qwen_1.8_feedback_dirty / trainer_state.json
terry69's picture
Model save
935f3da verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1661,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006020469596628537,
"grad_norm": 10.688992810202613,
"learning_rate": 5.98802395209581e-08,
"loss": 1.7931,
"step": 1
},
{
"epoch": 0.0030102347983142685,
"grad_norm": 11.087545625778352,
"learning_rate": 2.9940119760479047e-07,
"loss": 1.7944,
"step": 5
},
{
"epoch": 0.006020469596628537,
"grad_norm": 10.534678516725148,
"learning_rate": 5.988023952095809e-07,
"loss": 1.7878,
"step": 10
},
{
"epoch": 0.009030704394942806,
"grad_norm": 9.574983566804878,
"learning_rate": 8.982035928143713e-07,
"loss": 1.7696,
"step": 15
},
{
"epoch": 0.012040939193257074,
"grad_norm": 7.113460969940963,
"learning_rate": 1.1976047904191619e-06,
"loss": 1.6818,
"step": 20
},
{
"epoch": 0.015051173991571343,
"grad_norm": 2.6776966029838385,
"learning_rate": 1.4970059880239521e-06,
"loss": 1.5898,
"step": 25
},
{
"epoch": 0.018061408789885613,
"grad_norm": 2.2972572336933976,
"learning_rate": 1.7964071856287426e-06,
"loss": 1.4968,
"step": 30
},
{
"epoch": 0.02107164358819988,
"grad_norm": 2.4202996961757077,
"learning_rate": 2.095808383233533e-06,
"loss": 1.4052,
"step": 35
},
{
"epoch": 0.024081878386514148,
"grad_norm": 1.7252594455655237,
"learning_rate": 2.3952095808383237e-06,
"loss": 1.3382,
"step": 40
},
{
"epoch": 0.027092113184828417,
"grad_norm": 1.247660016611823,
"learning_rate": 2.694610778443114e-06,
"loss": 1.2973,
"step": 45
},
{
"epoch": 0.030102347983142687,
"grad_norm": 1.1871096528896479,
"learning_rate": 2.9940119760479042e-06,
"loss": 1.2516,
"step": 50
},
{
"epoch": 0.033112582781456956,
"grad_norm": 1.062407029099313,
"learning_rate": 3.2934131736526947e-06,
"loss": 1.2233,
"step": 55
},
{
"epoch": 0.036122817579771226,
"grad_norm": 1.074671822399871,
"learning_rate": 3.592814371257485e-06,
"loss": 1.2079,
"step": 60
},
{
"epoch": 0.03913305237808549,
"grad_norm": 1.0045673658364525,
"learning_rate": 3.892215568862276e-06,
"loss": 1.1883,
"step": 65
},
{
"epoch": 0.04214328717639976,
"grad_norm": 0.9990428932639726,
"learning_rate": 4.191616766467066e-06,
"loss": 1.192,
"step": 70
},
{
"epoch": 0.04515352197471403,
"grad_norm": 1.041260547807295,
"learning_rate": 4.4910179640718566e-06,
"loss": 1.1724,
"step": 75
},
{
"epoch": 0.048163756773028296,
"grad_norm": 1.0886242768356977,
"learning_rate": 4.7904191616766475e-06,
"loss": 1.1507,
"step": 80
},
{
"epoch": 0.051173991571342566,
"grad_norm": 1.0469933132322005,
"learning_rate": 5.0898203592814375e-06,
"loss": 1.1648,
"step": 85
},
{
"epoch": 0.054184226369656835,
"grad_norm": 1.0031768043333225,
"learning_rate": 5.389221556886228e-06,
"loss": 1.1557,
"step": 90
},
{
"epoch": 0.057194461167971104,
"grad_norm": 1.0235059168601668,
"learning_rate": 5.6886227544910184e-06,
"loss": 1.1433,
"step": 95
},
{
"epoch": 0.060204695966285374,
"grad_norm": 1.0810995112786321,
"learning_rate": 5.9880239520958085e-06,
"loss": 1.13,
"step": 100
},
{
"epoch": 0.06321493076459964,
"grad_norm": 1.0676305065720928,
"learning_rate": 6.2874251497005985e-06,
"loss": 1.1232,
"step": 105
},
{
"epoch": 0.06622516556291391,
"grad_norm": 1.1136702042740867,
"learning_rate": 6.586826347305389e-06,
"loss": 1.1148,
"step": 110
},
{
"epoch": 0.06923540036122817,
"grad_norm": 1.0798487239001604,
"learning_rate": 6.88622754491018e-06,
"loss": 1.1084,
"step": 115
},
{
"epoch": 0.07224563515954245,
"grad_norm": 1.12910359621852,
"learning_rate": 7.18562874251497e-06,
"loss": 1.1079,
"step": 120
},
{
"epoch": 0.07525586995785671,
"grad_norm": 1.0750989910445923,
"learning_rate": 7.485029940119761e-06,
"loss": 1.1073,
"step": 125
},
{
"epoch": 0.07826610475617098,
"grad_norm": 1.0718510986699241,
"learning_rate": 7.784431137724551e-06,
"loss": 1.0852,
"step": 130
},
{
"epoch": 0.08127633955448525,
"grad_norm": 1.093833315661369,
"learning_rate": 8.083832335329342e-06,
"loss": 1.1007,
"step": 135
},
{
"epoch": 0.08428657435279951,
"grad_norm": 1.089981936562526,
"learning_rate": 8.383233532934131e-06,
"loss": 1.0668,
"step": 140
},
{
"epoch": 0.08729680915111379,
"grad_norm": 1.0716347647949742,
"learning_rate": 8.682634730538922e-06,
"loss": 1.0671,
"step": 145
},
{
"epoch": 0.09030704394942805,
"grad_norm": 1.1760500582380917,
"learning_rate": 8.982035928143713e-06,
"loss": 1.052,
"step": 150
},
{
"epoch": 0.09331727874774233,
"grad_norm": 1.1394001820505688,
"learning_rate": 9.281437125748504e-06,
"loss": 1.0786,
"step": 155
},
{
"epoch": 0.09632751354605659,
"grad_norm": 1.1206173669500894,
"learning_rate": 9.580838323353295e-06,
"loss": 1.069,
"step": 160
},
{
"epoch": 0.09933774834437085,
"grad_norm": 1.1225031587454042,
"learning_rate": 9.880239520958084e-06,
"loss": 1.0407,
"step": 165
},
{
"epoch": 0.10234798314268513,
"grad_norm": 1.0802419173541655,
"learning_rate": 9.999900509954779e-06,
"loss": 1.0378,
"step": 170
},
{
"epoch": 0.1053582179409994,
"grad_norm": 1.1285826979669467,
"learning_rate": 9.999292529572152e-06,
"loss": 1.0404,
"step": 175
},
{
"epoch": 0.10836845273931367,
"grad_norm": 1.3030407887693638,
"learning_rate": 9.998131908181262e-06,
"loss": 1.0368,
"step": 180
},
{
"epoch": 0.11137868753762793,
"grad_norm": 1.13304407504489,
"learning_rate": 9.996418774081658e-06,
"loss": 1.0194,
"step": 185
},
{
"epoch": 0.11438892233594221,
"grad_norm": 1.148326778572453,
"learning_rate": 9.994153316649769e-06,
"loss": 1.0308,
"step": 190
},
{
"epoch": 0.11739915713425647,
"grad_norm": 1.0879087960219742,
"learning_rate": 9.991335786317964e-06,
"loss": 1.0101,
"step": 195
},
{
"epoch": 0.12040939193257075,
"grad_norm": 1.1573320998902041,
"learning_rate": 9.987966494546873e-06,
"loss": 0.9851,
"step": 200
},
{
"epoch": 0.12341962673088501,
"grad_norm": 1.1225156773539111,
"learning_rate": 9.984045813790959e-06,
"loss": 0.9975,
"step": 205
},
{
"epoch": 0.12642986152919927,
"grad_norm": 1.1163693583224434,
"learning_rate": 9.979574177457337e-06,
"loss": 0.9957,
"step": 210
},
{
"epoch": 0.12944009632751355,
"grad_norm": 1.1054539548885067,
"learning_rate": 9.974552079857873e-06,
"loss": 0.9944,
"step": 215
},
{
"epoch": 0.13245033112582782,
"grad_norm": 1.091493000603608,
"learning_rate": 9.968980076154533e-06,
"loss": 1.004,
"step": 220
},
{
"epoch": 0.13546056592414207,
"grad_norm": 1.1518094795338838,
"learning_rate": 9.962858782298023e-06,
"loss": 0.9779,
"step": 225
},
{
"epoch": 0.13847080072245635,
"grad_norm": 1.1074295258081415,
"learning_rate": 9.956188874959686e-06,
"loss": 0.9842,
"step": 230
},
{
"epoch": 0.14148103552077063,
"grad_norm": 1.0975825113307536,
"learning_rate": 9.948971091456715e-06,
"loss": 0.9764,
"step": 235
},
{
"epoch": 0.1444912703190849,
"grad_norm": 1.0992803599557273,
"learning_rate": 9.941206229670634e-06,
"loss": 0.9704,
"step": 240
},
{
"epoch": 0.14750150511739915,
"grad_norm": 1.105958751777608,
"learning_rate": 9.932895147959106e-06,
"loss": 0.9568,
"step": 245
},
{
"epoch": 0.15051173991571343,
"grad_norm": 1.1720050961084032,
"learning_rate": 9.924038765061042e-06,
"loss": 0.942,
"step": 250
},
{
"epoch": 0.1535219747140277,
"grad_norm": 1.116935163255732,
"learning_rate": 9.91463805999504e-06,
"loss": 0.9577,
"step": 255
},
{
"epoch": 0.15653220951234195,
"grad_norm": 1.1268188801983967,
"learning_rate": 9.904694071951167e-06,
"loss": 0.96,
"step": 260
},
{
"epoch": 0.15954244431065623,
"grad_norm": 1.0461466251094855,
"learning_rate": 9.894207900176074e-06,
"loss": 0.9424,
"step": 265
},
{
"epoch": 0.1625526791089705,
"grad_norm": 1.0867508676115896,
"learning_rate": 9.883180703851488e-06,
"loss": 0.948,
"step": 270
},
{
"epoch": 0.16556291390728478,
"grad_norm": 1.1549604201551043,
"learning_rate": 9.871613701966067e-06,
"loss": 0.9444,
"step": 275
},
{
"epoch": 0.16857314870559903,
"grad_norm": 1.0712495395377322,
"learning_rate": 9.859508173180653e-06,
"loss": 0.9245,
"step": 280
},
{
"epoch": 0.1715833835039133,
"grad_norm": 1.0255493028911156,
"learning_rate": 9.846865455686915e-06,
"loss": 0.9309,
"step": 285
},
{
"epoch": 0.17459361830222758,
"grad_norm": 1.0229903229977713,
"learning_rate": 9.833686947059436e-06,
"loss": 0.9247,
"step": 290
},
{
"epoch": 0.17760385310054183,
"grad_norm": 1.0193028798235328,
"learning_rate": 9.819974104101198e-06,
"loss": 0.9252,
"step": 295
},
{
"epoch": 0.1806140878988561,
"grad_norm": 1.026115519573417,
"learning_rate": 9.80572844268256e-06,
"loss": 0.9134,
"step": 300
},
{
"epoch": 0.18362432269717038,
"grad_norm": 1.0506692055629476,
"learning_rate": 9.790951537573686e-06,
"loss": 0.9415,
"step": 305
},
{
"epoch": 0.18663455749548466,
"grad_norm": 1.0790936545456895,
"learning_rate": 9.775645022270448e-06,
"loss": 0.9322,
"step": 310
},
{
"epoch": 0.1896447922937989,
"grad_norm": 1.025793289075722,
"learning_rate": 9.759810588813872e-06,
"loss": 0.928,
"step": 315
},
{
"epoch": 0.19265502709211318,
"grad_norm": 1.1339358401456807,
"learning_rate": 9.743449987603082e-06,
"loss": 0.9245,
"step": 320
},
{
"epoch": 0.19566526189042746,
"grad_norm": 1.1124339419780245,
"learning_rate": 9.726565027201813e-06,
"loss": 0.9422,
"step": 325
},
{
"epoch": 0.1986754966887417,
"grad_norm": 1.0501127986511691,
"learning_rate": 9.70915757413847e-06,
"loss": 0.934,
"step": 330
},
{
"epoch": 0.20168573148705599,
"grad_norm": 1.036530336541463,
"learning_rate": 9.691229552699817e-06,
"loss": 0.913,
"step": 335
},
{
"epoch": 0.20469596628537026,
"grad_norm": 1.0715370900298593,
"learning_rate": 9.672782944718234e-06,
"loss": 0.929,
"step": 340
},
{
"epoch": 0.20770620108368454,
"grad_norm": 1.0842780971098096,
"learning_rate": 9.65381978935266e-06,
"loss": 0.9225,
"step": 345
},
{
"epoch": 0.2107164358819988,
"grad_norm": 1.0776823644021207,
"learning_rate": 9.634342182863163e-06,
"loss": 0.8971,
"step": 350
},
{
"epoch": 0.21372667068031306,
"grad_norm": 1.0787360433631494,
"learning_rate": 9.614352278379217e-06,
"loss": 0.9315,
"step": 355
},
{
"epoch": 0.21673690547862734,
"grad_norm": 1.025638770495932,
"learning_rate": 9.593852285661684e-06,
"loss": 0.9302,
"step": 360
},
{
"epoch": 0.2197471402769416,
"grad_norm": 1.0128686388946724,
"learning_rate": 9.572844470858537e-06,
"loss": 0.9166,
"step": 365
},
{
"epoch": 0.22275737507525586,
"grad_norm": 1.1031191523153046,
"learning_rate": 9.551331156254358e-06,
"loss": 0.9229,
"step": 370
},
{
"epoch": 0.22576760987357014,
"grad_norm": 1.0053017279332288,
"learning_rate": 9.529314720013618e-06,
"loss": 0.9125,
"step": 375
},
{
"epoch": 0.22877784467188442,
"grad_norm": 1.0494040246870382,
"learning_rate": 9.506797595917787e-06,
"loss": 0.9195,
"step": 380
},
{
"epoch": 0.23178807947019867,
"grad_norm": 1.0465974471535013,
"learning_rate": 9.483782273096295e-06,
"loss": 0.9135,
"step": 385
},
{
"epoch": 0.23479831426851294,
"grad_norm": 1.0280771562627427,
"learning_rate": 9.460271295751373e-06,
"loss": 0.9283,
"step": 390
},
{
"epoch": 0.23780854906682722,
"grad_norm": 1.0317863254361297,
"learning_rate": 9.436267262876808e-06,
"loss": 0.9139,
"step": 395
},
{
"epoch": 0.2408187838651415,
"grad_norm": 1.006002207762667,
"learning_rate": 9.411772827970642e-06,
"loss": 0.911,
"step": 400
},
{
"epoch": 0.24382901866345574,
"grad_norm": 1.0330236444943743,
"learning_rate": 9.38679069874184e-06,
"loss": 0.9129,
"step": 405
},
{
"epoch": 0.24683925346177002,
"grad_norm": 1.0502639838743322,
"learning_rate": 9.36132363681097e-06,
"loss": 0.9161,
"step": 410
},
{
"epoch": 0.2498494882600843,
"grad_norm": 1.0378581674586218,
"learning_rate": 9.335374457404928e-06,
"loss": 0.9241,
"step": 415
},
{
"epoch": 0.25285972305839854,
"grad_norm": 1.0263844252114644,
"learning_rate": 9.308946029045726e-06,
"loss": 0.8987,
"step": 420
},
{
"epoch": 0.25586995785671285,
"grad_norm": 1.032423826827243,
"learning_rate": 9.282041273233402e-06,
"loss": 0.894,
"step": 425
},
{
"epoch": 0.2588801926550271,
"grad_norm": 1.026066371094123,
"learning_rate": 9.254663164123052e-06,
"loss": 0.8999,
"step": 430
},
{
"epoch": 0.26189042745334135,
"grad_norm": 1.0694165068256545,
"learning_rate": 9.226814728196072e-06,
"loss": 0.8997,
"step": 435
},
{
"epoch": 0.26490066225165565,
"grad_norm": 1.0389056305644437,
"learning_rate": 9.198499043925591e-06,
"loss": 0.903,
"step": 440
},
{
"epoch": 0.2679108970499699,
"grad_norm": 1.0691386715819782,
"learning_rate": 9.169719241436162e-06,
"loss": 0.9077,
"step": 445
},
{
"epoch": 0.27092113184828415,
"grad_norm": 1.0403573239959019,
"learning_rate": 9.14047850215775e-06,
"loss": 0.8856,
"step": 450
},
{
"epoch": 0.27393136664659845,
"grad_norm": 1.0321572558705492,
"learning_rate": 9.110780058474052e-06,
"loss": 0.8884,
"step": 455
},
{
"epoch": 0.2769416014449127,
"grad_norm": 1.0428857005082819,
"learning_rate": 9.080627193365155e-06,
"loss": 0.9014,
"step": 460
},
{
"epoch": 0.27995183624322695,
"grad_norm": 1.0172513904579585,
"learning_rate": 9.050023240044649e-06,
"loss": 0.9041,
"step": 465
},
{
"epoch": 0.28296207104154125,
"grad_norm": 1.036966322792995,
"learning_rate": 9.018971581591141e-06,
"loss": 0.8921,
"step": 470
},
{
"epoch": 0.2859723058398555,
"grad_norm": 1.0042423042637907,
"learning_rate": 8.987475650574289e-06,
"loss": 0.9183,
"step": 475
},
{
"epoch": 0.2889825406381698,
"grad_norm": 1.0812852684698064,
"learning_rate": 8.955538928675343e-06,
"loss": 0.8838,
"step": 480
},
{
"epoch": 0.29199277543648405,
"grad_norm": 1.0364976212069876,
"learning_rate": 8.923164946302274e-06,
"loss": 0.9004,
"step": 485
},
{
"epoch": 0.2950030102347983,
"grad_norm": 1.025560565799461,
"learning_rate": 8.890357282199504e-06,
"loss": 0.9082,
"step": 490
},
{
"epoch": 0.2980132450331126,
"grad_norm": 1.0120033023068284,
"learning_rate": 8.857119563052301e-06,
"loss": 0.8803,
"step": 495
},
{
"epoch": 0.30102347983142685,
"grad_norm": 1.0202841174928665,
"learning_rate": 8.823455463085873e-06,
"loss": 0.8996,
"step": 500
},
{
"epoch": 0.3040337146297411,
"grad_norm": 1.0422192444275267,
"learning_rate": 8.789368703659199e-06,
"loss": 0.9037,
"step": 505
},
{
"epoch": 0.3070439494280554,
"grad_norm": 1.039216841812148,
"learning_rate": 8.754863052853658e-06,
"loss": 0.8962,
"step": 510
},
{
"epoch": 0.31005418422636966,
"grad_norm": 1.03300361427746,
"learning_rate": 8.719942325056496e-06,
"loss": 0.8751,
"step": 515
},
{
"epoch": 0.3130644190246839,
"grad_norm": 1.056478959906002,
"learning_rate": 8.68461038053916e-06,
"loss": 0.8828,
"step": 520
},
{
"epoch": 0.3160746538229982,
"grad_norm": 1.091546086658953,
"learning_rate": 8.648871125030576e-06,
"loss": 0.8854,
"step": 525
},
{
"epoch": 0.31908488862131246,
"grad_norm": 1.0222370552684752,
"learning_rate": 8.612728509285395e-06,
"loss": 0.8798,
"step": 530
},
{
"epoch": 0.3220951234196267,
"grad_norm": 0.9917875099200865,
"learning_rate": 8.576186528647253e-06,
"loss": 0.8982,
"step": 535
},
{
"epoch": 0.325105358217941,
"grad_norm": 0.9754319321885829,
"learning_rate": 8.53924922260712e-06,
"loss": 0.8871,
"step": 540
},
{
"epoch": 0.32811559301625526,
"grad_norm": 1.0272701098817443,
"learning_rate": 8.501920674356755e-06,
"loss": 0.8713,
"step": 545
},
{
"epoch": 0.33112582781456956,
"grad_norm": 1.0815374085785938,
"learning_rate": 8.46420501033733e-06,
"loss": 0.8736,
"step": 550
},
{
"epoch": 0.3341360626128838,
"grad_norm": 1.073357925922567,
"learning_rate": 8.42610639978329e-06,
"loss": 0.8802,
"step": 555
},
{
"epoch": 0.33714629741119806,
"grad_norm": 1.0305445781682847,
"learning_rate": 8.387629054261454e-06,
"loss": 0.8758,
"step": 560
},
{
"epoch": 0.34015653220951236,
"grad_norm": 1.038417001227087,
"learning_rate": 8.348777227205462e-06,
"loss": 0.875,
"step": 565
},
{
"epoch": 0.3431667670078266,
"grad_norm": 1.072317877743011,
"learning_rate": 8.309555213445583e-06,
"loss": 0.8942,
"step": 570
},
{
"epoch": 0.34617700180614086,
"grad_norm": 1.0587298691073423,
"learning_rate": 8.269967348733947e-06,
"loss": 0.869,
"step": 575
},
{
"epoch": 0.34918723660445516,
"grad_norm": 1.0204733096808816,
"learning_rate": 8.230018009265255e-06,
"loss": 0.8814,
"step": 580
},
{
"epoch": 0.3521974714027694,
"grad_norm": 1.0052226002834497,
"learning_rate": 8.189711611193012e-06,
"loss": 0.877,
"step": 585
},
{
"epoch": 0.35520770620108366,
"grad_norm": 1.041633913402688,
"learning_rate": 8.149052610141357e-06,
"loss": 0.8655,
"step": 590
},
{
"epoch": 0.35821794099939797,
"grad_norm": 1.0277280895997796,
"learning_rate": 8.108045500712518e-06,
"loss": 0.8827,
"step": 595
},
{
"epoch": 0.3612281757977122,
"grad_norm": 1.0193255596334292,
"learning_rate": 8.066694815989961e-06,
"loss": 0.8842,
"step": 600
},
{
"epoch": 0.36423841059602646,
"grad_norm": 1.0832182052623744,
"learning_rate": 8.025005127037282e-06,
"loss": 0.8823,
"step": 605
},
{
"epoch": 0.36724864539434077,
"grad_norm": 1.0281183723809044,
"learning_rate": 7.982981042392907e-06,
"loss": 0.8714,
"step": 610
},
{
"epoch": 0.370258880192655,
"grad_norm": 1.038122520133344,
"learning_rate": 7.940627207560655e-06,
"loss": 0.8657,
"step": 615
},
{
"epoch": 0.3732691149909693,
"grad_norm": 1.024914409371675,
"learning_rate": 7.897948304496189e-06,
"loss": 0.8892,
"step": 620
},
{
"epoch": 0.37627934978928357,
"grad_norm": 1.0218762677764746,
"learning_rate": 7.854949051089467e-06,
"loss": 0.8908,
"step": 625
},
{
"epoch": 0.3792895845875978,
"grad_norm": 1.0227020273788703,
"learning_rate": 7.811634200643202e-06,
"loss": 0.8959,
"step": 630
},
{
"epoch": 0.3822998193859121,
"grad_norm": 1.0082689155687972,
"learning_rate": 7.768008541347423e-06,
"loss": 0.8526,
"step": 635
},
{
"epoch": 0.38531005418422637,
"grad_norm": 1.0462978268815175,
"learning_rate": 7.72407689575016e-06,
"loss": 0.8735,
"step": 640
},
{
"epoch": 0.3883202889825406,
"grad_norm": 1.0709673134969584,
"learning_rate": 7.67984412022434e-06,
"loss": 0.8689,
"step": 645
},
{
"epoch": 0.3913305237808549,
"grad_norm": 1.060848078120318,
"learning_rate": 7.635315104430959e-06,
"loss": 0.8661,
"step": 650
},
{
"epoch": 0.39434075857916917,
"grad_norm": 1.0044743257959272,
"learning_rate": 7.5904947707785434e-06,
"loss": 0.8502,
"step": 655
},
{
"epoch": 0.3973509933774834,
"grad_norm": 1.0265593650109908,
"learning_rate": 7.545388073879018e-06,
"loss": 0.8825,
"step": 660
},
{
"epoch": 0.4003612281757977,
"grad_norm": 1.0627184843788156,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8646,
"step": 665
},
{
"epoch": 0.40337146297411197,
"grad_norm": 1.0318543066146353,
"learning_rate": 7.454335566513603e-06,
"loss": 0.8701,
"step": 670
},
{
"epoch": 0.4063816977724263,
"grad_norm": 1.083589178780354,
"learning_rate": 7.408399821341787e-06,
"loss": 0.8764,
"step": 675
},
{
"epoch": 0.4093919325707405,
"grad_norm": 1.0109683298031107,
"learning_rate": 7.362197842398355e-06,
"loss": 0.8595,
"step": 680
},
{
"epoch": 0.4124021673690548,
"grad_norm": 1.047285422476373,
"learning_rate": 7.315734737027612e-06,
"loss": 0.8668,
"step": 685
},
{
"epoch": 0.4154124021673691,
"grad_norm": 1.0406374636538134,
"learning_rate": 7.2690156414397775e-06,
"loss": 0.8501,
"step": 690
},
{
"epoch": 0.4184226369656833,
"grad_norm": 1.0408378978282196,
"learning_rate": 7.22204572014322e-06,
"loss": 0.8541,
"step": 695
},
{
"epoch": 0.4214328717639976,
"grad_norm": 1.0597504963355342,
"learning_rate": 7.174830165373542e-06,
"loss": 0.8652,
"step": 700
},
{
"epoch": 0.4244431065623119,
"grad_norm": 1.0165015320978932,
"learning_rate": 7.127374196519616e-06,
"loss": 0.8552,
"step": 705
},
{
"epoch": 0.4274533413606261,
"grad_norm": 1.0245034045179864,
"learning_rate": 7.079683059546607e-06,
"loss": 0.8647,
"step": 710
},
{
"epoch": 0.4304635761589404,
"grad_norm": 1.0138247798670463,
"learning_rate": 7.031762026416074e-06,
"loss": 0.8508,
"step": 715
},
{
"epoch": 0.4334738109572547,
"grad_norm": 1.0798214937909605,
"learning_rate": 6.983616394503177e-06,
"loss": 0.8674,
"step": 720
},
{
"epoch": 0.43648404575556893,
"grad_norm": 1.0744753655006976,
"learning_rate": 6.9352514860110876e-06,
"loss": 0.8838,
"step": 725
},
{
"epoch": 0.4394942805538832,
"grad_norm": 1.0193393304874436,
"learning_rate": 6.886672647382653e-06,
"loss": 0.8556,
"step": 730
},
{
"epoch": 0.4425045153521975,
"grad_norm": 1.0241683235911196,
"learning_rate": 6.837885248709386e-06,
"loss": 0.8869,
"step": 735
},
{
"epoch": 0.44551475015051173,
"grad_norm": 1.0560789050729424,
"learning_rate": 6.788894683137822e-06,
"loss": 0.8554,
"step": 740
},
{
"epoch": 0.44852498494882603,
"grad_norm": 1.0320197701869642,
"learning_rate": 6.739706366273346e-06,
"loss": 0.8645,
"step": 745
},
{
"epoch": 0.4515352197471403,
"grad_norm": 0.9970260367357067,
"learning_rate": 6.690325735581532e-06,
"loss": 0.8502,
"step": 750
},
{
"epoch": 0.45454545454545453,
"grad_norm": 1.0077402138627367,
"learning_rate": 6.640758249787067e-06,
"loss": 0.867,
"step": 755
},
{
"epoch": 0.45755568934376883,
"grad_norm": 1.0638747456346818,
"learning_rate": 6.591009388270315e-06,
"loss": 0.8586,
"step": 760
},
{
"epoch": 0.4605659241420831,
"grad_norm": 1.0750019486293267,
"learning_rate": 6.54108465046161e-06,
"loss": 0.8629,
"step": 765
},
{
"epoch": 0.46357615894039733,
"grad_norm": 1.0281089034578543,
"learning_rate": 6.490989555233328e-06,
"loss": 0.8567,
"step": 770
},
{
"epoch": 0.46658639373871164,
"grad_norm": 1.0488620364160015,
"learning_rate": 6.440729640289809e-06,
"loss": 0.8528,
"step": 775
},
{
"epoch": 0.4695966285370259,
"grad_norm": 1.0249636337962247,
"learning_rate": 6.3903104615551956e-06,
"loss": 0.8458,
"step": 780
},
{
"epoch": 0.47260686333534013,
"grad_norm": 1.0419830712006013,
"learning_rate": 6.3397375925592675e-06,
"loss": 0.8307,
"step": 785
},
{
"epoch": 0.47561709813365444,
"grad_norm": 1.0164769677067202,
"learning_rate": 6.289016623821308e-06,
"loss": 0.8565,
"step": 790
},
{
"epoch": 0.4786273329319687,
"grad_norm": 1.0228613150621668,
"learning_rate": 6.2381531622321234e-06,
"loss": 0.8573,
"step": 795
},
{
"epoch": 0.481637567730283,
"grad_norm": 1.075852284931363,
"learning_rate": 6.18715283043422e-06,
"loss": 0.8494,
"step": 800
},
{
"epoch": 0.48464780252859724,
"grad_norm": 1.0349508902074944,
"learning_rate": 6.136021266200271e-06,
"loss": 0.8563,
"step": 805
},
{
"epoch": 0.4876580373269115,
"grad_norm": 1.0924038198527715,
"learning_rate": 6.084764121809878e-06,
"loss": 0.8479,
"step": 810
},
{
"epoch": 0.4906682721252258,
"grad_norm": 1.0471168501487311,
"learning_rate": 6.033387063424765e-06,
"loss": 0.8436,
"step": 815
},
{
"epoch": 0.49367850692354004,
"grad_norm": 1.0224660906723329,
"learning_rate": 5.9818957704624046e-06,
"loss": 0.8654,
"step": 820
},
{
"epoch": 0.4966887417218543,
"grad_norm": 1.0211655258926158,
"learning_rate": 5.930295934968197e-06,
"loss": 0.8463,
"step": 825
},
{
"epoch": 0.4996989765201686,
"grad_norm": 1.0195137398082676,
"learning_rate": 5.878593260986256e-06,
"loss": 0.8465,
"step": 830
},
{
"epoch": 0.5027092113184829,
"grad_norm": 1.0602817409712544,
"learning_rate": 5.8267934639288525e-06,
"loss": 0.8608,
"step": 835
},
{
"epoch": 0.5057194461167971,
"grad_norm": 1.061046137706525,
"learning_rate": 5.77490226994462e-06,
"loss": 0.8609,
"step": 840
},
{
"epoch": 0.5087296809151114,
"grad_norm": 1.051889967104703,
"learning_rate": 5.722925415285555e-06,
"loss": 0.8664,
"step": 845
},
{
"epoch": 0.5117399157134257,
"grad_norm": 1.072317194927684,
"learning_rate": 5.670868645672916e-06,
"loss": 0.8583,
"step": 850
},
{
"epoch": 0.5147501505117399,
"grad_norm": 1.0465774872570925,
"learning_rate": 5.618737715662067e-06,
"loss": 0.848,
"step": 855
},
{
"epoch": 0.5177603853100542,
"grad_norm": 1.029456210016313,
"learning_rate": 5.566538388006351e-06,
"loss": 0.8515,
"step": 860
},
{
"epoch": 0.5207706201083685,
"grad_norm": 1.0018249989098675,
"learning_rate": 5.514276433020044e-06,
"loss": 0.8426,
"step": 865
},
{
"epoch": 0.5237808549066827,
"grad_norm": 1.0794333658356523,
"learning_rate": 5.461957627940489e-06,
"loss": 0.8411,
"step": 870
},
{
"epoch": 0.526791089704997,
"grad_norm": 1.0658005957225543,
"learning_rate": 5.409587756289462e-06,
"loss": 0.8372,
"step": 875
},
{
"epoch": 0.5298013245033113,
"grad_norm": 1.0361101879365158,
"learning_rate": 5.357172607233831e-06,
"loss": 0.8421,
"step": 880
},
{
"epoch": 0.5328115593016255,
"grad_norm": 1.086943198789344,
"learning_rate": 5.304717974945596e-06,
"loss": 0.8403,
"step": 885
},
{
"epoch": 0.5358217940999398,
"grad_norm": 1.0481795809429144,
"learning_rate": 5.252229657961394e-06,
"loss": 0.8441,
"step": 890
},
{
"epoch": 0.5388320288982541,
"grad_norm": 1.0138631327859517,
"learning_rate": 5.199713458541495e-06,
"loss": 0.8506,
"step": 895
},
{
"epoch": 0.5418422636965683,
"grad_norm": 1.0256732385631195,
"learning_rate": 5.1471751820284e-06,
"loss": 0.8347,
"step": 900
},
{
"epoch": 0.5448524984948826,
"grad_norm": 1.0356944481021555,
"learning_rate": 5.094620636205096e-06,
"loss": 0.8384,
"step": 905
},
{
"epoch": 0.5478627332931969,
"grad_norm": 1.022741189017134,
"learning_rate": 5.042055630653042e-06,
"loss": 0.8394,
"step": 910
},
{
"epoch": 0.5508729680915111,
"grad_norm": 1.0919072566453618,
"learning_rate": 4.98948597610996e-06,
"loss": 0.834,
"step": 915
},
{
"epoch": 0.5538832028898254,
"grad_norm": 1.037713436516078,
"learning_rate": 4.936917483827483e-06,
"loss": 0.8284,
"step": 920
},
{
"epoch": 0.5568934376881397,
"grad_norm": 1.0400429976889096,
"learning_rate": 4.884355964928767e-06,
"loss": 0.8635,
"step": 925
},
{
"epoch": 0.5599036724864539,
"grad_norm": 0.9928760988451867,
"learning_rate": 4.831807229766101e-06,
"loss": 0.8255,
"step": 930
},
{
"epoch": 0.5629139072847682,
"grad_norm": 1.1246842464041702,
"learning_rate": 4.779277087278615e-06,
"loss": 0.8395,
"step": 935
},
{
"epoch": 0.5659241420830825,
"grad_norm": 1.0309925607948338,
"learning_rate": 4.7267713443501274e-06,
"loss": 0.8544,
"step": 940
},
{
"epoch": 0.5689343768813967,
"grad_norm": 1.0124528992084696,
"learning_rate": 4.67429580516724e-06,
"loss": 0.8344,
"step": 945
},
{
"epoch": 0.571944611679711,
"grad_norm": 1.0217717889794504,
"learning_rate": 4.6218562705777185e-06,
"loss": 0.8295,
"step": 950
},
{
"epoch": 0.5749548464780253,
"grad_norm": 1.0542480246959662,
"learning_rate": 4.5694585374492314e-06,
"loss": 0.8375,
"step": 955
},
{
"epoch": 0.5779650812763396,
"grad_norm": 1.046871667593415,
"learning_rate": 4.517108398028566e-06,
"loss": 0.849,
"step": 960
},
{
"epoch": 0.5809753160746538,
"grad_norm": 1.0378115278601967,
"learning_rate": 4.464811639301314e-06,
"loss": 0.8331,
"step": 965
},
{
"epoch": 0.5839855508729681,
"grad_norm": 1.0560995260768353,
"learning_rate": 4.412574042352156e-06,
"loss": 0.8472,
"step": 970
},
{
"epoch": 0.5869957856712824,
"grad_norm": 1.0687109447058267,
"learning_rate": 4.360401381725806e-06,
"loss": 0.8317,
"step": 975
},
{
"epoch": 0.5900060204695966,
"grad_norm": 1.0523892138754463,
"learning_rate": 4.308299424788667e-06,
"loss": 0.85,
"step": 980
},
{
"epoch": 0.5930162552679109,
"grad_norm": 1.0187828271643544,
"learning_rate": 4.256273931091284e-06,
"loss": 0.8395,
"step": 985
},
{
"epoch": 0.5960264900662252,
"grad_norm": 1.0220180896984419,
"learning_rate": 4.204330651731662e-06,
"loss": 0.8329,
"step": 990
},
{
"epoch": 0.5990367248645394,
"grad_norm": 1.0205596148401899,
"learning_rate": 4.152475328719517e-06,
"loss": 0.831,
"step": 995
},
{
"epoch": 0.6020469596628537,
"grad_norm": 1.040912209584121,
"learning_rate": 4.1007136943415325e-06,
"loss": 0.8201,
"step": 1000
},
{
"epoch": 0.605057194461168,
"grad_norm": 1.0001879021313027,
"learning_rate": 4.049051470527692e-06,
"loss": 0.8423,
"step": 1005
},
{
"epoch": 0.6080674292594822,
"grad_norm": 1.0398076808150207,
"learning_rate": 3.997494368218745e-06,
"loss": 0.8309,
"step": 1010
},
{
"epoch": 0.6110776640577965,
"grad_norm": 1.0959850286337298,
"learning_rate": 3.946048086734921e-06,
"loss": 0.8471,
"step": 1015
},
{
"epoch": 0.6140878988561108,
"grad_norm": 1.074686489630975,
"learning_rate": 3.894718313145873e-06,
"loss": 0.8361,
"step": 1020
},
{
"epoch": 0.617098133654425,
"grad_norm": 1.033345644977548,
"learning_rate": 3.843510721642036e-06,
"loss": 0.8228,
"step": 1025
},
{
"epoch": 0.6201083684527393,
"grad_norm": 1.0159701920636455,
"learning_rate": 3.7924309729073616e-06,
"loss": 0.8281,
"step": 1030
},
{
"epoch": 0.6231186032510536,
"grad_norm": 1.0886803763608572,
"learning_rate": 3.7414847134935716e-06,
"loss": 0.8282,
"step": 1035
},
{
"epoch": 0.6261288380493678,
"grad_norm": 1.0677088993104988,
"learning_rate": 3.6906775751959667e-06,
"loss": 0.8305,
"step": 1040
},
{
"epoch": 0.6291390728476821,
"grad_norm": 1.0672615560402277,
"learning_rate": 3.640015174430864e-06,
"loss": 0.8131,
"step": 1045
},
{
"epoch": 0.6321493076459964,
"grad_norm": 1.0365030160987536,
"learning_rate": 3.5895031116147355e-06,
"loss": 0.829,
"step": 1050
},
{
"epoch": 0.6351595424443106,
"grad_norm": 1.0441297603160167,
"learning_rate": 3.539146970545124e-06,
"loss": 0.8311,
"step": 1055
},
{
"epoch": 0.6381697772426249,
"grad_norm": 1.0678910916370306,
"learning_rate": 3.488952317783374e-06,
"loss": 0.8251,
"step": 1060
},
{
"epoch": 0.6411800120409392,
"grad_norm": 1.0182442454102976,
"learning_rate": 3.438924702039301e-06,
"loss": 0.824,
"step": 1065
},
{
"epoch": 0.6441902468392534,
"grad_norm": 1.0362344445465297,
"learning_rate": 3.389069653557805e-06,
"loss": 0.8324,
"step": 1070
},
{
"epoch": 0.6472004816375677,
"grad_norm": 1.0321140319577164,
"learning_rate": 3.3393926835075307e-06,
"loss": 0.8293,
"step": 1075
},
{
"epoch": 0.650210716435882,
"grad_norm": 1.0518576177949832,
"learning_rate": 3.289899283371657e-06,
"loss": 0.8164,
"step": 1080
},
{
"epoch": 0.6532209512341962,
"grad_norm": 1.117410399435437,
"learning_rate": 3.240594924340835e-06,
"loss": 0.8215,
"step": 1085
},
{
"epoch": 0.6562311860325105,
"grad_norm": 1.0503940857197025,
"learning_rate": 3.1914850567083866e-06,
"loss": 0.8276,
"step": 1090
},
{
"epoch": 0.6592414208308248,
"grad_norm": 1.0752717357492687,
"learning_rate": 3.1425751092678064e-06,
"loss": 0.8154,
"step": 1095
},
{
"epoch": 0.6622516556291391,
"grad_norm": 1.0233560268626287,
"learning_rate": 3.0938704887126425e-06,
"loss": 0.8094,
"step": 1100
},
{
"epoch": 0.6652618904274533,
"grad_norm": 1.011199570119972,
"learning_rate": 3.045376579038821e-06,
"loss": 0.812,
"step": 1105
},
{
"epoch": 0.6682721252257676,
"grad_norm": 1.0515169328232772,
"learning_rate": 2.9970987409494784e-06,
"loss": 0.8456,
"step": 1110
},
{
"epoch": 0.6712823600240819,
"grad_norm": 1.0464985847353594,
"learning_rate": 2.9490423112623646e-06,
"loss": 0.8166,
"step": 1115
},
{
"epoch": 0.6742925948223961,
"grad_norm": 1.0390545724673643,
"learning_rate": 2.9012126023198973e-06,
"loss": 0.8237,
"step": 1120
},
{
"epoch": 0.6773028296207104,
"grad_norm": 1.0422664657187024,
"learning_rate": 2.853614901401909e-06,
"loss": 0.8044,
"step": 1125
},
{
"epoch": 0.6803130644190247,
"grad_norm": 1.027410279635157,
"learning_rate": 2.806254470141174e-06,
"loss": 0.8261,
"step": 1130
},
{
"epoch": 0.6833232992173389,
"grad_norm": 1.0096212394827309,
"learning_rate": 2.759136543941773e-06,
"loss": 0.8002,
"step": 1135
},
{
"epoch": 0.6863335340156532,
"grad_norm": 1.0314561754682732,
"learning_rate": 2.712266331400332e-06,
"loss": 0.8084,
"step": 1140
},
{
"epoch": 0.6893437688139675,
"grad_norm": 1.0393657468292272,
"learning_rate": 2.66564901373027e-06,
"loss": 0.8203,
"step": 1145
},
{
"epoch": 0.6923540036122817,
"grad_norm": 1.0321818133370786,
"learning_rate": 2.6192897441890337e-06,
"loss": 0.8104,
"step": 1150
},
{
"epoch": 0.695364238410596,
"grad_norm": 1.0333683135026992,
"learning_rate": 2.573193647508426e-06,
"loss": 0.8135,
"step": 1155
},
{
"epoch": 0.6983744732089103,
"grad_norm": 0.9971600689374603,
"learning_rate": 2.5273658193281252e-06,
"loss": 0.8014,
"step": 1160
},
{
"epoch": 0.7013847080072245,
"grad_norm": 1.0714545318723498,
"learning_rate": 2.4818113256323745e-06,
"loss": 0.8178,
"step": 1165
},
{
"epoch": 0.7043949428055388,
"grad_norm": 1.0688985938039246,
"learning_rate": 2.4365352021899635e-06,
"loss": 0.8204,
"step": 1170
},
{
"epoch": 0.7074051776038531,
"grad_norm": 1.0354606458390736,
"learning_rate": 2.391542453997578e-06,
"loss": 0.8089,
"step": 1175
},
{
"epoch": 0.7104154124021673,
"grad_norm": 1.0579010566201743,
"learning_rate": 2.346838054726505e-06,
"loss": 0.8093,
"step": 1180
},
{
"epoch": 0.7134256472004816,
"grad_norm": 1.0289112089155492,
"learning_rate": 2.302426946172836e-06,
"loss": 0.8239,
"step": 1185
},
{
"epoch": 0.7164358819987959,
"grad_norm": 1.0382382418654816,
"learning_rate": 2.258314037711184e-06,
"loss": 0.8158,
"step": 1190
},
{
"epoch": 0.7194461167971101,
"grad_norm": 1.039809551524263,
"learning_rate": 2.214504205751971e-06,
"loss": 0.8324,
"step": 1195
},
{
"epoch": 0.7224563515954244,
"grad_norm": 1.0044596359240077,
"learning_rate": 2.1710022932023805e-06,
"loss": 0.8181,
"step": 1200
},
{
"epoch": 0.7254665863937387,
"grad_norm": 1.0210161092801369,
"learning_rate": 2.127813108931007e-06,
"loss": 0.819,
"step": 1205
},
{
"epoch": 0.7284768211920529,
"grad_norm": 1.0190275126749935,
"learning_rate": 2.084941427236245e-06,
"loss": 0.8215,
"step": 1210
},
{
"epoch": 0.7314870559903672,
"grad_norm": 1.0441883702294767,
"learning_rate": 2.04239198731855e-06,
"loss": 0.8108,
"step": 1215
},
{
"epoch": 0.7344972907886815,
"grad_norm": 1.0305419985469728,
"learning_rate": 2.000169492756523e-06,
"loss": 0.821,
"step": 1220
},
{
"epoch": 0.7375075255869958,
"grad_norm": 1.0474048248439474,
"learning_rate": 1.9582786109869713e-06,
"loss": 0.8173,
"step": 1225
},
{
"epoch": 0.74051776038531,
"grad_norm": 1.047276514055598,
"learning_rate": 1.9167239727889527e-06,
"loss": 0.8182,
"step": 1230
},
{
"epoch": 0.7435279951836243,
"grad_norm": 1.0650577544866475,
"learning_rate": 1.875510171771865e-06,
"loss": 0.8047,
"step": 1235
},
{
"epoch": 0.7465382299819386,
"grad_norm": 1.039137337049237,
"learning_rate": 1.8346417638676533e-06,
"loss": 0.8004,
"step": 1240
},
{
"epoch": 0.7495484647802528,
"grad_norm": 1.0943258261224829,
"learning_rate": 1.7941232668271863e-06,
"loss": 0.8223,
"step": 1245
},
{
"epoch": 0.7525586995785671,
"grad_norm": 1.049969345568919,
"learning_rate": 1.753959159720836e-06,
"loss": 0.8124,
"step": 1250
},
{
"epoch": 0.7555689343768814,
"grad_norm": 1.0323176464169084,
"learning_rate": 1.7141538824433506e-06,
"loss": 0.8163,
"step": 1255
},
{
"epoch": 0.7585791691751956,
"grad_norm": 1.0713783017354832,
"learning_rate": 1.6747118352230495e-06,
"loss": 0.812,
"step": 1260
},
{
"epoch": 0.7615894039735099,
"grad_norm": 1.0307746629765189,
"learning_rate": 1.6356373781354058e-06,
"loss": 0.7918,
"step": 1265
},
{
"epoch": 0.7645996387718242,
"grad_norm": 1.0450820274010493,
"learning_rate": 1.5969348306210692e-06,
"loss": 0.8095,
"step": 1270
},
{
"epoch": 0.7676098735701384,
"grad_norm": 1.0579929378641193,
"learning_rate": 1.5586084710083737e-06,
"loss": 0.8127,
"step": 1275
},
{
"epoch": 0.7706201083684527,
"grad_norm": 1.0235264382567837,
"learning_rate": 1.5206625360403943e-06,
"loss": 0.8056,
"step": 1280
},
{
"epoch": 0.773630343166767,
"grad_norm": 0.9995792397694476,
"learning_rate": 1.4831012204066114e-06,
"loss": 0.8165,
"step": 1285
},
{
"epoch": 0.7766405779650812,
"grad_norm": 1.0207367210720288,
"learning_rate": 1.445928676279199e-06,
"loss": 0.7994,
"step": 1290
},
{
"epoch": 0.7796508127633955,
"grad_norm": 1.0378813038666916,
"learning_rate": 1.4091490128540374e-06,
"loss": 0.7898,
"step": 1295
},
{
"epoch": 0.7826610475617098,
"grad_norm": 1.0170522177446832,
"learning_rate": 1.3727662958964627e-06,
"loss": 0.8059,
"step": 1300
},
{
"epoch": 0.785671282360024,
"grad_norm": 1.002759687451583,
"learning_rate": 1.3367845472918272e-06,
"loss": 0.8145,
"step": 1305
},
{
"epoch": 0.7886815171583383,
"grad_norm": 1.0717784571676352,
"learning_rate": 1.3012077446008969e-06,
"loss": 0.8078,
"step": 1310
},
{
"epoch": 0.7916917519566526,
"grad_norm": 1.0266680645611,
"learning_rate": 1.266039820620159e-06,
"loss": 0.8082,
"step": 1315
},
{
"epoch": 0.7947019867549668,
"grad_norm": 1.0418056632952235,
"learning_rate": 1.2312846629470826e-06,
"loss": 0.7941,
"step": 1320
},
{
"epoch": 0.7977122215532811,
"grad_norm": 1.0330602472030925,
"learning_rate": 1.1969461135503573e-06,
"loss": 0.8089,
"step": 1325
},
{
"epoch": 0.8007224563515954,
"grad_norm": 1.0326801178315057,
"learning_rate": 1.163027968345195e-06,
"loss": 0.8092,
"step": 1330
},
{
"epoch": 0.8037326911499096,
"grad_norm": 1.0596116976868275,
"learning_rate": 1.1295339767737125e-06,
"loss": 0.7975,
"step": 1335
},
{
"epoch": 0.8067429259482239,
"grad_norm": 1.0438113622873293,
"learning_rate": 1.0964678413904529e-06,
"loss": 0.8107,
"step": 1340
},
{
"epoch": 0.8097531607465382,
"grad_norm": 1.0219389795912814,
"learning_rate": 1.0638332174530953e-06,
"loss": 0.8192,
"step": 1345
},
{
"epoch": 0.8127633955448526,
"grad_norm": 1.036771538286032,
"learning_rate": 1.0316337125183817e-06,
"loss": 0.8067,
"step": 1350
},
{
"epoch": 0.8157736303431667,
"grad_norm": 1.025748011070849,
"learning_rate": 9.998728860433277e-07,
"loss": 0.804,
"step": 1355
},
{
"epoch": 0.818783865141481,
"grad_norm": 1.0266024772426974,
"learning_rate": 9.685542489917494e-07,
"loss": 0.7877,
"step": 1360
},
{
"epoch": 0.8217940999397954,
"grad_norm": 1.053826897283154,
"learning_rate": 9.376812634461418e-07,
"loss": 0.796,
"step": 1365
},
{
"epoch": 0.8248043347381095,
"grad_norm": 1.0434406635611175,
"learning_rate": 9.072573422249692e-07,
"loss": 0.8172,
"step": 1370
},
{
"epoch": 0.8278145695364238,
"grad_norm": 1.0537547493539212,
"learning_rate": 8.772858485054042e-07,
"loss": 0.8122,
"step": 1375
},
{
"epoch": 0.8308248043347382,
"grad_norm": 1.0237142125748788,
"learning_rate": 8.477700954515372e-07,
"loss": 0.8084,
"step": 1380
},
{
"epoch": 0.8338350391330523,
"grad_norm": 1.021758110729101,
"learning_rate": 8.187133458481416e-07,
"loss": 0.8096,
"step": 1385
},
{
"epoch": 0.8368452739313667,
"grad_norm": 1.038850029152452,
"learning_rate": 7.901188117399817e-07,
"loss": 0.8146,
"step": 1390
},
{
"epoch": 0.839855508729681,
"grad_norm": 1.056879030430651,
"learning_rate": 7.619896540767435e-07,
"loss": 0.8205,
"step": 1395
},
{
"epoch": 0.8428657435279951,
"grad_norm": 1.0410235485563226,
"learning_rate": 7.343289823636168e-07,
"loss": 0.813,
"step": 1400
},
{
"epoch": 0.8458759783263095,
"grad_norm": 1.059559749232041,
"learning_rate": 7.0713985431755e-07,
"loss": 0.811,
"step": 1405
},
{
"epoch": 0.8488862131246238,
"grad_norm": 1.0070532616145602,
"learning_rate": 6.804252755292429e-07,
"loss": 0.7892,
"step": 1410
},
{
"epoch": 0.851896447922938,
"grad_norm": 1.039370261590547,
"learning_rate": 6.541881991309013e-07,
"loss": 0.8055,
"step": 1415
},
{
"epoch": 0.8549066827212523,
"grad_norm": 1.0424718469075611,
"learning_rate": 6.284315254697726e-07,
"loss": 0.8089,
"step": 1420
},
{
"epoch": 0.8579169175195666,
"grad_norm": 1.032728933766913,
"learning_rate": 6.031581017875482e-07,
"loss": 0.8109,
"step": 1425
},
{
"epoch": 0.8609271523178808,
"grad_norm": 1.0241874679127938,
"learning_rate": 5.783707219056078e-07,
"loss": 0.8056,
"step": 1430
},
{
"epoch": 0.863937387116195,
"grad_norm": 1.0185177289645166,
"learning_rate": 5.540721259161774e-07,
"loss": 0.808,
"step": 1435
},
{
"epoch": 0.8669476219145094,
"grad_norm": 1.0446166216421613,
"learning_rate": 5.302649998794368e-07,
"loss": 0.8157,
"step": 1440
},
{
"epoch": 0.8699578567128236,
"grad_norm": 1.0138337149380894,
"learning_rate": 5.0695197552659e-07,
"loss": 0.7973,
"step": 1445
},
{
"epoch": 0.8729680915111379,
"grad_norm": 1.0269724233473905,
"learning_rate": 4.841356299689359e-07,
"loss": 0.8094,
"step": 1450
},
{
"epoch": 0.8759783263094522,
"grad_norm": 1.0545396346249256,
"learning_rate": 4.618184854129981e-07,
"loss": 0.8127,
"step": 1455
},
{
"epoch": 0.8789885611077664,
"grad_norm": 1.047715081854748,
"learning_rate": 4.4000300888169753e-07,
"loss": 0.8051,
"step": 1460
},
{
"epoch": 0.8819987959060807,
"grad_norm": 1.0395135440264711,
"learning_rate": 4.1869161194164565e-07,
"loss": 0.8084,
"step": 1465
},
{
"epoch": 0.885009030704395,
"grad_norm": 1.0503567677831138,
"learning_rate": 3.9788665043656083e-07,
"loss": 0.8034,
"step": 1470
},
{
"epoch": 0.8880192655027093,
"grad_norm": 1.0207611042877882,
"learning_rate": 3.775904242268391e-07,
"loss": 0.8212,
"step": 1475
},
{
"epoch": 0.8910295003010235,
"grad_norm": 1.0191123841112446,
"learning_rate": 3.578051769353219e-07,
"loss": 0.8124,
"step": 1480
},
{
"epoch": 0.8940397350993378,
"grad_norm": 1.0356740138027982,
"learning_rate": 3.385330956992816e-07,
"loss": 0.8072,
"step": 1485
},
{
"epoch": 0.8970499698976521,
"grad_norm": 1.0311377001693454,
"learning_rate": 3.1977631092863613e-07,
"loss": 0.8078,
"step": 1490
},
{
"epoch": 0.9000602046959663,
"grad_norm": 1.0192271510205593,
"learning_rate": 3.015368960704584e-07,
"loss": 0.792,
"step": 1495
},
{
"epoch": 0.9030704394942806,
"grad_norm": 1.0341426167681713,
"learning_rate": 2.8381686737975867e-07,
"loss": 0.8119,
"step": 1500
},
{
"epoch": 0.9060806742925949,
"grad_norm": 1.0298501396809976,
"learning_rate": 2.666181836966053e-07,
"loss": 0.8128,
"step": 1505
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.0220156234933062,
"learning_rate": 2.4994274622958726e-07,
"loss": 0.8122,
"step": 1510
},
{
"epoch": 0.9121011438892234,
"grad_norm": 1.0132135807820002,
"learning_rate": 2.3379239834564526e-07,
"loss": 0.8037,
"step": 1515
},
{
"epoch": 0.9151113786875377,
"grad_norm": 1.0712029435414339,
"learning_rate": 2.1816892536629775e-07,
"loss": 0.8126,
"step": 1520
},
{
"epoch": 0.9181216134858519,
"grad_norm": 1.0109661008831858,
"learning_rate": 2.0307405437029027e-07,
"loss": 0.7985,
"step": 1525
},
{
"epoch": 0.9211318482841662,
"grad_norm": 1.0503987675093023,
"learning_rate": 1.8850945400266994e-07,
"loss": 0.8154,
"step": 1530
},
{
"epoch": 0.9241420830824805,
"grad_norm": 1.0045462111154957,
"learning_rate": 1.7447673429033361e-07,
"loss": 0.7981,
"step": 1535
},
{
"epoch": 0.9271523178807947,
"grad_norm": 1.010463238088656,
"learning_rate": 1.6097744646404457e-07,
"loss": 0.8032,
"step": 1540
},
{
"epoch": 0.930162552679109,
"grad_norm": 1.041872770124601,
"learning_rate": 1.4801308278695636e-07,
"loss": 0.8071,
"step": 1545
},
{
"epoch": 0.9331727874774233,
"grad_norm": 1.0364425032179982,
"learning_rate": 1.3558507638965158e-07,
"loss": 0.8098,
"step": 1550
},
{
"epoch": 0.9361830222757375,
"grad_norm": 1.0561466427450865,
"learning_rate": 1.2369480111171784e-07,
"loss": 0.8052,
"step": 1555
},
{
"epoch": 0.9391932570740518,
"grad_norm": 1.0225495666991897,
"learning_rate": 1.1234357134987717e-07,
"loss": 0.7969,
"step": 1560
},
{
"epoch": 0.9422034918723661,
"grad_norm": 1.0242932807053133,
"learning_rate": 1.0153264191269052e-07,
"loss": 0.8135,
"step": 1565
},
{
"epoch": 0.9452137266706803,
"grad_norm": 1.042306867935529,
"learning_rate": 9.126320788184374e-08,
"loss": 0.8007,
"step": 1570
},
{
"epoch": 0.9482239614689946,
"grad_norm": 1.0312589182719134,
"learning_rate": 8.153640448003875e-08,
"loss": 0.7923,
"step": 1575
},
{
"epoch": 0.9512341962673089,
"grad_norm": 1.0317620067093864,
"learning_rate": 7.235330694550402e-08,
"loss": 0.8166,
"step": 1580
},
{
"epoch": 0.9542444310656231,
"grad_norm": 1.0205653409861422,
"learning_rate": 6.371493041313126e-08,
"loss": 0.7941,
"step": 1585
},
{
"epoch": 0.9572546658639374,
"grad_norm": 1.0236513464237393,
"learning_rate": 5.562222980225907e-08,
"loss": 0.803,
"step": 1590
},
{
"epoch": 0.9602649006622517,
"grad_norm": 1.0009271094917365,
"learning_rate": 4.807609971111238e-08,
"loss": 0.7929,
"step": 1595
},
{
"epoch": 0.963275135460566,
"grad_norm": 1.0278498971825467,
"learning_rate": 4.107737431791159e-08,
"loss": 0.815,
"step": 1600
},
{
"epoch": 0.9662853702588802,
"grad_norm": 1.0128171451294188,
"learning_rate": 3.462682728865685e-08,
"loss": 0.8038,
"step": 1605
},
{
"epoch": 0.9692956050571945,
"grad_norm": 1.0361890758527714,
"learning_rate": 2.8725171691605934e-08,
"loss": 0.8124,
"step": 1610
},
{
"epoch": 0.9723058398555088,
"grad_norm": 1.0128613768201065,
"learning_rate": 2.3373059918448958e-08,
"loss": 0.7996,
"step": 1615
},
{
"epoch": 0.975316074653823,
"grad_norm": 1.0528001409961834,
"learning_rate": 1.8571083612188845e-08,
"loss": 0.8132,
"step": 1620
},
{
"epoch": 0.9783263094521373,
"grad_norm": 1.0408652799513454,
"learning_rate": 1.431977360173975e-08,
"loss": 0.7996,
"step": 1625
},
{
"epoch": 0.9813365442504516,
"grad_norm": 1.0463227914292674,
"learning_rate": 1.0619599843249006e-08,
"loss": 0.8056,
"step": 1630
},
{
"epoch": 0.9843467790487658,
"grad_norm": 1.047367071297636,
"learning_rate": 7.470971368142011e-09,
"loss": 0.8129,
"step": 1635
},
{
"epoch": 0.9873570138470801,
"grad_norm": 1.0050037805593828,
"learning_rate": 4.874236237911723e-09,
"loss": 0.8035,
"step": 1640
},
{
"epoch": 0.9903672486453944,
"grad_norm": 1.044060059047643,
"learning_rate": 2.8296815056377824e-09,
"loss": 0.8164,
"step": 1645
},
{
"epoch": 0.9933774834437086,
"grad_norm": 1.0565007043243317,
"learning_rate": 1.3375331842574446e-09,
"loss": 0.8285,
"step": 1650
},
{
"epoch": 0.9963877182420229,
"grad_norm": 1.0168335467595218,
"learning_rate": 3.9795622158111945e-10,
"loss": 0.8084,
"step": 1655
},
{
"epoch": 0.9993979530403372,
"grad_norm": 1.0278311447859119,
"learning_rate": 1.1054482056405136e-11,
"loss": 0.8055,
"step": 1660
},
{
"epoch": 1.0,
"eval_runtime": 1.2973,
"eval_samples_per_second": 7.708,
"eval_steps_per_second": 2.313,
"step": 1661
},
{
"epoch": 1.0,
"step": 1661,
"total_flos": 161851279147008.0,
"train_loss": 0.8951905027058812,
"train_runtime": 4028.5705,
"train_samples_per_second": 6.596,
"train_steps_per_second": 0.412
}
],
"logging_steps": 5,
"max_steps": 1661,
"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": 161851279147008.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}