,
"learning_rate": 9.944615128773911e-06,
"loss": 0.7634,
"step": 345
},
{
"epoch": 0.14498757249378624,
"grad_norm": 2.740280537246316,
"learning_rate": 9.939118646184007e-06,
"loss": 0.7518,
"step": 350
},
{
"epoch": 0.14705882352941177,
"grad_norm": 2.796661144343301,
"learning_rate": 9.933363837856485e-06,
"loss": 0.7508,
"step": 355
},
{
"epoch": 0.1491300745650373,
"grad_norm": 2.739682595325342,
"learning_rate": 9.927351004779275e-06,
"loss": 0.74,
"step": 360
},
{
"epoch": 0.1512013256006628,
"grad_norm": 2.859085581091372,
"learning_rate": 9.921080461435522e-06,
"loss": 0.7528,
"step": 365
},
{
"epoch": 0.15327257663628832,
"grad_norm": 2.7680953782694586,
"learning_rate": 9.914552535787122e-06,
"loss": 0.7421,
"step": 370
},
{
"epoch": 0.15534382767191385,
"grad_norm": 2.8377090221372274,
"learning_rate": 9.90776756925758e-06,
"loss": 0.7455,
"step": 375
},
{
"epoch": 0.15741507870753935,
"grad_norm": 2.932524071556518,
"learning_rate": 9.900725916714157e-06,
"loss": 0.7335,
"step": 380
},
{
"epoch": 0.15948632974316487,
"grad_norm": 2.995608607614311,
"learning_rate": 9.893427946449297e-06,
"loss": 0.7451,
"step": 385
},
{
"epoch": 0.1615575807787904,
"grad_norm": 3.4887684282046254,
"learning_rate": 9.885874040161373e-06,
"loss": 0.7333,
"step": 390
},
{
"epoch": 0.1636288318144159,
"grad_norm": 2.685483854880427,
"learning_rate": 9.878064592934723e-06,
"loss": 0.7387,
"step": 395
},
{
"epoch": 0.16570008285004142,
"grad_norm": 2.6709003339090436,
"learning_rate": 9.87000001321898e-06,
"loss": 0.7332,
"step": 400
},
{
"epoch": 0.16777133388566695,
"grad_norm": 2.8316780066782177,
"learning_rate": 9.86168072280772e-06,
"loss": 0.7306,
"step": 405
},
{
"epoch": 0.16984258492129245,
"grad_norm": 2.800680145525587,
"learning_rate": 9.853107156816393e-06,
"loss": 0.7258,
"step": 410
},
{
"epoch": 0.17191383595691798,
"grad_norm": 3.015838299479087,
"learning_rate": 9.844279763659566e-06,
"loss": 0.716,
"step": 415
},
{
"epoch": 0.1739850869925435,
"grad_norm": 2.861365629083953,
"learning_rate": 9.835199005027477e-06,
"loss": 0.7256,
"step": 420
},
{
"epoch": 0.176056338028169,
"grad_norm": 2.8083228586917532,
"learning_rate": 9.825865355861878e-06,
"loss": 0.7249,
"step": 425
},
{
"epoch": 0.17812758906379453,
"grad_norm": 2.9188232580232696,
"learning_rate": 9.816279304331202e-06,
"loss": 0.728,
"step": 430
},
{
"epoch": 0.18019884009942005,
"grad_norm": 3.047823960774399,
"learning_rate": 9.806441351805025e-06,
"loss": 0.7158,
"step": 435
},
{
"epoch": 0.18227009113504558,
"grad_norm": 26.196165622685598,
"learning_rate": 9.79635201282785e-06,
"loss": 0.73,
"step": 440
},
{
"epoch": 0.18434134217067108,
"grad_norm": 2.941695755078049,
"learning_rate": 9.786011815092193e-06,
"loss": 0.709,
"step": 445
},
{
"epoch": 0.1864125932062966,
"grad_norm": 2.983234429221489,
"learning_rate": 9.775421299410977e-06,
"loss": 0.7113,
"step": 450
},
{
"epoch": 0.18848384424192213,
"grad_norm": 2.7152686163875512,
"learning_rate": 9.764581019689255e-06,
"loss": 0.7245,
"step": 455
},
{
"epoch": 0.19055509527754763,
"grad_norm": 3.140548600333453,
"learning_rate": 9.753491542895237e-06,
"loss": 0.6944,
"step": 460
},
{
"epoch": 0.19262634631317316,
"grad_norm": 2.7171623737634953,
"learning_rate": 9.742153449030639e-06,
"loss": 0.6997,
"step": 465
},
{
"epoch": 0.19469759734879868,
"grad_norm": 2.8513078067731144,
"learning_rate": 9.730567331100333e-06,
"loss": 0.6851,
"step": 470
},
{
"epoch": 0.19676884838442418,
"grad_norm": 2.7624587568061973,
"learning_rate": 9.71873379508136e-06,
"loss": 0.6879,
"step": 475
},
{
"epoch": 0.1988400994200497,
"grad_norm": 2.704414137904599,
"learning_rate": 9.706653459891207e-06,
"loss": 0.6869,
"step": 480
},
{
"epoch": 0.20091135045567524,
"grad_norm": 2.692949982244341,
"learning_rate": 9.694326957355452e-06,
"loss": 0.6939,
"step": 485
},
{
"epoch": 0.20298260149130073,
"grad_norm": 3.036163976932751,
"learning_rate": 9.681754932174719e-06,
"loss": 0.6873,
"step": 490
},
{
"epoch": 0.20505385252692626,
"grad_norm": 2.7169963354471163,
"learning_rate": 9.668938041890952e-06,
"loss": 0.6777,
"step": 495
},
{
"epoch": 0.2071251035625518,
"grad_norm": 2.938914484551871,
"learning_rate": 9.655876956853025e-06,
"loss": 0.6922,
"step": 500
},
{
"epoch": 0.2091963545981773,
"grad_norm": 2.933588065770229,
"learning_rate": 9.64257236018169e-06,
"loss": 0.6983,
"step": 505
},
{
"epoch": 0.2112676056338028,
"grad_norm": 2.6620735297970284,
"learning_rate": 9.629024947733836e-06,
"loss": 0.6801,
"step": 510
},
{
"epoch": 0.21333885666942834,
"grad_norm": 2.9246922115868244,
"learning_rate": 9.615235428066106e-06,
"loss": 0.6745,
"step": 515
},
{
"epoch": 0.21541010770505387,
"grad_norm": 2.7814126484359605,
"learning_rate": 9.601204522397826e-06,
"loss": 0.6902,
"step": 520
},
{
"epoch": 0.21748135874067936,
"grad_norm": 2.8666993290962917,
"learning_rate": 9.586932964573298e-06,
"loss": 0.6736,
"step": 525
},
{
"epoch": 0.2195526097763049,
"grad_norm": 2.8502393291510058,
"learning_rate": 9.572421501023403e-06,
"loss": 0.6726,
"step": 530
},
{
"epoch": 0.22162386081193042,
"grad_norm": 2.7285891092196493,
"learning_rate": 9.557670890726576e-06,
"loss": 0.7026,
"step": 535
},
{
"epoch": 0.22369511184755592,
"grad_norm": 2.7516541498998053,
"learning_rate": 9.5426819051691e-06,
"loss": 0.6498,
"step": 540
},
{
"epoch": 0.22576636288318144,
"grad_norm": 2.7688759610024336,
"learning_rate": 9.527455328304756e-06,
"loss": 0.6617,
"step": 545
},
{
"epoch": 0.22783761391880697,
"grad_norm": 2.7974559352399284,
"learning_rate": 9.511991956513828e-06,
"loss": 0.6577,
"step": 550
},
{
"epoch": 0.22990886495443247,
"grad_norm": 2.913490071476846,
"learning_rate": 9.496292598561445e-06,
"loss": 0.6574,
"step": 555
},
{
"epoch": 0.231980115990058,
"grad_norm": 2.800168111857422,
"learning_rate": 9.480358075555278e-06,
"loss": 0.6615,
"step": 560
},
{
"epoch": 0.23405136702568352,
"grad_norm": 2.8532660153153073,
"learning_rate": 9.464189220902603e-06,
"loss": 0.6401,
"step": 565
},
{
"epoch": 0.23612261806130902,
"grad_norm": 2.731343089226633,
"learning_rate": 9.447786880266706e-06,
"loss": 0.6765,
"step": 570
},
{
"epoch": 0.23819386909693455,
"grad_norm": 2.7368891548658745,
"learning_rate": 9.431151911522656e-06,
"loss": 0.6559,
"step": 575
},
{
"epoch": 0.24026512013256007,
"grad_norm": 2.8586019443103443,
"learning_rate": 9.414285184712432e-06,
"loss": 0.6549,
"step": 580
},
{
"epoch": 0.24233637116818557,
"grad_norm": 2.6893316866041292,
"learning_rate": 9.397187581999424e-06,
"loss": 0.6548,
"step": 585
},
{
"epoch": 0.2444076222038111,
"grad_norm": 2.960263188185504,
"learning_rate": 9.37985999762229e-06,
"loss": 0.6629,
"step": 590
},
{
"epoch": 0.24647887323943662,
"grad_norm": 2.794468346887519,
"learning_rate": 9.362303337848188e-06,
"loss": 0.6528,
"step": 595
},
{
"epoch": 0.24855012427506215,
"grad_norm": 2.7296543899015404,
"learning_rate": 9.344518520925377e-06,
"loss": 0.6539,
"step": 600
},
{
"epoch": 0.2506213753106877,
"grad_norm": 2.7814326323481793,
"learning_rate": 9.326506477035179e-06,
"loss": 0.6509,
"step": 605
},
{
"epoch": 0.2526926263463132,
"grad_norm": 2.7657465860310664,
"learning_rate": 9.308268148243355e-06,
"loss": 0.657,
"step": 610
},
{
"epoch": 0.2547638773819387,
"grad_norm": 2.8791689822910884,
"learning_rate": 9.289804488450805e-06,
"loss": 0.6363,
"step": 615
},
{
"epoch": 0.25683512841756423,
"grad_norm": 2.7171465550163045,
"learning_rate": 9.271116463343692e-06,
"loss": 0.6455,
"step": 620
},
{
"epoch": 0.25890637945318973,
"grad_norm": 2.73393442140209,
"learning_rate": 9.25220505034293e-06,
"loss": 0.6513,
"step": 625
},
{
"epoch": 0.2609776304888152,
"grad_norm": 2.632241124925423,
"learning_rate": 9.23307123855307e-06,
"loss": 0.6346,
"step": 630
},
{
"epoch": 0.2630488815244408,
"grad_norm": 2.993601522963101,
"learning_rate": 9.213716028710558e-06,
"loss": 0.6316,
"step": 635
},
{
"epoch": 0.2651201325600663,
"grad_norm": 2.76014350673819,
"learning_rate": 9.194140433131397e-06,
"loss": 0.6219,
"step": 640
},
{
"epoch": 0.2671913835956918,
"grad_norm": 2.759230217705747,
"learning_rate": 9.174345475658208e-06,
"loss": 0.6193,
"step": 645
},
{
"epoch": 0.26926263463131733,
"grad_norm": 2.740361207957431,
"learning_rate": 9.154332191606671e-06,
"loss": 0.6146,
"step": 650
},
{
"epoch": 0.27133388566694283,
"grad_norm": 2.7156945902866,
"learning_rate": 9.134101627711384e-06,
"loss": 0.6372,
"step": 655
},
{
"epoch": 0.27340513670256833,
"grad_norm": 2.580365026128783,
"learning_rate": 9.113654842071114e-06,
"loss": 0.6167,
"step": 660
},
{
"epoch": 0.2754763877381939,
"grad_norm": 2.863886132177369,
"learning_rate": 9.092992904093451e-06,
"loss": 0.6059,
"step": 665
},
{
"epoch": 0.2775476387738194,
"grad_norm": 2.6578998294664733,
"learning_rate": 9.072116894438885e-06,
"loss": 0.6168,
"step": 670
},
{
"epoch": 0.2796188898094449,
"grad_norm": 2.631660528391497,
"learning_rate": 9.051027904964279e-06,
"loss": 0.6146,
"step": 675
},
{
"epoch": 0.28169014084507044,
"grad_norm": 2.7672365380596164,
"learning_rate": 9.029727038665765e-06,
"loss": 0.6123,
"step": 680
},
{
"epoch": 0.28376139188069593,
"grad_norm": 3.0968185232658088,
"learning_rate": 9.008215409621053e-06,
"loss": 0.6142,
"step": 685
},
{
"epoch": 0.28583264291632143,
"grad_norm": 2.8569762938850514,
"learning_rate": 8.986494142931168e-06,
"loss": 0.602,
"step": 690
},
{
"epoch": 0.287903893951947,
"grad_norm": 2.7348379248853356,
"learning_rate": 8.964564374661597e-06,
"loss": 0.62,
"step": 695
},
{
"epoch": 0.2899751449875725,
"grad_norm": 2.6233125816324137,
"learning_rate": 8.94242725178288e-06,
"loss": 0.6135,
"step": 700
},
{
"epoch": 0.292046396023198,
"grad_norm": 2.929517704380803,
"learning_rate": 8.920083932110608e-06,
"loss": 0.6036,
"step": 705
},
{
"epoch": 0.29411764705882354,
"grad_norm": 2.6098554882359553,
"learning_rate": 8.89753558424488e-06,
"loss": 0.5931,
"step": 710
},
{
"epoch": 0.29618889809444904,
"grad_norm": 2.8924998708965397,
"learning_rate": 8.874783387509181e-06,
"loss": 0.5934,
"step": 715
},
{
"epoch": 0.2982601491300746,
"grad_norm": 2.596801412566951,
"learning_rate": 8.851828531888692e-06,
"loss": 0.6051,
"step": 720
},
{
"epoch": 0.3003314001657001,
"grad_norm": 2.619656208203822,
"learning_rate": 8.828672217968055e-06,
"loss": 0.5834,
"step": 725
},
{
"epoch": 0.3024026512013256,
"grad_norm": 2.791932552629316,
"learning_rate": 8.805315656868587e-06,
"loss": 0.5977,
"step": 730
},
{
"epoch": 0.30447390223695114,
"grad_norm": 2.6328719857151817,
"learning_rate": 8.781760070184933e-06,
"loss": 0.5786,
"step": 735
},
{
"epoch": 0.30654515327257664,
"grad_norm": 2.7134170911670306,
"learning_rate": 8.75800668992117e-06,
"loss": 0.6108,
"step": 740
},
{
"epoch": 0.30861640430820214,
"grad_norm": 2.60002217254436,
"learning_rate": 8.734056758426367e-06,
"loss": 0.6052,
"step": 745
},
{
"epoch": 0.3106876553438277,
"grad_norm": 2.714229727153767,
"learning_rate": 8.709911528329623e-06,
"loss": 0.5938,
"step": 750
},
{
"epoch": 0.3127589063794532,
"grad_norm": 2.7258279509398426,
"learning_rate": 8.685572262474538e-06,
"loss": 0.5964,
"step": 755
},
{
"epoch": 0.3148301574150787,
"grad_norm": 2.7585133476772636,
"learning_rate": 8.661040233853166e-06,
"loss": 0.5775,
"step": 760
},
{
"epoch": 0.31690140845070425,
"grad_norm": 2.8212453417581127,
"learning_rate": 8.636316725539445e-06,
"loss": 0.5866,
"step": 765
},
{
"epoch": 0.31897265948632975,
"grad_norm": 2.6417442757955247,
"learning_rate": 8.611403030622074e-06,
"loss": 0.58,
"step": 770
},
{
"epoch": 0.32104391052195524,
"grad_norm": 2.757161123839532,
"learning_rate": 8.586300452136895e-06,
"loss": 0.5707,
"step": 775
},
{
"epoch": 0.3231151615575808,
"grad_norm": 2.555247097752916,
"learning_rate": 8.561010302998734e-06,
"loss": 0.5925,
"step": 780
},
{
"epoch": 0.3251864125932063,
"grad_norm": 2.8366581801399655,
"learning_rate": 8.535533905932739e-06,
"loss": 0.5736,
"step": 785
},
{
"epoch": 0.3272576636288318,
"grad_norm": 3.056206842303661,
"learning_rate": 8.509872593405189e-06,
"loss": 0.5779,
"step": 790
},
{
"epoch": 0.32932891466445735,
"grad_norm": 2.53237830007663,
"learning_rate": 8.484027707553818e-06,
"loss": 0.5708,
"step": 795
},
{
"epoch": 0.33140016570008285,
"grad_norm": 2.8417713226772165,
"learning_rate": 8.458000600117604e-06,
"loss": 0.5814,
"step": 800
},
{
"epoch": 0.33347141673570835,
"grad_norm": 2.824242675911942,
"learning_rate": 8.43179263236608e-06,
"loss": 0.5631,
"step": 805
},
{
"epoch": 0.3355426677713339,
"grad_norm": 2.6595589053586792,
"learning_rate": 8.40540517502813e-06,
"loss": 0.5762,
"step": 810
},
{
"epoch": 0.3376139188069594,
"grad_norm": 2.5496305281888767,
"learning_rate": 8.378839608220304e-06,
"loss": 0.5708,
"step": 815
},
{
"epoch": 0.3396851698425849,
"grad_norm": 2.7369851726349457,
"learning_rate": 8.35209732137463e-06,
"loss": 0.5511,
"step": 820
},
{
"epoch": 0.34175642087821045,
"grad_norm": 2.7383476860613425,
"learning_rate": 8.32517971316595e-06,
"loss": 0.5746,
"step": 825
},
{
"epoch": 0.34382767191383595,
"grad_norm": 2.543223602393049,
"learning_rate": 8.298088191438753e-06,
"loss": 0.5545,
"step": 830
},
{
"epoch": 0.34589892294946145,
"grad_norm": 2.5789286430343945,
"learning_rate": 8.270824173133563e-06,
"loss": 0.5589,
"step": 835
},
{
"epoch": 0.347970173985087,
"grad_norm": 2.663964077331657,
"learning_rate": 8.243389084212808e-06,
"loss": 0.5844,
"step": 840
},
{
"epoch": 0.3500414250207125,
"grad_norm": 2.701713888365594,
"learning_rate": 8.215784359586257e-06,
"loss": 0.5335,
"step": 845
},
{
"epoch": 0.352112676056338,
"grad_norm": 2.5757700396462617,
"learning_rate": 8.188011443035962e-06,
"loss": 0.5691,
"step": 850
},
{
"epoch": 0.35418392709196356,
"grad_norm": 3.2685206527006967,
"learning_rate": 8.160071787140742e-06,
"loss": 0.5604,
"step": 855
},
{
"epoch": 0.35625517812758906,
"grad_norm": 2.6808681382303283,
"learning_rate": 8.131966853200226e-06,
"loss": 0.5589,
"step": 860
},
{
"epoch": 0.35832642916321455,
"grad_norm": 2.5865759561179695,
"learning_rate": 8.103698111158405e-06,
"loss": 0.5572,
"step": 865
},
{
"epoch": 0.3603976801988401,
"grad_norm": 2.621331974867394,
"learning_rate": 8.075267039526764e-06,
"loss": 0.5488,
"step": 870
},
{
"epoch": 0.3624689312344656,
"grad_norm": 2.728050169878493,
"learning_rate": 8.046675125306948e-06,
"loss": 0.5559,
"step": 875
},
{
"epoch": 0.36454018227009116,
"grad_norm": 2.706287204854472,
"learning_rate": 8.017923863912989e-06,
"loss": 0.5658,
"step": 880
},
{
"epoch": 0.36661143330571666,
"grad_norm": 2.6490252323634573,
"learning_rate": 7.989014759093095e-06,
"loss": 0.5414,
"step": 885
},
{
"epoch": 0.36868268434134216,
"grad_norm": 2.523456167360872,
"learning_rate": 7.959949322850994e-06,
"loss": 0.5441,
"step": 890
},
{
"epoch": 0.3707539353769677,
"grad_norm": 2.5740101976508964,
"learning_rate": 7.930729075366867e-06,
"loss": 0.5361,
"step": 895
},
{
"epoch": 0.3728251864125932,
"grad_norm": 2.7523704722735736,
"learning_rate": 7.901355544917827e-06,
"loss": 0.5459,
"step": 900
},
{
"epoch": 0.3748964374482187,
"grad_norm": 2.7136466139331716,
"learning_rate": 7.87183026779799e-06,
"loss": 0.5373,
"step": 905
},
{
"epoch": 0.37696768848384427,
"grad_norm": 2.5974139734998807,
"learning_rate": 7.842154788238124e-06,
"loss": 0.5163,
"step": 910
},
{
"epoch": 0.37903893951946976,
"grad_norm": 2.5980610883027624,
"learning_rate": 7.812330658324884e-06,
"loss": 0.5404,
"step": 915
},
{
"epoch": 0.38111019055509526,
"grad_norm": 2.9020873692936036,
"learning_rate": 7.782359437919644e-06,
"loss": 0.528,
"step": 920
},
{
"epoch": 0.3831814415907208,
"grad_norm": 2.8690861560512473,
"learning_rate": 7.75224269457689e-06,
"loss": 0.5265,
"step": 925
},
{
"epoch": 0.3852526926263463,
"grad_norm": 2.5570267026915765,
"learning_rate": 7.721982003462255e-06,
"loss": 0.5247,
"step": 930
},
{
"epoch": 0.3873239436619718,
"grad_norm": 2.6763850509883946,
"learning_rate": 7.691578947270122e-06,
"loss": 0.534,
"step": 935
},
{
"epoch": 0.38939519469759737,
"grad_norm": 2.5052520741352837,
"learning_rate": 7.661035116140856e-06,
"loss": 0.5215,
"step": 940
},
{
"epoch": 0.39146644573322287,
"grad_norm": 2.5675785241463687,
"learning_rate": 7.63035210757763e-06,
"loss": 0.5311,
"step": 945
},
{
"epoch": 0.39353769676884837,
"grad_norm": 2.564582103010566,
"learning_rate": 7.599531526362873e-06,
"loss": 0.523,
"step": 950
},
{
"epoch": 0.3956089478044739,
"grad_norm": 2.4030300519005827,
"learning_rate": 7.568574984474335e-06,
"loss": 0.5321,
"step": 955
},
{
"epoch": 0.3976801988400994,
"grad_norm": 2.55360688034303,
"learning_rate": 7.537484101000787e-06,
"loss": 0.5109,
"step": 960
},
{
"epoch": 0.3997514498757249,
"grad_norm": 2.731502989984868,
"learning_rate": 7.506260502057325e-06,
"loss": 0.5451,
"step": 965
},
{
"epoch": 0.40182270091135047,
"grad_norm": 2.6079554614607705,
"learning_rate": 7.474905820700334e-06,
"loss": 0.5368,
"step": 970
},
{
"epoch": 0.40389395194697597,
"grad_norm": 2.574046069630531,
"learning_rate": 7.443421696842066e-06,
"loss": 0.545,
"step": 975
},
{
"epoch": 0.40596520298260147,
"grad_norm": 3.0200475159212257,
"learning_rate": 7.411809777164873e-06,
"loss": 0.5291,
"step": 980
},
{
"epoch": 0.408036454018227,
"grad_norm": 2.7452681073557117,
"learning_rate": 7.380071715035089e-06,
"loss": 0.5049,
"step": 985
},
{
"epoch": 0.4101077050538525,
"grad_norm": 2.62214319189027,
"learning_rate": 7.3482091704165405e-06,
"loss": 0.5236,
"step": 990
},
{
"epoch": 0.412178956089478,
"grad_norm": 2.7249510221523847,
"learning_rate": 7.316223809783745e-06,
"loss": 0.5269,
"step": 995
},
{
"epoch": 0.4142502071251036,
"grad_norm": 2.630893735711103,
"learning_rate": 7.284117306034733e-06,
"loss": 0.5265,
"step": 1000
},
{
"epoch": 0.4163214581607291,
"grad_norm": 2.592181156178893,
"learning_rate": 7.2518913384035685e-06,
"loss": 0.5128,
"step": 1005
},
{
"epoch": 0.4183927091963546,
"grad_norm": 2.5689746215808635,
"learning_rate": 7.219547592372512e-06,
"loss": 0.5102,
"step": 1010
},
{
"epoch": 0.4204639602319801,
"grad_norm": 2.443889242394086,
"learning_rate": 7.187087759583869e-06,
"loss": 0.5106,
"step": 1015
},
{
"epoch": 0.4225352112676056,
"grad_norm": 2.4062918533464717,
"learning_rate": 7.15451353775151e-06,
"loss": 0.5161,
"step": 1020
},
{
"epoch": 0.4246064623032311,
"grad_norm": 2.488756641889923,
"learning_rate": 7.121826630572084e-06,
"loss": 0.5157,
"step": 1025
},
{
"epoch": 0.4266777133388567,
"grad_norm": 2.4678699964797435,
"learning_rate": 7.089028747635908e-06,
"loss": 0.5222,
"step": 1030
},
{
"epoch": 0.4287489643744822,
"grad_norm": 2.7985670814019645,
"learning_rate": 7.056121604337554e-06,
"loss": 0.5109,
"step": 1035
},
{
"epoch": 0.43082021541010773,
"grad_norm": 2.750263108942153,
"learning_rate": 7.023106921786118e-06,
"loss": 0.5087,
"step": 1040
},
{
"epoch": 0.43289146644573323,
"grad_norm": 2.527279053044518,
"learning_rate": 6.9899864267152275e-06,
"loss": 0.5103,
"step": 1045
},
{
"epoch": 0.43496271748135873,
"grad_norm": 2.673130489271966,
"learning_rate": 6.956761851392706e-06,
"loss": 0.5204,
"step": 1050
},
{
"epoch": 0.4370339685169843,
"grad_norm": 2.646174732179495,
"learning_rate": 6.9234349335299835e-06,
"loss": 0.4949,
"step": 1055
},
{
"epoch": 0.4391052195526098,
"grad_norm": 2.4710241904168826,
"learning_rate": 6.890007416191209e-06,
"loss": 0.4976,
"step": 1060
},
{
"epoch": 0.4411764705882353,
"grad_norm": 2.513490801734296,
"learning_rate": 6.8564810477020835e-06,
"loss": 0.4928,
"step": 1065
},
{
"epoch": 0.44324772162386084,
"grad_norm": 2.616872712286189,
"learning_rate": 6.822857581558423e-06,
"loss": 0.4851,
"step": 1070
},
{
"epoch": 0.44531897265948633,
"grad_norm": 2.5547631458988995,
"learning_rate": 6.789138776334441e-06,
"loss": 0.498,
"step": 1075
},
{
"epoch": 0.44739022369511183,
"grad_norm": 2.5426717042905294,
"learning_rate": 6.7553263955907755e-06,
"loss": 0.5021,
"step": 1080
},
{
"epoch": 0.4494614747307374,
"grad_norm": 2.762299627144321,
"learning_rate": 6.721422207782249e-06,
"loss": 0.4894,
"step": 1085
},
{
"epoch": 0.4515327257663629,
"grad_norm": 2.6282943687365314,
"learning_rate": 6.687427986165379e-06,
"loss": 0.4885,
"step": 1090
},
{
"epoch": 0.4536039768019884,
"grad_norm": 2.6227170032352505,
"learning_rate": 6.653345508705629e-06,
"loss": 0.4846,
"step": 1095
},
{
"epoch": 0.45567522783761394,
"grad_norm": 2.573980996203933,
"learning_rate": 6.6191765579844205e-06,
"loss": 0.4677,
"step": 1100
},
{
"epoch": 0.45774647887323944,
"grad_norm": 2.760609026653492,
"learning_rate": 6.584922921105894e-06,
"loss": 0.4704,
"step": 1105
},
{
"epoch": 0.45981772990886494,
"grad_norm": 2.5519488960897942,
"learning_rate": 6.550586389603451e-06,
"loss": 0.4763,
"step": 1110
},
{
"epoch": 0.4618889809444905,
"grad_norm": 2.4528803429641814,
"learning_rate": 6.5161687593460395e-06,
"loss": 0.4854,
"step": 1115
},
{
"epoch": 0.463960231980116,
"grad_norm": 2.7025880432763962,
"learning_rate": 6.481671830444243e-06,
"loss": 0.497,
"step": 1120
},
{
"epoch": 0.4660314830157415,
"grad_norm": 2.4770462760723895,
"learning_rate": 6.447097407156114e-06,
"loss": 0.4799,
"step": 1125
},
{
"epoch": 0.46810273405136704,
"grad_norm": 2.416009894405175,
"learning_rate": 6.412447297792818e-06,
"loss": 0.4772,
"step": 1130
},
{
"epoch": 0.47017398508699254,
"grad_norm": 2.422088326310343,
"learning_rate": 6.377723314624057e-06,
"loss": 0.4684,
"step": 1135
},
{
"epoch": 0.47224523612261804,
"grad_norm": 2.6486702874804497,
"learning_rate": 6.3429272737832726e-06,
"loss": 0.4737,
"step": 1140
},
{
"epoch": 0.4743164871582436,
"grad_norm": 2.474167276592783,
"learning_rate": 6.308060995172673e-06,
"loss": 0.4616,
"step": 1145
},
{
"epoch": 0.4763877381938691,
"grad_norm": 2.517931093717021,
"learning_rate": 6.273126302368037e-06,
"loss": 0.4656,
"step": 1150
},
{
"epoch": 0.4784589892294946,
"grad_norm": 2.6624977734074062,
"learning_rate": 6.238125022523343e-06,
"loss": 0.4747,
"step": 1155
},
{
"epoch": 0.48053024026512015,
"grad_norm": 2.501404850938584,
"learning_rate": 6.203058986275207e-06,
"loss": 0.4718,
"step": 1160
},
{
"epoch": 0.48260149130074564,
"grad_norm": 2.393771169197383,
"learning_rate": 6.1679300276471285e-06,
"loss": 0.4685,
"step": 1165
},
{
"epoch": 0.48467274233637114,
"grad_norm": 2.579946322003478,
"learning_rate": 6.132739983953579e-06,
"loss": 0.4616,
"step": 1170
},
{
"epoch": 0.4867439933719967,
"grad_norm": 2.5308648946074124,
"learning_rate": 6.097490695703896e-06,
"loss": 0.4624,
"step": 1175
},
{
"epoch": 0.4888152444076222,
"grad_norm": 2.4343897742629497,
"learning_rate": 6.062184006506027e-06,
"loss": 0.4576,
"step": 1180
},
{
"epoch": 0.4908864954432477,
"grad_norm": 2.5040006985626633,
"learning_rate": 6.026821762970102e-06,
"loss": 0.4503,
"step": 1185
},
{
"epoch": 0.49295774647887325,
"grad_norm": 2.637507097654403,
"learning_rate": 5.991405814611855e-06,
"loss": 0.4665,
"step": 1190
},
{
"epoch": 0.49502899751449875,
"grad_norm": 2.410738785571731,
"learning_rate": 5.955938013755888e-06,
"loss": 0.4745,
"step": 1195
},
{
"epoch": 0.4971002485501243,
"grad_norm": 2.5556725277525225,
"learning_rate": 5.920420215438794e-06,
"loss": 0.4609,
"step": 1200
},
{
"epoch": 0.4991714995857498,
"grad_norm": 2.450509349076648,
"learning_rate": 5.8848542773121285e-06,
"loss": 0.4692,
"step": 1205
},
{
"epoch": 0.5012427506213754,
"grad_norm": 2.633730260436242,
"learning_rate": 5.849242059545259e-06,
"loss": 0.4859,
"step": 1210
},
{
"epoch": 0.5033140016570008,
"grad_norm": 2.5117312210011784,
"learning_rate": 5.81358542472807e-06,
"loss": 0.4598,
"step": 1215
},
{
"epoch": 0.5053852526926264,
"grad_norm": 2.727727787839533,
"learning_rate": 5.777886237773542e-06,
"loss": 0.458,
"step": 1220
},
{
"epoch": 0.5074565037282519,
"grad_norm": 2.498652640274728,
"learning_rate": 5.742146365820223e-06,
"loss": 0.4568,
"step": 1225
},
{
"epoch": 0.5095277547638773,
"grad_norm": 2.4105214571197564,
"learning_rate": 5.706367678134562e-06,
"loss": 0.4529,
"step": 1230
},
{
"epoch": 0.5115990057995029,
"grad_norm": 2.482609496489558,
"learning_rate": 5.670552046013151e-06,
"loss": 0.4353,
"step": 1235
},
{
"epoch": 0.5136702568351285,
"grad_norm": 2.464474174865273,
"learning_rate": 5.634701342684852e-06,
"loss": 0.459,
"step": 1240
},
{
"epoch": 0.5157415078707539,
"grad_norm": 2.5816384699470394,
"learning_rate": 5.598817443212813e-06,
"loss": 0.4535,
"step": 1245
},
{
"epoch": 0.5178127589063795,
"grad_norm": 2.4360648057548056,
"learning_rate": 5.562902224396416e-06,
"loss": 0.4516,
"step": 1250
},
{
"epoch": 0.519884009942005,
"grad_norm": 2.451723183702262,
"learning_rate": 5.526957564673098e-06,
"loss": 0.4601,
"step": 1255
},
{
"epoch": 0.5219552609776305,
"grad_norm": 2.455817733692636,
"learning_rate": 5.49098534402012e-06,
"loss": 0.4622,
"step": 1260
},
{
"epoch": 0.524026512013256,
"grad_norm": 2.3800711048949017,
"learning_rate": 5.454987443856235e-06,
"loss": 0.4494,
"step": 1265
},
{
"epoch": 0.5260977630488816,
"grad_norm": 2.4719917969907708,
"learning_rate": 5.418965746943281e-06,
"loss": 0.4546,
"step": 1270
},
{
"epoch": 0.528169014084507,
"grad_norm": 2.5259312235567037,
"learning_rate": 5.3829221372877175e-06,
"loss": 0.4428,
"step": 1275
},
{
"epoch": 0.5302402651201326,
"grad_norm": 2.5192222564921867,
"learning_rate": 5.34685850004208e-06,
"loss": 0.4552,
"step": 1280
},
{
"epoch": 0.5323115161557581,
"grad_norm": 2.5214221566614383,
"learning_rate": 5.310776721406392e-06,
"loss": 0.4447,
"step": 1285
},
{
"epoch": 0.5343827671913836,
"grad_norm": 2.512402482362361,
"learning_rate": 5.2746786885295034e-06,
"loss": 0.4411,
"step": 1290
},
{
"epoch": 0.5364540182270091,
"grad_norm": 2.537503773050984,
"learning_rate": 5.238566289410396e-06,
"loss": 0.4397,
"step": 1295
},
{
"epoch": 0.5385252692626347,
"grad_norm": 2.3626937823064145,
"learning_rate": 5.2024414127994325e-06,
"loss": 0.4425,
"step": 1300
},
{
"epoch": 0.5405965202982601,
"grad_norm": 2.5522641550285416,
"learning_rate": 5.166305948099574e-06,
"loss": 0.4442,
"step": 1305
},
{
"epoch": 0.5426677713338857,
"grad_norm": 2.5568212604328453,
"learning_rate": 5.13016178526756e-06,
"loss": 0.4396,
"step": 1310
},
{
"epoch": 0.5447390223695112,
"grad_norm": 2.402388362583747,
"learning_rate": 5.094010814715062e-06,
"loss": 0.4347,
"step": 1315
},
{
"epoch": 0.5468102734051367,
"grad_norm": 2.6483619826882356,
"learning_rate": 5.057854927209804e-06,
"loss": 0.4275,
"step": 1320
},
{
"epoch": 0.5488815244407622,
"grad_norm": 2.5263379021082084,
"learning_rate": 5.0216960137766805e-06,
"loss": 0.4343,
"step": 1325
},
{
"epoch": 0.5509527754763878,
"grad_norm": 2.5196212568619796,
"learning_rate": 4.985535965598843e-06,
"loss": 0.4342,
"step": 1330
},
{
"epoch": 0.5530240265120132,
"grad_norm": 2.625850121251377,
"learning_rate": 4.949376673918802e-06,
"loss": 0.4408,
"step": 1335
},
{
"epoch": 0.5550952775476388,
"grad_norm": 2.3452199896145816,
"learning_rate": 4.913220029939491e-06,
"loss": 0.4257,
"step": 1340
},
{
"epoch": 0.5571665285832643,
"grad_norm": 2.574940433175934,
"learning_rate": 4.877067924725368e-06,
"loss": 0.4254,
"step": 1345
},
{
"epoch": 0.5592377796188898,
"grad_norm": 2.3655525367188197,
"learning_rate": 4.840922249103506e-06,
"loss": 0.42,
"step": 1350
},
{
"epoch": 0.5613090306545153,
"grad_norm": 2.5151014112590473,
"learning_rate": 4.804784893564697e-06,
"loss": 0.4325,
"step": 1355
},
{
"epoch": 0.5633802816901409,
"grad_norm": 2.4340664966958174,
"learning_rate": 4.7686577481645745e-06,
"loss": 0.4162,
"step": 1360
},
{
"epoch": 0.5654515327257663,
"grad_norm": 2.361074940083378,
"learning_rate": 4.732542702424759e-06,
"loss": 0.4273,
"step": 1365
},
{
"epoch": 0.5675227837613919,
"grad_norm": 2.4166634583502065,
"learning_rate": 4.696441645234042e-06,
"loss": 0.4247,
"step": 1370
},
{
"epoch": 0.5695940347970174,
"grad_norm": 2.416910012406247,
"learning_rate": 4.660356464749578e-06,
"loss": 0.4272,
"step": 1375
},
{
"epoch": 0.5716652858326429,
"grad_norm": 2.5029874954691294,
"learning_rate": 4.624289048298147e-06,
"loss": 0.4115,
"step": 1380
},
{
"epoch": 0.5737365368682684,
"grad_norm": 2.5064068268567343,
"learning_rate": 4.588241282277428e-06,
"loss": 0.4153,
"step": 1385
},
{
"epoch": 0.575807787903894,
"grad_norm": 2.497767560214899,
"learning_rate": 4.55221505205734e-06,
"loss": 0.4184,
"step": 1390
},
{
"epoch": 0.5778790389395194,
"grad_norm": 2.4311887305765882,
"learning_rate": 4.516212241881448e-06,
"loss": 0.4184,
"step": 1395
},
{
"epoch": 0.579950289975145,
"grad_norm": 2.463809168583045,
"learning_rate": 4.480234734768393e-06,
"loss": 0.4346,
"step": 1400
},
{
"epoch": 0.5820215410107705,
"grad_norm": 2.4645651971416744,
"learning_rate": 4.444284412413418e-06,
"loss": 0.4144,
"step": 1405
},
{
"epoch": 0.584092792046396,
"grad_norm": 2.4306509639257277,
"learning_rate": 4.408363155089952e-06,
"loss": 0.412,
"step": 1410
},
{
"epoch": 0.5861640430820215,
"grad_norm": 2.378781711346702,
"learning_rate": 4.3724728415512585e-06,
"loss": 0.4232,
"step": 1415
},
{
"epoch": 0.5882352941176471,
"grad_norm": 2.279100422257553,
"learning_rate": 4.3366153489321855e-06,
"loss": 0.425,
"step": 1420
},
{
"epoch": 0.5903065451532725,
"grad_norm": 2.3440151093069055,
"learning_rate": 4.30079255265098e-06,
"loss": 0.4109,
"step": 1425
},
{
"epoch": 0.5923777961888981,
"grad_norm": 2.461552106548784,
"learning_rate": 4.265006326311199e-06,
"loss": 0.4188,
"step": 1430
},
{
"epoch": 0.5944490472245236,
"grad_norm": 2.554483811668929,
"learning_rate": 4.229258541603723e-06,
"loss": 0.415,
"step": 1435
},
{
"epoch": 0.5965202982601492,
"grad_norm": 2.3890846762026685,
"learning_rate": 4.1935510682088545e-06,
"loss": 0.4105,
"step": 1440
},
{
"epoch": 0.5985915492957746,
"grad_norm": 2.4838854669224677,
"learning_rate": 4.157885773698535e-06,
"loss": 0.4017,
"step": 1445
},
{
"epoch": 0.6006628003314002,
"grad_norm": 2.5664440683068914,
"learning_rate": 4.122264523438668e-06,
"loss": 0.4021,
"step": 1450
},
{
"epoch": 0.6027340513670257,
"grad_norm": 2.5100602917679797,
"learning_rate": 4.086689180491554e-06,
"loss": 0.4087,
"step": 1455
},
{
"epoch": 0.6048053024026512,
"grad_norm": 2.464742428363488,
"learning_rate": 4.051161605518453e-06,
"loss": 0.4121,
"step": 1460
},
{
"epoch": 0.6068765534382767,
"grad_norm": 2.448216596422355,
"learning_rate": 4.015683656682255e-06,
"loss": 0.4171,
"step": 1465
},
{
"epoch": 0.6089478044739023,
"grad_norm": 2.463617402393949,
"learning_rate": 3.980257189550316e-06,
"loss": 0.4114,
"step": 1470
},
{
"epoch": 0.6110190555095277,
"grad_norm": 2.4607870559606693,
"learning_rate": 3.94488405699739e-06,
"loss": 0.4041,
"step": 1475
},
{
"epoch": 0.6130903065451533,
"grad_norm": 2.460386225566258,
"learning_rate": 3.909566109108727e-06,
"loss": 0.4019,
"step": 1480
},
{
"epoch": 0.6151615575807788,
"grad_norm": 2.4056278033306358,
"learning_rate": 3.874305193083313e-06,
"loss": 0.4088,
"step": 1485
},
{
"epoch": 0.6172328086164043,
"grad_norm": 2.7408409480377536,
"learning_rate": 3.839103153137247e-06,
"loss": 0.4043,
"step": 1490
},
{
"epoch": 0.6193040596520298,
"grad_norm": 2.4491154252670353,
"learning_rate": 3.803961830407297e-06,
"loss": 0.4068,
"step": 1495
},
{
"epoch": 0.6213753106876554,
"grad_norm": 2.57957185284119,
"learning_rate": 3.768883062854598e-06,
"loss": 0.3997,
"step": 1500
},
{
"epoch": 0.6234465617232808,
"grad_norm": 2.454129058379837,
"learning_rate": 3.7338686851685267e-06,
"loss": 0.4013,
"step": 1505
},
{
"epoch": 0.6255178127589064,
"grad_norm": 2.4197107098430006,
"learning_rate": 3.6989205286707398e-06,
"loss": 0.4017,
"step": 1510
},
{
"epoch": 0.6275890637945319,
"grad_norm": 2.6100868528373593,
"learning_rate": 3.664040421219393e-06,
"loss": 0.4042,
"step": 1515
},
{
"epoch": 0.6296603148301574,
"grad_norm": 2.466076344798711,
"learning_rate": 3.6292301871135425e-06,
"loss": 0.3953,
"step": 1520
},
{
"epoch": 0.6317315658657829,
"grad_norm": 2.295401269763307,
"learning_rate": 3.59449164699773e-06,
"loss": 0.4103,
"step": 1525
},
{
"epoch": 0.6338028169014085,
"grad_norm": 2.3840288567182686,
"learning_rate": 3.55982661776676e-06,
"loss": 0.3982,
"step": 1530
},
{
"epoch": 0.6358740679370339,
"grad_norm": 2.4506295124869912,
"learning_rate": 3.5252369124706697e-06,
"loss": 0.3987,
"step": 1535
},
{
"epoch": 0.6379453189726595,
"grad_norm": 2.3732585245856597,
"learning_rate": 3.4907243402199013e-06,
"loss": 0.391,
"step": 1540
},
{
"epoch": 0.640016570008285,
"grad_norm": 2.356804936776819,
"learning_rate": 3.4562907060906908e-06,
"loss": 0.4084,
"step": 1545
},
{
"epoch": 0.6420878210439105,
"grad_norm": 2.3210923465162647,
"learning_rate": 3.4219378110306523e-06,
"loss": 0.393,
"step": 1550
},
{
"epoch": 0.644159072079536,
"grad_norm": 2.3856628563586275,
"learning_rate": 3.3876674517645815e-06,
"loss": 0.3807,
"step": 1555
},
{
"epoch": 0.6462303231151616,
"grad_norm": 2.4513218289135192,
"learning_rate": 3.353481420700495e-06,
"loss": 0.392,
"step": 1560
},
{
"epoch": 0.648301574150787,
"grad_norm": 2.4846929435894363,
"learning_rate": 3.319381505835868e-06,
"loss": 0.389,
"step": 1565
},
{
"epoch": 0.6503728251864126,
"grad_norm": 2.401342535607237,
"learning_rate": 3.285369490664133e-06,
"loss": 0.3826,
"step": 1570
},
{
"epoch": 0.6524440762220381,
"grad_norm": 2.4190000807324408,
"learning_rate": 3.251447154081394e-06,
"loss": 0.395,
"step": 1575
},
{
"epoch": 0.6545153272576636,
"grad_norm": 2.3877772373335415,
"learning_rate": 3.2176162702933816e-06,
"loss": 0.4006,
"step": 1580
},
{
"epoch": 0.6565865782932891,
"grad_norm": 2.391815584391418,
"learning_rate": 3.183878608722669e-06,
"loss": 0.3836,
"step": 1585
},
{
"epoch": 0.6586578293289147,
"grad_norm": 2.5034068900739834,
"learning_rate": 3.150235933916115e-06,
"loss": 0.3808,
"step": 1590
},
{
"epoch": 0.6607290803645401,
"grad_norm": 2.774335520816645,
"learning_rate": 3.1166900054525873e-06,
"loss": 0.3912,
"step": 1595
},
{
"epoch": 0.6628003314001657,
"grad_norm": 2.4103041919857233,
"learning_rate": 3.0832425778509235e-06,
"loss": 0.3826,
"step": 1600
},
{
"epoch": 0.6648715824357913,
"grad_norm": 2.3239461500431853,
"learning_rate": 3.049895400478174e-06,
"loss": 0.3891,
"step": 1605
},
{
"epoch": 0.6669428334714167,
"grad_norm": 2.2530556952367693,
"learning_rate": 3.0166502174581012e-06,
"loss": 0.3909,
"step": 1610
},
{
"epoch": 0.6690140845070423,
"grad_norm": 2.4287687098232396,
"learning_rate": 2.983508767579956e-06,
"loss": 0.3875,
"step": 1615
},
{
"epoch": 0.6710853355426678,
"grad_norm": 2.3815469152902136,
"learning_rate": 2.950472784207544e-06,
"loss": 0.3897,
"step": 1620
},
{
"epoch": 0.6731565865782932,
"grad_norm": 2.516171148051071,
"learning_rate": 2.917543995188562e-06,
"loss": 0.3885,
"step": 1625
},
{
"epoch": 0.6752278376139188,
"grad_norm": 2.3576789549853205,
"learning_rate": 2.8847241227642255e-06,
"loss": 0.3873,
"step": 1630
},
{
"epoch": 0.6772990886495444,
"grad_norm": 2.385493122403789,
"learning_rate": 2.852014883479198e-06,
"loss": 0.3822,
"step": 1635
},
{
"epoch": 0.6793703396851698,
"grad_norm": 2.3142689230066043,
"learning_rate": 2.819417988091814e-06,
"loss": 0.3797,
"step": 1640
},
{
"epoch": 0.6814415907207954,
"grad_norm": 2.380402531887656,
"learning_rate": 2.786935141484586e-06,
"loss": 0.3794,
"step": 1645
},
{
"epoch": 0.6835128417564209,
"grad_norm": 2.382965883978087,
"learning_rate": 2.754568042575061e-06,
"loss": 0.363,
"step": 1650
},
{
"epoch": 0.6855840927920464,
"grad_norm": 2.4286550466790935,
"learning_rate": 2.7223183842269442e-06,
"loss": 0.3874,
"step": 1655
},
{
"epoch": 0.6876553438276719,
"grad_norm": 2.3689462455130808,
"learning_rate": 2.6901878531615677e-06,
"loss": 0.3801,
"step": 1660
},
{
"epoch": 0.6897265948632975,
"grad_norm": 2.435325781453067,
"learning_rate": 2.658178129869672e-06,
"loss": 0.3761,
"step": 1665
},
{
"epoch": 0.6917978458989229,
"grad_norm": 2.3644739422275833,
"learning_rate": 2.6262908885235046e-06,
"loss": 0.3731,
"step": 1670
},
{
"epoch": 0.6938690969345485,
"grad_norm": 2.4420478405497352,
"learning_rate": 2.594527796889265e-06,
"loss": 0.3721,
"step": 1675
},
{
"epoch": 0.695940347970174,
"grad_norm": 2.380326746264982,
"learning_rate": 2.5628905162398797e-06,
"loss": 0.3625,
"step": 1680
},
{
"epoch": 0.6980115990057995,
"grad_norm": 2.439689809561185,
"learning_rate": 2.531380701268108e-06,
"loss": 0.3677,
"step": 1685
},
{
"epoch": 0.700082850041425,
"grad_norm": 2.4792328667641823,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.3751,
"step": 1690
},
{
"epoch": 0.7021541010770506,
"grad_norm": 2.3577329201592527,
"learning_rate": 2.4687500537087027e-06,
"loss": 0.3639,
"step": 1695
},
{
"epoch": 0.704225352112676,
"grad_norm": 2.397062265682441,
"learning_rate": 2.4376324968286154e-06,
"loss": 0.371,
"step": 1700
},
{
"epoch": 0.7062966031483016,
"grad_norm": 2.5377567487355197,
"learning_rate": 2.40664895686991e-06,
"loss": 0.3692,
"step": 1705
},
{
"epoch": 0.7083678541839271,
"grad_norm": 2.476875205585498,
"learning_rate": 2.375801054333409e-06,
"loss": 0.3724,
"step": 1710
},
{
"epoch": 0.7104391052195526,
"grad_norm": 2.3221981126145916,
"learning_rate": 2.345090402625822e-06,
"loss": 0.3666,
"step": 1715
},
{
"epoch": 0.7125103562551781,
"grad_norm": 2.8524714154103408,
"learning_rate": 2.3145186079753685e-06,
"loss": 0.379,
"step": 1720
},
{
"epoch": 0.7145816072908037,
"grad_norm": 2.3860769129554447,
"learning_rate": 2.2840872693477694e-06,
"loss": 0.3641,
"step": 1725
},
{
"epoch": 0.7166528583264291,
"grad_norm": 2.3054527958287143,
"learning_rate": 2.253797978362617e-06,
"loss": 0.3672,
"step": 1730
},
{
"epoch": 0.7187241093620547,
"grad_norm": 2.3407707603025805,
"learning_rate": 2.2236523192101264e-06,
"loss": 0.3579,
"step": 1735
},
{
"epoch": 0.7207953603976802,
"grad_norm": 2.293792797634113,
"learning_rate": 2.193651868568285e-06,
"loss": 0.3591,
"step": 1740
},
{
"epoch": 0.7228666114333057,
"grad_norm": 2.4185753681052202,
"learning_rate": 2.16379819552038e-06,
"loss": 0.3681,
"step": 1745
},
{
"epoch": 0.7249378624689312,
"grad_norm": 2.524421200334089,
"learning_rate": 2.1340928614729445e-06,
"loss": 0.3745,
"step": 1750
},
{
"epoch": 0.7270091135045568,
"grad_norm": 2.455923080730161,
"learning_rate": 2.1045374200740863e-06,
"loss": 0.3604,
"step": 1755
},
{
"epoch": 0.7290803645401823,
"grad_norm": 2.40811466685782,
"learning_rate": 2.075133417132223e-06,
"loss": 0.3601,
"step": 1760
},
{
"epoch": 0.7311516155758078,
"grad_norm": 2.442010182183811,
"learning_rate": 2.045882390535248e-06,
"loss": 0.3644,
"step": 1765
},
{
"epoch": 0.7332228666114333,
"grad_norm": 2.498938310765272,
"learning_rate": 2.016785870170079e-06,
"loss": 0.3659,
"step": 1770
},
{
"epoch": 0.7352941176470589,
"grad_norm": 2.43600758234028,
"learning_rate": 1.987845377842656e-06,
"loss": 0.3567,
"step": 1775
},
{
"epoch": 0.7373653686826843,
"grad_norm": 2.2394000440439883,
"learning_rate": 1.9590624271983406e-06,
"loss": 0.3622,
"step": 1780
},
{
"epoch": 0.7394366197183099,
"grad_norm": 2.5234088238169705,
"learning_rate": 1.9304385236427505e-06,
"loss": 0.3531,
"step": 1785
},
{
"epoch": 0.7415078707539354,
"grad_norm": 2.4931659816433287,
"learning_rate": 1.9019751642630252e-06,
"loss": 0.3505,
"step": 1790
},
{
"epoch": 0.7435791217895609,
"grad_norm": 2.3924059427659916,
"learning_rate": 1.8736738377495196e-06,
"loss": 0.3599,
"step": 1795
},
{
"epoch": 0.7456503728251864,
"grad_norm": 2.388334766295047,
"learning_rate": 1.8455360243179537e-06,
"loss": 0.3603,
"step": 1800
},
{
"epoch": 0.747721623860812,
"grad_norm": 2.4055399882312947,
"learning_rate": 1.8175631956319823e-06,
"loss": 0.3499,
"step": 1805
},
{
"epoch": 0.7497928748964374,
"grad_norm": 2.3681998025448743,
"learning_rate": 1.7897568147262323e-06,
"loss": 0.3451,
"step": 1810
},
{
"epoch": 0.751864125932063,
"grad_norm": 2.3399190104028618,
"learning_rate": 1.7621183359297817e-06,
"loss": 0.3436,
"step": 1815
},
{
"epoch": 0.7539353769676885,
"grad_norm": 2.299763408013497,
"learning_rate": 1.7346492047900897e-06,
"loss": 0.3552,
"step": 1820
},
{
"epoch": 0.756006628003314,
"grad_norm": 2.304240086627427,
"learning_rate": 1.7073508579973996e-06,
"loss": 0.3501,
"step": 1825
},
{
"epoch": 0.7580778790389395,
"grad_norm": 2.471808773880512,
"learning_rate": 1.6802247233095914e-06,
"loss": 0.3575,
"step": 1830
},
{
"epoch": 0.7601491300745651,
"grad_norm": 2.1787659024138453,
"learning_rate": 1.6532722194775108e-06,
"loss": 0.3416,
"step": 1835
},
{
"epoch": 0.7622203811101905,
"grad_norm": 2.410186390448844,
"learning_rate": 1.626494756170765e-06,
"loss": 0.3573,
"step": 1840
},
{
"epoch": 0.7642916321458161,
"grad_norm": 2.453866507102359,
"learning_rate": 1.5998937339039889e-06,
"loss": 0.3512,
"step": 1845
},
{
"epoch": 0.7663628831814416,
"grad_norm": 2.373346324937054,
"learning_rate": 1.5734705439636017e-06,
"loss": 0.351,
"step": 1850
},
{
"epoch": 0.7684341342170671,
"grad_norm": 2.293953470114846,
"learning_rate": 1.5472265683350397e-06,
"loss": 0.3566,
"step": 1855
},
{
"epoch": 0.7705053852526926,
"grad_norm": 2.334589469178464,
"learning_rate": 1.5211631796304721e-06,
"loss": 0.344,
"step": 1860
},
{
"epoch": 0.7725766362883182,
"grad_norm": 2.1977473111231274,
"learning_rate": 1.495281741017016e-06,
"loss": 0.3466,
"step": 1865
},
{
"epoch": 0.7746478873239436,
"grad_norm": 2.447667991652811,
"learning_rate": 1.46958360614543e-06,
"loss": 0.3458,
"step": 1870
},
{
"epoch": 0.7767191383595692,
"grad_norm": 2.4293930001790343,
"learning_rate": 1.4440701190793278e-06,
"loss": 0.3454,
"step": 1875
},
{
"epoch": 0.7787903893951947,
"grad_norm": 2.4169714631028443,
"learning_rate": 1.4187426142248723e-06,
"loss": 0.3444,
"step": 1880
},
{
"epoch": 0.7808616404308202,
"grad_norm": 2.347141749048018,
"learning_rate": 1.3936024162609897e-06,
"loss": 0.3433,
"step": 1885
},
{
"epoch": 0.7829328914664457,
"grad_norm": 2.3150017921664934,
"learning_rate": 1.3686508400700787e-06,
"loss": 0.3424,
"step": 1890
},
{
"epoch": 0.7850041425020713,
"grad_norm": 2.433953451204477,
"learning_rate": 1.3438891906692447e-06,
"loss": 0.3402,
"step": 1895
},
{
"epoch": 0.7870753935376967,
"grad_norm": 2.4345942045059923,
"learning_rate": 1.3193187631420462e-06,
"loss": 0.3559,
"step": 1900
},
{
"epoch": 0.7891466445733223,
"grad_norm": 2.385776976242536,
"learning_rate": 1.2949408425707566e-06,
"loss": 0.3409,
"step": 1905
},
{
"epoch": 0.7912178956089478,
"grad_norm": 2.284820681480646,
"learning_rate": 1.2707567039691505e-06,
"loss": 0.3323,
"step": 1910
},
{
"epoch": 0.7932891466445733,
"grad_norm": 2.373599316713983,
"learning_rate": 1.2467676122158224e-06,
"loss": 0.3431,
"step": 1915
},
{
"epoch": 0.7953603976801988,
"grad_norm": 2.277308417116025,
"learning_rate": 1.222974821988024e-06,
"loss": 0.334,
"step": 1920
},
{
"epoch": 0.7974316487158244,
"grad_norm": 2.3778447080117577,
"learning_rate": 1.1993795776960498e-06,
"loss": 0.3326,
"step": 1925
},
{
"epoch": 0.7995028997514498,
"grad_norm": 2.452346766698329,
"learning_rate": 1.1759831134181504e-06,
"loss": 0.3317,
"step": 1930
},
{
"epoch": 0.8015741507870754,
"grad_norm": 2.4247770647591085,
"learning_rate": 1.1527866528359805e-06,
"loss": 0.3474,
"step": 1935
},
{
"epoch": 0.8036454018227009,
"grad_norm": 2.3076639437103292,
"learning_rate": 1.1297914091706086e-06,
"loss": 0.3423,
"step": 1940
},
{
"epoch": 0.8057166528583264,
"grad_norm": 2.346460611096731,
"learning_rate": 1.1069985851190524e-06,
"loss": 0.3373,
"step": 1945
},
{
"epoch": 0.8077879038939519,
"grad_norm": 2.297664318905517,
"learning_rate": 1.0844093727913868e-06,
"loss": 0.3452,
"step": 1950
},
{
"epoch": 0.8098591549295775,
"grad_norm": 2.3994880509321823,
"learning_rate": 1.062024953648384e-06,
"loss": 0.3303,
"step": 1955
},
{
"epoch": 0.8119304059652029,
"grad_norm": 2.3943126094691736,
"learning_rate": 1.039846498439727e-06,
"loss": 0.337,
"step": 1960
},
{
"epoch": 0.8140016570008285,
"grad_norm": 2.394969585780868,
"learning_rate": 1.0178751671427755e-06,
"loss": 0.333,
"step": 1965
},
{
"epoch": 0.816072908036454,
"grad_norm": 2.2584619131198327,
"learning_rate": 9.961121089018933e-07,
"loss": 0.343,
"step": 1970
},
{
"epoch": 0.8181441590720795,
"grad_norm": 2.5188681037163305,
"learning_rate": 9.745584619683524e-07,
"loss": 0.3386,
"step": 1975
},
{
"epoch": 0.820215410107705,
"grad_norm": 2.3562413171777856,
"learning_rate": 9.532153536407923e-07,
"loss": 0.3415,
"step": 1980
},
{
"epoch": 0.8222866611433306,
"grad_norm": 2.4322075663751153,
"learning_rate": 9.320839002062682e-07,
"loss": 0.3301,
"step": 1985
},
{
"epoch": 0.824357912178956,
"grad_norm": 2.4874846740356804,
"learning_rate": 9.111652068818621e-07,
"loss": 0.3433,
"step": 1990
},
{
"epoch": 0.8264291632145816,
"grad_norm": 2.312127292110117,
"learning_rate": 8.904603677568785e-07,
"loss": 0.3331,
"step": 1995
},
{
"epoch": 0.8285004142502072,
"grad_norm": 2.607402264179206,
"learning_rate": 8.699704657356195e-07,
"loss": 0.3333,
"step": 2000
},
{
"epoch": 0.8305716652858326,
"grad_norm": 2.493751130019948,
"learning_rate": 8.496965724807516e-07,
"loss": 0.3362,
"step": 2005
},
{
"epoch": 0.8326429163214581,
"grad_norm": 2.351799520241766,
"learning_rate": 8.296397483572515e-07,
"loss": 0.3361,
"step": 2010
},
{
"epoch": 0.8347141673570837,
"grad_norm": 2.3347796942844457,
"learning_rate": 8.098010423769503e-07,
"loss": 0.3427,
"step": 2015
},
{
"epoch": 0.8367854183927091,
"grad_norm": 2.416514153196825,
"learning_rate": 7.901814921436624e-07,
"loss": 0.3373,
"step": 2020
},
{
"epoch": 0.8388566694283347,
"grad_norm": 2.27725091962227,
"learning_rate": 7.70782123798921e-07,
"loss": 0.3375,
"step": 2025
},
{
"epoch": 0.8409279204639603,
"grad_norm": 2.24276056194008,
"learning_rate": 7.516039519683105e-07,
"loss": 0.3382,
"step": 2030
},
{
"epoch": 0.8429991714995857,
"grad_norm": 2.4106229979629545,
"learning_rate": 7.326479797083963e-07,
"loss": 0.3288,
"step": 2035
},
{
"epoch": 0.8450704225352113,
"grad_norm": 2.281186671589003,
"learning_rate": 7.139151984542636e-07,
"loss": 0.3269,
"step": 2040
},
{
"epoch": 0.8471416735708368,
"grad_norm": 2.441879212516951,
"learning_rate": 6.954065879676653e-07,
"loss": 0.3327,
"step": 2045
},
{
"epoch": 0.8492129246064622,
"grad_norm": 2.460206403334868,
"learning_rate": 6.771231162857722e-07,
"loss": 0.3375,
"step": 2050
},
{
"epoch": 0.8512841756420878,
"grad_norm": 2.288538904206948,
"learning_rate": 6.590657396705525e-07,
"loss": 0.3163,
"step": 2055
},
{
"epoch": 0.8533554266777134,
"grad_norm": 2.3886559210526426,
"learning_rate": 6.412354025587509e-07,
"loss": 0.3356,
"step": 2060
},
{
"epoch": 0.8554266777133389,
"grad_norm": 2.2517960802152337,
"learning_rate": 6.236330375124921e-07,
"loss": 0.341,
"step": 2065
},
{
"epoch": 0.8574979287489644,
"grad_norm": 2.33394312378329,
"learning_rate": 6.062595651705111e-07,
"loss": 0.3262,
"step": 2070
},
{
"epoch": 0.8595691797845899,
"grad_norm": 2.4559361971290077,
"learning_rate": 5.891158941999959e-07,
"loss": 0.3198,
"step": 2075
},
{
"epoch": 0.8616404308202155,
"grad_norm": 2.367468825376486,
"learning_rate": 5.722029212490666e-07,
"loss": 0.3264,
"step": 2080
},
{
"epoch": 0.8637116818558409,
"grad_norm": 2.4147043980300955,
"learning_rate": 5.555215308998779e-07,
"loss": 0.3352,
"step": 2085
},
{
"epoch": 0.8657829328914665,
"grad_norm": 2.4293612156766815,
"learning_rate": 5.390725956223531e-07,
"loss": 0.3209,
"step": 2090
},
{
"epoch": 0.867854183927092,
"grad_norm": 2.409157285292527,
"learning_rate": 5.22856975728554e-07,
"loss": 0.3249,
"step": 2095
},
{
"epoch": 0.8699254349627175,
"grad_norm": 2.233849266540862,
"learning_rate": 5.068755193276798e-07,
"loss": 0.3216,
"step": 2100
},
{
"epoch": 0.871996685998343,
"grad_norm": 2.3098128897942076,
"learning_rate": 4.911290622817161e-07,
"loss": 0.3133,
"step": 2105
},
{
"epoch": 0.8740679370339686,
"grad_norm": 2.382543737963039,
"learning_rate": 4.756184281617121e-07,
"loss": 0.3212,
"step": 2110
},
{
"epoch": 0.876139188069594,
"grad_norm": 2.67090524996084,
"learning_rate": 4.6034442820471037e-07,
"loss": 0.3304,
"step": 2115
},
{
"epoch": 0.8782104391052196,
"grad_norm": 2.342709610038623,
"learning_rate": 4.4530786127131575e-07,
"loss": 0.324,
"step": 2120
},
{
"epoch": 0.8802816901408451,
"grad_norm": 2.4993138899361953,
"learning_rate": 4.305095138039106e-07,
"loss": 0.328,
"step": 2125
},
{
"epoch": 0.8823529411764706,
"grad_norm": 2.3696872859071183,
"learning_rate": 4.159501597855287e-07,
"loss": 0.3258,
"step": 2130
},
{
"epoch": 0.8844241922120961,
"grad_norm": 2.4387312234421707,
"learning_rate": 4.0163056069936757e-07,
"loss": 0.3241,
"step": 2135
},
{
"epoch": 0.8864954432477217,
"grad_norm": 2.3989145869426536,
"learning_rate": 3.8755146548896784e-07,
"loss": 0.3275,
"step": 2140
},
{
"epoch": 0.8885666942833471,
"grad_norm": 2.4928751733374326,
"learning_rate": 3.737136105190337e-07,
"loss": 0.3293,
"step": 2145
},
{
"epoch": 0.8906379453189727,
"grad_norm": 2.2972269705841737,
"learning_rate": 3.6011771953693044e-07,
"loss": 0.3283,
"step": 2150
},
{
"epoch": 0.8927091963545982,
"grad_norm": 2.297085637993506,
"learning_rate": 3.4676450363481937e-07,
"loss": 0.3243,
"step": 2155
},
{
"epoch": 0.8947804473902237,
"grad_norm": 2.4352041283350756,
"learning_rate": 3.336546612124758e-07,
"loss": 0.3229,
"step": 2160
},
{
"epoch": 0.8968516984258492,
"grad_norm": 2.4453453128997693,
"learning_rate": 3.20788877940757e-07,
"loss": 0.3267,
"step": 2165
},
{
"epoch": 0.8989229494614748,
"grad_norm": 2.497196691348891,
"learning_rate": 3.081678267257404e-07,
"loss": 0.3245,
"step": 2170
},
{
"epoch": 0.9009942004971002,
"grad_norm": 2.294241552326414,
"learning_rate": 2.9579216767352815e-07,
"loss": 0.3174,
"step": 2175
},
{
"epoch": 0.9030654515327258,
"grad_norm": 2.5123348209381704,
"learning_rate": 2.836625480557265e-07,
"loss": 0.3185,
"step": 2180
},
{
"epoch": 0.9051367025683513,
"grad_norm": 2.253684937067122,
"learning_rate": 2.7177960227558863e-07,
"loss": 0.3144,
"step": 2185
},
{
"epoch": 0.9072079536039768,
"grad_norm": 2.1977246969912723,
"learning_rate": 2.601439518348331e-07,
"loss": 0.3265,
"step": 2190
},
{
"epoch": 0.9092792046396023,
"grad_norm": 2.3594137667020303,
"learning_rate": 2.487562053011422e-07,
"loss": 0.3182,
"step": 2195
},
{
"epoch": 0.9113504556752279,
"grad_norm": 2.4721070329083954,
"learning_rate": 2.376169582763288e-07,
"loss": 0.3204,
"step": 2200
},
{
"epoch": 0.9134217067108533,
"grad_norm": 2.5655843945707306,
"learning_rate": 2.2672679336518789e-07,
"loss": 0.3286,
"step": 2205
},
{
"epoch": 0.9154929577464789,
"grad_norm": 2.2713717970234155,
"learning_rate": 2.1608628014502364e-07,
"loss": 0.3186,
"step": 2210
},
{
"epoch": 0.9175642087821044,
"grad_norm": 2.319604155676886,
"learning_rate": 2.0569597513586004e-07,
"loss": 0.3189,
"step": 2215
},
{
"epoch": 0.9196354598177299,
"grad_norm": 2.582435571084389,
"learning_rate": 1.955564217713335e-07,
"loss": 0.3248,
"step": 2220
},
{
"epoch": 0.9217067108533554,
"grad_norm": 2.4094176417675532,
"learning_rate": 1.8566815037026897e-07,
"loss": 0.3371,
"step": 2225
},
{
"epoch": 0.923777961888981,
"grad_norm": 2.4509597114145913,
"learning_rate": 1.7603167810894662e-07,
"loss": 0.3203,
"step": 2230
},
{
"epoch": 0.9258492129246064,
"grad_norm": 2.363051351708879,
"learning_rate": 1.6664750899404892e-07,
"loss": 0.3101,
"step": 2235
},
{
"epoch": 0.927920463960232,
"grad_norm": 2.420545920556074,
"learning_rate": 1.5751613383630128e-07,
"loss": 0.3218,
"step": 2240
},
{
"epoch": 0.9299917149958575,
"grad_norm": 2.3859750154616397,
"learning_rate": 1.4863803022480362e-07,
"loss": 0.3118,
"step": 2245
},
{
"epoch": 0.932062966031483,
"grad_norm": 2.1948107719137955,
"learning_rate": 1.4001366250204762e-07,
"loss": 0.3109,
"step": 2250
},
{
"epoch": 0.9341342170671085,
"grad_norm": 2.27270194473527,
"learning_rate": 1.3164348173963392e-07,
"loss": 0.3193,
"step": 2255
},
{
"epoch": 0.9362054681027341,
"grad_norm": 2.3657310934258833,
"learning_rate": 1.235279257146804e-07,
"loss": 0.3203,
"step": 2260
},
{
"epoch": 0.9382767191383595,
"grad_norm": 2.338711000044841,
"learning_rate": 1.1566741888692168e-07,
"loss": 0.3177,
"step": 2265
},
{
"epoch": 0.9403479701739851,
"grad_norm": 2.5901775745271376,
"learning_rate": 1.080623723765134e-07,
"loss": 0.3178,
"step": 2270
},
{
"epoch": 0.9424192212096106,
"grad_norm": 2.540579688066845,
"learning_rate": 1.0071318394252849e-07,
"loss": 0.3199,
"step": 2275
},
{
"epoch": 0.9444904722452361,
"grad_norm": 2.46660297908585,
"learning_rate": 9.362023796215036e-08,
"loss": 0.3183,
"step": 2280
},
{
"epoch": 0.9465617232808616,
"grad_norm": 2.56480438463466,
"learning_rate": 8.678390541057512e-08,
"loss": 0.3127,
"step": 2285
},
{
"epoch": 0.9486329743164872,
"grad_norm": 2.4346321810323923,
"learning_rate": 8.020454384160437e-08,
"loss": 0.3217,
"step": 2290
},
{
"epoch": 0.9507042253521126,
"grad_norm": 2.4276401722046734,
"learning_rate": 7.388249736894615e-08,
"loss": 0.3095,
"step": 2295
},
{
"epoch": 0.9527754763877382,
"grad_norm": 2.329219841891159,
"learning_rate": 6.78180966482156e-08,
"loss": 0.319,
"step": 2300
},
{
"epoch": 0.9548467274233637,
"grad_norm": 2.6042724587430257,
"learning_rate": 6.201165885964311e-08,
"loss": 0.317,
"step": 2305
},
{
"epoch": 0.9569179784589892,
"grad_norm": 2.2303445759455967,
"learning_rate": 5.646348769148491e-08,
"loss": 0.3217,
"step": 2310
},
{
"epoch": 0.9589892294946147,
"grad_norm": 2.4607182859035728,
"learning_rate": 5.117387332413737e-08,
"loss": 0.3379,
"step": 2315
},
{
"epoch": 0.9610604805302403,
"grad_norm": 2.3872878040403775,
"learning_rate": 4.6143092414961396e-08,
"loss": 0.3232,
"step": 2320
},
{
"epoch": 0.9631317315658657,
"grad_norm": 2.369941738658177,
"learning_rate": 4.1371408083815126e-08,
"loss": 0.313,
"step": 2325
},
{
"epoch": 0.9652029826014913,
"grad_norm": 2.413042323401412,
"learning_rate": 3.685906989928656e-08,
"loss": 0.3324,
"step": 2330
},
{
"epoch": 0.9672742336371168,
"grad_norm": 2.3918886450019747,
"learning_rate": 3.2606313865646276e-08,
"loss": 0.3213,
"step": 2335
},
{
"epoch": 0.9693454846727423,
"grad_norm": 2.212556778563815,
"learning_rate": 2.861336241050061e-08,
"loss": 0.3136,
"step": 2340
},
{
"epoch": 0.9714167357083678,
"grad_norm": 2.3753322114534514,
"learning_rate": 2.488042437315985e-08,
"loss": 0.3196,
"step": 2345
},
{
"epoch": 0.9734879867439934,
"grad_norm": 2.635822549535026,
"learning_rate": 2.1407694993714755e-08,
"loss": 0.3201,
"step": 2350
},
{
"epoch": 0.9755592377796188,
"grad_norm": 2.4223063593868774,
"learning_rate": 1.8195355902824174e-08,
"loss": 0.3177,
"step": 2355
},
{
"epoch": 0.9776304888152444,
"grad_norm": 2.5265938379888864,
"learning_rate": 1.5243575112218744e-08,
"loss": 0.3208,
"step": 2360
},
{
"epoch": 0.97970173985087,
"grad_norm": 2.344598334063139,
"learning_rate": 1.2552507005909042e-08,
"loss": 0.3103,
"step": 2365
},
{
"epoch": 0.9817729908864954,
"grad_norm": 2.3574452536812145,
"learning_rate": 1.0122292332114814e-08,
"loss": 0.3191,
"step": 2370
},
{
"epoch": 0.9838442419221209,
"grad_norm": 2.3406496363432128,
"learning_rate": 7.953058195900864e-09,
"loss": 0.317,
"step": 2375
},
{
"epoch": 0.9859154929577465,
"grad_norm": 2.2200414240286155,
"learning_rate": 6.044918052531268e-09,
"loss": 0.3079,
"step": 2380
},
{
"epoch": 0.987986743993372,
"grad_norm": 2.345671675412989,
"learning_rate": 4.397971701533554e-09,
"loss": 0.3113,
"step": 2385
},
{
"epoch": 0.9900579950289975,
"grad_norm": 2.4132441952700687,
"learning_rate": 3.0123052814812203e-09,
"loss": 0.3207,
"step": 2390
},
{
"epoch": 0.992129246064623,
"grad_norm": 2.599715525455967,
"learning_rate": 1.887991265486222e-09,
"loss": 0.318,
"step": 2395
},
{
"epoch": 0.9942004971002486,
"grad_norm": 2.371853977706459,
"learning_rate": 1.025088457409229e-09,
"loss": 0.3283,
"step": 2400
},
{
"epoch": 0.996271748135874,
"grad_norm": 2.4606089652599876,
"learning_rate": 4.2364198878597216e-10,
"loss": 0.3211,
"step": 2405
},
{
"epoch": 0.9983429991714996,
"grad_norm": 2.471628534592773,
"learning_rate": 8.368331646302353e-11,
"loss": 0.3281,
"step": 2410
},
{
"epoch": 1.0,
"eval_runtime": 2.6867,
"eval_samples_per_second": 3.722,
"eval_steps_per_second": 1.117,
"step": 2414
},
{
"epoch": 1.0,
"step": 2414,
"total_flos": 252721244405760.0,
"train_loss": 0.0,
"train_runtime": 0.0095,
"train_samples_per_second": 4074976.188,
"train_steps_per_second": 254692.606
}
],
"logging_steps": 5,
"max_steps": 2414,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 252721244405760.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}