terry69's picture
Model save
30b6464 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": 25.254537940927257,
"learning_rate": 4.132231404958678e-08,
"loss": 1.412,
"step": 1
},
{
"epoch": 0.002071251035625518,
"grad_norm": 24.193702154969035,
"learning_rate": 2.066115702479339e-07,
"loss": 1.4065,
"step": 5
},
{
"epoch": 0.004142502071251036,
"grad_norm": 16.11013133526923,
"learning_rate": 4.132231404958678e-07,
"loss": 1.3868,
"step": 10
},
{
"epoch": 0.006213753106876553,
"grad_norm": 8.07327351576777,
"learning_rate": 6.198347107438018e-07,
"loss": 1.2625,
"step": 15
},
{
"epoch": 0.008285004142502071,
"grad_norm": 10.509388738181952,
"learning_rate": 8.264462809917356e-07,
"loss": 1.1356,
"step": 20
},
{
"epoch": 0.010356255178127589,
"grad_norm": 4.703358240756608,
"learning_rate": 1.0330578512396695e-06,
"loss": 1.0286,
"step": 25
},
{
"epoch": 0.012427506213753107,
"grad_norm": 3.2512662119009605,
"learning_rate": 1.2396694214876035e-06,
"loss": 0.9808,
"step": 30
},
{
"epoch": 0.014498757249378625,
"grad_norm": 3.337004320452114,
"learning_rate": 1.4462809917355372e-06,
"loss": 0.957,
"step": 35
},
{
"epoch": 0.016570008285004142,
"grad_norm": 3.0169658450065566,
"learning_rate": 1.6528925619834712e-06,
"loss": 0.9333,
"step": 40
},
{
"epoch": 0.018641259320629662,
"grad_norm": 3.2086765864897906,
"learning_rate": 1.859504132231405e-06,
"loss": 0.9112,
"step": 45
},
{
"epoch": 0.020712510356255178,
"grad_norm": 3.088796251871216,
"learning_rate": 2.066115702479339e-06,
"loss": 0.8961,
"step": 50
},
{
"epoch": 0.022783761391880698,
"grad_norm": 3.026962043558522,
"learning_rate": 2.2727272727272728e-06,
"loss": 0.9017,
"step": 55
},
{
"epoch": 0.024855012427506214,
"grad_norm": 3.07234506723115,
"learning_rate": 2.479338842975207e-06,
"loss": 0.8766,
"step": 60
},
{
"epoch": 0.026926263463131733,
"grad_norm": 3.0659611385007164,
"learning_rate": 2.6859504132231405e-06,
"loss": 0.87,
"step": 65
},
{
"epoch": 0.02899751449875725,
"grad_norm": 3.1227749224040635,
"learning_rate": 2.8925619834710743e-06,
"loss": 0.8845,
"step": 70
},
{
"epoch": 0.03106876553438277,
"grad_norm": 3.0880074832241178,
"learning_rate": 3.0991735537190086e-06,
"loss": 0.8678,
"step": 75
},
{
"epoch": 0.033140016570008285,
"grad_norm": 3.120775124837908,
"learning_rate": 3.3057851239669424e-06,
"loss": 0.8682,
"step": 80
},
{
"epoch": 0.035211267605633804,
"grad_norm": 3.095558490573068,
"learning_rate": 3.5123966942148763e-06,
"loss": 0.8607,
"step": 85
},
{
"epoch": 0.037282518641259324,
"grad_norm": 3.1250007713844927,
"learning_rate": 3.71900826446281e-06,
"loss": 0.8606,
"step": 90
},
{
"epoch": 0.03935376967688484,
"grad_norm": 3.097218644088177,
"learning_rate": 3.925619834710744e-06,
"loss": 0.8574,
"step": 95
},
{
"epoch": 0.041425020712510356,
"grad_norm": 3.0441673467000037,
"learning_rate": 4.132231404958678e-06,
"loss": 0.8515,
"step": 100
},
{
"epoch": 0.043496271748135876,
"grad_norm": 3.161190915018127,
"learning_rate": 4.338842975206612e-06,
"loss": 0.8519,
"step": 105
},
{
"epoch": 0.045567522783761395,
"grad_norm": 3.227268013960282,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.8621,
"step": 110
},
{
"epoch": 0.04763877381938691,
"grad_norm": 3.0897116118246157,
"learning_rate": 4.75206611570248e-06,
"loss": 0.8427,
"step": 115
},
{
"epoch": 0.04971002485501243,
"grad_norm": 3.2844976068028258,
"learning_rate": 4.958677685950414e-06,
"loss": 0.8353,
"step": 120
},
{
"epoch": 0.05178127589063795,
"grad_norm": 3.1192188656455224,
"learning_rate": 5.165289256198347e-06,
"loss": 0.8437,
"step": 125
},
{
"epoch": 0.053852526926263466,
"grad_norm": 3.2679637301243853,
"learning_rate": 5.371900826446281e-06,
"loss": 0.8358,
"step": 130
},
{
"epoch": 0.05592377796188898,
"grad_norm": 3.0780914240453874,
"learning_rate": 5.578512396694216e-06,
"loss": 0.825,
"step": 135
},
{
"epoch": 0.0579950289975145,
"grad_norm": 3.25382551367048,
"learning_rate": 5.785123966942149e-06,
"loss": 0.8216,
"step": 140
},
{
"epoch": 0.06006628003314002,
"grad_norm": 3.0253311819425894,
"learning_rate": 5.991735537190083e-06,
"loss": 0.8269,
"step": 145
},
{
"epoch": 0.06213753106876554,
"grad_norm": 3.1474397410989443,
"learning_rate": 6.198347107438017e-06,
"loss": 0.8309,
"step": 150
},
{
"epoch": 0.06420878210439106,
"grad_norm": 3.1460550164248082,
"learning_rate": 6.404958677685951e-06,
"loss": 0.8261,
"step": 155
},
{
"epoch": 0.06628003314001657,
"grad_norm": 3.205531279756332,
"learning_rate": 6.611570247933885e-06,
"loss": 0.8193,
"step": 160
},
{
"epoch": 0.06835128417564208,
"grad_norm": 3.0584320167014716,
"learning_rate": 6.818181818181818e-06,
"loss": 0.8231,
"step": 165
},
{
"epoch": 0.07042253521126761,
"grad_norm": 3.364436381312218,
"learning_rate": 7.0247933884297525e-06,
"loss": 0.8381,
"step": 170
},
{
"epoch": 0.07249378624689312,
"grad_norm": 3.0607684762570657,
"learning_rate": 7.231404958677687e-06,
"loss": 0.8125,
"step": 175
},
{
"epoch": 0.07456503728251865,
"grad_norm": 3.1769839559478217,
"learning_rate": 7.43801652892562e-06,
"loss": 0.8181,
"step": 180
},
{
"epoch": 0.07663628831814416,
"grad_norm": 3.2485320001525775,
"learning_rate": 7.644628099173555e-06,
"loss": 0.8175,
"step": 185
},
{
"epoch": 0.07870753935376967,
"grad_norm": 3.0285674774818703,
"learning_rate": 7.851239669421489e-06,
"loss": 0.8369,
"step": 190
},
{
"epoch": 0.0807787903893952,
"grad_norm": 3.0776318925046935,
"learning_rate": 8.057851239669421e-06,
"loss": 0.8132,
"step": 195
},
{
"epoch": 0.08285004142502071,
"grad_norm": 2.8691855717883925,
"learning_rate": 8.264462809917356e-06,
"loss": 0.816,
"step": 200
},
{
"epoch": 0.08492129246064622,
"grad_norm": 2.9476869958534966,
"learning_rate": 8.47107438016529e-06,
"loss": 0.8012,
"step": 205
},
{
"epoch": 0.08699254349627175,
"grad_norm": 3.3571704988688427,
"learning_rate": 8.677685950413224e-06,
"loss": 0.8247,
"step": 210
},
{
"epoch": 0.08906379453189726,
"grad_norm": 3.4831864378671065,
"learning_rate": 8.884297520661158e-06,
"loss": 0.8058,
"step": 215
},
{
"epoch": 0.09113504556752279,
"grad_norm": 3.1293898894830185,
"learning_rate": 9.090909090909091e-06,
"loss": 0.8192,
"step": 220
},
{
"epoch": 0.0932062966031483,
"grad_norm": 2.922367620897215,
"learning_rate": 9.297520661157025e-06,
"loss": 0.8026,
"step": 225
},
{
"epoch": 0.09527754763877382,
"grad_norm": 3.0278336776621226,
"learning_rate": 9.50413223140496e-06,
"loss": 0.8008,
"step": 230
},
{
"epoch": 0.09734879867439934,
"grad_norm": 2.8373067338877553,
"learning_rate": 9.710743801652894e-06,
"loss": 0.8112,
"step": 235
},
{
"epoch": 0.09942004971002485,
"grad_norm": 2.918119441347903,
"learning_rate": 9.917355371900828e-06,
"loss": 0.7919,
"step": 240
},
{
"epoch": 0.10149130074565037,
"grad_norm": 2.992444281817911,
"learning_rate": 9.999952928077044e-06,
"loss": 0.789,
"step": 245
},
{
"epoch": 0.1035625517812759,
"grad_norm": 3.0703947004506893,
"learning_rate": 9.999665269535307e-06,
"loss": 0.8105,
"step": 250
},
{
"epoch": 0.1056338028169014,
"grad_norm": 2.996368047379752,
"learning_rate": 9.99911611854702e-06,
"loss": 0.8113,
"step": 255
},
{
"epoch": 0.10770505385252693,
"grad_norm": 2.9207997990477814,
"learning_rate": 9.998305503833872e-06,
"loss": 0.7946,
"step": 260
},
{
"epoch": 0.10977630488815245,
"grad_norm": 2.985671898690861,
"learning_rate": 9.997233467792626e-06,
"loss": 0.7814,
"step": 265
},
{
"epoch": 0.11184755592377796,
"grad_norm": 2.8011598076545816,
"learning_rate": 9.995900066492902e-06,
"loss": 0.7753,
"step": 270
},
{
"epoch": 0.11391880695940348,
"grad_norm": 2.885294953704069,
"learning_rate": 9.994305369674242e-06,
"loss": 0.781,
"step": 275
},
{
"epoch": 0.115990057995029,
"grad_norm": 3.143622544253548,
"learning_rate": 9.992449460742464e-06,
"loss": 0.7823,
"step": 280
},
{
"epoch": 0.11806130903065451,
"grad_norm": 3.146418619593177,
"learning_rate": 9.9903324367653e-06,
"loss": 0.7977,
"step": 285
},
{
"epoch": 0.12013256006628004,
"grad_norm": 2.973066010365782,
"learning_rate": 9.98795440846732e-06,
"loss": 0.7823,
"step": 290
},
{
"epoch": 0.12220381110190555,
"grad_norm": 2.9253645728220925,
"learning_rate": 9.985315500224135e-06,
"loss": 0.7657,
"step": 295
},
{
"epoch": 0.12427506213753108,
"grad_norm": 2.7951944746822677,
"learning_rate": 9.982415850055902e-06,
"loss": 0.7708,
"step": 300
},
{
"epoch": 0.1263463131731566,
"grad_norm": 3.1659773297780966,
"learning_rate": 9.979255609620095e-06,
"loss": 0.776,
"step": 305
},
{
"epoch": 0.12841756420878211,
"grad_norm": 2.9879593945151792,
"learning_rate": 9.975834944203581e-06,
"loss": 0.7734,
"step": 310
},
{
"epoch": 0.1304888152444076,
"grad_norm": 2.985895349891211,
"learning_rate": 9.972154032713973e-06,
"loss": 0.7612,
"step": 315
},
{
"epoch": 0.13256006628003314,
"grad_norm": 2.9529864978776037,
"learning_rate": 9.968213067670265e-06,
"loss": 0.7614,
"step": 320
},
{
"epoch": 0.13463131731565867,
"grad_norm": 2.7267524949676485,
"learning_rate": 9.964012255192776e-06,
"loss": 0.7666,
"step": 325
},
{
"epoch": 0.13670256835128416,
"grad_norm": 2.8509365131119813,
"learning_rate": 9.959551814992364e-06,
"loss": 0.7621,
"step": 330
},
{
"epoch": 0.1387738193869097,
"grad_norm": 2.9770598788165823,
"learning_rate": 9.954831980358928e-06,
"loss": 0.7648,
"step": 335
},
{
"epoch": 0.14084507042253522,
"grad_norm": 2.968871762580154,
"learning_rate": 9.949852998149217e-06,
"loss": 0.7584,
"step": 340
},
{
"epoch": 0.14291632145816072,
"grad_norm": 2.8765123534484474,
"learning_rate": 9.944615128773911e-06,
"loss": 0.7366,
"step": 345
},
{
"epoch": 0.14498757249378624,
"grad_norm": 2.877358345539652,
"learning_rate": 9.939118646184007e-06,
"loss": 0.7645,
"step": 350
},
{
"epoch": 0.14705882352941177,
"grad_norm": 2.944453042289488,
"learning_rate": 9.933363837856485e-06,
"loss": 0.7506,
"step": 355
},
{
"epoch": 0.1491300745650373,
"grad_norm": 2.8714409602766957,
"learning_rate": 9.927351004779275e-06,
"loss": 0.7388,
"step": 360
},
{
"epoch": 0.1512013256006628,
"grad_norm": 2.7538932235071742,
"learning_rate": 9.921080461435522e-06,
"loss": 0.7381,
"step": 365
},
{
"epoch": 0.15327257663628832,
"grad_norm": 2.802085235529085,
"learning_rate": 9.914552535787122e-06,
"loss": 0.7527,
"step": 370
},
{
"epoch": 0.15534382767191385,
"grad_norm": 3.081520608158383,
"learning_rate": 9.90776756925758e-06,
"loss": 0.7336,
"step": 375
},
{
"epoch": 0.15741507870753935,
"grad_norm": 2.8672122064055356,
"learning_rate": 9.900725916714157e-06,
"loss": 0.7213,
"step": 380
},
{
"epoch": 0.15948632974316487,
"grad_norm": 2.774256549974597,
"learning_rate": 9.893427946449297e-06,
"loss": 0.7407,
"step": 385
},
{
"epoch": 0.1615575807787904,
"grad_norm": 2.6603866290733182,
"learning_rate": 9.885874040161373e-06,
"loss": 0.7402,
"step": 390
},
{
"epoch": 0.1636288318144159,
"grad_norm": 2.9061817991256738,
"learning_rate": 9.878064592934723e-06,
"loss": 0.7422,
"step": 395
},
{
"epoch": 0.16570008285004142,
"grad_norm": 2.783323806110707,
"learning_rate": 9.87000001321898e-06,
"loss": 0.7349,
"step": 400
},
{
"epoch": 0.16777133388566695,
"grad_norm": 2.8382691908256783,
"learning_rate": 9.86168072280772e-06,
"loss": 0.7091,
"step": 405
},
{
"epoch": 0.16984258492129245,
"grad_norm": 2.9931327370685987,
"learning_rate": 9.853107156816393e-06,
"loss": 0.7227,
"step": 410
},
{
"epoch": 0.17191383595691798,
"grad_norm": 2.9117297947111767,
"learning_rate": 9.844279763659566e-06,
"loss": 0.7151,
"step": 415
},
{
"epoch": 0.1739850869925435,
"grad_norm": 2.604556349211621,
"learning_rate": 9.835199005027477e-06,
"loss": 0.7137,
"step": 420
},
{
"epoch": 0.176056338028169,
"grad_norm": 2.6512301633245494,
"learning_rate": 9.825865355861878e-06,
"loss": 0.7109,
"step": 425
},
{
"epoch": 0.17812758906379453,
"grad_norm": 2.8407538606614464,
"learning_rate": 9.816279304331202e-06,
"loss": 0.6934,
"step": 430
},
{
"epoch": 0.18019884009942005,
"grad_norm": 2.8439513245733754,
"learning_rate": 9.806441351805025e-06,
"loss": 0.7234,
"step": 435
},
{
"epoch": 0.18227009113504558,
"grad_norm": 2.798364319791846,
"learning_rate": 9.79635201282785e-06,
"loss": 0.7144,
"step": 440
},
{
"epoch": 0.18434134217067108,
"grad_norm": 2.829699537398248,
"learning_rate": 9.786011815092193e-06,
"loss": 0.715,
"step": 445
},
{
"epoch": 0.1864125932062966,
"grad_norm": 2.793592254800858,
"learning_rate": 9.775421299410977e-06,
"loss": 0.7027,
"step": 450
},
{
"epoch": 0.18848384424192213,
"grad_norm": 2.725093759335724,
"learning_rate": 9.764581019689255e-06,
"loss": 0.7123,
"step": 455
},
{
"epoch": 0.19055509527754763,
"grad_norm": 2.9951737192270556,
"learning_rate": 9.753491542895237e-06,
"loss": 0.7124,
"step": 460
},
{
"epoch": 0.19262634631317316,
"grad_norm": 2.723762962665785,
"learning_rate": 9.742153449030639e-06,
"loss": 0.6891,
"step": 465
},
{
"epoch": 0.19469759734879868,
"grad_norm": 2.8097737563843395,
"learning_rate": 9.730567331100333e-06,
"loss": 0.7115,
"step": 470
},
{
"epoch": 0.19676884838442418,
"grad_norm": 2.7032549520221556,
"learning_rate": 9.71873379508136e-06,
"loss": 0.7091,
"step": 475
},
{
"epoch": 0.1988400994200497,
"grad_norm": 2.6572300829667665,
"learning_rate": 9.706653459891207e-06,
"loss": 0.68,
"step": 480
},
{
"epoch": 0.20091135045567524,
"grad_norm": 2.7902518845602833,
"learning_rate": 9.694326957355452e-06,
"loss": 0.695,
"step": 485
},
{
"epoch": 0.20298260149130073,
"grad_norm": 2.8595070173120876,
"learning_rate": 9.681754932174719e-06,
"loss": 0.6746,
"step": 490
},
{
"epoch": 0.20505385252692626,
"grad_norm": 2.775119823880149,
"learning_rate": 9.668938041890952e-06,
"loss": 0.6651,
"step": 495
},
{
"epoch": 0.2071251035625518,
"grad_norm": 2.8300747272322955,
"learning_rate": 9.655876956853025e-06,
"loss": 0.6837,
"step": 500
},
{
"epoch": 0.2091963545981773,
"grad_norm": 2.68862934210806,
"learning_rate": 9.64257236018169e-06,
"loss": 0.6845,
"step": 505
},
{
"epoch": 0.2112676056338028,
"grad_norm": 2.9380613735304575,
"learning_rate": 9.629024947733836e-06,
"loss": 0.6699,
"step": 510
},
{
"epoch": 0.21333885666942834,
"grad_norm": 2.900170863879882,
"learning_rate": 9.615235428066106e-06,
"loss": 0.6836,
"step": 515
},
{
"epoch": 0.21541010770505387,
"grad_norm": 2.7447728191832046,
"learning_rate": 9.601204522397826e-06,
"loss": 0.6753,
"step": 520
},
{
"epoch": 0.21748135874067936,
"grad_norm": 2.6332156758147995,
"learning_rate": 9.586932964573298e-06,
"loss": 0.6795,
"step": 525
},
{
"epoch": 0.2195526097763049,
"grad_norm": 3.0019737177126133,
"learning_rate": 9.572421501023403e-06,
"loss": 0.6658,
"step": 530
},
{
"epoch": 0.22162386081193042,
"grad_norm": 2.691180491826932,
"learning_rate": 9.557670890726576e-06,
"loss": 0.6612,
"step": 535
},
{
"epoch": 0.22369511184755592,
"grad_norm": 2.692761284388416,
"learning_rate": 9.5426819051691e-06,
"loss": 0.6745,
"step": 540
},
{
"epoch": 0.22576636288318144,
"grad_norm": 2.885575545800491,
"learning_rate": 9.527455328304756e-06,
"loss": 0.6437,
"step": 545
},
{
"epoch": 0.22783761391880697,
"grad_norm": 3.7475436767857206,
"learning_rate": 9.511991956513828e-06,
"loss": 0.6567,
"step": 550
},
{
"epoch": 0.22990886495443247,
"grad_norm": 2.892045312721076,
"learning_rate": 9.496292598561445e-06,
"loss": 0.6613,
"step": 555
},
{
"epoch": 0.231980115990058,
"grad_norm": 2.8887320639722462,
"learning_rate": 9.480358075555278e-06,
"loss": 0.6552,
"step": 560
},
{
"epoch": 0.23405136702568352,
"grad_norm": 2.7432063566041176,
"learning_rate": 9.464189220902603e-06,
"loss": 0.6584,
"step": 565
},
{
"epoch": 0.23612261806130902,
"grad_norm": 2.717134474546673,
"learning_rate": 9.447786880266706e-06,
"loss": 0.6567,
"step": 570
},
{
"epoch": 0.23819386909693455,
"grad_norm": 3.50010105808707,
"learning_rate": 9.431151911522656e-06,
"loss": 0.6521,
"step": 575
},
{
"epoch": 0.24026512013256007,
"grad_norm": 3.1415449817336794,
"learning_rate": 9.414285184712432e-06,
"loss": 0.6539,
"step": 580
},
{
"epoch": 0.24233637116818557,
"grad_norm": 2.731218864869034,
"learning_rate": 9.397187581999424e-06,
"loss": 0.626,
"step": 585
},
{
"epoch": 0.2444076222038111,
"grad_norm": 2.788585498299461,
"learning_rate": 9.37985999762229e-06,
"loss": 0.6409,
"step": 590
},
{
"epoch": 0.24647887323943662,
"grad_norm": 2.863758065279955,
"learning_rate": 9.362303337848188e-06,
"loss": 0.6447,
"step": 595
},
{
"epoch": 0.24855012427506215,
"grad_norm": 2.8640982790187204,
"learning_rate": 9.344518520925377e-06,
"loss": 0.6383,
"step": 600
},
{
"epoch": 0.2506213753106877,
"grad_norm": 2.8277305534887764,
"learning_rate": 9.326506477035179e-06,
"loss": 0.639,
"step": 605
},
{
"epoch": 0.2526926263463132,
"grad_norm": 2.664289854153118,
"learning_rate": 9.308268148243355e-06,
"loss": 0.6412,
"step": 610
},
{
"epoch": 0.2547638773819387,
"grad_norm": 2.744294057010904,
"learning_rate": 9.289804488450805e-06,
"loss": 0.6383,
"step": 615
},
{
"epoch": 0.25683512841756423,
"grad_norm": 2.762679931681209,
"learning_rate": 9.271116463343692e-06,
"loss": 0.6472,
"step": 620
},
{
"epoch": 0.25890637945318973,
"grad_norm": 2.7936718883733906,
"learning_rate": 9.25220505034293e-06,
"loss": 0.6283,
"step": 625
},
{
"epoch": 0.2609776304888152,
"grad_norm": 2.702644364966462,
"learning_rate": 9.23307123855307e-06,
"loss": 0.6354,
"step": 630
},
{
"epoch": 0.2630488815244408,
"grad_norm": 2.6524584260807864,
"learning_rate": 9.213716028710558e-06,
"loss": 0.6116,
"step": 635
},
{
"epoch": 0.2651201325600663,
"grad_norm": 2.5335412029402367,
"learning_rate": 9.194140433131397e-06,
"loss": 0.6194,
"step": 640
},
{
"epoch": 0.2671913835956918,
"grad_norm": 2.7919689300708783,
"learning_rate": 9.174345475658208e-06,
"loss": 0.6209,
"step": 645
},
{
"epoch": 0.26926263463131733,
"grad_norm": 2.707149083323794,
"learning_rate": 9.154332191606671e-06,
"loss": 0.6136,
"step": 650
},
{
"epoch": 0.27133388566694283,
"grad_norm": 2.6221689387749048,
"learning_rate": 9.134101627711384e-06,
"loss": 0.6402,
"step": 655
},
{
"epoch": 0.27340513670256833,
"grad_norm": 2.6574146052486536,
"learning_rate": 9.113654842071114e-06,
"loss": 0.6039,
"step": 660
},
{
"epoch": 0.2754763877381939,
"grad_norm": 3.0914329797904925,
"learning_rate": 9.092992904093451e-06,
"loss": 0.6371,
"step": 665
},
{
"epoch": 0.2775476387738194,
"grad_norm": 4.080535779383502,
"learning_rate": 9.072116894438885e-06,
"loss": 0.6001,
"step": 670
},
{
"epoch": 0.2796188898094449,
"grad_norm": 3.1310996031052802,
"learning_rate": 9.051027904964279e-06,
"loss": 0.6097,
"step": 675
},
{
"epoch": 0.28169014084507044,
"grad_norm": 2.694121266388182,
"learning_rate": 9.029727038665765e-06,
"loss": 0.6018,
"step": 680
},
{
"epoch": 0.28376139188069593,
"grad_norm": 2.7099859959705452,
"learning_rate": 9.008215409621053e-06,
"loss": 0.5987,
"step": 685
},
{
"epoch": 0.28583264291632143,
"grad_norm": 2.7290311424672087,
"learning_rate": 8.986494142931168e-06,
"loss": 0.5943,
"step": 690
},
{
"epoch": 0.287903893951947,
"grad_norm": 2.7513013645930817,
"learning_rate": 8.964564374661597e-06,
"loss": 0.6106,
"step": 695
},
{
"epoch": 0.2899751449875725,
"grad_norm": 2.722972539639557,
"learning_rate": 8.94242725178288e-06,
"loss": 0.6126,
"step": 700
},
{
"epoch": 0.292046396023198,
"grad_norm": 2.7848304561579025,
"learning_rate": 8.920083932110608e-06,
"loss": 0.6568,
"step": 705
},
{
"epoch": 0.29411764705882354,
"grad_norm": 2.9491408257282874,
"learning_rate": 8.89753558424488e-06,
"loss": 0.6214,
"step": 710
},
{
"epoch": 0.29618889809444904,
"grad_norm": 2.690336696229449,
"learning_rate": 8.874783387509181e-06,
"loss": 0.6117,
"step": 715
},
{
"epoch": 0.2982601491300746,
"grad_norm": 2.670802651297672,
"learning_rate": 8.851828531888692e-06,
"loss": 0.5951,
"step": 720
},
{
"epoch": 0.3003314001657001,
"grad_norm": 2.638042315197902,
"learning_rate": 8.828672217968055e-06,
"loss": 0.5997,
"step": 725
},
{
"epoch": 0.3024026512013256,
"grad_norm": 2.9789844819810822,
"learning_rate": 8.805315656868587e-06,
"loss": 0.5808,
"step": 730
},
{
"epoch": 0.30447390223695114,
"grad_norm": 2.7137901378908134,
"learning_rate": 8.781760070184933e-06,
"loss": 0.5877,
"step": 735
},
{
"epoch": 0.30654515327257664,
"grad_norm": 2.6525222438395053,
"learning_rate": 8.75800668992117e-06,
"loss": 0.602,
"step": 740
},
{
"epoch": 0.30861640430820214,
"grad_norm": 2.7137413029317887,
"learning_rate": 8.734056758426367e-06,
"loss": 0.5843,
"step": 745
},
{
"epoch": 0.3106876553438277,
"grad_norm": 2.7143182881620405,
"learning_rate": 8.709911528329623e-06,
"loss": 0.5833,
"step": 750
},
{
"epoch": 0.3127589063794532,
"grad_norm": 2.8294993842812333,
"learning_rate": 8.685572262474538e-06,
"loss": 0.6022,
"step": 755
},
{
"epoch": 0.3148301574150787,
"grad_norm": 2.6660531901712585,
"learning_rate": 8.661040233853166e-06,
"loss": 0.5883,
"step": 760
},
{
"epoch": 0.31690140845070425,
"grad_norm": 2.5753297715004893,
"learning_rate": 8.636316725539445e-06,
"loss": 0.5813,
"step": 765
},
{
"epoch": 0.31897265948632975,
"grad_norm": 2.598139002034116,
"learning_rate": 8.611403030622074e-06,
"loss": 0.5819,
"step": 770
},
{
"epoch": 0.32104391052195524,
"grad_norm": 2.809509102764727,
"learning_rate": 8.586300452136895e-06,
"loss": 0.5666,
"step": 775
},
{
"epoch": 0.3231151615575808,
"grad_norm": 2.728970191068225,
"learning_rate": 8.561010302998734e-06,
"loss": 0.572,
"step": 780
},
{
"epoch": 0.3251864125932063,
"grad_norm": 2.7938578148190367,
"learning_rate": 8.535533905932739e-06,
"loss": 0.5711,
"step": 785
},
{
"epoch": 0.3272576636288318,
"grad_norm": 2.686665176609409,
"learning_rate": 8.509872593405189e-06,
"loss": 0.5746,
"step": 790
},
{
"epoch": 0.32932891466445735,
"grad_norm": 2.96114679262856,
"learning_rate": 8.484027707553818e-06,
"loss": 0.5759,
"step": 795
},
{
"epoch": 0.33140016570008285,
"grad_norm": 2.682476049310721,
"learning_rate": 8.458000600117604e-06,
"loss": 0.5554,
"step": 800
},
{
"epoch": 0.33347141673570835,
"grad_norm": 2.8062866639540305,
"learning_rate": 8.43179263236608e-06,
"loss": 0.5727,
"step": 805
},
{
"epoch": 0.3355426677713339,
"grad_norm": 2.5032565502374697,
"learning_rate": 8.40540517502813e-06,
"loss": 0.5606,
"step": 810
},
{
"epoch": 0.3376139188069594,
"grad_norm": 2.588877805073754,
"learning_rate": 8.378839608220304e-06,
"loss": 0.5733,
"step": 815
},
{
"epoch": 0.3396851698425849,
"grad_norm": 2.5556233186653166,
"learning_rate": 8.35209732137463e-06,
"loss": 0.5598,
"step": 820
},
{
"epoch": 0.34175642087821045,
"grad_norm": 3.09195320870581,
"learning_rate": 8.32517971316595e-06,
"loss": 0.5566,
"step": 825
},
{
"epoch": 0.34382767191383595,
"grad_norm": 2.5718134472178678,
"learning_rate": 8.298088191438753e-06,
"loss": 0.5567,
"step": 830
},
{
"epoch": 0.34589892294946145,
"grad_norm": 2.7405529612153896,
"learning_rate": 8.270824173133563e-06,
"loss": 0.5801,
"step": 835
},
{
"epoch": 0.347970173985087,
"grad_norm": 2.5641097869776255,
"learning_rate": 8.243389084212808e-06,
"loss": 0.5492,
"step": 840
},
{
"epoch": 0.3500414250207125,
"grad_norm": 2.9233335723519307,
"learning_rate": 8.215784359586257e-06,
"loss": 0.5646,
"step": 845
},
{
"epoch": 0.352112676056338,
"grad_norm": 2.653078447889723,
"learning_rate": 8.188011443035962e-06,
"loss": 0.5501,
"step": 850
},
{
"epoch": 0.35418392709196356,
"grad_norm": 2.6410656716688594,
"learning_rate": 8.160071787140742e-06,
"loss": 0.5647,
"step": 855
},
{
"epoch": 0.35625517812758906,
"grad_norm": 2.604977685510966,
"learning_rate": 8.131966853200226e-06,
"loss": 0.5736,
"step": 860
},
{
"epoch": 0.35832642916321455,
"grad_norm": 2.6151064854302284,
"learning_rate": 8.103698111158405e-06,
"loss": 0.5472,
"step": 865
},
{
"epoch": 0.3603976801988401,
"grad_norm": 2.7202636772410065,
"learning_rate": 8.075267039526764e-06,
"loss": 0.5501,
"step": 870
},
{
"epoch": 0.3624689312344656,
"grad_norm": 2.776388856485309,
"learning_rate": 8.046675125306948e-06,
"loss": 0.5468,
"step": 875
},
{
"epoch": 0.36454018227009116,
"grad_norm": 2.6113291789143016,
"learning_rate": 8.017923863912989e-06,
"loss": 0.5289,
"step": 880
},
{
"epoch": 0.36661143330571666,
"grad_norm": 2.6617516582114775,
"learning_rate": 7.989014759093095e-06,
"loss": 0.5313,
"step": 885
},
{
"epoch": 0.36868268434134216,
"grad_norm": 2.529021574604729,
"learning_rate": 7.959949322850994e-06,
"loss": 0.5227,
"step": 890
},
{
"epoch": 0.3707539353769677,
"grad_norm": 2.6639798694022043,
"learning_rate": 7.930729075366867e-06,
"loss": 0.5255,
"step": 895
},
{
"epoch": 0.3728251864125932,
"grad_norm": 2.676931390694889,
"learning_rate": 7.901355544917827e-06,
"loss": 0.5242,
"step": 900
},
{
"epoch": 0.3748964374482187,
"grad_norm": 2.6352389372687086,
"learning_rate": 7.87183026779799e-06,
"loss": 0.526,
"step": 905
},
{
"epoch": 0.37696768848384427,
"grad_norm": 2.6249152377324987,
"learning_rate": 7.842154788238124e-06,
"loss": 0.5397,
"step": 910
},
{
"epoch": 0.37903893951946976,
"grad_norm": 2.5946185632885,
"learning_rate": 7.812330658324884e-06,
"loss": 0.5268,
"step": 915
},
{
"epoch": 0.38111019055509526,
"grad_norm": 2.8257139288588147,
"learning_rate": 7.782359437919644e-06,
"loss": 0.5148,
"step": 920
},
{
"epoch": 0.3831814415907208,
"grad_norm": 2.7146819691914152,
"learning_rate": 7.75224269457689e-06,
"loss": 0.5476,
"step": 925
},
{
"epoch": 0.3852526926263463,
"grad_norm": 2.689176461961057,
"learning_rate": 7.721982003462255e-06,
"loss": 0.5367,
"step": 930
},
{
"epoch": 0.3873239436619718,
"grad_norm": 2.6217087065252085,
"learning_rate": 7.691578947270122e-06,
"loss": 0.5265,
"step": 935
},
{
"epoch": 0.38939519469759737,
"grad_norm": 2.611186684583373,
"learning_rate": 7.661035116140856e-06,
"loss": 0.5326,
"step": 940
},
{
"epoch": 0.39146644573322287,
"grad_norm": 2.727433020851049,
"learning_rate": 7.63035210757763e-06,
"loss": 0.5451,
"step": 945
},
{
"epoch": 0.39353769676884837,
"grad_norm": 2.6720882907191705,
"learning_rate": 7.599531526362873e-06,
"loss": 0.5172,
"step": 950
},
{
"epoch": 0.3956089478044739,
"grad_norm": 2.659996298294283,
"learning_rate": 7.568574984474335e-06,
"loss": 0.5373,
"step": 955
},
{
"epoch": 0.3976801988400994,
"grad_norm": 2.5093912943242955,
"learning_rate": 7.537484101000787e-06,
"loss": 0.5067,
"step": 960
},
{
"epoch": 0.3997514498757249,
"grad_norm": 2.8059615971815037,
"learning_rate": 7.506260502057325e-06,
"loss": 0.5406,
"step": 965
},
{
"epoch": 0.40182270091135047,
"grad_norm": 2.524377317993043,
"learning_rate": 7.474905820700334e-06,
"loss": 0.5256,
"step": 970
},
{
"epoch": 0.40389395194697597,
"grad_norm": 2.7745824727223076,
"learning_rate": 7.443421696842066e-06,
"loss": 0.5119,
"step": 975
},
{
"epoch": 0.40596520298260147,
"grad_norm": 2.8050272533475504,
"learning_rate": 7.411809777164873e-06,
"loss": 0.5256,
"step": 980
},
{
"epoch": 0.408036454018227,
"grad_norm": 2.602889185147068,
"learning_rate": 7.380071715035089e-06,
"loss": 0.5152,
"step": 985
},
{
"epoch": 0.4101077050538525,
"grad_norm": 2.538757080202685,
"learning_rate": 7.3482091704165405e-06,
"loss": 0.5253,
"step": 990
},
{
"epoch": 0.412178956089478,
"grad_norm": 2.6357438468997594,
"learning_rate": 7.316223809783745e-06,
"loss": 0.5101,
"step": 995
},
{
"epoch": 0.4142502071251036,
"grad_norm": 2.613789468715128,
"learning_rate": 7.284117306034733e-06,
"loss": 0.5155,
"step": 1000
},
{
"epoch": 0.4163214581607291,
"grad_norm": 2.4899090799020818,
"learning_rate": 7.2518913384035685e-06,
"loss": 0.5084,
"step": 1005
},
{
"epoch": 0.4183927091963546,
"grad_norm": 2.6062511498745295,
"learning_rate": 7.219547592372512e-06,
"loss": 0.5213,
"step": 1010
},
{
"epoch": 0.4204639602319801,
"grad_norm": 2.5868254269917372,
"learning_rate": 7.187087759583869e-06,
"loss": 0.5025,
"step": 1015
},
{
"epoch": 0.4225352112676056,
"grad_norm": 2.60436348601176,
"learning_rate": 7.15451353775151e-06,
"loss": 0.5157,
"step": 1020
},
{
"epoch": 0.4246064623032311,
"grad_norm": 2.595575075152848,
"learning_rate": 7.121826630572084e-06,
"loss": 0.4977,
"step": 1025
},
{
"epoch": 0.4266777133388567,
"grad_norm": 2.730306168768171,
"learning_rate": 7.089028747635908e-06,
"loss": 0.5067,
"step": 1030
},
{
"epoch": 0.4287489643744822,
"grad_norm": 2.429565594815639,
"learning_rate": 7.056121604337554e-06,
"loss": 0.5042,
"step": 1035
},
{
"epoch": 0.43082021541010773,
"grad_norm": 2.6008211508268686,
"learning_rate": 7.023106921786118e-06,
"loss": 0.4954,
"step": 1040
},
{
"epoch": 0.43289146644573323,
"grad_norm": 2.646586187294066,
"learning_rate": 6.9899864267152275e-06,
"loss": 0.4904,
"step": 1045
},
{
"epoch": 0.43496271748135873,
"grad_norm": 2.3779944382404743,
"learning_rate": 6.956761851392706e-06,
"loss": 0.4889,
"step": 1050
},
{
"epoch": 0.4370339685169843,
"grad_norm": 2.4397871611252513,
"learning_rate": 6.9234349335299835e-06,
"loss": 0.4956,
"step": 1055
},
{
"epoch": 0.4391052195526098,
"grad_norm": 2.8653046484616693,
"learning_rate": 6.890007416191209e-06,
"loss": 0.4886,
"step": 1060
},
{
"epoch": 0.4411764705882353,
"grad_norm": 2.5842292091537886,
"learning_rate": 6.8564810477020835e-06,
"loss": 0.504,
"step": 1065
},
{
"epoch": 0.44324772162386084,
"grad_norm": 2.6494637460436143,
"learning_rate": 6.822857581558423e-06,
"loss": 0.4849,
"step": 1070
},
{
"epoch": 0.44531897265948633,
"grad_norm": 2.4996836650425895,
"learning_rate": 6.789138776334441e-06,
"loss": 0.4888,
"step": 1075
},
{
"epoch": 0.44739022369511183,
"grad_norm": 2.6825342299727777,
"learning_rate": 6.7553263955907755e-06,
"loss": 0.4783,
"step": 1080
},
{
"epoch": 0.4494614747307374,
"grad_norm": 2.531980161923532,
"learning_rate": 6.721422207782249e-06,
"loss": 0.4931,
"step": 1085
},
{
"epoch": 0.4515327257663629,
"grad_norm": 2.57473990498951,
"learning_rate": 6.687427986165379e-06,
"loss": 0.4783,
"step": 1090
},
{
"epoch": 0.4536039768019884,
"grad_norm": 2.572163087103638,
"learning_rate": 6.653345508705629e-06,
"loss": 0.4807,
"step": 1095
},
{
"epoch": 0.45567522783761394,
"grad_norm": 2.5561287573335405,
"learning_rate": 6.6191765579844205e-06,
"loss": 0.4848,
"step": 1100
},
{
"epoch": 0.45774647887323944,
"grad_norm": 2.5714524421545613,
"learning_rate": 6.584922921105894e-06,
"loss": 0.4899,
"step": 1105
},
{
"epoch": 0.45981772990886494,
"grad_norm": 2.6961085116928536,
"learning_rate": 6.550586389603451e-06,
"loss": 0.4857,
"step": 1110
},
{
"epoch": 0.4618889809444905,
"grad_norm": 2.7844667811693236,
"learning_rate": 6.5161687593460395e-06,
"loss": 0.4793,
"step": 1115
},
{
"epoch": 0.463960231980116,
"grad_norm": 2.6758872490089303,
"learning_rate": 6.481671830444243e-06,
"loss": 0.4879,
"step": 1120
},
{
"epoch": 0.4660314830157415,
"grad_norm": 2.56835061152998,
"learning_rate": 6.447097407156114e-06,
"loss": 0.4775,
"step": 1125
},
{
"epoch": 0.46810273405136704,
"grad_norm": 2.563350080923348,
"learning_rate": 6.412447297792818e-06,
"loss": 0.469,
"step": 1130
},
{
"epoch": 0.47017398508699254,
"grad_norm": 2.4364058841521246,
"learning_rate": 6.377723314624057e-06,
"loss": 0.4747,
"step": 1135
},
{
"epoch": 0.47224523612261804,
"grad_norm": 2.5436107294297785,
"learning_rate": 6.3429272737832726e-06,
"loss": 0.4553,
"step": 1140
},
{
"epoch": 0.4743164871582436,
"grad_norm": 2.5165858809477304,
"learning_rate": 6.308060995172673e-06,
"loss": 0.4708,
"step": 1145
},
{
"epoch": 0.4763877381938691,
"grad_norm": 2.5128351080872995,
"learning_rate": 6.273126302368037e-06,
"loss": 0.4898,
"step": 1150
},
{
"epoch": 0.4784589892294946,
"grad_norm": 2.603860292694571,
"learning_rate": 6.238125022523343e-06,
"loss": 0.474,
"step": 1155
},
{
"epoch": 0.48053024026512015,
"grad_norm": 2.5011591305662453,
"learning_rate": 6.203058986275207e-06,
"loss": 0.4627,
"step": 1160
},
{
"epoch": 0.48260149130074564,
"grad_norm": 2.5762888844460092,
"learning_rate": 6.1679300276471285e-06,
"loss": 0.4588,
"step": 1165
},
{
"epoch": 0.48467274233637114,
"grad_norm": 2.613848629581562,
"learning_rate": 6.132739983953579e-06,
"loss": 0.4565,
"step": 1170
},
{
"epoch": 0.4867439933719967,
"grad_norm": 2.4865376595756636,
"learning_rate": 6.097490695703896e-06,
"loss": 0.4494,
"step": 1175
},
{
"epoch": 0.4888152444076222,
"grad_norm": 2.5110745710063944,
"learning_rate": 6.062184006506027e-06,
"loss": 0.4632,
"step": 1180
},
{
"epoch": 0.4908864954432477,
"grad_norm": 2.71807275477703,
"learning_rate": 6.026821762970102e-06,
"loss": 0.4577,
"step": 1185
},
{
"epoch": 0.49295774647887325,
"grad_norm": 2.5543522417001316,
"learning_rate": 5.991405814611855e-06,
"loss": 0.4695,
"step": 1190
},
{
"epoch": 0.49502899751449875,
"grad_norm": 2.6490160914028507,
"learning_rate": 5.955938013755888e-06,
"loss": 0.4306,
"step": 1195
},
{
"epoch": 0.4971002485501243,
"grad_norm": 2.65287059904939,
"learning_rate": 5.920420215438794e-06,
"loss": 0.4643,
"step": 1200
},
{
"epoch": 0.4991714995857498,
"grad_norm": 2.8005444354959406,
"learning_rate": 5.8848542773121285e-06,
"loss": 0.4365,
"step": 1205
},
{
"epoch": 0.5012427506213754,
"grad_norm": 2.6038026851582545,
"learning_rate": 5.849242059545259e-06,
"loss": 0.4729,
"step": 1210
},
{
"epoch": 0.5033140016570008,
"grad_norm": 2.769549146198858,
"learning_rate": 5.81358542472807e-06,
"loss": 0.4495,
"step": 1215
},
{
"epoch": 0.5053852526926264,
"grad_norm": 2.5235652990477044,
"learning_rate": 5.777886237773542e-06,
"loss": 0.4663,
"step": 1220
},
{
"epoch": 0.5074565037282519,
"grad_norm": 2.3657865773053866,
"learning_rate": 5.742146365820223e-06,
"loss": 0.4458,
"step": 1225
},
{
"epoch": 0.5095277547638773,
"grad_norm": 2.5324699506355492,
"learning_rate": 5.706367678134562e-06,
"loss": 0.4429,
"step": 1230
},
{
"epoch": 0.5115990057995029,
"grad_norm": 2.616108468496064,
"learning_rate": 5.670552046013151e-06,
"loss": 0.4471,
"step": 1235
},
{
"epoch": 0.5136702568351285,
"grad_norm": 2.668938167187185,
"learning_rate": 5.634701342684852e-06,
"loss": 0.4677,
"step": 1240
},
{
"epoch": 0.5157415078707539,
"grad_norm": 2.470348244118417,
"learning_rate": 5.598817443212813e-06,
"loss": 0.4519,
"step": 1245
},
{
"epoch": 0.5178127589063795,
"grad_norm": 2.472745570372645,
"learning_rate": 5.562902224396416e-06,
"loss": 0.4373,
"step": 1250
},
{
"epoch": 0.519884009942005,
"grad_norm": 2.4462368506462324,
"learning_rate": 5.526957564673098e-06,
"loss": 0.4411,
"step": 1255
},
{
"epoch": 0.5219552609776305,
"grad_norm": 2.5799972464923115,
"learning_rate": 5.49098534402012e-06,
"loss": 0.4548,
"step": 1260
},
{
"epoch": 0.524026512013256,
"grad_norm": 2.4840179248507406,
"learning_rate": 5.454987443856235e-06,
"loss": 0.4614,
"step": 1265
},
{
"epoch": 0.5260977630488816,
"grad_norm": 2.546787702828046,
"learning_rate": 5.418965746943281e-06,
"loss": 0.4421,
"step": 1270
},
{
"epoch": 0.528169014084507,
"grad_norm": 2.6463641954544825,
"learning_rate": 5.3829221372877175e-06,
"loss": 0.4473,
"step": 1275
},
{
"epoch": 0.5302402651201326,
"grad_norm": 2.5500320208273677,
"learning_rate": 5.34685850004208e-06,
"loss": 0.449,
"step": 1280
},
{
"epoch": 0.5323115161557581,
"grad_norm": 2.336913107994826,
"learning_rate": 5.310776721406392e-06,
"loss": 0.4338,
"step": 1285
},
{
"epoch": 0.5343827671913836,
"grad_norm": 2.5373015768646847,
"learning_rate": 5.2746786885295034e-06,
"loss": 0.4481,
"step": 1290
},
{
"epoch": 0.5364540182270091,
"grad_norm": 2.4870998061462664,
"learning_rate": 5.238566289410396e-06,
"loss": 0.4259,
"step": 1295
},
{
"epoch": 0.5385252692626347,
"grad_norm": 2.3888415610915885,
"learning_rate": 5.2024414127994325e-06,
"loss": 0.4213,
"step": 1300
},
{
"epoch": 0.5405965202982601,
"grad_norm": 2.4893805872738817,
"learning_rate": 5.166305948099574e-06,
"loss": 0.4327,
"step": 1305
},
{
"epoch": 0.5426677713338857,
"grad_norm": 2.5964136277509784,
"learning_rate": 5.13016178526756e-06,
"loss": 0.4355,
"step": 1310
},
{
"epoch": 0.5447390223695112,
"grad_norm": 2.579500300302397,
"learning_rate": 5.094010814715062e-06,
"loss": 0.4241,
"step": 1315
},
{
"epoch": 0.5468102734051367,
"grad_norm": 2.4639072073106147,
"learning_rate": 5.057854927209804e-06,
"loss": 0.4281,
"step": 1320
},
{
"epoch": 0.5488815244407622,
"grad_norm": 2.57018141822179,
"learning_rate": 5.0216960137766805e-06,
"loss": 0.4261,
"step": 1325
},
{
"epoch": 0.5509527754763878,
"grad_norm": 2.499348596384133,
"learning_rate": 4.985535965598843e-06,
"loss": 0.4355,
"step": 1330
},
{
"epoch": 0.5530240265120132,
"grad_norm": 2.6128115671413936,
"learning_rate": 4.949376673918802e-06,
"loss": 0.4147,
"step": 1335
},
{
"epoch": 0.5550952775476388,
"grad_norm": 2.25777210775929,
"learning_rate": 4.913220029939491e-06,
"loss": 0.421,
"step": 1340
},
{
"epoch": 0.5571665285832643,
"grad_norm": 2.5101423479639746,
"learning_rate": 4.877067924725368e-06,
"loss": 0.4379,
"step": 1345
},
{
"epoch": 0.5592377796188898,
"grad_norm": 2.500950985421387,
"learning_rate": 4.840922249103506e-06,
"loss": 0.4221,
"step": 1350
},
{
"epoch": 0.5613090306545153,
"grad_norm": 2.6429313222768878,
"learning_rate": 4.804784893564697e-06,
"loss": 0.4249,
"step": 1355
},
{
"epoch": 0.5633802816901409,
"grad_norm": 2.4319897802858597,
"learning_rate": 4.7686577481645745e-06,
"loss": 0.4262,
"step": 1360
},
{
"epoch": 0.5654515327257663,
"grad_norm": 2.369903742738793,
"learning_rate": 4.732542702424759e-06,
"loss": 0.4246,
"step": 1365
},
{
"epoch": 0.5675227837613919,
"grad_norm": 2.431383395924861,
"learning_rate": 4.696441645234042e-06,
"loss": 0.4204,
"step": 1370
},
{
"epoch": 0.5695940347970174,
"grad_norm": 2.5537370398776766,
"learning_rate": 4.660356464749578e-06,
"loss": 0.4259,
"step": 1375
},
{
"epoch": 0.5716652858326429,
"grad_norm": 2.4956485931438532,
"learning_rate": 4.624289048298147e-06,
"loss": 0.4369,
"step": 1380
},
{
"epoch": 0.5737365368682684,
"grad_norm": 2.4023558739234225,
"learning_rate": 4.588241282277428e-06,
"loss": 0.4115,
"step": 1385
},
{
"epoch": 0.575807787903894,
"grad_norm": 2.502714953160222,
"learning_rate": 4.55221505205734e-06,
"loss": 0.408,
"step": 1390
},
{
"epoch": 0.5778790389395194,
"grad_norm": 2.4724438182116195,
"learning_rate": 4.516212241881448e-06,
"loss": 0.4281,
"step": 1395
},
{
"epoch": 0.579950289975145,
"grad_norm": 2.630497078397546,
"learning_rate": 4.480234734768393e-06,
"loss": 0.409,
"step": 1400
},
{
"epoch": 0.5820215410107705,
"grad_norm": 2.5092319075068623,
"learning_rate": 4.444284412413418e-06,
"loss": 0.4127,
"step": 1405
},
{
"epoch": 0.584092792046396,
"grad_norm": 2.7865552515532905,
"learning_rate": 4.408363155089952e-06,
"loss": 0.4099,
"step": 1410
},
{
"epoch": 0.5861640430820215,
"grad_norm": 2.462422528338733,
"learning_rate": 4.3724728415512585e-06,
"loss": 0.4228,
"step": 1415
},
{
"epoch": 0.5882352941176471,
"grad_norm": 2.4828719332673503,
"learning_rate": 4.3366153489321855e-06,
"loss": 0.4192,
"step": 1420
},
{
"epoch": 0.5903065451532725,
"grad_norm": 2.483350370620534,
"learning_rate": 4.30079255265098e-06,
"loss": 0.4198,
"step": 1425
},
{
"epoch": 0.5923777961888981,
"grad_norm": 2.5471823188847456,
"learning_rate": 4.265006326311199e-06,
"loss": 0.4086,
"step": 1430
},
{
"epoch": 0.5944490472245236,
"grad_norm": 2.4124476470616596,
"learning_rate": 4.229258541603723e-06,
"loss": 0.4105,
"step": 1435
},
{
"epoch": 0.5965202982601492,
"grad_norm": 2.4133805554148178,
"learning_rate": 4.1935510682088545e-06,
"loss": 0.4098,
"step": 1440
},
{
"epoch": 0.5985915492957746,
"grad_norm": 2.44838319944855,
"learning_rate": 4.157885773698535e-06,
"loss": 0.3979,
"step": 1445
},
{
"epoch": 0.6006628003314002,
"grad_norm": 2.485596270259576,
"learning_rate": 4.122264523438668e-06,
"loss": 0.4032,
"step": 1450
},
{
"epoch": 0.6027340513670257,
"grad_norm": 2.439794587785727,
"learning_rate": 4.086689180491554e-06,
"loss": 0.4066,
"step": 1455
},
{
"epoch": 0.6048053024026512,
"grad_norm": 2.389465339801465,
"learning_rate": 4.051161605518453e-06,
"loss": 0.4007,
"step": 1460
},
{
"epoch": 0.6068765534382767,
"grad_norm": 2.3894667965740832,
"learning_rate": 4.015683656682255e-06,
"loss": 0.3894,
"step": 1465
},
{
"epoch": 0.6089478044739023,
"grad_norm": 2.4148034073879754,
"learning_rate": 3.980257189550316e-06,
"loss": 0.4096,
"step": 1470
},
{
"epoch": 0.6110190555095277,
"grad_norm": 2.4550838501981938,
"learning_rate": 3.94488405699739e-06,
"loss": 0.3932,
"step": 1475
},
{
"epoch": 0.6130903065451533,
"grad_norm": 2.3557776341003147,
"learning_rate": 3.909566109108727e-06,
"loss": 0.4019,
"step": 1480
},
{
"epoch": 0.6151615575807788,
"grad_norm": 2.355520809882233,
"learning_rate": 3.874305193083313e-06,
"loss": 0.3991,
"step": 1485
},
{
"epoch": 0.6172328086164043,
"grad_norm": 2.503937287372334,
"learning_rate": 3.839103153137247e-06,
"loss": 0.3947,
"step": 1490
},
{
"epoch": 0.6193040596520298,
"grad_norm": 2.487168542242338,
"learning_rate": 3.803961830407297e-06,
"loss": 0.3972,
"step": 1495
},
{
"epoch": 0.6213753106876554,
"grad_norm": 2.504277056661968,
"learning_rate": 3.768883062854598e-06,
"loss": 0.388,
"step": 1500
},
{
"epoch": 0.6234465617232808,
"grad_norm": 2.503259597551036,
"learning_rate": 3.7338686851685267e-06,
"loss": 0.4004,
"step": 1505
},
{
"epoch": 0.6255178127589064,
"grad_norm": 2.2995387816904627,
"learning_rate": 3.6989205286707398e-06,
"loss": 0.4009,
"step": 1510
},
{
"epoch": 0.6275890637945319,
"grad_norm": 2.4628767249815313,
"learning_rate": 3.664040421219393e-06,
"loss": 0.3851,
"step": 1515
},
{
"epoch": 0.6296603148301574,
"grad_norm": 2.4028998705697515,
"learning_rate": 3.6292301871135425e-06,
"loss": 0.3972,
"step": 1520
},
{
"epoch": 0.6317315658657829,
"grad_norm": 2.401801780874264,
"learning_rate": 3.59449164699773e-06,
"loss": 0.3948,
"step": 1525
},
{
"epoch": 0.6338028169014085,
"grad_norm": 2.3696780113753793,
"learning_rate": 3.55982661776676e-06,
"loss": 0.3968,
"step": 1530
},
{
"epoch": 0.6358740679370339,
"grad_norm": 2.3030397433363134,
"learning_rate": 3.5252369124706697e-06,
"loss": 0.3863,
"step": 1535
},
{
"epoch": 0.6379453189726595,
"grad_norm": 2.504126684793789,
"learning_rate": 3.4907243402199013e-06,
"loss": 0.3936,
"step": 1540
},
{
"epoch": 0.640016570008285,
"grad_norm": 2.5523957871004015,
"learning_rate": 3.4562907060906908e-06,
"loss": 0.3709,
"step": 1545
},
{
"epoch": 0.6420878210439105,
"grad_norm": 2.3017120628972374,
"learning_rate": 3.4219378110306523e-06,
"loss": 0.3752,
"step": 1550
},
{
"epoch": 0.644159072079536,
"grad_norm": 2.559688464851059,
"learning_rate": 3.3876674517645815e-06,
"loss": 0.3828,
"step": 1555
},
{
"epoch": 0.6462303231151616,
"grad_norm": 2.342011913376553,
"learning_rate": 3.353481420700495e-06,
"loss": 0.3922,
"step": 1560
},
{
"epoch": 0.648301574150787,
"grad_norm": 2.545312173063789,
"learning_rate": 3.319381505835868e-06,
"loss": 0.3816,
"step": 1565
},
{
"epoch": 0.6503728251864126,
"grad_norm": 2.335753689140659,
"learning_rate": 3.285369490664133e-06,
"loss": 0.3903,
"step": 1570
},
{
"epoch": 0.6524440762220381,
"grad_norm": 2.537247293228224,
"learning_rate": 3.251447154081394e-06,
"loss": 0.3827,
"step": 1575
},
{
"epoch": 0.6545153272576636,
"grad_norm": 2.4660063836088204,
"learning_rate": 3.2176162702933816e-06,
"loss": 0.3819,
"step": 1580
},
{
"epoch": 0.6565865782932891,
"grad_norm": 2.424767956890677,
"learning_rate": 3.183878608722669e-06,
"loss": 0.375,
"step": 1585
},
{
"epoch": 0.6586578293289147,
"grad_norm": 2.4445768322421846,
"learning_rate": 3.150235933916115e-06,
"loss": 0.3761,
"step": 1590
},
{
"epoch": 0.6607290803645401,
"grad_norm": 2.4065267541974023,
"learning_rate": 3.1166900054525873e-06,
"loss": 0.3875,
"step": 1595
},
{
"epoch": 0.6628003314001657,
"grad_norm": 2.6400582565507222,
"learning_rate": 3.0832425778509235e-06,
"loss": 0.3712,
"step": 1600
},
{
"epoch": 0.6648715824357913,
"grad_norm": 2.340827833071,
"learning_rate": 3.049895400478174e-06,
"loss": 0.3702,
"step": 1605
},
{
"epoch": 0.6669428334714167,
"grad_norm": 2.3156204896430213,
"learning_rate": 3.0166502174581012e-06,
"loss": 0.3798,
"step": 1610
},
{
"epoch": 0.6690140845070423,
"grad_norm": 2.4083373430976636,
"learning_rate": 2.983508767579956e-06,
"loss": 0.3821,
"step": 1615
},
{
"epoch": 0.6710853355426678,
"grad_norm": 2.4477331076378355,
"learning_rate": 2.950472784207544e-06,
"loss": 0.371,
"step": 1620
},
{
"epoch": 0.6731565865782932,
"grad_norm": 2.478115835624783,
"learning_rate": 2.917543995188562e-06,
"loss": 0.3673,
"step": 1625
},
{
"epoch": 0.6752278376139188,
"grad_norm": 2.378992834820742,
"learning_rate": 2.8847241227642255e-06,
"loss": 0.3758,
"step": 1630
},
{
"epoch": 0.6772990886495444,
"grad_norm": 2.400853068447095,
"learning_rate": 2.852014883479198e-06,
"loss": 0.3861,
"step": 1635
},
{
"epoch": 0.6793703396851698,
"grad_norm": 2.255045212372985,
"learning_rate": 2.819417988091814e-06,
"loss": 0.3548,
"step": 1640
},
{
"epoch": 0.6814415907207954,
"grad_norm": 2.372208019621044,
"learning_rate": 2.786935141484586e-06,
"loss": 0.3739,
"step": 1645
},
{
"epoch": 0.6835128417564209,
"grad_norm": 2.3127468876840673,
"learning_rate": 2.754568042575061e-06,
"loss": 0.3726,
"step": 1650
},
{
"epoch": 0.6855840927920464,
"grad_norm": 2.3802477192557876,
"learning_rate": 2.7223183842269442e-06,
"loss": 0.3676,
"step": 1655
},
{
"epoch": 0.6876553438276719,
"grad_norm": 2.4100679029748506,
"learning_rate": 2.6901878531615677e-06,
"loss": 0.3696,
"step": 1660
},
{
"epoch": 0.6897265948632975,
"grad_norm": 2.3141281544473973,
"learning_rate": 2.658178129869672e-06,
"loss": 0.3654,
"step": 1665
},
{
"epoch": 0.6917978458989229,
"grad_norm": 2.330190053747853,
"learning_rate": 2.6262908885235046e-06,
"loss": 0.3674,
"step": 1670
},
{
"epoch": 0.6938690969345485,
"grad_norm": 2.406593544141906,
"learning_rate": 2.594527796889265e-06,
"loss": 0.3732,
"step": 1675
},
{
"epoch": 0.695940347970174,
"grad_norm": 2.3060033537071014,
"learning_rate": 2.5628905162398797e-06,
"loss": 0.3755,
"step": 1680
},
{
"epoch": 0.6980115990057995,
"grad_norm": 2.2903429406172795,
"learning_rate": 2.531380701268108e-06,
"loss": 0.3728,
"step": 1685
},
{
"epoch": 0.700082850041425,
"grad_norm": 2.318618148906904,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.3635,
"step": 1690
},
{
"epoch": 0.7021541010770506,
"grad_norm": 2.583792008951684,
"learning_rate": 2.4687500537087027e-06,
"loss": 0.3712,
"step": 1695
},
{
"epoch": 0.704225352112676,
"grad_norm": 2.450885698030727,
"learning_rate": 2.4376324968286154e-06,
"loss": 0.3689,
"step": 1700
},
{
"epoch": 0.7062966031483016,
"grad_norm": 2.450057815835257,
"learning_rate": 2.40664895686991e-06,
"loss": 0.3577,
"step": 1705
},
{
"epoch": 0.7083678541839271,
"grad_norm": 2.467490669219595,
"learning_rate": 2.375801054333409e-06,
"loss": 0.3641,
"step": 1710
},
{
"epoch": 0.7104391052195526,
"grad_norm": 2.425436035747948,
"learning_rate": 2.345090402625822e-06,
"loss": 0.3732,
"step": 1715
},
{
"epoch": 0.7125103562551781,
"grad_norm": 2.276164641153553,
"learning_rate": 2.3145186079753685e-06,
"loss": 0.3577,
"step": 1720
},
{
"epoch": 0.7145816072908037,
"grad_norm": 2.476709370280348,
"learning_rate": 2.2840872693477694e-06,
"loss": 0.3583,
"step": 1725
},
{
"epoch": 0.7166528583264291,
"grad_norm": 2.400537878647997,
"learning_rate": 2.253797978362617e-06,
"loss": 0.3702,
"step": 1730
},
{
"epoch": 0.7187241093620547,
"grad_norm": 2.4691432195074556,
"learning_rate": 2.2236523192101264e-06,
"loss": 0.3693,
"step": 1735
},
{
"epoch": 0.7207953603976802,
"grad_norm": 2.2562039371338534,
"learning_rate": 2.193651868568285e-06,
"loss": 0.3438,
"step": 1740
},
{
"epoch": 0.7228666114333057,
"grad_norm": 2.330896067895607,
"learning_rate": 2.16379819552038e-06,
"loss": 0.3588,
"step": 1745
},
{
"epoch": 0.7249378624689312,
"grad_norm": 2.4113082649258177,
"learning_rate": 2.1340928614729445e-06,
"loss": 0.3635,
"step": 1750
},
{
"epoch": 0.7270091135045568,
"grad_norm": 2.3387727892633015,
"learning_rate": 2.1045374200740863e-06,
"loss": 0.365,
"step": 1755
},
{
"epoch": 0.7290803645401823,
"grad_norm": 2.2613900004920833,
"learning_rate": 2.075133417132223e-06,
"loss": 0.3576,
"step": 1760
},
{
"epoch": 0.7311516155758078,
"grad_norm": 2.4609121314188718,
"learning_rate": 2.045882390535248e-06,
"loss": 0.3619,
"step": 1765
},
{
"epoch": 0.7332228666114333,
"grad_norm": 2.3602122614686105,
"learning_rate": 2.016785870170079e-06,
"loss": 0.3539,
"step": 1770
},
{
"epoch": 0.7352941176470589,
"grad_norm": 2.499152077189355,
"learning_rate": 1.987845377842656e-06,
"loss": 0.3482,
"step": 1775
},
{
"epoch": 0.7373653686826843,
"grad_norm": 2.638979206767168,
"learning_rate": 1.9590624271983406e-06,
"loss": 0.3566,
"step": 1780
},
{
"epoch": 0.7394366197183099,
"grad_norm": 2.6474175258765893,
"learning_rate": 1.9304385236427505e-06,
"loss": 0.3534,
"step": 1785
},
{
"epoch": 0.7415078707539354,
"grad_norm": 2.5229724473640855,
"learning_rate": 1.9019751642630252e-06,
"loss": 0.3624,
"step": 1790
},
{
"epoch": 0.7435791217895609,
"grad_norm": 2.4578052419784,
"learning_rate": 1.8736738377495196e-06,
"loss": 0.3558,
"step": 1795
},
{
"epoch": 0.7456503728251864,
"grad_norm": 2.378939474149635,
"learning_rate": 1.8455360243179537e-06,
"loss": 0.3578,
"step": 1800
},
{
"epoch": 0.747721623860812,
"grad_norm": 2.467646645867323,
"learning_rate": 1.8175631956319823e-06,
"loss": 0.3475,
"step": 1805
},
{
"epoch": 0.7497928748964374,
"grad_norm": 2.392811340109784,
"learning_rate": 1.7897568147262323e-06,
"loss": 0.3418,
"step": 1810
},
{
"epoch": 0.751864125932063,
"grad_norm": 2.275851835773417,
"learning_rate": 1.7621183359297817e-06,
"loss": 0.3514,
"step": 1815
},
{
"epoch": 0.7539353769676885,
"grad_norm": 2.2869067582923654,
"learning_rate": 1.7346492047900897e-06,
"loss": 0.3479,
"step": 1820
},
{
"epoch": 0.756006628003314,
"grad_norm": 2.5045250724638723,
"learning_rate": 1.7073508579973996e-06,
"loss": 0.3576,
"step": 1825
},
{
"epoch": 0.7580778790389395,
"grad_norm": 2.429925756949897,
"learning_rate": 1.6802247233095914e-06,
"loss": 0.3421,
"step": 1830
},
{
"epoch": 0.7601491300745651,
"grad_norm": 2.4031855180861212,
"learning_rate": 1.6532722194775108e-06,
"loss": 0.3501,
"step": 1835
},
{
"epoch": 0.7622203811101905,
"grad_norm": 2.3314549850392683,
"learning_rate": 1.626494756170765e-06,
"loss": 0.3369,
"step": 1840
},
{
"epoch": 0.7642916321458161,
"grad_norm": 2.746700588878786,
"learning_rate": 1.5998937339039889e-06,
"loss": 0.3416,
"step": 1845
},
{
"epoch": 0.7663628831814416,
"grad_norm": 2.4445354861344772,
"learning_rate": 1.5734705439636017e-06,
"loss": 0.3393,
"step": 1850
},
{
"epoch": 0.7684341342170671,
"grad_norm": 2.5158277625792005,
"learning_rate": 1.5472265683350397e-06,
"loss": 0.3343,
"step": 1855
},
{
"epoch": 0.7705053852526926,
"grad_norm": 2.405933939237092,
"learning_rate": 1.5211631796304721e-06,
"loss": 0.3536,
"step": 1860
},
{
"epoch": 0.7725766362883182,
"grad_norm": 2.320015008905018,
"learning_rate": 1.495281741017016e-06,
"loss": 0.3367,
"step": 1865
},
{
"epoch": 0.7746478873239436,
"grad_norm": 2.371461313517718,
"learning_rate": 1.46958360614543e-06,
"loss": 0.3481,
"step": 1870
},
{
"epoch": 0.7767191383595692,
"grad_norm": 2.4734969724980274,
"learning_rate": 1.4440701190793278e-06,
"loss": 0.3481,
"step": 1875
},
{
"epoch": 0.7787903893951947,
"grad_norm": 2.498590911272468,
"learning_rate": 1.4187426142248723e-06,
"loss": 0.3486,
"step": 1880
},
{
"epoch": 0.7808616404308202,
"grad_norm": 2.4210457478662124,
"learning_rate": 1.3936024162609897e-06,
"loss": 0.3505,
"step": 1885
},
{
"epoch": 0.7829328914664457,
"grad_norm": 2.27079739762161,
"learning_rate": 1.3686508400700787e-06,
"loss": 0.3414,
"step": 1890
},
{
"epoch": 0.7850041425020713,
"grad_norm": 2.3590027996904306,
"learning_rate": 1.3438891906692447e-06,
"loss": 0.3344,
"step": 1895
},
{
"epoch": 0.7870753935376967,
"grad_norm": 2.374602364344923,
"learning_rate": 1.3193187631420462e-06,
"loss": 0.3429,
"step": 1900
},
{
"epoch": 0.7891466445733223,
"grad_norm": 2.2916676051455,
"learning_rate": 1.2949408425707566e-06,
"loss": 0.3233,
"step": 1905
},
{
"epoch": 0.7912178956089478,
"grad_norm": 2.1300366849765844,
"learning_rate": 1.2707567039691505e-06,
"loss": 0.3289,
"step": 1910
},
{
"epoch": 0.7932891466445733,
"grad_norm": 2.34807621465065,
"learning_rate": 1.2467676122158224e-06,
"loss": 0.352,
"step": 1915
},
{
"epoch": 0.7953603976801988,
"grad_norm": 2.3203613666379783,
"learning_rate": 1.222974821988024e-06,
"loss": 0.3278,
"step": 1920
},
{
"epoch": 0.7974316487158244,
"grad_norm": 2.457411519439677,
"learning_rate": 1.1993795776960498e-06,
"loss": 0.345,
"step": 1925
},
{
"epoch": 0.7995028997514498,
"grad_norm": 2.4363442851068005,
"learning_rate": 1.1759831134181504e-06,
"loss": 0.3403,
"step": 1930
},
{
"epoch": 0.8015741507870754,
"grad_norm": 2.340545464189151,
"learning_rate": 1.1527866528359805e-06,
"loss": 0.3331,
"step": 1935
},
{
"epoch": 0.8036454018227009,
"grad_norm": 2.4620962627646326,
"learning_rate": 1.1297914091706086e-06,
"loss": 0.3385,
"step": 1940
},
{
"epoch": 0.8057166528583264,
"grad_norm": 2.2807194426974373,
"learning_rate": 1.1069985851190524e-06,
"loss": 0.3387,
"step": 1945
},
{
"epoch": 0.8077879038939519,
"grad_norm": 2.433026421478007,
"learning_rate": 1.0844093727913868e-06,
"loss": 0.3177,
"step": 1950
},
{
"epoch": 0.8098591549295775,
"grad_norm": 2.204766152464149,
"learning_rate": 1.062024953648384e-06,
"loss": 0.3243,
"step": 1955
},
{
"epoch": 0.8119304059652029,
"grad_norm": 2.3848362304936357,
"learning_rate": 1.039846498439727e-06,
"loss": 0.3289,
"step": 1960
},
{
"epoch": 0.8140016570008285,
"grad_norm": 2.382898053034431,
"learning_rate": 1.0178751671427755e-06,
"loss": 0.332,
"step": 1965
},
{
"epoch": 0.816072908036454,
"grad_norm": 2.3611571385246526,
"learning_rate": 9.961121089018933e-07,
"loss": 0.3332,
"step": 1970
},
{
"epoch": 0.8181441590720795,
"grad_norm": 2.557662051747092,
"learning_rate": 9.745584619683524e-07,
"loss": 0.3306,
"step": 1975
},
{
"epoch": 0.820215410107705,
"grad_norm": 2.397306089083263,
"learning_rate": 9.532153536407923e-07,
"loss": 0.3258,
"step": 1980
},
{
"epoch": 0.8222866611433306,
"grad_norm": 2.41185664034236,
"learning_rate": 9.320839002062682e-07,
"loss": 0.3397,
"step": 1985
},
{
"epoch": 0.824357912178956,
"grad_norm": 2.295751060263886,
"learning_rate": 9.111652068818621e-07,
"loss": 0.3391,
"step": 1990
},
{
"epoch": 0.8264291632145816,
"grad_norm": 2.504888586749888,
"learning_rate": 8.904603677568785e-07,
"loss": 0.3356,
"step": 1995
},
{
"epoch": 0.8285004142502072,
"grad_norm": 2.4202401247094927,
"learning_rate": 8.699704657356195e-07,
"loss": 0.321,
"step": 2000
},
{
"epoch": 0.8305716652858326,
"grad_norm": 2.307984596975984,
"learning_rate": 8.496965724807516e-07,
"loss": 0.3325,
"step": 2005
},
{
"epoch": 0.8326429163214581,
"grad_norm": 2.295235220088202,
"learning_rate": 8.296397483572515e-07,
"loss": 0.3372,
"step": 2010
},
{
"epoch": 0.8347141673570837,
"grad_norm": 2.2076463931848362,
"learning_rate": 8.098010423769503e-07,
"loss": 0.32,
"step": 2015
},
{
"epoch": 0.8367854183927091,
"grad_norm": 2.5220317282827445,
"learning_rate": 7.901814921436624e-07,
"loss": 0.329,
"step": 2020
},
{
"epoch": 0.8388566694283347,
"grad_norm": 2.333027676634786,
"learning_rate": 7.70782123798921e-07,
"loss": 0.3348,
"step": 2025
},
{
"epoch": 0.8409279204639603,
"grad_norm": 2.510269587950094,
"learning_rate": 7.516039519683105e-07,
"loss": 0.3341,
"step": 2030
},
{
"epoch": 0.8429991714995857,
"grad_norm": 2.3997783375586885,
"learning_rate": 7.326479797083963e-07,
"loss": 0.3357,
"step": 2035
},
{
"epoch": 0.8450704225352113,
"grad_norm": 2.297591159261384,
"learning_rate": 7.139151984542636e-07,
"loss": 0.3343,
"step": 2040
},
{
"epoch": 0.8471416735708368,
"grad_norm": 2.3763676660234605,
"learning_rate": 6.954065879676653e-07,
"loss": 0.3263,
"step": 2045
},
{
"epoch": 0.8492129246064622,
"grad_norm": 2.274408055139544,
"learning_rate": 6.771231162857722e-07,
"loss": 0.3194,
"step": 2050
},
{
"epoch": 0.8512841756420878,
"grad_norm": 2.326030943789442,
"learning_rate": 6.590657396705525e-07,
"loss": 0.3274,
"step": 2055
},
{
"epoch": 0.8533554266777134,
"grad_norm": 2.343414027503624,
"learning_rate": 6.412354025587509e-07,
"loss": 0.3214,
"step": 2060
},
{
"epoch": 0.8554266777133389,
"grad_norm": 2.268590241297634,
"learning_rate": 6.236330375124921e-07,
"loss": 0.3258,
"step": 2065
},
{
"epoch": 0.8574979287489644,
"grad_norm": 2.299226303968449,
"learning_rate": 6.062595651705111e-07,
"loss": 0.3184,
"step": 2070
},
{
"epoch": 0.8595691797845899,
"grad_norm": 2.319340602566474,
"learning_rate": 5.891158941999959e-07,
"loss": 0.3273,
"step": 2075
},
{
"epoch": 0.8616404308202155,
"grad_norm": 2.23014016114103,
"learning_rate": 5.722029212490666e-07,
"loss": 0.3251,
"step": 2080
},
{
"epoch": 0.8637116818558409,
"grad_norm": 2.228713177211921,
"learning_rate": 5.555215308998779e-07,
"loss": 0.3194,
"step": 2085
},
{
"epoch": 0.8657829328914665,
"grad_norm": 2.4149113998611753,
"learning_rate": 5.390725956223531e-07,
"loss": 0.3127,
"step": 2090
},
{
"epoch": 0.867854183927092,
"grad_norm": 2.31201046602113,
"learning_rate": 5.22856975728554e-07,
"loss": 0.3205,
"step": 2095
},
{
"epoch": 0.8699254349627175,
"grad_norm": 2.3321678684665983,
"learning_rate": 5.068755193276798e-07,
"loss": 0.3206,
"step": 2100
},
{
"epoch": 0.871996685998343,
"grad_norm": 2.3618505787616155,
"learning_rate": 4.911290622817161e-07,
"loss": 0.3351,
"step": 2105
},
{
"epoch": 0.8740679370339686,
"grad_norm": 2.496406107634527,
"learning_rate": 4.756184281617121e-07,
"loss": 0.3277,
"step": 2110
},
{
"epoch": 0.876139188069594,
"grad_norm": 2.5051865056946343,
"learning_rate": 4.6034442820471037e-07,
"loss": 0.3302,
"step": 2115
},
{
"epoch": 0.8782104391052196,
"grad_norm": 2.434716634912814,
"learning_rate": 4.4530786127131575e-07,
"loss": 0.3309,
"step": 2120
},
{
"epoch": 0.8802816901408451,
"grad_norm": 2.450158523247205,
"learning_rate": 4.305095138039106e-07,
"loss": 0.3219,
"step": 2125
},
{
"epoch": 0.8823529411764706,
"grad_norm": 2.421404693121894,
"learning_rate": 4.159501597855287e-07,
"loss": 0.3151,
"step": 2130
},
{
"epoch": 0.8844241922120961,
"grad_norm": 2.352690164019225,
"learning_rate": 4.0163056069936757e-07,
"loss": 0.3287,
"step": 2135
},
{
"epoch": 0.8864954432477217,
"grad_norm": 2.6282570060805015,
"learning_rate": 3.8755146548896784e-07,
"loss": 0.3302,
"step": 2140
},
{
"epoch": 0.8885666942833471,
"grad_norm": 2.367645825806905,
"learning_rate": 3.737136105190337e-07,
"loss": 0.3243,
"step": 2145
},
{
"epoch": 0.8906379453189727,
"grad_norm": 2.59816978411163,
"learning_rate": 3.6011771953693044e-07,
"loss": 0.3196,
"step": 2150
},
{
"epoch": 0.8927091963545982,
"grad_norm": 2.425575544989018,
"learning_rate": 3.4676450363481937e-07,
"loss": 0.3238,
"step": 2155
},
{
"epoch": 0.8947804473902237,
"grad_norm": 2.192389337916798,
"learning_rate": 3.336546612124758e-07,
"loss": 0.3138,
"step": 2160
},
{
"epoch": 0.8968516984258492,
"grad_norm": 2.4581010644364434,
"learning_rate": 3.20788877940757e-07,
"loss": 0.3188,
"step": 2165
},
{
"epoch": 0.8989229494614748,
"grad_norm": 2.450187860216197,
"learning_rate": 3.081678267257404e-07,
"loss": 0.3171,
"step": 2170
},
{
"epoch": 0.9009942004971002,
"grad_norm": 2.3936000969068902,
"learning_rate": 2.9579216767352815e-07,
"loss": 0.3152,
"step": 2175
},
{
"epoch": 0.9030654515327258,
"grad_norm": 2.452612362284696,
"learning_rate": 2.836625480557265e-07,
"loss": 0.319,
"step": 2180
},
{
"epoch": 0.9051367025683513,
"grad_norm": 2.3271669592010116,
"learning_rate": 2.7177960227558863e-07,
"loss": 0.3164,
"step": 2185
},
{
"epoch": 0.9072079536039768,
"grad_norm": 2.3614252712491863,
"learning_rate": 2.601439518348331e-07,
"loss": 0.318,
"step": 2190
},
{
"epoch": 0.9092792046396023,
"grad_norm": 2.1964726679308444,
"learning_rate": 2.487562053011422e-07,
"loss": 0.3205,
"step": 2195
},
{
"epoch": 0.9113504556752279,
"grad_norm": 2.219387790222263,
"learning_rate": 2.376169582763288e-07,
"loss": 0.3182,
"step": 2200
},
{
"epoch": 0.9134217067108533,
"grad_norm": 2.211089888272885,
"learning_rate": 2.2672679336518789e-07,
"loss": 0.3187,
"step": 2205
},
{
"epoch": 0.9154929577464789,
"grad_norm": 2.332902382859145,
"learning_rate": 2.1608628014502364e-07,
"loss": 0.3248,
"step": 2210
},
{
"epoch": 0.9175642087821044,
"grad_norm": 2.5038230064732665,
"learning_rate": 2.0569597513586004e-07,
"loss": 0.319,
"step": 2215
},
{
"epoch": 0.9196354598177299,
"grad_norm": 2.4550598502748953,
"learning_rate": 1.955564217713335e-07,
"loss": 0.3135,
"step": 2220
},
{
"epoch": 0.9217067108533554,
"grad_norm": 2.618416462513135,
"learning_rate": 1.8566815037026897e-07,
"loss": 0.325,
"step": 2225
},
{
"epoch": 0.923777961888981,
"grad_norm": 2.3741506182685264,
"learning_rate": 1.7603167810894662e-07,
"loss": 0.3274,
"step": 2230
},
{
"epoch": 0.9258492129246064,
"grad_norm": 2.396530079213212,
"learning_rate": 1.6664750899404892e-07,
"loss": 0.3107,
"step": 2235
},
{
"epoch": 0.927920463960232,
"grad_norm": 2.3232747321047733,
"learning_rate": 1.5751613383630128e-07,
"loss": 0.315,
"step": 2240
},
{
"epoch": 0.9299917149958575,
"grad_norm": 2.4194051550958027,
"learning_rate": 1.4863803022480362e-07,
"loss": 0.3279,
"step": 2245
},
{
"epoch": 0.932062966031483,
"grad_norm": 2.410482682790985,
"learning_rate": 1.4001366250204762e-07,
"loss": 0.319,
"step": 2250
},
{
"epoch": 0.9341342170671085,
"grad_norm": 2.3790218958415177,
"learning_rate": 1.3164348173963392e-07,
"loss": 0.3202,
"step": 2255
},
{
"epoch": 0.9362054681027341,
"grad_norm": 2.278784018257007,
"learning_rate": 1.235279257146804e-07,
"loss": 0.3214,
"step": 2260
},
{
"epoch": 0.9382767191383595,
"grad_norm": 2.567614509309173,
"learning_rate": 1.1566741888692168e-07,
"loss": 0.3277,
"step": 2265
},
{
"epoch": 0.9403479701739851,
"grad_norm": 2.351104210359246,
"learning_rate": 1.080623723765134e-07,
"loss": 0.3206,
"step": 2270
},
{
"epoch": 0.9424192212096106,
"grad_norm": 2.3029933899655877,
"learning_rate": 1.0071318394252849e-07,
"loss": 0.3129,
"step": 2275
},
{
"epoch": 0.9444904722452361,
"grad_norm": 2.4691002097043895,
"learning_rate": 9.362023796215036e-08,
"loss": 0.3211,
"step": 2280
},
{
"epoch": 0.9465617232808616,
"grad_norm": 2.3679710372577847,
"learning_rate": 8.678390541057512e-08,
"loss": 0.313,
"step": 2285
},
{
"epoch": 0.9486329743164872,
"grad_norm": 2.5040520456195234,
"learning_rate": 8.020454384160437e-08,
"loss": 0.3173,
"step": 2290
},
{
"epoch": 0.9507042253521126,
"grad_norm": 2.2587473827290263,
"learning_rate": 7.388249736894615e-08,
"loss": 0.3123,
"step": 2295
},
{
"epoch": 0.9527754763877382,
"grad_norm": 2.3523433019856976,
"learning_rate": 6.78180966482156e-08,
"loss": 0.3134,
"step": 2300
},
{
"epoch": 0.9548467274233637,
"grad_norm": 2.3711725145374305,
"learning_rate": 6.201165885964311e-08,
"loss": 0.3243,
"step": 2305
},
{
"epoch": 0.9569179784589892,
"grad_norm": 2.490897999507605,
"learning_rate": 5.646348769148491e-08,
"loss": 0.3164,
"step": 2310
},
{
"epoch": 0.9589892294946147,
"grad_norm": 2.411849128078695,
"learning_rate": 5.117387332413737e-08,
"loss": 0.3203,
"step": 2315
},
{
"epoch": 0.9610604805302403,
"grad_norm": 2.2961096086453514,
"learning_rate": 4.6143092414961396e-08,
"loss": 0.2994,
"step": 2320
},
{
"epoch": 0.9631317315658657,
"grad_norm": 2.4739902376996405,
"learning_rate": 4.1371408083815126e-08,
"loss": 0.3265,
"step": 2325
},
{
"epoch": 0.9652029826014913,
"grad_norm": 2.4945733973368274,
"learning_rate": 3.685906989928656e-08,
"loss": 0.3101,
"step": 2330
},
{
"epoch": 0.9672742336371168,
"grad_norm": 2.3837106401498485,
"learning_rate": 3.2606313865646276e-08,
"loss": 0.3112,
"step": 2335
},
{
"epoch": 0.9693454846727423,
"grad_norm": 2.3694096519216297,
"learning_rate": 2.861336241050061e-08,
"loss": 0.313,
"step": 2340
},
{
"epoch": 0.9714167357083678,
"grad_norm": 2.3987583338206577,
"learning_rate": 2.488042437315985e-08,
"loss": 0.3239,
"step": 2345
},
{
"epoch": 0.9734879867439934,
"grad_norm": 2.370749504361077,
"learning_rate": 2.1407694993714755e-08,
"loss": 0.3045,
"step": 2350
},
{
"epoch": 0.9755592377796188,
"grad_norm": 2.3065291126618943,
"learning_rate": 1.8195355902824174e-08,
"loss": 0.3232,
"step": 2355
},
{
"epoch": 0.9776304888152444,
"grad_norm": 2.4870976284021307,
"learning_rate": 1.5243575112218744e-08,
"loss": 0.3172,
"step": 2360
},
{
"epoch": 0.97970173985087,
"grad_norm": 2.290255184865737,
"learning_rate": 1.2552507005909042e-08,
"loss": 0.3158,
"step": 2365
},
{
"epoch": 0.9817729908864954,
"grad_norm": 2.382468298424662,
"learning_rate": 1.0122292332114814e-08,
"loss": 0.3151,
"step": 2370
},
{
"epoch": 0.9838442419221209,
"grad_norm": 2.357004572247049,
"learning_rate": 7.953058195900864e-09,
"loss": 0.3229,
"step": 2375
},
{
"epoch": 0.9859154929577465,
"grad_norm": 2.2427558508252843,
"learning_rate": 6.044918052531268e-09,
"loss": 0.3099,
"step": 2380
},
{
"epoch": 0.987986743993372,
"grad_norm": 2.3295263573420555,
"learning_rate": 4.397971701533554e-09,
"loss": 0.3156,
"step": 2385
},
{
"epoch": 0.9900579950289975,
"grad_norm": 2.2718544101734635,
"learning_rate": 3.0123052814812203e-09,
"loss": 0.316,
"step": 2390
},
{
"epoch": 0.992129246064623,
"grad_norm": 2.3606235616126976,
"learning_rate": 1.887991265486222e-09,
"loss": 0.3108,
"step": 2395
},
{
"epoch": 0.9942004971002486,
"grad_norm": 2.3142676503036657,
"learning_rate": 1.025088457409229e-09,
"loss": 0.3189,
"step": 2400
},
{
"epoch": 0.996271748135874,
"grad_norm": 2.3866448821206276,
"learning_rate": 4.2364198878597216e-10,
"loss": 0.3175,
"step": 2405
},
{
"epoch": 0.9983429991714996,
"grad_norm": 2.3282211619772304,
"learning_rate": 8.368331646302353e-11,
"loss": 0.318,
"step": 2410
},
{
"epoch": 1.0,
"eval_runtime": 3.5501,
"eval_samples_per_second": 2.817,
"eval_steps_per_second": 0.845,
"step": 2414
},
{
"epoch": 1.0,
"step": 2414,
"total_flos": 252721244405760.0,
"train_loss": 0.5123422154732351,
"train_runtime": 21578.342,
"train_samples_per_second": 1.789,
"train_steps_per_second": 0.112
}
],
"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
}