qwen_0.5_feedback_dirty / trainer_state.json
terry69's picture
Model save
af1d59a 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": 27.23306162664463,
"learning_rate": 5.98802395209581e-08,
"loss": 1.971,
"step": 1
},
{
"epoch": 0.0030102347983142685,
"grad_norm": 13.650791380879662,
"learning_rate": 2.9940119760479047e-07,
"loss": 1.9968,
"step": 5
},
{
"epoch": 0.006020469596628537,
"grad_norm": 12.306747766476132,
"learning_rate": 5.988023952095809e-07,
"loss": 1.9812,
"step": 10
},
{
"epoch": 0.009030704394942806,
"grad_norm": 9.02007086489199,
"learning_rate": 8.982035928143713e-07,
"loss": 1.9655,
"step": 15
},
{
"epoch": 0.012040939193257074,
"grad_norm": 7.401304214197773,
"learning_rate": 1.1976047904191619e-06,
"loss": 1.9352,
"step": 20
},
{
"epoch": 0.015051173991571343,
"grad_norm": 10.329535150502213,
"learning_rate": 1.4970059880239521e-06,
"loss": 1.892,
"step": 25
},
{
"epoch": 0.018061408789885613,
"grad_norm": 3.0948017863977086,
"learning_rate": 1.7964071856287426e-06,
"loss": 1.8184,
"step": 30
},
{
"epoch": 0.02107164358819988,
"grad_norm": 3.6935226911228827,
"learning_rate": 2.095808383233533e-06,
"loss": 1.7491,
"step": 35
},
{
"epoch": 0.024081878386514148,
"grad_norm": 6.79455681089959,
"learning_rate": 2.3952095808383237e-06,
"loss": 1.7084,
"step": 40
},
{
"epoch": 0.027092113184828417,
"grad_norm": 1.7513542710333883,
"learning_rate": 2.694610778443114e-06,
"loss": 1.6927,
"step": 45
},
{
"epoch": 0.030102347983142687,
"grad_norm": 5.378774459518788,
"learning_rate": 2.9940119760479042e-06,
"loss": 1.6141,
"step": 50
},
{
"epoch": 0.033112582781456956,
"grad_norm": 1.7236309463564579,
"learning_rate": 3.2934131736526947e-06,
"loss": 1.5602,
"step": 55
},
{
"epoch": 0.036122817579771226,
"grad_norm": 4.263330975488616,
"learning_rate": 3.592814371257485e-06,
"loss": 1.5176,
"step": 60
},
{
"epoch": 0.03913305237808549,
"grad_norm": 3.2624764680298513,
"learning_rate": 3.892215568862276e-06,
"loss": 1.4552,
"step": 65
},
{
"epoch": 0.04214328717639976,
"grad_norm": 1.114188393507732,
"learning_rate": 4.191616766467066e-06,
"loss": 1.4621,
"step": 70
},
{
"epoch": 0.04515352197471403,
"grad_norm": 18.7316085941672,
"learning_rate": 4.4910179640718566e-06,
"loss": 1.4325,
"step": 75
},
{
"epoch": 0.048163756773028296,
"grad_norm": 0.9868351913311059,
"learning_rate": 4.7904191616766475e-06,
"loss": 1.3936,
"step": 80
},
{
"epoch": 0.051173991571342566,
"grad_norm": 1.7253367860105293,
"learning_rate": 5.0898203592814375e-06,
"loss": 1.4142,
"step": 85
},
{
"epoch": 0.054184226369656835,
"grad_norm": 0.9242524570526937,
"learning_rate": 5.389221556886228e-06,
"loss": 1.3915,
"step": 90
},
{
"epoch": 0.057194461167971104,
"grad_norm": 0.9116834724175251,
"learning_rate": 5.6886227544910184e-06,
"loss": 1.3879,
"step": 95
},
{
"epoch": 0.060204695966285374,
"grad_norm": 0.9760286306029387,
"learning_rate": 5.9880239520958085e-06,
"loss": 1.3615,
"step": 100
},
{
"epoch": 0.06321493076459964,
"grad_norm": 0.9533636268140245,
"learning_rate": 6.2874251497005985e-06,
"loss": 1.3516,
"step": 105
},
{
"epoch": 0.06622516556291391,
"grad_norm": 0.9318286267990689,
"learning_rate": 6.586826347305389e-06,
"loss": 1.3374,
"step": 110
},
{
"epoch": 0.06923540036122817,
"grad_norm": 0.9861214247946927,
"learning_rate": 6.88622754491018e-06,
"loss": 1.3381,
"step": 115
},
{
"epoch": 0.07224563515954245,
"grad_norm": 1.0087362740253691,
"learning_rate": 7.18562874251497e-06,
"loss": 1.3287,
"step": 120
},
{
"epoch": 0.07525586995785671,
"grad_norm": 3.800322225138513,
"learning_rate": 7.485029940119761e-06,
"loss": 1.3317,
"step": 125
},
{
"epoch": 0.07826610475617098,
"grad_norm": 3.4474877346813013,
"learning_rate": 7.784431137724551e-06,
"loss": 1.3036,
"step": 130
},
{
"epoch": 0.08127633955448525,
"grad_norm": 0.9416140469616731,
"learning_rate": 8.083832335329342e-06,
"loss": 1.3208,
"step": 135
},
{
"epoch": 0.08428657435279951,
"grad_norm": 0.930813958919029,
"learning_rate": 8.383233532934131e-06,
"loss": 1.2844,
"step": 140
},
{
"epoch": 0.08729680915111379,
"grad_norm": 0.9070530897870428,
"learning_rate": 8.682634730538922e-06,
"loss": 1.2861,
"step": 145
},
{
"epoch": 0.09030704394942805,
"grad_norm": 1.0007264664595041,
"learning_rate": 8.982035928143713e-06,
"loss": 1.2707,
"step": 150
},
{
"epoch": 0.09331727874774233,
"grad_norm": 0.9553156573254931,
"learning_rate": 9.281437125748504e-06,
"loss": 1.3042,
"step": 155
},
{
"epoch": 0.09632751354605659,
"grad_norm": 0.957996113896581,
"learning_rate": 9.580838323353295e-06,
"loss": 1.2973,
"step": 160
},
{
"epoch": 0.09933774834437085,
"grad_norm": 0.9845982064473081,
"learning_rate": 9.880239520958084e-06,
"loss": 1.2698,
"step": 165
},
{
"epoch": 0.10234798314268513,
"grad_norm": 0.9872519714050891,
"learning_rate": 9.999900509954779e-06,
"loss": 1.2678,
"step": 170
},
{
"epoch": 0.1053582179409994,
"grad_norm": 1.0192244610172112,
"learning_rate": 9.999292529572152e-06,
"loss": 1.2685,
"step": 175
},
{
"epoch": 0.10836845273931367,
"grad_norm": 0.9727786532787158,
"learning_rate": 9.998131908181262e-06,
"loss": 1.2697,
"step": 180
},
{
"epoch": 0.11137868753762793,
"grad_norm": 0.961370749302959,
"learning_rate": 9.996418774081658e-06,
"loss": 1.2525,
"step": 185
},
{
"epoch": 0.11438892233594221,
"grad_norm": 1.0654272615306832,
"learning_rate": 9.994153316649769e-06,
"loss": 1.2731,
"step": 190
},
{
"epoch": 0.11739915713425647,
"grad_norm": 0.9755470408109456,
"learning_rate": 9.991335786317964e-06,
"loss": 1.2556,
"step": 195
},
{
"epoch": 0.12040939193257075,
"grad_norm": 1.0036625167312827,
"learning_rate": 9.987966494546873e-06,
"loss": 1.2258,
"step": 200
},
{
"epoch": 0.12341962673088501,
"grad_norm": 1.003865146067013,
"learning_rate": 9.984045813790959e-06,
"loss": 1.2469,
"step": 205
},
{
"epoch": 0.12642986152919927,
"grad_norm": 0.9786952100452682,
"learning_rate": 9.979574177457337e-06,
"loss": 1.2468,
"step": 210
},
{
"epoch": 0.12944009632751355,
"grad_norm": 0.9740050481925707,
"learning_rate": 9.974552079857873e-06,
"loss": 1.2427,
"step": 215
},
{
"epoch": 0.13245033112582782,
"grad_norm": 0.9630667645499639,
"learning_rate": 9.968980076154533e-06,
"loss": 1.2521,
"step": 220
},
{
"epoch": 0.13546056592414207,
"grad_norm": 0.9815376328795034,
"learning_rate": 9.962858782298023e-06,
"loss": 1.232,
"step": 225
},
{
"epoch": 0.13847080072245635,
"grad_norm": 1.1558477232615034,
"learning_rate": 9.956188874959686e-06,
"loss": 1.2306,
"step": 230
},
{
"epoch": 0.14148103552077063,
"grad_norm": 1.010662989571516,
"learning_rate": 9.948971091456715e-06,
"loss": 1.2316,
"step": 235
},
{
"epoch": 0.1444912703190849,
"grad_norm": 1.016349108965321,
"learning_rate": 9.941206229670634e-06,
"loss": 1.2202,
"step": 240
},
{
"epoch": 0.14750150511739915,
"grad_norm": 1.012113158589424,
"learning_rate": 9.932895147959106e-06,
"loss": 1.2042,
"step": 245
},
{
"epoch": 0.15051173991571343,
"grad_norm": 1.0179802871184889,
"learning_rate": 9.924038765061042e-06,
"loss": 1.1872,
"step": 250
},
{
"epoch": 0.1535219747140277,
"grad_norm": 1.0631543044499743,
"learning_rate": 9.91463805999504e-06,
"loss": 1.2117,
"step": 255
},
{
"epoch": 0.15653220951234195,
"grad_norm": 1.0508741168922897,
"learning_rate": 9.904694071951167e-06,
"loss": 1.2122,
"step": 260
},
{
"epoch": 0.15954244431065623,
"grad_norm": 1.0059019047260795,
"learning_rate": 9.894207900176074e-06,
"loss": 1.1955,
"step": 265
},
{
"epoch": 0.1625526791089705,
"grad_norm": 1.0033418286179807,
"learning_rate": 9.883180703851488e-06,
"loss": 1.2012,
"step": 270
},
{
"epoch": 0.16556291390728478,
"grad_norm": 0.9877620622300878,
"learning_rate": 9.871613701966067e-06,
"loss": 1.1805,
"step": 275
},
{
"epoch": 0.16857314870559903,
"grad_norm": 0.9400638251546493,
"learning_rate": 9.859508173180653e-06,
"loss": 1.1719,
"step": 280
},
{
"epoch": 0.1715833835039133,
"grad_norm": 0.9874311482394339,
"learning_rate": 9.846865455686915e-06,
"loss": 1.1755,
"step": 285
},
{
"epoch": 0.17459361830222758,
"grad_norm": 0.9914705153634205,
"learning_rate": 9.833686947059436e-06,
"loss": 1.166,
"step": 290
},
{
"epoch": 0.17760385310054183,
"grad_norm": 1.023017013404828,
"learning_rate": 9.819974104101198e-06,
"loss": 1.1611,
"step": 295
},
{
"epoch": 0.1806140878988561,
"grad_norm": 0.9792763961764498,
"learning_rate": 9.80572844268256e-06,
"loss": 1.1493,
"step": 300
},
{
"epoch": 0.18362432269717038,
"grad_norm": 1.0359591597037352,
"learning_rate": 9.790951537573686e-06,
"loss": 1.1832,
"step": 305
},
{
"epoch": 0.18663455749548466,
"grad_norm": 1.453908493308493,
"learning_rate": 9.775645022270448e-06,
"loss": 1.1641,
"step": 310
},
{
"epoch": 0.1896447922937989,
"grad_norm": 0.9821000141264724,
"learning_rate": 9.759810588813872e-06,
"loss": 1.1648,
"step": 315
},
{
"epoch": 0.19265502709211318,
"grad_norm": 1.1094696031738442,
"learning_rate": 9.743449987603082e-06,
"loss": 1.1602,
"step": 320
},
{
"epoch": 0.19566526189042746,
"grad_norm": 1.0229747825382478,
"learning_rate": 9.726565027201813e-06,
"loss": 1.1732,
"step": 325
},
{
"epoch": 0.1986754966887417,
"grad_norm": 1.0330731927842094,
"learning_rate": 9.70915757413847e-06,
"loss": 1.1613,
"step": 330
},
{
"epoch": 0.20168573148705599,
"grad_norm": 1.0671526485809508,
"learning_rate": 9.691229552699817e-06,
"loss": 1.1404,
"step": 335
},
{
"epoch": 0.20469596628537026,
"grad_norm": 1.1599170579317017,
"learning_rate": 9.672782944718234e-06,
"loss": 1.1652,
"step": 340
},
{
"epoch": 0.20770620108368454,
"grad_norm": 1.078143063135002,
"learning_rate": 9.65381978935266e-06,
"loss": 1.1393,
"step": 345
},
{
"epoch": 0.2107164358819988,
"grad_norm": 1.051353820041791,
"learning_rate": 9.634342182863163e-06,
"loss": 1.1159,
"step": 350
},
{
"epoch": 0.21372667068031306,
"grad_norm": 1.0456430688838831,
"learning_rate": 9.614352278379217e-06,
"loss": 1.149,
"step": 355
},
{
"epoch": 0.21673690547862734,
"grad_norm": 1.091953950241735,
"learning_rate": 9.593852285661684e-06,
"loss": 1.1552,
"step": 360
},
{
"epoch": 0.2197471402769416,
"grad_norm": 1.015428888405563,
"learning_rate": 9.572844470858537e-06,
"loss": 1.1369,
"step": 365
},
{
"epoch": 0.22275737507525586,
"grad_norm": 1.1341410310607158,
"learning_rate": 9.551331156254358e-06,
"loss": 1.138,
"step": 370
},
{
"epoch": 0.22576760987357014,
"grad_norm": 1.0098698732632645,
"learning_rate": 9.529314720013618e-06,
"loss": 1.1249,
"step": 375
},
{
"epoch": 0.22877784467188442,
"grad_norm": 1.0306055734848716,
"learning_rate": 9.506797595917787e-06,
"loss": 1.1309,
"step": 380
},
{
"epoch": 0.23178807947019867,
"grad_norm": 1.0832575075897741,
"learning_rate": 9.483782273096295e-06,
"loss": 1.1296,
"step": 385
},
{
"epoch": 0.23479831426851294,
"grad_norm": 1.0215735515542592,
"learning_rate": 9.460271295751373e-06,
"loss": 1.1436,
"step": 390
},
{
"epoch": 0.23780854906682722,
"grad_norm": 1.0127022199777589,
"learning_rate": 9.436267262876808e-06,
"loss": 1.1202,
"step": 395
},
{
"epoch": 0.2408187838651415,
"grad_norm": 1.0677989226321623,
"learning_rate": 9.411772827970642e-06,
"loss": 1.1182,
"step": 400
},
{
"epoch": 0.24382901866345574,
"grad_norm": 1.0863425275828655,
"learning_rate": 9.38679069874184e-06,
"loss": 1.1192,
"step": 405
},
{
"epoch": 0.24683925346177002,
"grad_norm": 1.003581452419569,
"learning_rate": 9.36132363681097e-06,
"loss": 1.1195,
"step": 410
},
{
"epoch": 0.2498494882600843,
"grad_norm": 1.0173137664236545,
"learning_rate": 9.335374457404928e-06,
"loss": 1.127,
"step": 415
},
{
"epoch": 0.25285972305839854,
"grad_norm": 1.0256692516120043,
"learning_rate": 9.308946029045726e-06,
"loss": 1.1002,
"step": 420
},
{
"epoch": 0.25586995785671285,
"grad_norm": 1.1681651650487053,
"learning_rate": 9.282041273233402e-06,
"loss": 1.0926,
"step": 425
},
{
"epoch": 0.2588801926550271,
"grad_norm": 1.046473420004503,
"learning_rate": 9.254663164123052e-06,
"loss": 1.0937,
"step": 430
},
{
"epoch": 0.26189042745334135,
"grad_norm": 1.0077718278045176,
"learning_rate": 9.226814728196072e-06,
"loss": 1.096,
"step": 435
},
{
"epoch": 0.26490066225165565,
"grad_norm": 1.003746070586757,
"learning_rate": 9.198499043925591e-06,
"loss": 1.1047,
"step": 440
},
{
"epoch": 0.2679108970499699,
"grad_norm": 1.007686883808698,
"learning_rate": 9.169719241436162e-06,
"loss": 1.0996,
"step": 445
},
{
"epoch": 0.27092113184828415,
"grad_norm": 1.002375870866833,
"learning_rate": 9.14047850215775e-06,
"loss": 1.082,
"step": 450
},
{
"epoch": 0.27393136664659845,
"grad_norm": 1.0070740266122291,
"learning_rate": 9.110780058474052e-06,
"loss": 1.0766,
"step": 455
},
{
"epoch": 0.2769416014449127,
"grad_norm": 1.030435067694461,
"learning_rate": 9.080627193365155e-06,
"loss": 1.0948,
"step": 460
},
{
"epoch": 0.27995183624322695,
"grad_norm": 0.9984477878066329,
"learning_rate": 9.050023240044649e-06,
"loss": 1.0944,
"step": 465
},
{
"epoch": 0.28296207104154125,
"grad_norm": 0.9909219592815518,
"learning_rate": 9.018971581591141e-06,
"loss": 1.0817,
"step": 470
},
{
"epoch": 0.2859723058398555,
"grad_norm": 0.9836671240022327,
"learning_rate": 8.987475650574289e-06,
"loss": 1.1055,
"step": 475
},
{
"epoch": 0.2889825406381698,
"grad_norm": 1.052915518162781,
"learning_rate": 8.955538928675343e-06,
"loss": 1.0675,
"step": 480
},
{
"epoch": 0.29199277543648405,
"grad_norm": 1.0046123233442037,
"learning_rate": 8.923164946302274e-06,
"loss": 1.0846,
"step": 485
},
{
"epoch": 0.2950030102347983,
"grad_norm": 1.0927507536443555,
"learning_rate": 8.890357282199504e-06,
"loss": 1.0938,
"step": 490
},
{
"epoch": 0.2980132450331126,
"grad_norm": 1.0365967534347835,
"learning_rate": 8.857119563052301e-06,
"loss": 1.0638,
"step": 495
},
{
"epoch": 0.30102347983142685,
"grad_norm": 0.9900367247286117,
"learning_rate": 8.823455463085873e-06,
"loss": 1.0784,
"step": 500
},
{
"epoch": 0.3040337146297411,
"grad_norm": 1.0187836052415793,
"learning_rate": 8.789368703659199e-06,
"loss": 1.086,
"step": 505
},
{
"epoch": 0.3070439494280554,
"grad_norm": 1.4170045215112093,
"learning_rate": 8.754863052853658e-06,
"loss": 1.0786,
"step": 510
},
{
"epoch": 0.31005418422636966,
"grad_norm": 1.0563490229181656,
"learning_rate": 8.719942325056496e-06,
"loss": 1.0552,
"step": 515
},
{
"epoch": 0.3130644190246839,
"grad_norm": 0.9581250571752027,
"learning_rate": 8.68461038053916e-06,
"loss": 1.0611,
"step": 520
},
{
"epoch": 0.3160746538229982,
"grad_norm": 1.0882591426901487,
"learning_rate": 8.648871125030576e-06,
"loss": 1.0674,
"step": 525
},
{
"epoch": 0.31908488862131246,
"grad_norm": 0.9748747047047535,
"learning_rate": 8.612728509285395e-06,
"loss": 1.0569,
"step": 530
},
{
"epoch": 0.3220951234196267,
"grad_norm": 1.0007083161058548,
"learning_rate": 8.576186528647253e-06,
"loss": 1.0773,
"step": 535
},
{
"epoch": 0.325105358217941,
"grad_norm": 1.0024226084870622,
"learning_rate": 8.53924922260712e-06,
"loss": 1.0688,
"step": 540
},
{
"epoch": 0.32811559301625526,
"grad_norm": 1.0647162244790471,
"learning_rate": 8.501920674356755e-06,
"loss": 1.0471,
"step": 545
},
{
"epoch": 0.33112582781456956,
"grad_norm": 0.993432183224602,
"learning_rate": 8.46420501033733e-06,
"loss": 1.0615,
"step": 550
},
{
"epoch": 0.3341360626128838,
"grad_norm": 1.0547538639367902,
"learning_rate": 8.42610639978329e-06,
"loss": 1.0568,
"step": 555
},
{
"epoch": 0.33714629741119806,
"grad_norm": 0.9966572739245336,
"learning_rate": 8.387629054261454e-06,
"loss": 1.0523,
"step": 560
},
{
"epoch": 0.34015653220951236,
"grad_norm": 0.9680858165661955,
"learning_rate": 8.348777227205462e-06,
"loss": 1.0499,
"step": 565
},
{
"epoch": 0.3431667670078266,
"grad_norm": 0.9422881419321228,
"learning_rate": 8.309555213445583e-06,
"loss": 1.0745,
"step": 570
},
{
"epoch": 0.34617700180614086,
"grad_norm": 0.9684939972353577,
"learning_rate": 8.269967348733947e-06,
"loss": 1.046,
"step": 575
},
{
"epoch": 0.34918723660445516,
"grad_norm": 0.9621359710421508,
"learning_rate": 8.230018009265255e-06,
"loss": 1.0602,
"step": 580
},
{
"epoch": 0.3521974714027694,
"grad_norm": 0.9496961252830911,
"learning_rate": 8.189711611193012e-06,
"loss": 1.0517,
"step": 585
},
{
"epoch": 0.35520770620108366,
"grad_norm": 0.9723471751859206,
"learning_rate": 8.149052610141357e-06,
"loss": 1.044,
"step": 590
},
{
"epoch": 0.35821794099939797,
"grad_norm": 0.9399615087193514,
"learning_rate": 8.108045500712518e-06,
"loss": 1.0588,
"step": 595
},
{
"epoch": 0.3612281757977122,
"grad_norm": 0.9845428819310491,
"learning_rate": 8.066694815989961e-06,
"loss": 1.061,
"step": 600
},
{
"epoch": 0.36423841059602646,
"grad_norm": 0.9891275737820486,
"learning_rate": 8.025005127037282e-06,
"loss": 1.0615,
"step": 605
},
{
"epoch": 0.36724864539434077,
"grad_norm": 0.9954042733317254,
"learning_rate": 7.982981042392907e-06,
"loss": 1.0497,
"step": 610
},
{
"epoch": 0.370258880192655,
"grad_norm": 1.000537961876559,
"learning_rate": 7.940627207560655e-06,
"loss": 1.0441,
"step": 615
},
{
"epoch": 0.3732691149909693,
"grad_norm": 0.9496111907275188,
"learning_rate": 7.897948304496189e-06,
"loss": 1.0648,
"step": 620
},
{
"epoch": 0.37627934978928357,
"grad_norm": 0.9393968941122551,
"learning_rate": 7.854949051089467e-06,
"loss": 1.0676,
"step": 625
},
{
"epoch": 0.3792895845875978,
"grad_norm": 0.9380308472022797,
"learning_rate": 7.811634200643202e-06,
"loss": 1.0739,
"step": 630
},
{
"epoch": 0.3822998193859121,
"grad_norm": 0.9486486114086516,
"learning_rate": 7.768008541347423e-06,
"loss": 1.0221,
"step": 635
},
{
"epoch": 0.38531005418422637,
"grad_norm": 68.18970323635645,
"learning_rate": 7.72407689575016e-06,
"loss": 1.0515,
"step": 640
},
{
"epoch": 0.3883202889825406,
"grad_norm": 0.9508633036813551,
"learning_rate": 7.67984412022434e-06,
"loss": 1.0467,
"step": 645
},
{
"epoch": 0.3913305237808549,
"grad_norm": 0.9916855461468728,
"learning_rate": 7.635315104430959e-06,
"loss": 1.0367,
"step": 650
},
{
"epoch": 0.39434075857916917,
"grad_norm": 0.9446909197494837,
"learning_rate": 7.5904947707785434e-06,
"loss": 1.0305,
"step": 655
},
{
"epoch": 0.3973509933774834,
"grad_norm": 0.9210322564666947,
"learning_rate": 7.545388073879018e-06,
"loss": 1.0591,
"step": 660
},
{
"epoch": 0.4003612281757977,
"grad_norm": 0.9306461151238373,
"learning_rate": 7.500000000000001e-06,
"loss": 1.0413,
"step": 665
},
{
"epoch": 0.40337146297411197,
"grad_norm": 0.9658878999866418,
"learning_rate": 7.454335566513603e-06,
"loss": 1.0465,
"step": 670
},
{
"epoch": 0.4063816977724263,
"grad_norm": 0.9583821484101821,
"learning_rate": 7.408399821341787e-06,
"loss": 1.0544,
"step": 675
},
{
"epoch": 0.4093919325707405,
"grad_norm": 0.9127670551716062,
"learning_rate": 7.362197842398355e-06,
"loss": 1.0346,
"step": 680
},
{
"epoch": 0.4124021673690548,
"grad_norm": 0.9278308537424353,
"learning_rate": 7.315734737027612e-06,
"loss": 1.0453,
"step": 685
},
{
"epoch": 0.4154124021673691,
"grad_norm": 0.9723444779061409,
"learning_rate": 7.2690156414397775e-06,
"loss": 1.0258,
"step": 690
},
{
"epoch": 0.4184226369656833,
"grad_norm": 0.9347486140545276,
"learning_rate": 7.22204572014322e-06,
"loss": 1.0338,
"step": 695
},
{
"epoch": 0.4214328717639976,
"grad_norm": 3.85387878818612,
"learning_rate": 7.174830165373542e-06,
"loss": 1.0431,
"step": 700
},
{
"epoch": 0.4244431065623119,
"grad_norm": 0.9143452453786128,
"learning_rate": 7.127374196519616e-06,
"loss": 1.0336,
"step": 705
},
{
"epoch": 0.4274533413606261,
"grad_norm": 0.9482641660012492,
"learning_rate": 7.079683059546607e-06,
"loss": 1.0454,
"step": 710
},
{
"epoch": 0.4304635761589404,
"grad_norm": 1.04293445699686,
"learning_rate": 7.031762026416074e-06,
"loss": 1.0261,
"step": 715
},
{
"epoch": 0.4334738109572547,
"grad_norm": 0.9875257232657468,
"learning_rate": 6.983616394503177e-06,
"loss": 1.0448,
"step": 720
},
{
"epoch": 0.43648404575556893,
"grad_norm": 2.4930481995359015,
"learning_rate": 6.9352514860110876e-06,
"loss": 1.0688,
"step": 725
},
{
"epoch": 0.4394942805538832,
"grad_norm": 0.9536222920261503,
"learning_rate": 6.886672647382653e-06,
"loss": 1.0338,
"step": 730
},
{
"epoch": 0.4425045153521975,
"grad_norm": 1.079837034688785,
"learning_rate": 6.837885248709386e-06,
"loss": 1.0675,
"step": 735
},
{
"epoch": 0.44551475015051173,
"grad_norm": 0.9267764883427583,
"learning_rate": 6.788894683137822e-06,
"loss": 1.0309,
"step": 740
},
{
"epoch": 0.44852498494882603,
"grad_norm": 1.0722868967903327,
"learning_rate": 6.739706366273346e-06,
"loss": 1.0475,
"step": 745
},
{
"epoch": 0.4515352197471403,
"grad_norm": 1.1196709849520698,
"learning_rate": 6.690325735581532e-06,
"loss": 1.0303,
"step": 750
},
{
"epoch": 0.45454545454545453,
"grad_norm": 0.9174528976707828,
"learning_rate": 6.640758249787067e-06,
"loss": 1.0494,
"step": 755
},
{
"epoch": 0.45755568934376883,
"grad_norm": 0.9480777677094587,
"learning_rate": 6.591009388270315e-06,
"loss": 1.0421,
"step": 760
},
{
"epoch": 0.4605659241420831,
"grad_norm": 2.212409600409066,
"learning_rate": 6.54108465046161e-06,
"loss": 1.0426,
"step": 765
},
{
"epoch": 0.46357615894039733,
"grad_norm": 1.0682566847564687,
"learning_rate": 6.490989555233328e-06,
"loss": 1.0404,
"step": 770
},
{
"epoch": 0.46658639373871164,
"grad_norm": 10.206202788584402,
"learning_rate": 6.440729640289809e-06,
"loss": 1.0292,
"step": 775
},
{
"epoch": 0.4695966285370259,
"grad_norm": 1.1020753443992015,
"learning_rate": 6.3903104615551956e-06,
"loss": 1.0232,
"step": 780
},
{
"epoch": 0.47260686333534013,
"grad_norm": 5.167899676012411,
"learning_rate": 6.3397375925592675e-06,
"loss": 1.006,
"step": 785
},
{
"epoch": 0.47561709813365444,
"grad_norm": 1.192166577504938,
"learning_rate": 6.289016623821308e-06,
"loss": 1.0419,
"step": 790
},
{
"epoch": 0.4786273329319687,
"grad_norm": 1.367111758271087,
"learning_rate": 6.2381531622321234e-06,
"loss": 1.0426,
"step": 795
},
{
"epoch": 0.481637567730283,
"grad_norm": 0.9109840034743182,
"learning_rate": 6.18715283043422e-06,
"loss": 1.0314,
"step": 800
},
{
"epoch": 0.48464780252859724,
"grad_norm": 0.8741141489406814,
"learning_rate": 6.136021266200271e-06,
"loss": 1.0328,
"step": 805
},
{
"epoch": 0.4876580373269115,
"grad_norm": 0.9172189163013432,
"learning_rate": 6.084764121809878e-06,
"loss": 1.0325,
"step": 810
},
{
"epoch": 0.4906682721252258,
"grad_norm": 0.9369840939625793,
"learning_rate": 6.033387063424765e-06,
"loss": 1.0228,
"step": 815
},
{
"epoch": 0.49367850692354004,
"grad_norm": 1.432152664217959,
"learning_rate": 5.9818957704624046e-06,
"loss": 1.0477,
"step": 820
},
{
"epoch": 0.4966887417218543,
"grad_norm": 0.8878345436889384,
"learning_rate": 5.930295934968197e-06,
"loss": 1.0238,
"step": 825
},
{
"epoch": 0.4996989765201686,
"grad_norm": 0.8962519472927415,
"learning_rate": 5.878593260986256e-06,
"loss": 1.0194,
"step": 830
},
{
"epoch": 0.5027092113184829,
"grad_norm": 1.499274394693842,
"learning_rate": 5.8267934639288525e-06,
"loss": 1.0432,
"step": 835
},
{
"epoch": 0.5057194461167971,
"grad_norm": 0.9364157701378347,
"learning_rate": 5.77490226994462e-06,
"loss": 1.0456,
"step": 840
},
{
"epoch": 0.5087296809151114,
"grad_norm": 0.9248923318440813,
"learning_rate": 5.722925415285555e-06,
"loss": 1.0528,
"step": 845
},
{
"epoch": 0.5117399157134257,
"grad_norm": 0.9622767369752785,
"learning_rate": 5.670868645672916e-06,
"loss": 1.0433,
"step": 850
},
{
"epoch": 0.5147501505117399,
"grad_norm": 1.170025668084364,
"learning_rate": 5.618737715662067e-06,
"loss": 1.03,
"step": 855
},
{
"epoch": 0.5177603853100542,
"grad_norm": 0.9043466850630412,
"learning_rate": 5.566538388006351e-06,
"loss": 1.039,
"step": 860
},
{
"epoch": 0.5207706201083685,
"grad_norm": 0.8849608499162146,
"learning_rate": 5.514276433020044e-06,
"loss": 1.0249,
"step": 865
},
{
"epoch": 0.5237808549066827,
"grad_norm": 0.9464553636560172,
"learning_rate": 5.461957627940489e-06,
"loss": 1.0198,
"step": 870
},
{
"epoch": 0.526791089704997,
"grad_norm": 0.9179569227254959,
"learning_rate": 5.409587756289462e-06,
"loss": 1.0205,
"step": 875
},
{
"epoch": 0.5298013245033113,
"grad_norm": 1.9045126443295857,
"learning_rate": 5.357172607233831e-06,
"loss": 1.0182,
"step": 880
},
{
"epoch": 0.5328115593016255,
"grad_norm": 1.4051118790220352,
"learning_rate": 5.304717974945596e-06,
"loss": 1.0221,
"step": 885
},
{
"epoch": 0.5358217940999398,
"grad_norm": 0.9078454332904633,
"learning_rate": 5.252229657961394e-06,
"loss": 1.0206,
"step": 890
},
{
"epoch": 0.5388320288982541,
"grad_norm": 0.8875050113196231,
"learning_rate": 5.199713458541495e-06,
"loss": 1.0352,
"step": 895
},
{
"epoch": 0.5418422636965683,
"grad_norm": 0.9480042480699051,
"learning_rate": 5.1471751820284e-06,
"loss": 1.0134,
"step": 900
},
{
"epoch": 0.5448524984948826,
"grad_norm": 0.902719752286893,
"learning_rate": 5.094620636205096e-06,
"loss": 1.01,
"step": 905
},
{
"epoch": 0.5478627332931969,
"grad_norm": 0.8905457235226512,
"learning_rate": 5.042055630653042e-06,
"loss": 1.0206,
"step": 910
},
{
"epoch": 0.5508729680915111,
"grad_norm": 0.9206243693271725,
"learning_rate": 4.98948597610996e-06,
"loss": 1.0121,
"step": 915
},
{
"epoch": 0.5538832028898254,
"grad_norm": 0.882467337650688,
"learning_rate": 4.936917483827483e-06,
"loss": 0.9989,
"step": 920
},
{
"epoch": 0.5568934376881397,
"grad_norm": 0.8674487532731974,
"learning_rate": 4.884355964928767e-06,
"loss": 1.0463,
"step": 925
},
{
"epoch": 0.5599036724864539,
"grad_norm": 0.8791260040749262,
"learning_rate": 4.831807229766101e-06,
"loss": 1.0047,
"step": 930
},
{
"epoch": 0.5629139072847682,
"grad_norm": 0.9167695803034189,
"learning_rate": 4.779277087278615e-06,
"loss": 1.03,
"step": 935
},
{
"epoch": 0.5659241420830825,
"grad_norm": 0.8934690190804779,
"learning_rate": 4.7267713443501274e-06,
"loss": 1.0381,
"step": 940
},
{
"epoch": 0.5689343768813967,
"grad_norm": 1.1473340162402226,
"learning_rate": 4.67429580516724e-06,
"loss": 1.0146,
"step": 945
},
{
"epoch": 0.571944611679711,
"grad_norm": 0.8911747505115302,
"learning_rate": 4.6218562705777185e-06,
"loss": 1.0142,
"step": 950
},
{
"epoch": 0.5749548464780253,
"grad_norm": 0.9543234233344876,
"learning_rate": 4.5694585374492314e-06,
"loss": 1.0243,
"step": 955
},
{
"epoch": 0.5779650812763396,
"grad_norm": 0.9223284966528281,
"learning_rate": 4.517108398028566e-06,
"loss": 1.0417,
"step": 960
},
{
"epoch": 0.5809753160746538,
"grad_norm": 0.8812509821809876,
"learning_rate": 4.464811639301314e-06,
"loss": 1.0195,
"step": 965
},
{
"epoch": 0.5839855508729681,
"grad_norm": 0.9091706304207465,
"learning_rate": 4.412574042352156e-06,
"loss": 1.0337,
"step": 970
},
{
"epoch": 0.5869957856712824,
"grad_norm": 0.8926473999654263,
"learning_rate": 4.360401381725806e-06,
"loss": 1.0221,
"step": 975
},
{
"epoch": 0.5900060204695966,
"grad_norm": 1.6530772893621817,
"learning_rate": 4.308299424788667e-06,
"loss": 1.0382,
"step": 980
},
{
"epoch": 0.5930162552679109,
"grad_norm": 0.9083816640002393,
"learning_rate": 4.256273931091284e-06,
"loss": 1.0239,
"step": 985
},
{
"epoch": 0.5960264900662252,
"grad_norm": 0.9029678699407463,
"learning_rate": 4.204330651731662e-06,
"loss": 1.0191,
"step": 990
},
{
"epoch": 0.5990367248645394,
"grad_norm": 0.8880707813892602,
"learning_rate": 4.152475328719517e-06,
"loss": 1.0205,
"step": 995
},
{
"epoch": 0.6020469596628537,
"grad_norm": 0.9115699732773067,
"learning_rate": 4.1007136943415325e-06,
"loss": 1.0072,
"step": 1000
},
{
"epoch": 0.605057194461168,
"grad_norm": 0.8460435318858166,
"learning_rate": 4.049051470527692e-06,
"loss": 1.0262,
"step": 1005
},
{
"epoch": 0.6080674292594822,
"grad_norm": 0.992395161503029,
"learning_rate": 3.997494368218745e-06,
"loss": 1.0165,
"step": 1010
},
{
"epoch": 0.6110776640577965,
"grad_norm": 1.859721509427287,
"learning_rate": 3.946048086734921e-06,
"loss": 1.0347,
"step": 1015
},
{
"epoch": 0.6140878988561108,
"grad_norm": 0.8921204238192161,
"learning_rate": 3.894718313145873e-06,
"loss": 1.0256,
"step": 1020
},
{
"epoch": 0.617098133654425,
"grad_norm": 1.2681625671865207,
"learning_rate": 3.843510721642036e-06,
"loss": 1.0081,
"step": 1025
},
{
"epoch": 0.6201083684527393,
"grad_norm": 0.8915300481916899,
"learning_rate": 3.7924309729073616e-06,
"loss": 1.0155,
"step": 1030
},
{
"epoch": 0.6231186032510536,
"grad_norm": 0.9396438819521465,
"learning_rate": 3.7414847134935716e-06,
"loss": 1.0046,
"step": 1035
},
{
"epoch": 0.6261288380493678,
"grad_norm": 3.922062228338965,
"learning_rate": 3.6906775751959667e-06,
"loss": 1.022,
"step": 1040
},
{
"epoch": 0.6291390728476821,
"grad_norm": 0.9172947691983047,
"learning_rate": 3.640015174430864e-06,
"loss": 0.9998,
"step": 1045
},
{
"epoch": 0.6321493076459964,
"grad_norm": 0.889016554505559,
"learning_rate": 3.5895031116147355e-06,
"loss": 1.0143,
"step": 1050
},
{
"epoch": 0.6351595424443106,
"grad_norm": 1.0135918728882847,
"learning_rate": 3.539146970545124e-06,
"loss": 1.02,
"step": 1055
},
{
"epoch": 0.6381697772426249,
"grad_norm": 0.8701415123045054,
"learning_rate": 3.488952317783374e-06,
"loss": 1.0127,
"step": 1060
},
{
"epoch": 0.6411800120409392,
"grad_norm": 0.8880221243740891,
"learning_rate": 3.438924702039301e-06,
"loss": 1.0089,
"step": 1065
},
{
"epoch": 0.6441902468392534,
"grad_norm": 0.8735477996007547,
"learning_rate": 3.389069653557805e-06,
"loss": 1.0216,
"step": 1070
},
{
"epoch": 0.6472004816375677,
"grad_norm": 0.9072044456513535,
"learning_rate": 3.3393926835075307e-06,
"loss": 1.0115,
"step": 1075
},
{
"epoch": 0.650210716435882,
"grad_norm": 0.8883607705827921,
"learning_rate": 3.289899283371657e-06,
"loss": 1.0014,
"step": 1080
},
{
"epoch": 0.6532209512341962,
"grad_norm": 0.9518629250264209,
"learning_rate": 3.240594924340835e-06,
"loss": 1.0094,
"step": 1085
},
{
"epoch": 0.6562311860325105,
"grad_norm": 0.877493524698432,
"learning_rate": 3.1914850567083866e-06,
"loss": 1.019,
"step": 1090
},
{
"epoch": 0.6592414208308248,
"grad_norm": 0.9145761162077933,
"learning_rate": 3.1425751092678064e-06,
"loss": 1.005,
"step": 1095
},
{
"epoch": 0.6622516556291391,
"grad_norm": 1.475049622537514,
"learning_rate": 3.0938704887126425e-06,
"loss": 0.9926,
"step": 1100
},
{
"epoch": 0.6652618904274533,
"grad_norm": 0.8572700406837115,
"learning_rate": 3.045376579038821e-06,
"loss": 1.0021,
"step": 1105
},
{
"epoch": 0.6682721252257676,
"grad_norm": 0.852655010706633,
"learning_rate": 2.9970987409494784e-06,
"loss": 1.0387,
"step": 1110
},
{
"epoch": 0.6712823600240819,
"grad_norm": 0.9146351266641374,
"learning_rate": 2.9490423112623646e-06,
"loss": 1.0042,
"step": 1115
},
{
"epoch": 0.6742925948223961,
"grad_norm": 0.9064467261792086,
"learning_rate": 2.9012126023198973e-06,
"loss": 1.017,
"step": 1120
},
{
"epoch": 0.6773028296207104,
"grad_norm": 1.3125167951045396,
"learning_rate": 2.853614901401909e-06,
"loss": 0.9869,
"step": 1125
},
{
"epoch": 0.6803130644190247,
"grad_norm": 0.8839117074053637,
"learning_rate": 2.806254470141174e-06,
"loss": 1.0214,
"step": 1130
},
{
"epoch": 0.6833232992173389,
"grad_norm": 0.8715135185909122,
"learning_rate": 2.759136543941773e-06,
"loss": 0.9828,
"step": 1135
},
{
"epoch": 0.6863335340156532,
"grad_norm": 0.8908461313314309,
"learning_rate": 2.712266331400332e-06,
"loss": 1.0032,
"step": 1140
},
{
"epoch": 0.6893437688139675,
"grad_norm": 0.8884047218150805,
"learning_rate": 2.66564901373027e-06,
"loss": 1.0124,
"step": 1145
},
{
"epoch": 0.6923540036122817,
"grad_norm": 0.9119096888988255,
"learning_rate": 2.6192897441890337e-06,
"loss": 1.003,
"step": 1150
},
{
"epoch": 0.695364238410596,
"grad_norm": 4.230701533700929,
"learning_rate": 2.573193647508426e-06,
"loss": 1.0037,
"step": 1155
},
{
"epoch": 0.6983744732089103,
"grad_norm": 0.8664627590837305,
"learning_rate": 2.5273658193281252e-06,
"loss": 0.9871,
"step": 1160
},
{
"epoch": 0.7013847080072245,
"grad_norm": 1.3686260181680083,
"learning_rate": 2.4818113256323745e-06,
"loss": 1.0031,
"step": 1165
},
{
"epoch": 0.7043949428055388,
"grad_norm": 0.8810846166712105,
"learning_rate": 2.4365352021899635e-06,
"loss": 1.0075,
"step": 1170
},
{
"epoch": 0.7074051776038531,
"grad_norm": 0.9013744748771672,
"learning_rate": 2.391542453997578e-06,
"loss": 0.9983,
"step": 1175
},
{
"epoch": 0.7104154124021673,
"grad_norm": 0.8824175367501911,
"learning_rate": 2.346838054726505e-06,
"loss": 1.0049,
"step": 1180
},
{
"epoch": 0.7134256472004816,
"grad_norm": 0.8647677789614462,
"learning_rate": 2.302426946172836e-06,
"loss": 1.0199,
"step": 1185
},
{
"epoch": 0.7164358819987959,
"grad_norm": 0.8974669950167197,
"learning_rate": 2.258314037711184e-06,
"loss": 1.0078,
"step": 1190
},
{
"epoch": 0.7194461167971101,
"grad_norm": 0.8942689905651798,
"learning_rate": 2.214504205751971e-06,
"loss": 1.0248,
"step": 1195
},
{
"epoch": 0.7224563515954244,
"grad_norm": 0.8613593925636255,
"learning_rate": 2.1710022932023805e-06,
"loss": 1.0041,
"step": 1200
},
{
"epoch": 0.7254665863937387,
"grad_norm": 0.9273768242975111,
"learning_rate": 2.127813108931007e-06,
"loss": 1.0093,
"step": 1205
},
{
"epoch": 0.7284768211920529,
"grad_norm": 0.8689116817146251,
"learning_rate": 2.084941427236245e-06,
"loss": 1.0165,
"step": 1210
},
{
"epoch": 0.7314870559903672,
"grad_norm": 0.925626300168374,
"learning_rate": 2.04239198731855e-06,
"loss": 1.0004,
"step": 1215
},
{
"epoch": 0.7344972907886815,
"grad_norm": 1.0155037341491024,
"learning_rate": 2.000169492756523e-06,
"loss": 1.0147,
"step": 1220
},
{
"epoch": 0.7375075255869958,
"grad_norm": 0.9027560092938075,
"learning_rate": 1.9582786109869713e-06,
"loss": 1.0131,
"step": 1225
},
{
"epoch": 0.74051776038531,
"grad_norm": 0.8889230240612882,
"learning_rate": 1.9167239727889527e-06,
"loss": 1.0106,
"step": 1230
},
{
"epoch": 0.7435279951836243,
"grad_norm": 0.8859832869083346,
"learning_rate": 1.875510171771865e-06,
"loss": 0.9935,
"step": 1235
},
{
"epoch": 0.7465382299819386,
"grad_norm": 0.8935698851896953,
"learning_rate": 1.8346417638676533e-06,
"loss": 0.9845,
"step": 1240
},
{
"epoch": 0.7495484647802528,
"grad_norm": 0.9329442227378806,
"learning_rate": 1.7941232668271863e-06,
"loss": 1.0111,
"step": 1245
},
{
"epoch": 0.7525586995785671,
"grad_norm": 0.8933360465539598,
"learning_rate": 1.753959159720836e-06,
"loss": 1.0013,
"step": 1250
},
{
"epoch": 0.7555689343768814,
"grad_norm": 0.8850416820407446,
"learning_rate": 1.7141538824433506e-06,
"loss": 1.0092,
"step": 1255
},
{
"epoch": 0.7585791691751956,
"grad_norm": 0.8911148955200473,
"learning_rate": 1.6747118352230495e-06,
"loss": 1.0035,
"step": 1260
},
{
"epoch": 0.7615894039735099,
"grad_norm": 0.8485797781055803,
"learning_rate": 1.6356373781354058e-06,
"loss": 0.9819,
"step": 1265
},
{
"epoch": 0.7645996387718242,
"grad_norm": 0.8845205626085196,
"learning_rate": 1.5969348306210692e-06,
"loss": 1.0027,
"step": 1270
},
{
"epoch": 0.7676098735701384,
"grad_norm": 0.8858624981584671,
"learning_rate": 1.5586084710083737e-06,
"loss": 1.0093,
"step": 1275
},
{
"epoch": 0.7706201083684527,
"grad_norm": 0.8814363753611782,
"learning_rate": 1.5206625360403943e-06,
"loss": 0.9979,
"step": 1280
},
{
"epoch": 0.773630343166767,
"grad_norm": 0.8491013076963814,
"learning_rate": 1.4831012204066114e-06,
"loss": 1.0148,
"step": 1285
},
{
"epoch": 0.7766405779650812,
"grad_norm": 0.8677146375124508,
"learning_rate": 1.445928676279199e-06,
"loss": 0.9916,
"step": 1290
},
{
"epoch": 0.7796508127633955,
"grad_norm": 0.8646340907999152,
"learning_rate": 1.4091490128540374e-06,
"loss": 0.9836,
"step": 1295
},
{
"epoch": 0.7826610475617098,
"grad_norm": 0.8451931482695525,
"learning_rate": 1.3727662958964627e-06,
"loss": 0.9995,
"step": 1300
},
{
"epoch": 0.785671282360024,
"grad_norm": 0.8554438998915223,
"learning_rate": 1.3367845472918272e-06,
"loss": 1.0065,
"step": 1305
},
{
"epoch": 0.7886815171583383,
"grad_norm": 0.8789308792137018,
"learning_rate": 1.3012077446008969e-06,
"loss": 1.0012,
"step": 1310
},
{
"epoch": 0.7916917519566526,
"grad_norm": 0.8781828397403413,
"learning_rate": 1.266039820620159e-06,
"loss": 1.0042,
"step": 1315
},
{
"epoch": 0.7947019867549668,
"grad_norm": 0.8625507845317792,
"learning_rate": 1.2312846629470826e-06,
"loss": 0.9886,
"step": 1320
},
{
"epoch": 0.7977122215532811,
"grad_norm": 0.8959832830725274,
"learning_rate": 1.1969461135503573e-06,
"loss": 1.0059,
"step": 1325
},
{
"epoch": 0.8007224563515954,
"grad_norm": 0.8678075235569929,
"learning_rate": 1.163027968345195e-06,
"loss": 1.0094,
"step": 1330
},
{
"epoch": 0.8037326911499096,
"grad_norm": 1.0055210740771279,
"learning_rate": 1.1295339767737125e-06,
"loss": 0.9887,
"step": 1335
},
{
"epoch": 0.8067429259482239,
"grad_norm": 1.6253290809007488,
"learning_rate": 1.0964678413904529e-06,
"loss": 1.0028,
"step": 1340
},
{
"epoch": 0.8097531607465382,
"grad_norm": 0.9127445612020656,
"learning_rate": 1.0638332174530953e-06,
"loss": 1.0114,
"step": 1345
},
{
"epoch": 0.8127633955448526,
"grad_norm": 0.8928296605602871,
"learning_rate": 1.0316337125183817e-06,
"loss": 0.9991,
"step": 1350
},
{
"epoch": 0.8157736303431667,
"grad_norm": 1.1302003367983464,
"learning_rate": 9.998728860433277e-07,
"loss": 1.0002,
"step": 1355
},
{
"epoch": 0.818783865141481,
"grad_norm": 0.8807431712720114,
"learning_rate": 9.685542489917494e-07,
"loss": 0.9819,
"step": 1360
},
{
"epoch": 0.8217940999397954,
"grad_norm": 0.8894964118409194,
"learning_rate": 9.376812634461418e-07,
"loss": 0.9923,
"step": 1365
},
{
"epoch": 0.8248043347381095,
"grad_norm": 0.8950990656247966,
"learning_rate": 9.072573422249692e-07,
"loss": 1.0165,
"step": 1370
},
{
"epoch": 0.8278145695364238,
"grad_norm": 0.8600942880069391,
"learning_rate": 8.772858485054042e-07,
"loss": 1.0055,
"step": 1375
},
{
"epoch": 0.8308248043347382,
"grad_norm": 1.0176570771410658,
"learning_rate": 8.477700954515372e-07,
"loss": 1.0034,
"step": 1380
},
{
"epoch": 0.8338350391330523,
"grad_norm": 1.1332975568419288,
"learning_rate": 8.187133458481416e-07,
"loss": 1.0037,
"step": 1385
},
{
"epoch": 0.8368452739313667,
"grad_norm": 0.8771143983027159,
"learning_rate": 7.901188117399817e-07,
"loss": 1.0134,
"step": 1390
},
{
"epoch": 0.839855508729681,
"grad_norm": 0.8769447007569386,
"learning_rate": 7.619896540767435e-07,
"loss": 1.0137,
"step": 1395
},
{
"epoch": 0.8428657435279951,
"grad_norm": 0.8715300102195939,
"learning_rate": 7.343289823636168e-07,
"loss": 1.0119,
"step": 1400
},
{
"epoch": 0.8458759783263095,
"grad_norm": 3.8854548450704884,
"learning_rate": 7.0713985431755e-07,
"loss": 1.0101,
"step": 1405
},
{
"epoch": 0.8488862131246238,
"grad_norm": 0.8550203433574886,
"learning_rate": 6.804252755292429e-07,
"loss": 0.9804,
"step": 1410
},
{
"epoch": 0.851896447922938,
"grad_norm": 0.8574614831593009,
"learning_rate": 6.541881991309013e-07,
"loss": 0.9992,
"step": 1415
},
{
"epoch": 0.8549066827212523,
"grad_norm": 0.8801899228066352,
"learning_rate": 6.284315254697726e-07,
"loss": 1.0055,
"step": 1420
},
{
"epoch": 0.8579169175195666,
"grad_norm": 0.8646087529710637,
"learning_rate": 6.031581017875482e-07,
"loss": 1.0097,
"step": 1425
},
{
"epoch": 0.8609271523178808,
"grad_norm": 0.8697784834913861,
"learning_rate": 5.783707219056078e-07,
"loss": 1.0004,
"step": 1430
},
{
"epoch": 0.863937387116195,
"grad_norm": 0.8738950667485581,
"learning_rate": 5.540721259161774e-07,
"loss": 1.0078,
"step": 1435
},
{
"epoch": 0.8669476219145094,
"grad_norm": 0.8961144717239811,
"learning_rate": 5.302649998794368e-07,
"loss": 1.0113,
"step": 1440
},
{
"epoch": 0.8699578567128236,
"grad_norm": 0.870935027097703,
"learning_rate": 5.0695197552659e-07,
"loss": 0.9904,
"step": 1445
},
{
"epoch": 0.8729680915111379,
"grad_norm": 0.8567929575500429,
"learning_rate": 4.841356299689359e-07,
"loss": 1.0028,
"step": 1450
},
{
"epoch": 0.8759783263094522,
"grad_norm": 0.9123895694513533,
"learning_rate": 4.618184854129981e-07,
"loss": 1.0126,
"step": 1455
},
{
"epoch": 0.8789885611077664,
"grad_norm": 1.0746686361139868,
"learning_rate": 4.4000300888169753e-07,
"loss": 1.0017,
"step": 1460
},
{
"epoch": 0.8819987959060807,
"grad_norm": 0.8759725943448924,
"learning_rate": 4.1869161194164565e-07,
"loss": 1.0031,
"step": 1465
},
{
"epoch": 0.885009030704395,
"grad_norm": 1.4852691417263835,
"learning_rate": 3.9788665043656083e-07,
"loss": 0.9965,
"step": 1470
},
{
"epoch": 0.8880192655027093,
"grad_norm": 0.8929495276279715,
"learning_rate": 3.775904242268391e-07,
"loss": 1.0236,
"step": 1475
},
{
"epoch": 0.8910295003010235,
"grad_norm": 0.8758884812335123,
"learning_rate": 3.578051769353219e-07,
"loss": 1.0118,
"step": 1480
},
{
"epoch": 0.8940397350993378,
"grad_norm": 1.9572396797440648,
"learning_rate": 3.385330956992816e-07,
"loss": 0.9987,
"step": 1485
},
{
"epoch": 0.8970499698976521,
"grad_norm": 0.8813248265767825,
"learning_rate": 3.1977631092863613e-07,
"loss": 1.004,
"step": 1490
},
{
"epoch": 0.9000602046959663,
"grad_norm": 0.8785988861657655,
"learning_rate": 3.015368960704584e-07,
"loss": 0.9875,
"step": 1495
},
{
"epoch": 0.9030704394942806,
"grad_norm": 0.8712193660484964,
"learning_rate": 2.8381686737975867e-07,
"loss": 1.0108,
"step": 1500
},
{
"epoch": 0.9060806742925949,
"grad_norm": 0.8779863820284991,
"learning_rate": 2.666181836966053e-07,
"loss": 1.0141,
"step": 1505
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.8640378775819231,
"learning_rate": 2.4994274622958726e-07,
"loss": 1.0136,
"step": 1510
},
{
"epoch": 0.9121011438892234,
"grad_norm": 0.8607419097086582,
"learning_rate": 2.3379239834564526e-07,
"loss": 1.0044,
"step": 1515
},
{
"epoch": 0.9151113786875377,
"grad_norm": 0.8927580080579623,
"learning_rate": 2.1816892536629775e-07,
"loss": 1.0114,
"step": 1520
},
{
"epoch": 0.9181216134858519,
"grad_norm": 0.8535555555413261,
"learning_rate": 2.0307405437029027e-07,
"loss": 0.9961,
"step": 1525
},
{
"epoch": 0.9211318482841662,
"grad_norm": 0.9049922865987879,
"learning_rate": 1.8850945400266994e-07,
"loss": 1.0163,
"step": 1530
},
{
"epoch": 0.9241420830824805,
"grad_norm": 0.8499334715851968,
"learning_rate": 1.7447673429033361e-07,
"loss": 0.9946,
"step": 1535
},
{
"epoch": 0.9271523178807947,
"grad_norm": 0.8909209774091112,
"learning_rate": 1.6097744646404457e-07,
"loss": 0.9961,
"step": 1540
},
{
"epoch": 0.930162552679109,
"grad_norm": 0.8560970950116495,
"learning_rate": 1.4801308278695636e-07,
"loss": 1.0019,
"step": 1545
},
{
"epoch": 0.9331727874774233,
"grad_norm": 0.883934931226354,
"learning_rate": 1.3558507638965158e-07,
"loss": 1.0061,
"step": 1550
},
{
"epoch": 0.9361830222757375,
"grad_norm": 0.8961538667036166,
"learning_rate": 1.2369480111171784e-07,
"loss": 1.0054,
"step": 1555
},
{
"epoch": 0.9391932570740518,
"grad_norm": 0.9001828543117812,
"learning_rate": 1.1234357134987717e-07,
"loss": 0.9876,
"step": 1560
},
{
"epoch": 0.9422034918723661,
"grad_norm": 0.9606731439347472,
"learning_rate": 1.0153264191269052e-07,
"loss": 1.0106,
"step": 1565
},
{
"epoch": 0.9452137266706803,
"grad_norm": 0.9005969806614522,
"learning_rate": 9.126320788184374e-08,
"loss": 0.9934,
"step": 1570
},
{
"epoch": 0.9482239614689946,
"grad_norm": 0.8701900370996407,
"learning_rate": 8.153640448003875e-08,
"loss": 0.9845,
"step": 1575
},
{
"epoch": 0.9512341962673089,
"grad_norm": 0.8767879074096734,
"learning_rate": 7.235330694550402e-08,
"loss": 1.0158,
"step": 1580
},
{
"epoch": 0.9542444310656231,
"grad_norm": 0.8630669248508293,
"learning_rate": 6.371493041313126e-08,
"loss": 0.9907,
"step": 1585
},
{
"epoch": 0.9572546658639374,
"grad_norm": 3.3701456799509244,
"learning_rate": 5.562222980225907e-08,
"loss": 0.9962,
"step": 1590
},
{
"epoch": 0.9602649006622517,
"grad_norm": 0.8469748656240412,
"learning_rate": 4.807609971111238e-08,
"loss": 0.9891,
"step": 1595
},
{
"epoch": 0.963275135460566,
"grad_norm": 0.9099505574681345,
"learning_rate": 4.107737431791159e-08,
"loss": 1.0185,
"step": 1600
},
{
"epoch": 0.9662853702588802,
"grad_norm": 0.8973725037838026,
"learning_rate": 3.462682728865685e-08,
"loss": 0.9981,
"step": 1605
},
{
"epoch": 0.9692956050571945,
"grad_norm": 0.9061860450514891,
"learning_rate": 2.8725171691605934e-08,
"loss": 1.0174,
"step": 1610
},
{
"epoch": 0.9723058398555088,
"grad_norm": 0.981943649150158,
"learning_rate": 2.3373059918448958e-08,
"loss": 1.0,
"step": 1615
},
{
"epoch": 0.975316074653823,
"grad_norm": 0.8890699624303203,
"learning_rate": 1.8571083612188845e-08,
"loss": 1.0094,
"step": 1620
},
{
"epoch": 0.9783263094521373,
"grad_norm": 0.8705689131489066,
"learning_rate": 1.431977360173975e-08,
"loss": 0.9934,
"step": 1625
},
{
"epoch": 0.9813365442504516,
"grad_norm": 0.910606622312982,
"learning_rate": 1.0619599843249006e-08,
"loss": 1.001,
"step": 1630
},
{
"epoch": 0.9843467790487658,
"grad_norm": 0.8711153771031834,
"learning_rate": 7.470971368142011e-09,
"loss": 1.0114,
"step": 1635
},
{
"epoch": 0.9873570138470801,
"grad_norm": 0.8635302129219209,
"learning_rate": 4.874236237911723e-09,
"loss": 1.0028,
"step": 1640
},
{
"epoch": 0.9903672486453944,
"grad_norm": 0.8822503527540938,
"learning_rate": 2.8296815056377824e-09,
"loss": 1.0149,
"step": 1645
},
{
"epoch": 0.9933774834437086,
"grad_norm": 0.8786588529013926,
"learning_rate": 1.3375331842574446e-09,
"loss": 1.0336,
"step": 1650
},
{
"epoch": 0.9963877182420229,
"grad_norm": 3.6164785189545476,
"learning_rate": 3.9795622158111945e-10,
"loss": 1.0063,
"step": 1655
},
{
"epoch": 0.9993979530403372,
"grad_norm": 0.8926911293866052,
"learning_rate": 1.1054482056405136e-11,
"loss": 1.002,
"step": 1660
},
{
"epoch": 1.0,
"eval_runtime": 0.728,
"eval_samples_per_second": 13.736,
"eval_steps_per_second": 4.121,
"step": 1661
},
{
"epoch": 1.0,
"step": 1661,
"total_flos": 80925639573504.0,
"train_loss": 1.0968368504149306,
"train_runtime": 1531.4087,
"train_samples_per_second": 17.352,
"train_steps_per_second": 1.085
}
],
"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": 80925639573504.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}