terry69's picture
Model save
f586159 verified
raw
history blame contribute delete
No virus
66 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1869,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005350454788657035,
"grad_norm": 23.32975499538448,
"learning_rate": 5.3475935828877005e-08,
"loss": 1.3233,
"step": 1
},
{
"epoch": 0.002675227394328518,
"grad_norm": 22.00335872083755,
"learning_rate": 2.6737967914438503e-07,
"loss": 1.3186,
"step": 5
},
{
"epoch": 0.005350454788657036,
"grad_norm": 15.4065659696612,
"learning_rate": 5.347593582887701e-07,
"loss": 1.2822,
"step": 10
},
{
"epoch": 0.008025682182985553,
"grad_norm": 11.502912896830692,
"learning_rate": 8.021390374331551e-07,
"loss": 1.1274,
"step": 15
},
{
"epoch": 0.010700909577314071,
"grad_norm": 9.500638744694992,
"learning_rate": 1.0695187165775401e-06,
"loss": 1.0332,
"step": 20
},
{
"epoch": 0.01337613697164259,
"grad_norm": 3.5746284262621084,
"learning_rate": 1.3368983957219254e-06,
"loss": 0.908,
"step": 25
},
{
"epoch": 0.016051364365971106,
"grad_norm": 3.476299775893293,
"learning_rate": 1.6042780748663103e-06,
"loss": 0.8905,
"step": 30
},
{
"epoch": 0.018726591760299626,
"grad_norm": 2.9586982678137983,
"learning_rate": 1.8716577540106954e-06,
"loss": 0.8566,
"step": 35
},
{
"epoch": 0.021401819154628143,
"grad_norm": 2.921422317023882,
"learning_rate": 2.1390374331550802e-06,
"loss": 0.8463,
"step": 40
},
{
"epoch": 0.024077046548956663,
"grad_norm": 2.8850607045290033,
"learning_rate": 2.4064171122994653e-06,
"loss": 0.8306,
"step": 45
},
{
"epoch": 0.02675227394328518,
"grad_norm": 3.204350090952928,
"learning_rate": 2.673796791443851e-06,
"loss": 0.813,
"step": 50
},
{
"epoch": 0.029427501337613696,
"grad_norm": 3.082541545860967,
"learning_rate": 2.9411764705882355e-06,
"loss": 0.8103,
"step": 55
},
{
"epoch": 0.03210272873194221,
"grad_norm": 3.03497970770737,
"learning_rate": 3.2085561497326205e-06,
"loss": 0.7921,
"step": 60
},
{
"epoch": 0.034777956126270736,
"grad_norm": 3.102342967951045,
"learning_rate": 3.4759358288770056e-06,
"loss": 0.7957,
"step": 65
},
{
"epoch": 0.03745318352059925,
"grad_norm": 2.9578938537372044,
"learning_rate": 3.7433155080213907e-06,
"loss": 0.7839,
"step": 70
},
{
"epoch": 0.04012841091492777,
"grad_norm": 3.0373015894615283,
"learning_rate": 4.010695187165775e-06,
"loss": 0.773,
"step": 75
},
{
"epoch": 0.042803638309256285,
"grad_norm": 3.076413211099657,
"learning_rate": 4.2780748663101604e-06,
"loss": 0.7879,
"step": 80
},
{
"epoch": 0.0454788657035848,
"grad_norm": 3.116251894957883,
"learning_rate": 4.5454545454545455e-06,
"loss": 0.7564,
"step": 85
},
{
"epoch": 0.048154093097913325,
"grad_norm": 3.293157475867486,
"learning_rate": 4.812834224598931e-06,
"loss": 0.7447,
"step": 90
},
{
"epoch": 0.05082932049224184,
"grad_norm": 2.9781238058172015,
"learning_rate": 5.0802139037433165e-06,
"loss": 0.7333,
"step": 95
},
{
"epoch": 0.05350454788657036,
"grad_norm": 2.954596587676888,
"learning_rate": 5.347593582887702e-06,
"loss": 0.7373,
"step": 100
},
{
"epoch": 0.056179775280898875,
"grad_norm": 2.8773857530480833,
"learning_rate": 5.614973262032086e-06,
"loss": 0.7303,
"step": 105
},
{
"epoch": 0.05885500267522739,
"grad_norm": 2.9567813017811084,
"learning_rate": 5.882352941176471e-06,
"loss": 0.7279,
"step": 110
},
{
"epoch": 0.061530230069555915,
"grad_norm": 2.9858797822378955,
"learning_rate": 6.149732620320856e-06,
"loss": 0.7214,
"step": 115
},
{
"epoch": 0.06420545746388442,
"grad_norm": 3.012701625430249,
"learning_rate": 6.417112299465241e-06,
"loss": 0.717,
"step": 120
},
{
"epoch": 0.06688068485821295,
"grad_norm": 2.9708435052571955,
"learning_rate": 6.684491978609626e-06,
"loss": 0.7248,
"step": 125
},
{
"epoch": 0.06955591225254147,
"grad_norm": 2.8861878079531236,
"learning_rate": 6.951871657754011e-06,
"loss": 0.72,
"step": 130
},
{
"epoch": 0.07223113964686999,
"grad_norm": 2.887798712701533,
"learning_rate": 7.219251336898396e-06,
"loss": 0.7172,
"step": 135
},
{
"epoch": 0.0749063670411985,
"grad_norm": 2.9349544237124094,
"learning_rate": 7.486631016042781e-06,
"loss": 0.711,
"step": 140
},
{
"epoch": 0.07758159443552702,
"grad_norm": 3.0198425274374943,
"learning_rate": 7.754010695187166e-06,
"loss": 0.7007,
"step": 145
},
{
"epoch": 0.08025682182985554,
"grad_norm": 2.844285698601967,
"learning_rate": 8.02139037433155e-06,
"loss": 0.7163,
"step": 150
},
{
"epoch": 0.08293204922418405,
"grad_norm": 2.906658583579405,
"learning_rate": 8.288770053475937e-06,
"loss": 0.7135,
"step": 155
},
{
"epoch": 0.08560727661851257,
"grad_norm": 2.775520126426031,
"learning_rate": 8.556149732620321e-06,
"loss": 0.6986,
"step": 160
},
{
"epoch": 0.08828250401284109,
"grad_norm": 2.69467913452044,
"learning_rate": 8.823529411764707e-06,
"loss": 0.6965,
"step": 165
},
{
"epoch": 0.0909577314071696,
"grad_norm": 2.7282250830391357,
"learning_rate": 9.090909090909091e-06,
"loss": 0.7034,
"step": 170
},
{
"epoch": 0.09363295880149813,
"grad_norm": 2.892650307618601,
"learning_rate": 9.358288770053477e-06,
"loss": 0.7118,
"step": 175
},
{
"epoch": 0.09630818619582665,
"grad_norm": 2.6740856111003164,
"learning_rate": 9.625668449197861e-06,
"loss": 0.6921,
"step": 180
},
{
"epoch": 0.09898341359015517,
"grad_norm": 2.833027188962373,
"learning_rate": 9.893048128342247e-06,
"loss": 0.7014,
"step": 185
},
{
"epoch": 0.10165864098448368,
"grad_norm": 3.0007083274918993,
"learning_rate": 9.999921507322408e-06,
"loss": 0.6943,
"step": 190
},
{
"epoch": 0.1043338683788122,
"grad_norm": 2.699937414637709,
"learning_rate": 9.999441838772916e-06,
"loss": 0.6879,
"step": 195
},
{
"epoch": 0.10700909577314072,
"grad_norm": 2.8840558663751246,
"learning_rate": 9.99852615049999e-06,
"loss": 0.6934,
"step": 200
},
{
"epoch": 0.10968432316746923,
"grad_norm": 2.6349082313416026,
"learning_rate": 9.997174522364177e-06,
"loss": 0.6857,
"step": 205
},
{
"epoch": 0.11235955056179775,
"grad_norm": 2.7831287150153137,
"learning_rate": 9.995387072245939e-06,
"loss": 0.7051,
"step": 210
},
{
"epoch": 0.11503477795612627,
"grad_norm": 2.869759215432631,
"learning_rate": 9.993163956035381e-06,
"loss": 0.698,
"step": 215
},
{
"epoch": 0.11771000535045478,
"grad_norm": 2.6687295223771885,
"learning_rate": 9.990505367618647e-06,
"loss": 0.6688,
"step": 220
},
{
"epoch": 0.12038523274478331,
"grad_norm": 2.8932052543699442,
"learning_rate": 9.987411538861023e-06,
"loss": 0.6748,
"step": 225
},
{
"epoch": 0.12306046013911183,
"grad_norm": 2.6718609572340806,
"learning_rate": 9.9838827395867e-06,
"loss": 0.6905,
"step": 230
},
{
"epoch": 0.12573568753344033,
"grad_norm": 2.812043760617457,
"learning_rate": 9.979919277555247e-06,
"loss": 0.6768,
"step": 235
},
{
"epoch": 0.12841091492776885,
"grad_norm": 3.211381659481926,
"learning_rate": 9.97552149843478e-06,
"loss": 0.6787,
"step": 240
},
{
"epoch": 0.13108614232209737,
"grad_norm": 3.1485034799180593,
"learning_rate": 9.970689785771798e-06,
"loss": 0.6815,
"step": 245
},
{
"epoch": 0.1337613697164259,
"grad_norm": 2.7247090563062417,
"learning_rate": 9.96542456095775e-06,
"loss": 0.6819,
"step": 250
},
{
"epoch": 0.13643659711075443,
"grad_norm": 2.7085417921956387,
"learning_rate": 9.95972628319227e-06,
"loss": 0.6681,
"step": 255
},
{
"epoch": 0.13911182450508294,
"grad_norm": 2.691457151596662,
"learning_rate": 9.953595449443134e-06,
"loss": 0.6579,
"step": 260
},
{
"epoch": 0.14178705189941146,
"grad_norm": 2.584599998752218,
"learning_rate": 9.947032594402917e-06,
"loss": 0.679,
"step": 265
},
{
"epoch": 0.14446227929373998,
"grad_norm": 2.6045453255332554,
"learning_rate": 9.940038290442362e-06,
"loss": 0.6556,
"step": 270
},
{
"epoch": 0.1471375066880685,
"grad_norm": 2.4412081165922848,
"learning_rate": 9.932613147560464e-06,
"loss": 0.6703,
"step": 275
},
{
"epoch": 0.149812734082397,
"grad_norm": 2.5230097371528437,
"learning_rate": 9.924757813331256e-06,
"loss": 0.6744,
"step": 280
},
{
"epoch": 0.15248796147672553,
"grad_norm": 2.585452773043538,
"learning_rate": 9.916472972847353e-06,
"loss": 0.669,
"step": 285
},
{
"epoch": 0.15516318887105404,
"grad_norm": 2.645498442811453,
"learning_rate": 9.907759348660186e-06,
"loss": 0.6398,
"step": 290
},
{
"epoch": 0.15783841626538256,
"grad_norm": 2.6829258576996553,
"learning_rate": 9.898617700716988e-06,
"loss": 0.6721,
"step": 295
},
{
"epoch": 0.16051364365971107,
"grad_norm": 2.365791638696942,
"learning_rate": 9.889048826294527e-06,
"loss": 0.6501,
"step": 300
},
{
"epoch": 0.1631888710540396,
"grad_norm": 2.6788731378294535,
"learning_rate": 9.879053559929556e-06,
"loss": 0.6633,
"step": 305
},
{
"epoch": 0.1658640984483681,
"grad_norm": 2.655949788782083,
"learning_rate": 9.868632773346044e-06,
"loss": 0.6512,
"step": 310
},
{
"epoch": 0.16853932584269662,
"grad_norm": 2.525125627328949,
"learning_rate": 9.857787375379144e-06,
"loss": 0.6541,
"step": 315
},
{
"epoch": 0.17121455323702514,
"grad_norm": 2.467491746753525,
"learning_rate": 9.84651831189593e-06,
"loss": 0.6579,
"step": 320
},
{
"epoch": 0.17388978063135366,
"grad_norm": 2.516277938731903,
"learning_rate": 9.834826565712901e-06,
"loss": 0.6577,
"step": 325
},
{
"epoch": 0.17656500802568217,
"grad_norm": 2.397791799699944,
"learning_rate": 9.822713156510278e-06,
"loss": 0.6496,
"step": 330
},
{
"epoch": 0.1792402354200107,
"grad_norm": 2.5057966279445854,
"learning_rate": 9.81017914074306e-06,
"loss": 0.6517,
"step": 335
},
{
"epoch": 0.1819154628143392,
"grad_norm": 2.4923621059461105,
"learning_rate": 9.797225611548896e-06,
"loss": 0.6688,
"step": 340
},
{
"epoch": 0.18459069020866772,
"grad_norm": 2.5049630606580613,
"learning_rate": 9.783853698652737e-06,
"loss": 0.6607,
"step": 345
},
{
"epoch": 0.18726591760299627,
"grad_norm": 2.358221426077587,
"learning_rate": 9.770064568268329e-06,
"loss": 0.652,
"step": 350
},
{
"epoch": 0.18994114499732478,
"grad_norm": 2.6263829483361367,
"learning_rate": 9.75585942299648e-06,
"loss": 0.6447,
"step": 355
},
{
"epoch": 0.1926163723916533,
"grad_norm": 2.4355704032991707,
"learning_rate": 9.741239501720197e-06,
"loss": 0.6447,
"step": 360
},
{
"epoch": 0.19529159978598182,
"grad_norm": 2.4808141511939796,
"learning_rate": 9.726206079496619e-06,
"loss": 0.6562,
"step": 365
},
{
"epoch": 0.19796682718031033,
"grad_norm": 2.7306912925058127,
"learning_rate": 9.71076046744583e-06,
"loss": 0.6355,
"step": 370
},
{
"epoch": 0.20064205457463885,
"grad_norm": 2.4644682995293317,
"learning_rate": 9.694904012636509e-06,
"loss": 0.6333,
"step": 375
},
{
"epoch": 0.20331728196896737,
"grad_norm": 2.4720584580787865,
"learning_rate": 9.678638097968435e-06,
"loss": 0.6195,
"step": 380
},
{
"epoch": 0.20599250936329588,
"grad_norm": 2.3633575413378503,
"learning_rate": 9.661964142051896e-06,
"loss": 0.6409,
"step": 385
},
{
"epoch": 0.2086677367576244,
"grad_norm": 2.3057026800286713,
"learning_rate": 9.644883599083959e-06,
"loss": 0.6246,
"step": 390
},
{
"epoch": 0.21134296415195292,
"grad_norm": 2.4532088121133246,
"learning_rate": 9.627397958721638e-06,
"loss": 0.642,
"step": 395
},
{
"epoch": 0.21401819154628143,
"grad_norm": 2.332423943377658,
"learning_rate": 9.609508745951988e-06,
"loss": 0.6226,
"step": 400
},
{
"epoch": 0.21669341894060995,
"grad_norm": 2.4978626441486833,
"learning_rate": 9.591217520959095e-06,
"loss": 0.6133,
"step": 405
},
{
"epoch": 0.21936864633493847,
"grad_norm": 2.2954192266671383,
"learning_rate": 9.572525878988014e-06,
"loss": 0.6148,
"step": 410
},
{
"epoch": 0.22204387372926698,
"grad_norm": 2.8095485182952564,
"learning_rate": 9.55343545020564e-06,
"loss": 0.6268,
"step": 415
},
{
"epoch": 0.2247191011235955,
"grad_norm": 2.3665476960634195,
"learning_rate": 9.533947899558521e-06,
"loss": 0.6244,
"step": 420
},
{
"epoch": 0.22739432851792402,
"grad_norm": 2.6753913761867523,
"learning_rate": 9.514064926627684e-06,
"loss": 0.6323,
"step": 425
},
{
"epoch": 0.23006955591225253,
"grad_norm": 2.5112745956277926,
"learning_rate": 9.49378826548037e-06,
"loss": 0.6195,
"step": 430
},
{
"epoch": 0.23274478330658105,
"grad_norm": 2.413468337663542,
"learning_rate": 9.473119684518834e-06,
"loss": 0.6074,
"step": 435
},
{
"epoch": 0.23542001070090957,
"grad_norm": 2.6553211159808567,
"learning_rate": 9.452060986326088e-06,
"loss": 0.6241,
"step": 440
},
{
"epoch": 0.23809523809523808,
"grad_norm": 2.659385798106578,
"learning_rate": 9.430614007508712e-06,
"loss": 0.6158,
"step": 445
},
{
"epoch": 0.24077046548956663,
"grad_norm": 2.2931792547740466,
"learning_rate": 9.408780618536664e-06,
"loss": 0.6018,
"step": 450
},
{
"epoch": 0.24344569288389514,
"grad_norm": 2.4514111664427043,
"learning_rate": 9.386562723580155e-06,
"loss": 0.6158,
"step": 455
},
{
"epoch": 0.24612092027822366,
"grad_norm": 2.420489212347836,
"learning_rate": 9.363962260343577e-06,
"loss": 0.5934,
"step": 460
},
{
"epoch": 0.24879614767255218,
"grad_norm": 2.5033351510386055,
"learning_rate": 9.340981199896515e-06,
"loss": 0.5992,
"step": 465
},
{
"epoch": 0.25147137506688066,
"grad_norm": 2.601071476421057,
"learning_rate": 9.317621546501827e-06,
"loss": 0.5921,
"step": 470
},
{
"epoch": 0.2541466024612092,
"grad_norm": 2.580094514637511,
"learning_rate": 9.293885337440869e-06,
"loss": 0.5966,
"step": 475
},
{
"epoch": 0.2568218298555377,
"grad_norm": 2.443217288458462,
"learning_rate": 9.26977464283579e-06,
"loss": 0.5945,
"step": 480
},
{
"epoch": 0.2594970572498662,
"grad_norm": 2.4281259437423177,
"learning_rate": 9.245291565469007e-06,
"loss": 0.5974,
"step": 485
},
{
"epoch": 0.26217228464419473,
"grad_norm": 2.6683703968935135,
"learning_rate": 9.220438240599813e-06,
"loss": 0.6115,
"step": 490
},
{
"epoch": 0.26484751203852325,
"grad_norm": 2.447016408813491,
"learning_rate": 9.19521683577814e-06,
"loss": 0.5878,
"step": 495
},
{
"epoch": 0.2675227394328518,
"grad_norm": 2.355863764903897,
"learning_rate": 9.169629550655532e-06,
"loss": 0.5758,
"step": 500
},
{
"epoch": 0.27019796682718034,
"grad_norm": 2.4204699917498806,
"learning_rate": 9.143678616793299e-06,
"loss": 0.5952,
"step": 505
},
{
"epoch": 0.27287319422150885,
"grad_norm": 2.4163211236401847,
"learning_rate": 9.117366297467899e-06,
"loss": 0.595,
"step": 510
},
{
"epoch": 0.27554842161583737,
"grad_norm": 2.502428814553625,
"learning_rate": 9.090694887473539e-06,
"loss": 0.5897,
"step": 515
},
{
"epoch": 0.2782236490101659,
"grad_norm": 2.568793263794682,
"learning_rate": 9.063666712922054e-06,
"loss": 0.5699,
"step": 520
},
{
"epoch": 0.2808988764044944,
"grad_norm": 2.382473697312008,
"learning_rate": 9.036284131040027e-06,
"loss": 0.589,
"step": 525
},
{
"epoch": 0.2835741037988229,
"grad_norm": 2.263455139276105,
"learning_rate": 9.008549529963202e-06,
"loss": 0.5864,
"step": 530
},
{
"epoch": 0.28624933119315144,
"grad_norm": 2.4693065814392527,
"learning_rate": 8.98046532852822e-06,
"loss": 0.5667,
"step": 535
},
{
"epoch": 0.28892455858747995,
"grad_norm": 2.4610990146399003,
"learning_rate": 8.952033976061651e-06,
"loss": 0.5978,
"step": 540
},
{
"epoch": 0.29159978598180847,
"grad_norm": 2.642358939686009,
"learning_rate": 8.923257952166391e-06,
"loss": 0.5821,
"step": 545
},
{
"epoch": 0.294275013376137,
"grad_norm": 2.5217057546849087,
"learning_rate": 8.894139766505391e-06,
"loss": 0.5793,
"step": 550
},
{
"epoch": 0.2969502407704655,
"grad_norm": 2.2661838032050468,
"learning_rate": 8.864681958582795e-06,
"loss": 0.567,
"step": 555
},
{
"epoch": 0.299625468164794,
"grad_norm": 2.442984986637873,
"learning_rate": 8.834887097522452e-06,
"loss": 0.558,
"step": 560
},
{
"epoch": 0.30230069555912253,
"grad_norm": 2.4181112148582184,
"learning_rate": 8.80475778184386e-06,
"loss": 0.5797,
"step": 565
},
{
"epoch": 0.30497592295345105,
"grad_norm": 2.4733057765728694,
"learning_rate": 8.774296639235527e-06,
"loss": 0.5729,
"step": 570
},
{
"epoch": 0.30765115034777957,
"grad_norm": 2.4046309203890264,
"learning_rate": 8.743506326325814e-06,
"loss": 0.5725,
"step": 575
},
{
"epoch": 0.3103263777421081,
"grad_norm": 2.3380230010121177,
"learning_rate": 8.712389528451236e-06,
"loss": 0.5461,
"step": 580
},
{
"epoch": 0.3130016051364366,
"grad_norm": 2.5038468197885835,
"learning_rate": 8.680948959422266e-06,
"loss": 0.5785,
"step": 585
},
{
"epoch": 0.3156768325307651,
"grad_norm": 3.1445703294824767,
"learning_rate": 8.649187361286641e-06,
"loss": 0.568,
"step": 590
},
{
"epoch": 0.31835205992509363,
"grad_norm": 2.406292242894352,
"learning_rate": 8.617107504090239e-06,
"loss": 0.557,
"step": 595
},
{
"epoch": 0.32102728731942215,
"grad_norm": 2.370295664665951,
"learning_rate": 8.584712185635477e-06,
"loss": 0.5601,
"step": 600
},
{
"epoch": 0.32370251471375067,
"grad_norm": 2.7526828810884303,
"learning_rate": 8.552004231237308e-06,
"loss": 0.5592,
"step": 605
},
{
"epoch": 0.3263777421080792,
"grad_norm": 2.5217279078466603,
"learning_rate": 8.518986493476819e-06,
"loss": 0.5424,
"step": 610
},
{
"epoch": 0.3290529695024077,
"grad_norm": 2.4256807544144117,
"learning_rate": 8.485661851952443e-06,
"loss": 0.5558,
"step": 615
},
{
"epoch": 0.3317281968967362,
"grad_norm": 2.3341524059148084,
"learning_rate": 8.452033213028822e-06,
"loss": 0.5621,
"step": 620
},
{
"epoch": 0.33440342429106473,
"grad_norm": 2.3741670788238065,
"learning_rate": 8.418103509583323e-06,
"loss": 0.5625,
"step": 625
},
{
"epoch": 0.33707865168539325,
"grad_norm": 2.3488707698974465,
"learning_rate": 8.383875700750272e-06,
"loss": 0.5564,
"step": 630
},
{
"epoch": 0.33975387907972177,
"grad_norm": 2.2944826885447456,
"learning_rate": 8.349352771662848e-06,
"loss": 0.5448,
"step": 635
},
{
"epoch": 0.3424291064740503,
"grad_norm": 2.340765525681543,
"learning_rate": 8.314537733192762e-06,
"loss": 0.5464,
"step": 640
},
{
"epoch": 0.3451043338683788,
"grad_norm": 2.4912590904296383,
"learning_rate": 8.279433621687658e-06,
"loss": 0.5508,
"step": 645
},
{
"epoch": 0.3477795612627073,
"grad_norm": 2.4869220531713974,
"learning_rate": 8.2440434987063e-06,
"loss": 0.539,
"step": 650
},
{
"epoch": 0.35045478865703583,
"grad_norm": 2.3322530356260986,
"learning_rate": 8.208370450751568e-06,
"loss": 0.5245,
"step": 655
},
{
"epoch": 0.35313001605136435,
"grad_norm": 2.4501780549096113,
"learning_rate": 8.172417589001275e-06,
"loss": 0.5458,
"step": 660
},
{
"epoch": 0.35580524344569286,
"grad_norm": 2.429282446394144,
"learning_rate": 8.136188049036817e-06,
"loss": 0.5457,
"step": 665
},
{
"epoch": 0.3584804708400214,
"grad_norm": 2.5498468100287046,
"learning_rate": 8.099684990569717e-06,
"loss": 0.5253,
"step": 670
},
{
"epoch": 0.3611556982343499,
"grad_norm": 2.3702949455068683,
"learning_rate": 8.06291159716606e-06,
"loss": 0.5426,
"step": 675
},
{
"epoch": 0.3638309256286784,
"grad_norm": 2.546455034570349,
"learning_rate": 8.025871075968828e-06,
"loss": 0.5285,
"step": 680
},
{
"epoch": 0.36650615302300693,
"grad_norm": 2.286543746187905,
"learning_rate": 7.988566657418202e-06,
"loss": 0.5249,
"step": 685
},
{
"epoch": 0.36918138041733545,
"grad_norm": 2.3795001406889416,
"learning_rate": 7.951001594969827e-06,
"loss": 0.5382,
"step": 690
},
{
"epoch": 0.37185660781166396,
"grad_norm": 2.5367851021499273,
"learning_rate": 7.91317916481106e-06,
"loss": 0.5313,
"step": 695
},
{
"epoch": 0.37453183520599254,
"grad_norm": 2.4790821311175866,
"learning_rate": 7.875102665575241e-06,
"loss": 0.5364,
"step": 700
},
{
"epoch": 0.37720706260032105,
"grad_norm": 2.436587808572014,
"learning_rate": 7.83677541805401e-06,
"loss": 0.5285,
"step": 705
},
{
"epoch": 0.37988228999464957,
"grad_norm": 2.37913475078827,
"learning_rate": 7.798200764907691e-06,
"loss": 0.5199,
"step": 710
},
{
"epoch": 0.3825575173889781,
"grad_norm": 2.5005240559845134,
"learning_rate": 7.759382070373755e-06,
"loss": 0.5358,
"step": 715
},
{
"epoch": 0.3852327447833066,
"grad_norm": 2.4480000456926625,
"learning_rate": 7.720322719973433e-06,
"loss": 0.5344,
"step": 720
},
{
"epoch": 0.3879079721776351,
"grad_norm": 2.2698845818992455,
"learning_rate": 7.68102612021643e-06,
"loss": 0.5279,
"step": 725
},
{
"epoch": 0.39058319957196364,
"grad_norm": 2.4186655398319252,
"learning_rate": 7.641495698303844e-06,
"loss": 0.5169,
"step": 730
},
{
"epoch": 0.39325842696629215,
"grad_norm": 2.5271248633702674,
"learning_rate": 7.60173490182926e-06,
"loss": 0.5232,
"step": 735
},
{
"epoch": 0.39593365436062067,
"grad_norm": 2.2703784304673005,
"learning_rate": 7.5617471984780885e-06,
"loss": 0.528,
"step": 740
},
{
"epoch": 0.3986088817549492,
"grad_norm": 2.532874522915725,
"learning_rate": 7.521536075725106e-06,
"loss": 0.5249,
"step": 745
},
{
"epoch": 0.4012841091492777,
"grad_norm": 2.4292067983633077,
"learning_rate": 7.481105040530334e-06,
"loss": 0.5369,
"step": 750
},
{
"epoch": 0.4039593365436062,
"grad_norm": 2.3933411598642698,
"learning_rate": 7.440457619033155e-06,
"loss": 0.5205,
"step": 755
},
{
"epoch": 0.40663456393793473,
"grad_norm": 2.6309076094473935,
"learning_rate": 7.3995973562448065e-06,
"loss": 0.4936,
"step": 760
},
{
"epoch": 0.40930979133226325,
"grad_norm": 2.4382597285863916,
"learning_rate": 7.358527815739192e-06,
"loss": 0.5397,
"step": 765
},
{
"epoch": 0.41198501872659177,
"grad_norm": 2.493128592305897,
"learning_rate": 7.317252579342096e-06,
"loss": 0.5062,
"step": 770
},
{
"epoch": 0.4146602461209203,
"grad_norm": 2.39571739461516,
"learning_rate": 7.275775246818802e-06,
"loss": 0.5077,
"step": 775
},
{
"epoch": 0.4173354735152488,
"grad_norm": 2.3257482836528798,
"learning_rate": 7.23409943556014e-06,
"loss": 0.5051,
"step": 780
},
{
"epoch": 0.4200107009095773,
"grad_norm": 2.3506716083497623,
"learning_rate": 7.192228780266997e-06,
"loss": 0.5051,
"step": 785
},
{
"epoch": 0.42268592830390583,
"grad_norm": 2.3896789190082948,
"learning_rate": 7.150166932633328e-06,
"loss": 0.4925,
"step": 790
},
{
"epoch": 0.42536115569823435,
"grad_norm": 2.304737494822103,
"learning_rate": 7.1079175610276775e-06,
"loss": 0.516,
"step": 795
},
{
"epoch": 0.42803638309256287,
"grad_norm": 2.376368349197412,
"learning_rate": 7.065484350173242e-06,
"loss": 0.508,
"step": 800
},
{
"epoch": 0.4307116104868914,
"grad_norm": 2.390599123508061,
"learning_rate": 7.022871000826519e-06,
"loss": 0.4957,
"step": 805
},
{
"epoch": 0.4333868378812199,
"grad_norm": 2.4503319503961496,
"learning_rate": 6.980081229454545e-06,
"loss": 0.4932,
"step": 810
},
{
"epoch": 0.4360620652755484,
"grad_norm": 2.388736493290777,
"learning_rate": 6.937118767910771e-06,
"loss": 0.5031,
"step": 815
},
{
"epoch": 0.43873729266987693,
"grad_norm": 2.34192919036682,
"learning_rate": 6.893987363109595e-06,
"loss": 0.4964,
"step": 820
},
{
"epoch": 0.44141252006420545,
"grad_norm": 2.525102607168378,
"learning_rate": 6.850690776699574e-06,
"loss": 0.4976,
"step": 825
},
{
"epoch": 0.44408774745853397,
"grad_norm": 2.582188474917445,
"learning_rate": 6.807232784735363e-06,
"loss": 0.4798,
"step": 830
},
{
"epoch": 0.4467629748528625,
"grad_norm": 2.3217690049430937,
"learning_rate": 6.763617177348394e-06,
"loss": 0.4882,
"step": 835
},
{
"epoch": 0.449438202247191,
"grad_norm": 2.3691948839930643,
"learning_rate": 6.719847758416316e-06,
"loss": 0.4921,
"step": 840
},
{
"epoch": 0.4521134296415195,
"grad_norm": 2.384102404640162,
"learning_rate": 6.675928345231248e-06,
"loss": 0.4872,
"step": 845
},
{
"epoch": 0.45478865703584803,
"grad_norm": 2.396616687246066,
"learning_rate": 6.631862768166861e-06,
"loss": 0.4982,
"step": 850
},
{
"epoch": 0.45746388443017655,
"grad_norm": 2.493144015288168,
"learning_rate": 6.587654870344318e-06,
"loss": 0.4927,
"step": 855
},
{
"epoch": 0.46013911182450506,
"grad_norm": 2.4553160494373296,
"learning_rate": 6.543308507297094e-06,
"loss": 0.4924,
"step": 860
},
{
"epoch": 0.4628143392188336,
"grad_norm": 2.3226795438252354,
"learning_rate": 6.498827546634733e-06,
"loss": 0.4709,
"step": 865
},
{
"epoch": 0.4654895666131621,
"grad_norm": 2.281005549264743,
"learning_rate": 6.454215867705526e-06,
"loss": 0.4983,
"step": 870
},
{
"epoch": 0.4681647940074906,
"grad_norm": 2.4555385388069473,
"learning_rate": 6.409477361258188e-06,
"loss": 0.4753,
"step": 875
},
{
"epoch": 0.47084002140181913,
"grad_norm": 2.4255588194409583,
"learning_rate": 6.364615929102531e-06,
"loss": 0.4763,
"step": 880
},
{
"epoch": 0.47351524879614765,
"grad_norm": 2.406190699304918,
"learning_rate": 6.319635483769164e-06,
"loss": 0.4843,
"step": 885
},
{
"epoch": 0.47619047619047616,
"grad_norm": 2.386893315693651,
"learning_rate": 6.274539948168279e-06,
"loss": 0.4716,
"step": 890
},
{
"epoch": 0.47886570358480474,
"grad_norm": 2.2728664616613887,
"learning_rate": 6.229333255247511e-06,
"loss": 0.467,
"step": 895
},
{
"epoch": 0.48154093097913325,
"grad_norm": 2.4446705313074064,
"learning_rate": 6.184019347648939e-06,
"loss": 0.4683,
"step": 900
},
{
"epoch": 0.48421615837346177,
"grad_norm": 2.297929824104002,
"learning_rate": 6.138602177365218e-06,
"loss": 0.4814,
"step": 905
},
{
"epoch": 0.4868913857677903,
"grad_norm": 2.2398605800909226,
"learning_rate": 6.093085705394934e-06,
"loss": 0.4761,
"step": 910
},
{
"epoch": 0.4895666131621188,
"grad_norm": 2.4670515565573083,
"learning_rate": 6.04747390139713e-06,
"loss": 0.4705,
"step": 915
},
{
"epoch": 0.4922418405564473,
"grad_norm": 2.436162832733763,
"learning_rate": 6.001770743345108e-06,
"loss": 0.4689,
"step": 920
},
{
"epoch": 0.49491706795077584,
"grad_norm": 2.7004239367895053,
"learning_rate": 5.9559802171794955e-06,
"loss": 0.464,
"step": 925
},
{
"epoch": 0.49759229534510435,
"grad_norm": 2.352197390213324,
"learning_rate": 5.9101063164606165e-06,
"loss": 0.4588,
"step": 930
},
{
"epoch": 0.5002675227394329,
"grad_norm": 2.386178979724787,
"learning_rate": 5.864153042020191e-06,
"loss": 0.4593,
"step": 935
},
{
"epoch": 0.5029427501337613,
"grad_norm": 2.271248403727243,
"learning_rate": 5.818124401612416e-06,
"loss": 0.4554,
"step": 940
},
{
"epoch": 0.5056179775280899,
"grad_norm": 2.34018371964795,
"learning_rate": 5.7720244095644305e-06,
"loss": 0.4703,
"step": 945
},
{
"epoch": 0.5082932049224184,
"grad_norm": 2.270516604487789,
"learning_rate": 5.725857086426216e-06,
"loss": 0.454,
"step": 950
},
{
"epoch": 0.5109684323167469,
"grad_norm": 2.3872382871657325,
"learning_rate": 5.679626458619947e-06,
"loss": 0.4457,
"step": 955
},
{
"epoch": 0.5136436597110754,
"grad_norm": 2.520375771831823,
"learning_rate": 5.633336558088823e-06,
"loss": 0.4598,
"step": 960
},
{
"epoch": 0.516318887105404,
"grad_norm": 2.2353434501523575,
"learning_rate": 5.586991421945445e-06,
"loss": 0.4711,
"step": 965
},
{
"epoch": 0.5189941144997324,
"grad_norm": 2.2812982114304563,
"learning_rate": 5.540595092119709e-06,
"loss": 0.4439,
"step": 970
},
{
"epoch": 0.521669341894061,
"grad_norm": 2.392550181275633,
"learning_rate": 5.494151615006307e-06,
"loss": 0.4604,
"step": 975
},
{
"epoch": 0.5243445692883895,
"grad_norm": 2.233410991408843,
"learning_rate": 5.44766504111181e-06,
"loss": 0.4674,
"step": 980
},
{
"epoch": 0.527019796682718,
"grad_norm": 2.4198668089222153,
"learning_rate": 5.401139424701427e-06,
"loss": 0.4553,
"step": 985
},
{
"epoch": 0.5296950240770465,
"grad_norm": 2.377473414386296,
"learning_rate": 5.354578823445404e-06,
"loss": 0.4477,
"step": 990
},
{
"epoch": 0.5323702514713751,
"grad_norm": 2.3622932708922213,
"learning_rate": 5.307987298065145e-06,
"loss": 0.4339,
"step": 995
},
{
"epoch": 0.5350454788657036,
"grad_norm": 2.332131787995006,
"learning_rate": 5.26136891197906e-06,
"loss": 0.4404,
"step": 1000
},
{
"epoch": 0.5377207062600321,
"grad_norm": 2.4251937487497783,
"learning_rate": 5.214727730948181e-06,
"loss": 0.4566,
"step": 1005
},
{
"epoch": 0.5403959336543607,
"grad_norm": 2.3957788984396937,
"learning_rate": 5.1680678227215705e-06,
"loss": 0.4405,
"step": 1010
},
{
"epoch": 0.5430711610486891,
"grad_norm": 2.3847750525666873,
"learning_rate": 5.121393256681561e-06,
"loss": 0.4446,
"step": 1015
},
{
"epoch": 0.5457463884430177,
"grad_norm": 2.2698162072049533,
"learning_rate": 5.07470810348884e-06,
"loss": 0.4354,
"step": 1020
},
{
"epoch": 0.5484216158373462,
"grad_norm": 2.436311734783881,
"learning_rate": 5.02801643472745e-06,
"loss": 0.4414,
"step": 1025
},
{
"epoch": 0.5510968432316747,
"grad_norm": 2.2916945959636643,
"learning_rate": 4.98132232254967e-06,
"loss": 0.4391,
"step": 1030
},
{
"epoch": 0.5537720706260032,
"grad_norm": 2.4013743839662602,
"learning_rate": 4.934629839320885e-06,
"loss": 0.4371,
"step": 1035
},
{
"epoch": 0.5564472980203318,
"grad_norm": 2.3988828791639314,
"learning_rate": 4.88794305726441e-06,
"loss": 0.4375,
"step": 1040
},
{
"epoch": 0.5591225254146602,
"grad_norm": 2.330590397904805,
"learning_rate": 4.841266048106343e-06,
"loss": 0.434,
"step": 1045
},
{
"epoch": 0.5617977528089888,
"grad_norm": 2.3366142392042124,
"learning_rate": 4.794602882720448e-06,
"loss": 0.4281,
"step": 1050
},
{
"epoch": 0.5644729802033173,
"grad_norm": 2.3883999122895503,
"learning_rate": 4.747957630773124e-06,
"loss": 0.4317,
"step": 1055
},
{
"epoch": 0.5671482075976458,
"grad_norm": 2.3510366869048323,
"learning_rate": 4.701334360368473e-06,
"loss": 0.4272,
"step": 1060
},
{
"epoch": 0.5698234349919743,
"grad_norm": 2.3683109214556315,
"learning_rate": 4.654737137693508e-06,
"loss": 0.4381,
"step": 1065
},
{
"epoch": 0.5724986623863029,
"grad_norm": 2.2171740005409304,
"learning_rate": 4.6081700266635195e-06,
"loss": 0.4227,
"step": 1070
},
{
"epoch": 0.5751738897806313,
"grad_norm": 2.3232281982068783,
"learning_rate": 4.561637088567654e-06,
"loss": 0.4217,
"step": 1075
},
{
"epoch": 0.5778491171749599,
"grad_norm": 2.4194928745884514,
"learning_rate": 4.51514238171471e-06,
"loss": 0.4255,
"step": 1080
},
{
"epoch": 0.5805243445692884,
"grad_norm": 2.202777210726338,
"learning_rate": 4.468689961079195e-06,
"loss": 0.4327,
"step": 1085
},
{
"epoch": 0.5831995719636169,
"grad_norm": 2.2294786297275926,
"learning_rate": 4.4222838779476866e-06,
"loss": 0.4352,
"step": 1090
},
{
"epoch": 0.5858747993579454,
"grad_norm": 2.3255340067920933,
"learning_rate": 4.375928179565494e-06,
"loss": 0.418,
"step": 1095
},
{
"epoch": 0.588550026752274,
"grad_norm": 2.3996404743348205,
"learning_rate": 4.329626908783685e-06,
"loss": 0.4189,
"step": 1100
},
{
"epoch": 0.5912252541466024,
"grad_norm": 2.303291302129872,
"learning_rate": 4.2833841037065e-06,
"loss": 0.4077,
"step": 1105
},
{
"epoch": 0.593900481540931,
"grad_norm": 2.301388594139086,
"learning_rate": 4.237203797339169e-06,
"loss": 0.4104,
"step": 1110
},
{
"epoch": 0.5965757089352595,
"grad_norm": 2.160810901865994,
"learning_rate": 4.191090017236177e-06,
"loss": 0.4235,
"step": 1115
},
{
"epoch": 0.599250936329588,
"grad_norm": 2.2158867728636413,
"learning_rate": 4.145046785150013e-06,
"loss": 0.4046,
"step": 1120
},
{
"epoch": 0.6019261637239165,
"grad_norm": 2.2735600817905923,
"learning_rate": 4.09907811668041e-06,
"loss": 0.408,
"step": 1125
},
{
"epoch": 0.6046013911182451,
"grad_norm": 2.3803804980296444,
"learning_rate": 4.0531880209241356e-06,
"loss": 0.4136,
"step": 1130
},
{
"epoch": 0.6072766185125735,
"grad_norm": 2.266085861670805,
"learning_rate": 4.0073805001253405e-06,
"loss": 0.4051,
"step": 1135
},
{
"epoch": 0.6099518459069021,
"grad_norm": 2.1954654483630796,
"learning_rate": 3.961659549326512e-06,
"loss": 0.4262,
"step": 1140
},
{
"epoch": 0.6126270733012306,
"grad_norm": 2.228408225766588,
"learning_rate": 3.916029156020044e-06,
"loss": 0.4029,
"step": 1145
},
{
"epoch": 0.6153023006955591,
"grad_norm": 2.3199361261442144,
"learning_rate": 3.870493299800484e-06,
"loss": 0.4168,
"step": 1150
},
{
"epoch": 0.6179775280898876,
"grad_norm": 2.2619585349946796,
"learning_rate": 3.82505595201745e-06,
"loss": 0.408,
"step": 1155
},
{
"epoch": 0.6206527554842162,
"grad_norm": 2.4880856896093544,
"learning_rate": 3.7797210754292766e-06,
"loss": 0.4128,
"step": 1160
},
{
"epoch": 0.6233279828785446,
"grad_norm": 2.212702659150222,
"learning_rate": 3.7344926238574074e-06,
"loss": 0.4121,
"step": 1165
},
{
"epoch": 0.6260032102728732,
"grad_norm": 2.262392758862919,
"learning_rate": 3.6893745418415692e-06,
"loss": 0.3935,
"step": 1170
},
{
"epoch": 0.6286784376672017,
"grad_norm": 2.2203307618409185,
"learning_rate": 3.6443707642957526e-06,
"loss": 0.3996,
"step": 1175
},
{
"epoch": 0.6313536650615302,
"grad_norm": 2.183173739658585,
"learning_rate": 3.5994852161650386e-06,
"loss": 0.389,
"step": 1180
},
{
"epoch": 0.6340288924558587,
"grad_norm": 2.3980317993943814,
"learning_rate": 3.5547218120832807e-06,
"loss": 0.3936,
"step": 1185
},
{
"epoch": 0.6367041198501873,
"grad_norm": 2.3294327156692654,
"learning_rate": 3.5100844560317028e-06,
"loss": 0.4095,
"step": 1190
},
{
"epoch": 0.6393793472445158,
"grad_norm": 2.3318739063082936,
"learning_rate": 3.465577040998417e-06,
"loss": 0.4036,
"step": 1195
},
{
"epoch": 0.6420545746388443,
"grad_norm": 2.13554424918326,
"learning_rate": 3.4212034486388972e-06,
"loss": 0.385,
"step": 1200
},
{
"epoch": 0.6447298020331729,
"grad_norm": 2.3059775180831887,
"learning_rate": 3.376967548937457e-06,
"loss": 0.3959,
"step": 1205
},
{
"epoch": 0.6474050294275013,
"grad_norm": 2.2003509075548124,
"learning_rate": 3.332873199869719e-06,
"loss": 0.3918,
"step": 1210
},
{
"epoch": 0.6500802568218299,
"grad_norm": 2.2638155453662576,
"learning_rate": 3.2889242470661553e-06,
"loss": 0.3964,
"step": 1215
},
{
"epoch": 0.6527554842161584,
"grad_norm": 2.2328176385540126,
"learning_rate": 3.245124523476699e-06,
"loss": 0.3872,
"step": 1220
},
{
"epoch": 0.6554307116104869,
"grad_norm": 2.2612712629675364,
"learning_rate": 3.2014778490364484e-06,
"loss": 0.4019,
"step": 1225
},
{
"epoch": 0.6581059390048154,
"grad_norm": 2.183300299715255,
"learning_rate": 3.157988030332526e-06,
"loss": 0.3986,
"step": 1230
},
{
"epoch": 0.660781166399144,
"grad_norm": 2.2008867618272245,
"learning_rate": 3.1146588602720884e-06,
"loss": 0.3793,
"step": 1235
},
{
"epoch": 0.6634563937934724,
"grad_norm": 2.3131230110616396,
"learning_rate": 3.0714941177515307e-06,
"loss": 0.389,
"step": 1240
},
{
"epoch": 0.666131621187801,
"grad_norm": 2.3622281138792105,
"learning_rate": 3.0284975673269175e-06,
"loss": 0.3906,
"step": 1245
},
{
"epoch": 0.6688068485821295,
"grad_norm": 2.2764800784114696,
"learning_rate": 2.9856729588856615e-06,
"loss": 0.3917,
"step": 1250
},
{
"epoch": 0.671482075976458,
"grad_norm": 2.1993222173811064,
"learning_rate": 2.9430240273194844e-06,
"loss": 0.3829,
"step": 1255
},
{
"epoch": 0.6741573033707865,
"grad_norm": 2.2083725272220414,
"learning_rate": 2.9005544921986774e-06,
"loss": 0.3778,
"step": 1260
},
{
"epoch": 0.6768325307651151,
"grad_norm": 2.214779231475582,
"learning_rate": 2.858268057447712e-06,
"loss": 0.3901,
"step": 1265
},
{
"epoch": 0.6795077581594435,
"grad_norm": 2.2254433327654097,
"learning_rate": 2.8161684110221987e-06,
"loss": 0.387,
"step": 1270
},
{
"epoch": 0.6821829855537721,
"grad_norm": 2.31973974930244,
"learning_rate": 2.7742592245872523e-06,
"loss": 0.3852,
"step": 1275
},
{
"epoch": 0.6848582129481006,
"grad_norm": 2.212647002572668,
"learning_rate": 2.7325441531972685e-06,
"loss": 0.3753,
"step": 1280
},
{
"epoch": 0.6875334403424291,
"grad_norm": 2.1846100608702606,
"learning_rate": 2.691026834977161e-06,
"loss": 0.3995,
"step": 1285
},
{
"epoch": 0.6902086677367576,
"grad_norm": 2.229995641206547,
"learning_rate": 2.649710890805055e-06,
"loss": 0.3828,
"step": 1290
},
{
"epoch": 0.6928838951310862,
"grad_norm": 2.1422721805468354,
"learning_rate": 2.6085999239965094e-06,
"loss": 0.3692,
"step": 1295
},
{
"epoch": 0.6955591225254146,
"grad_norm": 2.318213911947721,
"learning_rate": 2.567697519990249e-06,
"loss": 0.3796,
"step": 1300
},
{
"epoch": 0.6982343499197432,
"grad_norm": 2.2119506235831827,
"learning_rate": 2.52700724603547e-06,
"loss": 0.3775,
"step": 1305
},
{
"epoch": 0.7009095773140717,
"grad_norm": 2.224473043074625,
"learning_rate": 2.4865326508807274e-06,
"loss": 0.37,
"step": 1310
},
{
"epoch": 0.7035848047084002,
"grad_norm": 2.079689857122922,
"learning_rate": 2.446277264464431e-06,
"loss": 0.3771,
"step": 1315
},
{
"epoch": 0.7062600321027287,
"grad_norm": 2.277862216469956,
"learning_rate": 2.406244597606994e-06,
"loss": 0.3673,
"step": 1320
},
{
"epoch": 0.7089352594970573,
"grad_norm": 2.1871262495624952,
"learning_rate": 2.3664381417046362e-06,
"loss": 0.372,
"step": 1325
},
{
"epoch": 0.7116104868913857,
"grad_norm": 2.5202258537627293,
"learning_rate": 2.3268613684248846e-06,
"loss": 0.381,
"step": 1330
},
{
"epoch": 0.7142857142857143,
"grad_norm": 2.2472346697981296,
"learning_rate": 2.287517729403802e-06,
"loss": 0.368,
"step": 1335
},
{
"epoch": 0.7169609416800428,
"grad_norm": 2.221291336010394,
"learning_rate": 2.2484106559449527e-06,
"loss": 0.363,
"step": 1340
},
{
"epoch": 0.7196361690743713,
"grad_norm": 2.183019746627642,
"learning_rate": 2.2095435587201487e-06,
"loss": 0.3705,
"step": 1345
},
{
"epoch": 0.7223113964686998,
"grad_norm": 2.1399656648022933,
"learning_rate": 2.1709198274719908e-06,
"loss": 0.3654,
"step": 1350
},
{
"epoch": 0.7249866238630284,
"grad_norm": 2.0861165370537105,
"learning_rate": 2.1325428307182357e-06,
"loss": 0.3669,
"step": 1355
},
{
"epoch": 0.7276618512573568,
"grad_norm": 2.164652924106045,
"learning_rate": 2.0944159154580225e-06,
"loss": 0.3696,
"step": 1360
},
{
"epoch": 0.7303370786516854,
"grad_norm": 2.372293297792043,
"learning_rate": 2.056542406879957e-06,
"loss": 0.3818,
"step": 1365
},
{
"epoch": 0.7330123060460139,
"grad_norm": 2.090955405220056,
"learning_rate": 2.018925608072118e-06,
"loss": 0.3542,
"step": 1370
},
{
"epoch": 0.7356875334403424,
"grad_norm": 2.2442087175829193,
"learning_rate": 1.981568799733979e-06,
"loss": 0.3657,
"step": 1375
},
{
"epoch": 0.7383627608346709,
"grad_norm": 2.2516703366046578,
"learning_rate": 1.9444752398902874e-06,
"loss": 0.3615,
"step": 1380
},
{
"epoch": 0.7410379882289995,
"grad_norm": 2.259664285200144,
"learning_rate": 1.907648163606925e-06,
"loss": 0.3587,
"step": 1385
},
{
"epoch": 0.7437132156233279,
"grad_norm": 2.4360594395654473,
"learning_rate": 1.871090782708756e-06,
"loss": 0.3603,
"step": 1390
},
{
"epoch": 0.7463884430176565,
"grad_norm": 2.247000162892966,
"learning_rate": 1.834806285499519e-06,
"loss": 0.3542,
"step": 1395
},
{
"epoch": 0.7490636704119851,
"grad_norm": 2.0975932964206425,
"learning_rate": 1.7987978364837649e-06,
"loss": 0.3557,
"step": 1400
},
{
"epoch": 0.7517388978063135,
"grad_norm": 2.104100339362327,
"learning_rate": 1.7630685760908623e-06,
"loss": 0.358,
"step": 1405
},
{
"epoch": 0.7544141252006421,
"grad_norm": 2.2966685174118324,
"learning_rate": 1.727621620401112e-06,
"loss": 0.3672,
"step": 1410
},
{
"epoch": 0.7570893525949706,
"grad_norm": 2.2434262781603422,
"learning_rate": 1.6924600608739843e-06,
"loss": 0.3632,
"step": 1415
},
{
"epoch": 0.7597645799892991,
"grad_norm": 2.3695918803743052,
"learning_rate": 1.6575869640784998e-06,
"loss": 0.3674,
"step": 1420
},
{
"epoch": 0.7624398073836276,
"grad_norm": 2.2896558172173096,
"learning_rate": 1.6230053714257821e-06,
"loss": 0.3615,
"step": 1425
},
{
"epoch": 0.7651150347779562,
"grad_norm": 2.132919841878146,
"learning_rate": 1.588718298903803e-06,
"loss": 0.3564,
"step": 1430
},
{
"epoch": 0.7677902621722846,
"grad_norm": 2.179194206405239,
"learning_rate": 1.554728736814356e-06,
"loss": 0.352,
"step": 1435
},
{
"epoch": 0.7704654895666132,
"grad_norm": 2.2259000819781063,
"learning_rate": 1.5210396495122481e-06,
"loss": 0.3505,
"step": 1440
},
{
"epoch": 0.7731407169609417,
"grad_norm": 2.1237345266350847,
"learning_rate": 1.4876539751467806e-06,
"loss": 0.3521,
"step": 1445
},
{
"epoch": 0.7758159443552702,
"grad_norm": 2.2438319275859304,
"learning_rate": 1.45457462540549e-06,
"loss": 0.3467,
"step": 1450
},
{
"epoch": 0.7784911717495987,
"grad_norm": 2.1311046073289326,
"learning_rate": 1.4218044852602176e-06,
"loss": 0.3517,
"step": 1455
},
{
"epoch": 0.7811663991439273,
"grad_norm": 2.151731010721367,
"learning_rate": 1.3893464127154976e-06,
"loss": 0.36,
"step": 1460
},
{
"epoch": 0.7838416265382557,
"grad_norm": 2.1371565977438203,
"learning_rate": 1.3572032385592999e-06,
"loss": 0.3466,
"step": 1465
},
{
"epoch": 0.7865168539325843,
"grad_norm": 2.1766415440270532,
"learning_rate": 1.325377766116146e-06,
"loss": 0.3488,
"step": 1470
},
{
"epoch": 0.7891920813269128,
"grad_norm": 2.02619887138196,
"learning_rate": 1.293872771002625e-06,
"loss": 0.3486,
"step": 1475
},
{
"epoch": 0.7918673087212413,
"grad_norm": 2.3827191729249737,
"learning_rate": 1.2626910008853154e-06,
"loss": 0.3703,
"step": 1480
},
{
"epoch": 0.7945425361155698,
"grad_norm": 2.2615580929872117,
"learning_rate": 1.231835175241155e-06,
"loss": 0.3562,
"step": 1485
},
{
"epoch": 0.7972177635098984,
"grad_norm": 2.229708858675527,
"learning_rate": 1.2013079851202642e-06,
"loss": 0.3624,
"step": 1490
},
{
"epoch": 0.7998929909042268,
"grad_norm": 2.333874491979822,
"learning_rate": 1.1711120929112507e-06,
"loss": 0.3404,
"step": 1495
},
{
"epoch": 0.8025682182985554,
"grad_norm": 1.947457938033558,
"learning_rate": 1.141250132109009e-06,
"loss": 0.3535,
"step": 1500
},
{
"epoch": 0.8052434456928839,
"grad_norm": 2.0935289084516615,
"learning_rate": 1.1117247070850534e-06,
"loss": 0.3406,
"step": 1505
},
{
"epoch": 0.8079186730872124,
"grad_norm": 2.2325175076090065,
"learning_rate": 1.0825383928603656e-06,
"loss": 0.3526,
"step": 1510
},
{
"epoch": 0.8105939004815409,
"grad_norm": 2.0095290479997647,
"learning_rate": 1.0536937348808341e-06,
"loss": 0.3409,
"step": 1515
},
{
"epoch": 0.8132691278758695,
"grad_norm": 1.9836588566813402,
"learning_rate": 1.0251932487952437e-06,
"loss": 0.339,
"step": 1520
},
{
"epoch": 0.8159443552701979,
"grad_norm": 2.26784995989598,
"learning_rate": 9.97039420235884e-07,
"loss": 0.3471,
"step": 1525
},
{
"epoch": 0.8186195826645265,
"grad_norm": 2.1193196211762846,
"learning_rate": 9.692347046017647e-07,
"loss": 0.344,
"step": 1530
},
{
"epoch": 0.821294810058855,
"grad_norm": 2.211413032254883,
"learning_rate": 9.417815268444719e-07,
"loss": 0.3441,
"step": 1535
},
{
"epoch": 0.8239700374531835,
"grad_norm": 1.9471769083672645,
"learning_rate": 9.146822812566819e-07,
"loss": 0.3502,
"step": 1540
},
{
"epoch": 0.826645264847512,
"grad_norm": 2.1918923218693105,
"learning_rate": 8.879393312633405e-07,
"loss": 0.3451,
"step": 1545
},
{
"epoch": 0.8293204922418406,
"grad_norm": 2.1371349511781506,
"learning_rate": 8.615550092155478e-07,
"loss": 0.3435,
"step": 1550
},
{
"epoch": 0.831995719636169,
"grad_norm": 2.0845002217133923,
"learning_rate": 8.355316161871369e-07,
"loss": 0.3466,
"step": 1555
},
{
"epoch": 0.8346709470304976,
"grad_norm": 2.25020629433754,
"learning_rate": 8.098714217739928e-07,
"loss": 0.3419,
"step": 1560
},
{
"epoch": 0.8373461744248261,
"grad_norm": 2.0744989514612127,
"learning_rate": 7.845766638961172e-07,
"loss": 0.3473,
"step": 1565
},
{
"epoch": 0.8400214018191546,
"grad_norm": 3.639837300779382,
"learning_rate": 7.596495486024402e-07,
"loss": 0.3515,
"step": 1570
},
{
"epoch": 0.8426966292134831,
"grad_norm": 2.0191662732122104,
"learning_rate": 7.350922498784335e-07,
"loss": 0.3519,
"step": 1575
},
{
"epoch": 0.8453718566078117,
"grad_norm": 2.228031235416842,
"learning_rate": 7.109069094565024e-07,
"loss": 0.3448,
"step": 1580
},
{
"epoch": 0.8480470840021401,
"grad_norm": 2.523017960304099,
"learning_rate": 6.870956366291998e-07,
"loss": 0.3388,
"step": 1585
},
{
"epoch": 0.8507223113964687,
"grad_norm": 2.060587000595492,
"learning_rate": 6.636605080652686e-07,
"loss": 0.3487,
"step": 1590
},
{
"epoch": 0.8533975387907973,
"grad_norm": 1.9701444425040686,
"learning_rate": 6.406035676285244e-07,
"loss": 0.3301,
"step": 1595
},
{
"epoch": 0.8560727661851257,
"grad_norm": 2.0633572735895873,
"learning_rate": 6.179268261996052e-07,
"loss": 0.3367,
"step": 1600
},
{
"epoch": 0.8587479935794543,
"grad_norm": 2.1729206325004204,
"learning_rate": 5.956322615005928e-07,
"loss": 0.3463,
"step": 1605
},
{
"epoch": 0.8614232209737828,
"grad_norm": 2.1697102226952585,
"learning_rate": 5.737218179225318e-07,
"loss": 0.3427,
"step": 1610
},
{
"epoch": 0.8640984483681113,
"grad_norm": 2.2114503479333942,
"learning_rate": 5.521974063558477e-07,
"loss": 0.3452,
"step": 1615
},
{
"epoch": 0.8667736757624398,
"grad_norm": 2.0058575813876645,
"learning_rate": 5.310609040236963e-07,
"loss": 0.3471,
"step": 1620
},
{
"epoch": 0.8694489031567684,
"grad_norm": 1.978522418614209,
"learning_rate": 5.103141543182389e-07,
"loss": 0.333,
"step": 1625
},
{
"epoch": 0.8721241305510968,
"grad_norm": 2.1706890891020127,
"learning_rate": 4.89958966639878e-07,
"loss": 0.3399,
"step": 1630
},
{
"epoch": 0.8747993579454254,
"grad_norm": 2.102475974621354,
"learning_rate": 4.6999711623944787e-07,
"loss": 0.3295,
"step": 1635
},
{
"epoch": 0.8774745853397539,
"grad_norm": 2.1350237073132243,
"learning_rate": 4.504303440633928e-07,
"loss": 0.3443,
"step": 1640
},
{
"epoch": 0.8801498127340824,
"grad_norm": 2.0600510975030195,
"learning_rate": 4.3126035660193076e-07,
"loss": 0.3365,
"step": 1645
},
{
"epoch": 0.8828250401284109,
"grad_norm": 2.236173654716049,
"learning_rate": 4.124888257402243e-07,
"loss": 0.3387,
"step": 1650
},
{
"epoch": 0.8855002675227395,
"grad_norm": 1.996336345890085,
"learning_rate": 3.9411738861256934e-07,
"loss": 0.3403,
"step": 1655
},
{
"epoch": 0.8881754949170679,
"grad_norm": 1.9330066477895274,
"learning_rate": 3.7614764745961377e-07,
"loss": 0.3391,
"step": 1660
},
{
"epoch": 0.8908507223113965,
"grad_norm": 2.1574712219106478,
"learning_rate": 3.585811694886232e-07,
"loss": 0.3337,
"step": 1665
},
{
"epoch": 0.893525949705725,
"grad_norm": 2.1699077488308403,
"learning_rate": 3.4141948673679593e-07,
"loss": 0.3344,
"step": 1670
},
{
"epoch": 0.8962011771000535,
"grad_norm": 2.128864116785053,
"learning_rate": 3.2466409593764734e-07,
"loss": 0.3363,
"step": 1675
},
{
"epoch": 0.898876404494382,
"grad_norm": 2.118472484473068,
"learning_rate": 3.083164583904802e-07,
"loss": 0.3415,
"step": 1680
},
{
"epoch": 0.9015516318887106,
"grad_norm": 2.0200831474062886,
"learning_rate": 2.923779998329318e-07,
"loss": 0.3224,
"step": 1685
},
{
"epoch": 0.904226859283039,
"grad_norm": 2.144747824231417,
"learning_rate": 2.76850110316636e-07,
"loss": 0.3343,
"step": 1690
},
{
"epoch": 0.9069020866773676,
"grad_norm": 2.122020641145892,
"learning_rate": 2.617341440859883e-07,
"loss": 0.3465,
"step": 1695
},
{
"epoch": 0.9095773140716961,
"grad_norm": 2.072786373493989,
"learning_rate": 2.470314194600376e-07,
"loss": 0.3231,
"step": 1700
},
{
"epoch": 0.9122525414660246,
"grad_norm": 1.9359758916982563,
"learning_rate": 2.3274321871751436e-07,
"loss": 0.3353,
"step": 1705
},
{
"epoch": 0.9149277688603531,
"grad_norm": 1.9922992143020988,
"learning_rate": 2.1887078798499272e-07,
"loss": 0.3413,
"step": 1710
},
{
"epoch": 0.9176029962546817,
"grad_norm": 2.334136482586755,
"learning_rate": 2.0541533712821527e-07,
"loss": 0.3354,
"step": 1715
},
{
"epoch": 0.9202782236490101,
"grad_norm": 2.2163195568166145,
"learning_rate": 1.923780396465741e-07,
"loss": 0.3412,
"step": 1720
},
{
"epoch": 0.9229534510433387,
"grad_norm": 1.9231313598051207,
"learning_rate": 1.7976003257076823e-07,
"loss": 0.3385,
"step": 1725
},
{
"epoch": 0.9256286784376672,
"grad_norm": 2.0705877113088844,
"learning_rate": 1.6756241636363413e-07,
"loss": 0.3298,
"step": 1730
},
{
"epoch": 0.9283039058319957,
"grad_norm": 2.057376654150935,
"learning_rate": 1.557862548241762e-07,
"loss": 0.3327,
"step": 1735
},
{
"epoch": 0.9309791332263242,
"grad_norm": 2.107808663106814,
"learning_rate": 1.4443257499478447e-07,
"loss": 0.3416,
"step": 1740
},
{
"epoch": 0.9336543606206528,
"grad_norm": 1.9829198272876933,
"learning_rate": 1.3350236707166508e-07,
"loss": 0.3414,
"step": 1745
},
{
"epoch": 0.9363295880149812,
"grad_norm": 2.056040338556245,
"learning_rate": 1.229965843184805e-07,
"loss": 0.3413,
"step": 1750
},
{
"epoch": 0.9390048154093098,
"grad_norm": 2.337178376808903,
"learning_rate": 1.1291614298321097e-07,
"loss": 0.3409,
"step": 1755
},
{
"epoch": 0.9416800428036383,
"grad_norm": 2.1520247260228746,
"learning_rate": 1.0326192221824738e-07,
"loss": 0.3232,
"step": 1760
},
{
"epoch": 0.9443552701979668,
"grad_norm": 2.142605939615871,
"learning_rate": 9.403476400371425e-08,
"loss": 0.3397,
"step": 1765
},
{
"epoch": 0.9470304975922953,
"grad_norm": 2.0450418837244366,
"learning_rate": 8.523547307404179e-08,
"loss": 0.3368,
"step": 1770
},
{
"epoch": 0.9497057249866239,
"grad_norm": 2.0614174232017923,
"learning_rate": 7.686481684777758e-08,
"loss": 0.3321,
"step": 1775
},
{
"epoch": 0.9523809523809523,
"grad_norm": 2.1361641337619015,
"learning_rate": 6.89235253606596e-08,
"loss": 0.3154,
"step": 1780
},
{
"epoch": 0.9550561797752809,
"grad_norm": 2.0514520677857244,
"learning_rate": 6.141229120194714e-08,
"loss": 0.3386,
"step": 1785
},
{
"epoch": 0.9577314071696095,
"grad_norm": 2.069777118733485,
"learning_rate": 5.4331769454016306e-08,
"loss": 0.3319,
"step": 1790
},
{
"epoch": 0.9604066345639379,
"grad_norm": 2.136465423921526,
"learning_rate": 4.76825776352291e-08,
"loss": 0.338,
"step": 1795
},
{
"epoch": 0.9630818619582665,
"grad_norm": 2.0522216771136934,
"learning_rate": 4.1465295646076484e-08,
"loss": 0.3298,
"step": 1800
},
{
"epoch": 0.965757089352595,
"grad_norm": 2.1070855069769094,
"learning_rate": 3.568046571860384e-08,
"loss": 0.3291,
"step": 1805
},
{
"epoch": 0.9684323167469235,
"grad_norm": 2.2197166876041576,
"learning_rate": 3.0328592369120443e-08,
"loss": 0.3373,
"step": 1810
},
{
"epoch": 0.971107544141252,
"grad_norm": 1.8711860549385024,
"learning_rate": 2.541014235419914e-08,
"loss": 0.3314,
"step": 1815
},
{
"epoch": 0.9737827715355806,
"grad_norm": 2.0398927890227805,
"learning_rate": 2.0925544629967763e-08,
"loss": 0.337,
"step": 1820
},
{
"epoch": 0.976457998929909,
"grad_norm": 2.0934381794140253,
"learning_rate": 1.6875190314700197e-08,
"loss": 0.3348,
"step": 1825
},
{
"epoch": 0.9791332263242376,
"grad_norm": 2.0969584920178517,
"learning_rate": 1.3259432654703641e-08,
"loss": 0.3338,
"step": 1830
},
{
"epoch": 0.9818084537185661,
"grad_norm": 1.859307599341638,
"learning_rate": 1.0078586993511052e-08,
"loss": 0.3373,
"step": 1835
},
{
"epoch": 0.9844836811128946,
"grad_norm": 2.085094787953033,
"learning_rate": 7.332930744380906e-09,
"loss": 0.331,
"step": 1840
},
{
"epoch": 0.9871589085072231,
"grad_norm": 2.0189271533665263,
"learning_rate": 5.0227033660987804e-09,
"loss": 0.3259,
"step": 1845
},
{
"epoch": 0.9898341359015517,
"grad_norm": 2.087692299427238,
"learning_rate": 3.1481063420985e-09,
"loss": 0.3324,
"step": 1850
},
{
"epoch": 0.9925093632958801,
"grad_norm": 2.1232755165842856,
"learning_rate": 1.7093031628850899e-09,
"loss": 0.328,
"step": 1855
},
{
"epoch": 0.9951845906902087,
"grad_norm": 2.2098470869075864,
"learning_rate": 7.064193117806151e-10,
"loss": 0.3408,
"step": 1860
},
{
"epoch": 0.9978598180845372,
"grad_norm": 2.177061948752331,
"learning_rate": 1.3954225397516673e-10,
"loss": 0.3345,
"step": 1865
},
{
"epoch": 1.0,
"eval_runtime": 3.4105,
"eval_samples_per_second": 2.932,
"eval_steps_per_second": 0.88,
"step": 1869
},
{
"epoch": 1.0,
"step": 1869,
"total_flos": 195665288232960.0,
"train_loss": 0.4974420365322842,
"train_runtime": 16839.356,
"train_samples_per_second": 1.776,
"train_steps_per_second": 0.111
}
],
"logging_steps": 5,
"max_steps": 1869,
"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": 195665288232960.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}