qwen_feedback / trainer_state.json
terry69's picture
Model save
d59b163 verified
raw
history blame
58.9 kB
{
"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": 41.92205807920962,
"learning_rate": 5.98802395209581e-08,
"loss": 2.0696,
"step": 1
},
{
"epoch": 0.0030102347983142685,
"grad_norm": 41.95143902657454,
"learning_rate": 2.9940119760479047e-07,
"loss": 2.0656,
"step": 5
},
{
"epoch": 0.006020469596628537,
"grad_norm": 27.748468191825033,
"learning_rate": 5.988023952095809e-07,
"loss": 1.9946,
"step": 10
},
{
"epoch": 0.009030704394942806,
"grad_norm": 15.587842890269123,
"learning_rate": 8.982035928143713e-07,
"loss": 1.7623,
"step": 15
},
{
"epoch": 0.012040939193257074,
"grad_norm": 5.238640819714065,
"learning_rate": 1.1976047904191619e-06,
"loss": 1.5059,
"step": 20
},
{
"epoch": 0.015051173991571343,
"grad_norm": 4.531923782640025,
"learning_rate": 1.4970059880239521e-06,
"loss": 1.3322,
"step": 25
},
{
"epoch": 0.018061408789885613,
"grad_norm": 2.9213704309561614,
"learning_rate": 1.7964071856287426e-06,
"loss": 1.255,
"step": 30
},
{
"epoch": 0.02107164358819988,
"grad_norm": 1.9235170352981705,
"learning_rate": 2.095808383233533e-06,
"loss": 1.1734,
"step": 35
},
{
"epoch": 0.024081878386514148,
"grad_norm": 1.4658856095298773,
"learning_rate": 2.3952095808383237e-06,
"loss": 1.1022,
"step": 40
},
{
"epoch": 0.027092113184828417,
"grad_norm": 1.4116919266440267,
"learning_rate": 2.694610778443114e-06,
"loss": 1.0786,
"step": 45
},
{
"epoch": 0.030102347983142687,
"grad_norm": 1.2856710577917903,
"learning_rate": 2.9940119760479042e-06,
"loss": 1.0395,
"step": 50
},
{
"epoch": 0.033112582781456956,
"grad_norm": 1.191081117611533,
"learning_rate": 3.2934131736526947e-06,
"loss": 1.0493,
"step": 55
},
{
"epoch": 0.036122817579771226,
"grad_norm": 1.2407336303580925,
"learning_rate": 3.592814371257485e-06,
"loss": 1.0163,
"step": 60
},
{
"epoch": 0.03913305237808549,
"grad_norm": 1.2517904434964362,
"learning_rate": 3.892215568862276e-06,
"loss": 1.0031,
"step": 65
},
{
"epoch": 0.04214328717639976,
"grad_norm": 1.1898573918091981,
"learning_rate": 4.191616766467066e-06,
"loss": 0.9945,
"step": 70
},
{
"epoch": 0.04515352197471403,
"grad_norm": 1.2209003810778682,
"learning_rate": 4.4910179640718566e-06,
"loss": 0.9881,
"step": 75
},
{
"epoch": 0.048163756773028296,
"grad_norm": 1.2515970343619347,
"learning_rate": 4.7904191616766475e-06,
"loss": 0.9797,
"step": 80
},
{
"epoch": 0.051173991571342566,
"grad_norm": 1.192783734551512,
"learning_rate": 5.0898203592814375e-06,
"loss": 0.9627,
"step": 85
},
{
"epoch": 0.054184226369656835,
"grad_norm": 1.223525320981026,
"learning_rate": 5.389221556886228e-06,
"loss": 0.9511,
"step": 90
},
{
"epoch": 0.057194461167971104,
"grad_norm": 1.2516482707774526,
"learning_rate": 5.6886227544910184e-06,
"loss": 0.9366,
"step": 95
},
{
"epoch": 0.060204695966285374,
"grad_norm": 1.2648194370111776,
"learning_rate": 5.9880239520958085e-06,
"loss": 0.942,
"step": 100
},
{
"epoch": 0.06321493076459964,
"grad_norm": 1.301961795790511,
"learning_rate": 6.2874251497005985e-06,
"loss": 0.9239,
"step": 105
},
{
"epoch": 0.06622516556291391,
"grad_norm": 1.318910357881307,
"learning_rate": 6.586826347305389e-06,
"loss": 0.9295,
"step": 110
},
{
"epoch": 0.06923540036122817,
"grad_norm": 1.4146859865499117,
"learning_rate": 6.88622754491018e-06,
"loss": 0.9184,
"step": 115
},
{
"epoch": 0.07224563515954245,
"grad_norm": 1.3765402810382459,
"learning_rate": 7.18562874251497e-06,
"loss": 0.9104,
"step": 120
},
{
"epoch": 0.07525586995785671,
"grad_norm": 1.329497309221002,
"learning_rate": 7.485029940119761e-06,
"loss": 0.8829,
"step": 125
},
{
"epoch": 0.07826610475617098,
"grad_norm": 1.2582931017325312,
"learning_rate": 7.784431137724551e-06,
"loss": 0.8772,
"step": 130
},
{
"epoch": 0.08127633955448525,
"grad_norm": 1.2761906563557504,
"learning_rate": 8.083832335329342e-06,
"loss": 0.8863,
"step": 135
},
{
"epoch": 0.08428657435279951,
"grad_norm": 1.305371614001374,
"learning_rate": 8.383233532934131e-06,
"loss": 0.8736,
"step": 140
},
{
"epoch": 0.08729680915111379,
"grad_norm": 1.2522121501580266,
"learning_rate": 8.682634730538922e-06,
"loss": 0.87,
"step": 145
},
{
"epoch": 0.09030704394942805,
"grad_norm": 1.2869728645207454,
"learning_rate": 8.982035928143713e-06,
"loss": 0.8669,
"step": 150
},
{
"epoch": 0.09331727874774233,
"grad_norm": 1.2991051636455195,
"learning_rate": 9.281437125748504e-06,
"loss": 0.8576,
"step": 155
},
{
"epoch": 0.09632751354605659,
"grad_norm": 1.2786814630152046,
"learning_rate": 9.580838323353295e-06,
"loss": 0.8496,
"step": 160
},
{
"epoch": 0.09933774834437085,
"grad_norm": 1.2277883256106235,
"learning_rate": 9.880239520958084e-06,
"loss": 0.8393,
"step": 165
},
{
"epoch": 0.10234798314268513,
"grad_norm": 1.2307747824783823,
"learning_rate": 9.999900509954779e-06,
"loss": 0.8226,
"step": 170
},
{
"epoch": 0.1053582179409994,
"grad_norm": 1.2949808017913165,
"learning_rate": 9.999292529572152e-06,
"loss": 0.8414,
"step": 175
},
{
"epoch": 0.10836845273931367,
"grad_norm": 1.3249580865142179,
"learning_rate": 9.998131908181262e-06,
"loss": 0.8137,
"step": 180
},
{
"epoch": 0.11137868753762793,
"grad_norm": 1.371692604105339,
"learning_rate": 9.996418774081658e-06,
"loss": 0.8237,
"step": 185
},
{
"epoch": 0.11438892233594221,
"grad_norm": 1.2829546119130226,
"learning_rate": 9.994153316649769e-06,
"loss": 0.8205,
"step": 190
},
{
"epoch": 0.11739915713425647,
"grad_norm": 1.342998197089531,
"learning_rate": 9.991335786317964e-06,
"loss": 0.8239,
"step": 195
},
{
"epoch": 0.12040939193257075,
"grad_norm": 1.2976871743628908,
"learning_rate": 9.987966494546873e-06,
"loss": 0.8146,
"step": 200
},
{
"epoch": 0.12341962673088501,
"grad_norm": 1.2301463331443308,
"learning_rate": 9.984045813790959e-06,
"loss": 0.8283,
"step": 205
},
{
"epoch": 0.12642986152919927,
"grad_norm": 1.2412905970181487,
"learning_rate": 9.979574177457337e-06,
"loss": 0.8124,
"step": 210
},
{
"epoch": 0.12944009632751355,
"grad_norm": 1.283058093189711,
"learning_rate": 9.974552079857873e-06,
"loss": 0.8084,
"step": 215
},
{
"epoch": 0.13245033112582782,
"grad_norm": 1.2963566163720668,
"learning_rate": 9.968980076154533e-06,
"loss": 0.8188,
"step": 220
},
{
"epoch": 0.13546056592414207,
"grad_norm": 1.307056994934901,
"learning_rate": 9.962858782298023e-06,
"loss": 0.8244,
"step": 225
},
{
"epoch": 0.13847080072245635,
"grad_norm": 1.2138422771404471,
"learning_rate": 9.956188874959686e-06,
"loss": 0.8046,
"step": 230
},
{
"epoch": 0.14148103552077063,
"grad_norm": 1.24733957721358,
"learning_rate": 9.948971091456715e-06,
"loss": 0.8058,
"step": 235
},
{
"epoch": 0.1444912703190849,
"grad_norm": 1.3355706431877798,
"learning_rate": 9.941206229670634e-06,
"loss": 0.8117,
"step": 240
},
{
"epoch": 0.14750150511739915,
"grad_norm": 1.2561128971062672,
"learning_rate": 9.932895147959106e-06,
"loss": 0.798,
"step": 245
},
{
"epoch": 0.15051173991571343,
"grad_norm": 1.270591111220033,
"learning_rate": 9.924038765061042e-06,
"loss": 0.7829,
"step": 250
},
{
"epoch": 0.1535219747140277,
"grad_norm": 1.3107584214302355,
"learning_rate": 9.91463805999504e-06,
"loss": 0.7941,
"step": 255
},
{
"epoch": 0.15653220951234195,
"grad_norm": 1.2232660141148983,
"learning_rate": 9.904694071951167e-06,
"loss": 0.792,
"step": 260
},
{
"epoch": 0.15954244431065623,
"grad_norm": 1.2409348272552818,
"learning_rate": 9.894207900176074e-06,
"loss": 0.7951,
"step": 265
},
{
"epoch": 0.1625526791089705,
"grad_norm": 1.278980886871779,
"learning_rate": 9.883180703851488e-06,
"loss": 0.7945,
"step": 270
},
{
"epoch": 0.16556291390728478,
"grad_norm": 1.291936431929788,
"learning_rate": 9.871613701966067e-06,
"loss": 0.8016,
"step": 275
},
{
"epoch": 0.16857314870559903,
"grad_norm": 1.2668111499149215,
"learning_rate": 9.859508173180653e-06,
"loss": 0.7926,
"step": 280
},
{
"epoch": 0.1715833835039133,
"grad_norm": 1.250463723228362,
"learning_rate": 9.846865455686915e-06,
"loss": 0.7879,
"step": 285
},
{
"epoch": 0.17459361830222758,
"grad_norm": 1.3112862736174495,
"learning_rate": 9.833686947059436e-06,
"loss": 0.7929,
"step": 290
},
{
"epoch": 0.17760385310054183,
"grad_norm": 1.1820374232575113,
"learning_rate": 9.819974104101198e-06,
"loss": 0.7912,
"step": 295
},
{
"epoch": 0.1806140878988561,
"grad_norm": 1.1994367974309528,
"learning_rate": 9.80572844268256e-06,
"loss": 0.7709,
"step": 300
},
{
"epoch": 0.18362432269717038,
"grad_norm": 1.2106571499448653,
"learning_rate": 9.790951537573686e-06,
"loss": 0.7819,
"step": 305
},
{
"epoch": 0.18663455749548466,
"grad_norm": 1.2411822431871422,
"learning_rate": 9.775645022270448e-06,
"loss": 0.7798,
"step": 310
},
{
"epoch": 0.1896447922937989,
"grad_norm": 1.2173832414288233,
"learning_rate": 9.759810588813872e-06,
"loss": 0.7698,
"step": 315
},
{
"epoch": 0.19265502709211318,
"grad_norm": 1.206540724864329,
"learning_rate": 9.743449987603082e-06,
"loss": 0.7721,
"step": 320
},
{
"epoch": 0.19566526189042746,
"grad_norm": 1.211521580157664,
"learning_rate": 9.726565027201813e-06,
"loss": 0.7873,
"step": 325
},
{
"epoch": 0.1986754966887417,
"grad_norm": 1.19240172696107,
"learning_rate": 9.70915757413847e-06,
"loss": 0.7699,
"step": 330
},
{
"epoch": 0.20168573148705599,
"grad_norm": 1.1585014114964425,
"learning_rate": 9.691229552699817e-06,
"loss": 0.7662,
"step": 335
},
{
"epoch": 0.20469596628537026,
"grad_norm": 1.2388838685746641,
"learning_rate": 9.672782944718234e-06,
"loss": 0.7721,
"step": 340
},
{
"epoch": 0.20770620108368454,
"grad_norm": 1.1676447847746696,
"learning_rate": 9.65381978935266e-06,
"loss": 0.7926,
"step": 345
},
{
"epoch": 0.2107164358819988,
"grad_norm": 1.1880515427609326,
"learning_rate": 9.634342182863163e-06,
"loss": 0.7666,
"step": 350
},
{
"epoch": 0.21372667068031306,
"grad_norm": 1.3394374470348678,
"learning_rate": 9.614352278379217e-06,
"loss": 0.7751,
"step": 355
},
{
"epoch": 0.21673690547862734,
"grad_norm": 1.2148169383488967,
"learning_rate": 9.593852285661684e-06,
"loss": 0.7736,
"step": 360
},
{
"epoch": 0.2197471402769416,
"grad_norm": 1.2418535249704257,
"learning_rate": 9.572844470858537e-06,
"loss": 0.7667,
"step": 365
},
{
"epoch": 0.22275737507525586,
"grad_norm": 1.1836007116668403,
"learning_rate": 9.551331156254358e-06,
"loss": 0.7587,
"step": 370
},
{
"epoch": 0.22576760987357014,
"grad_norm": 1.2576722953575195,
"learning_rate": 9.529314720013618e-06,
"loss": 0.7649,
"step": 375
},
{
"epoch": 0.22877784467188442,
"grad_norm": 1.3127260473166227,
"learning_rate": 9.506797595917787e-06,
"loss": 0.7556,
"step": 380
},
{
"epoch": 0.23178807947019867,
"grad_norm": 1.2123856962626416,
"learning_rate": 9.483782273096295e-06,
"loss": 0.749,
"step": 385
},
{
"epoch": 0.23479831426851294,
"grad_norm": 1.2872889149850437,
"learning_rate": 9.460271295751373e-06,
"loss": 0.7574,
"step": 390
},
{
"epoch": 0.23780854906682722,
"grad_norm": 1.263319415183585,
"learning_rate": 9.436267262876808e-06,
"loss": 0.7587,
"step": 395
},
{
"epoch": 0.2408187838651415,
"grad_norm": 1.1849312644592684,
"learning_rate": 9.411772827970642e-06,
"loss": 0.7326,
"step": 400
},
{
"epoch": 0.24382901866345574,
"grad_norm": 1.1766218556431003,
"learning_rate": 9.38679069874184e-06,
"loss": 0.754,
"step": 405
},
{
"epoch": 0.24683925346177002,
"grad_norm": 1.290526908514049,
"learning_rate": 9.36132363681097e-06,
"loss": 0.7492,
"step": 410
},
{
"epoch": 0.2498494882600843,
"grad_norm": 1.1826527637981796,
"learning_rate": 9.335374457404928e-06,
"loss": 0.7458,
"step": 415
},
{
"epoch": 0.25285972305839854,
"grad_norm": 1.178690042871564,
"learning_rate": 9.308946029045726e-06,
"loss": 0.7409,
"step": 420
},
{
"epoch": 0.25586995785671285,
"grad_norm": 1.2484570253176477,
"learning_rate": 9.282041273233402e-06,
"loss": 0.7417,
"step": 425
},
{
"epoch": 0.2588801926550271,
"grad_norm": 1.273386773114662,
"learning_rate": 9.254663164123052e-06,
"loss": 0.7499,
"step": 430
},
{
"epoch": 0.26189042745334135,
"grad_norm": 1.1934961011332912,
"learning_rate": 9.226814728196072e-06,
"loss": 0.7149,
"step": 435
},
{
"epoch": 0.26490066225165565,
"grad_norm": 1.3353496110689795,
"learning_rate": 9.198499043925591e-06,
"loss": 0.7453,
"step": 440
},
{
"epoch": 0.2679108970499699,
"grad_norm": 1.2809286469157408,
"learning_rate": 9.169719241436162e-06,
"loss": 0.7326,
"step": 445
},
{
"epoch": 0.27092113184828415,
"grad_norm": 1.3365855596712244,
"learning_rate": 9.14047850215775e-06,
"loss": 0.7415,
"step": 450
},
{
"epoch": 0.27393136664659845,
"grad_norm": 1.2482839002741117,
"learning_rate": 9.110780058474052e-06,
"loss": 0.7363,
"step": 455
},
{
"epoch": 0.2769416014449127,
"grad_norm": 1.193302574955275,
"learning_rate": 9.080627193365155e-06,
"loss": 0.7354,
"step": 460
},
{
"epoch": 0.27995183624322695,
"grad_norm": 1.1845235529692022,
"learning_rate": 9.050023240044649e-06,
"loss": 0.7323,
"step": 465
},
{
"epoch": 0.28296207104154125,
"grad_norm": 1.1721388192198845,
"learning_rate": 9.018971581591141e-06,
"loss": 0.7375,
"step": 470
},
{
"epoch": 0.2859723058398555,
"grad_norm": 1.2075020325950543,
"learning_rate": 8.987475650574289e-06,
"loss": 0.731,
"step": 475
},
{
"epoch": 0.2889825406381698,
"grad_norm": 1.328392371779474,
"learning_rate": 8.955538928675343e-06,
"loss": 0.7232,
"step": 480
},
{
"epoch": 0.29199277543648405,
"grad_norm": 1.1739151729528425,
"learning_rate": 8.923164946302274e-06,
"loss": 0.7321,
"step": 485
},
{
"epoch": 0.2950030102347983,
"grad_norm": 1.1939351821984825,
"learning_rate": 8.890357282199504e-06,
"loss": 0.7252,
"step": 490
},
{
"epoch": 0.2980132450331126,
"grad_norm": 1.2277194033736811,
"learning_rate": 8.857119563052301e-06,
"loss": 0.7282,
"step": 495
},
{
"epoch": 0.30102347983142685,
"grad_norm": 1.2708327611399552,
"learning_rate": 8.823455463085873e-06,
"loss": 0.6969,
"step": 500
},
{
"epoch": 0.3040337146297411,
"grad_norm": 1.2115563428539284,
"learning_rate": 8.789368703659199e-06,
"loss": 0.7158,
"step": 505
},
{
"epoch": 0.3070439494280554,
"grad_norm": 1.1978355348696688,
"learning_rate": 8.754863052853658e-06,
"loss": 0.7171,
"step": 510
},
{
"epoch": 0.31005418422636966,
"grad_norm": 1.3124481912397998,
"learning_rate": 8.719942325056496e-06,
"loss": 0.7246,
"step": 515
},
{
"epoch": 0.3130644190246839,
"grad_norm": 1.2314941185008341,
"learning_rate": 8.68461038053916e-06,
"loss": 0.7263,
"step": 520
},
{
"epoch": 0.3160746538229982,
"grad_norm": 1.224013137534039,
"learning_rate": 8.648871125030576e-06,
"loss": 0.7083,
"step": 525
},
{
"epoch": 0.31908488862131246,
"grad_norm": 1.227965310988214,
"learning_rate": 8.612728509285395e-06,
"loss": 0.7095,
"step": 530
},
{
"epoch": 0.3220951234196267,
"grad_norm": 1.1959993758173042,
"learning_rate": 8.576186528647253e-06,
"loss": 0.694,
"step": 535
},
{
"epoch": 0.325105358217941,
"grad_norm": 1.2534772083993118,
"learning_rate": 8.53924922260712e-06,
"loss": 0.7168,
"step": 540
},
{
"epoch": 0.32811559301625526,
"grad_norm": 1.2274144164892773,
"learning_rate": 8.501920674356755e-06,
"loss": 0.7198,
"step": 545
},
{
"epoch": 0.33112582781456956,
"grad_norm": 1.210111783105508,
"learning_rate": 8.46420501033733e-06,
"loss": 0.7109,
"step": 550
},
{
"epoch": 0.3341360626128838,
"grad_norm": 1.2312597757641623,
"learning_rate": 8.42610639978329e-06,
"loss": 0.699,
"step": 555
},
{
"epoch": 0.33714629741119806,
"grad_norm": 1.1910413172047565,
"learning_rate": 8.387629054261454e-06,
"loss": 0.7048,
"step": 560
},
{
"epoch": 0.34015653220951236,
"grad_norm": 1.2182074145817383,
"learning_rate": 8.348777227205462e-06,
"loss": 0.7089,
"step": 565
},
{
"epoch": 0.3431667670078266,
"grad_norm": 1.22978245210155,
"learning_rate": 8.309555213445583e-06,
"loss": 0.6895,
"step": 570
},
{
"epoch": 0.34617700180614086,
"grad_norm": 1.2174592755837728,
"learning_rate": 8.269967348733947e-06,
"loss": 0.6912,
"step": 575
},
{
"epoch": 0.34918723660445516,
"grad_norm": 1.1889473916563937,
"learning_rate": 8.230018009265255e-06,
"loss": 0.7013,
"step": 580
},
{
"epoch": 0.3521974714027694,
"grad_norm": 1.236378472808724,
"learning_rate": 8.189711611193012e-06,
"loss": 0.6798,
"step": 585
},
{
"epoch": 0.35520770620108366,
"grad_norm": 1.2136777205264142,
"learning_rate": 8.149052610141357e-06,
"loss": 0.6916,
"step": 590
},
{
"epoch": 0.35821794099939797,
"grad_norm": 1.346765531962112,
"learning_rate": 8.108045500712518e-06,
"loss": 0.7018,
"step": 595
},
{
"epoch": 0.3612281757977122,
"grad_norm": 1.3636064667620953,
"learning_rate": 8.066694815989961e-06,
"loss": 0.674,
"step": 600
},
{
"epoch": 0.36423841059602646,
"grad_norm": 1.2275626464611498,
"learning_rate": 8.025005127037282e-06,
"loss": 0.6954,
"step": 605
},
{
"epoch": 0.36724864539434077,
"grad_norm": 1.262810130773417,
"learning_rate": 7.982981042392907e-06,
"loss": 0.6891,
"step": 610
},
{
"epoch": 0.370258880192655,
"grad_norm": 1.3116340438880467,
"learning_rate": 7.940627207560655e-06,
"loss": 0.7057,
"step": 615
},
{
"epoch": 0.3732691149909693,
"grad_norm": 1.2968792501944857,
"learning_rate": 7.897948304496189e-06,
"loss": 0.6816,
"step": 620
},
{
"epoch": 0.37627934978928357,
"grad_norm": 1.3837518855998683,
"learning_rate": 7.854949051089467e-06,
"loss": 0.6991,
"step": 625
},
{
"epoch": 0.3792895845875978,
"grad_norm": 1.2584131007142432,
"learning_rate": 7.811634200643202e-06,
"loss": 0.6839,
"step": 630
},
{
"epoch": 0.3822998193859121,
"grad_norm": 1.179717893419111,
"learning_rate": 7.768008541347423e-06,
"loss": 0.6781,
"step": 635
},
{
"epoch": 0.38531005418422637,
"grad_norm": 1.1579172841485028,
"learning_rate": 7.72407689575016e-06,
"loss": 0.6776,
"step": 640
},
{
"epoch": 0.3883202889825406,
"grad_norm": 1.2457079092742271,
"learning_rate": 7.67984412022434e-06,
"loss": 0.6885,
"step": 645
},
{
"epoch": 0.3913305237808549,
"grad_norm": 1.1752744144417815,
"learning_rate": 7.635315104430959e-06,
"loss": 0.6626,
"step": 650
},
{
"epoch": 0.39434075857916917,
"grad_norm": 1.2508542445727817,
"learning_rate": 7.5904947707785434e-06,
"loss": 0.6722,
"step": 655
},
{
"epoch": 0.3973509933774834,
"grad_norm": 1.2550002014037438,
"learning_rate": 7.545388073879018e-06,
"loss": 0.6921,
"step": 660
},
{
"epoch": 0.4003612281757977,
"grad_norm": 1.1364322698328715,
"learning_rate": 7.500000000000001e-06,
"loss": 0.6504,
"step": 665
},
{
"epoch": 0.40337146297411197,
"grad_norm": 1.223430718844965,
"learning_rate": 7.454335566513603e-06,
"loss": 0.6627,
"step": 670
},
{
"epoch": 0.4063816977724263,
"grad_norm": 1.325519859064528,
"learning_rate": 7.408399821341787e-06,
"loss": 0.6832,
"step": 675
},
{
"epoch": 0.4093919325707405,
"grad_norm": 1.2864653106496968,
"learning_rate": 7.362197842398355e-06,
"loss": 0.6766,
"step": 680
},
{
"epoch": 0.4124021673690548,
"grad_norm": 1.1824581903822915,
"learning_rate": 7.315734737027612e-06,
"loss": 0.6656,
"step": 685
},
{
"epoch": 0.4154124021673691,
"grad_norm": 1.2157872690010714,
"learning_rate": 7.2690156414397775e-06,
"loss": 0.6709,
"step": 690
},
{
"epoch": 0.4184226369656833,
"grad_norm": 1.19882231252741,
"learning_rate": 7.22204572014322e-06,
"loss": 0.6603,
"step": 695
},
{
"epoch": 0.4214328717639976,
"grad_norm": 1.2166274243887607,
"learning_rate": 7.174830165373542e-06,
"loss": 0.6511,
"step": 700
},
{
"epoch": 0.4244431065623119,
"grad_norm": 1.2085721212618408,
"learning_rate": 7.127374196519616e-06,
"loss": 0.6745,
"step": 705
},
{
"epoch": 0.4274533413606261,
"grad_norm": 1.2438528141488636,
"learning_rate": 7.079683059546607e-06,
"loss": 0.6691,
"step": 710
},
{
"epoch": 0.4304635761589404,
"grad_norm": 1.2395889341588784,
"learning_rate": 7.031762026416074e-06,
"loss": 0.664,
"step": 715
},
{
"epoch": 0.4334738109572547,
"grad_norm": 1.209939412026367,
"learning_rate": 6.983616394503177e-06,
"loss": 0.6545,
"step": 720
},
{
"epoch": 0.43648404575556893,
"grad_norm": 1.2085222788343624,
"learning_rate": 6.9352514860110876e-06,
"loss": 0.6652,
"step": 725
},
{
"epoch": 0.4394942805538832,
"grad_norm": 1.215874760750364,
"learning_rate": 6.886672647382653e-06,
"loss": 0.6506,
"step": 730
},
{
"epoch": 0.4425045153521975,
"grad_norm": 1.2270402546979953,
"learning_rate": 6.837885248709386e-06,
"loss": 0.6555,
"step": 735
},
{
"epoch": 0.44551475015051173,
"grad_norm": 1.168121733189555,
"learning_rate": 6.788894683137822e-06,
"loss": 0.6496,
"step": 740
},
{
"epoch": 0.44852498494882603,
"grad_norm": 1.2729494639482952,
"learning_rate": 6.739706366273346e-06,
"loss": 0.6744,
"step": 745
},
{
"epoch": 0.4515352197471403,
"grad_norm": 1.1910337115343361,
"learning_rate": 6.690325735581532e-06,
"loss": 0.6649,
"step": 750
},
{
"epoch": 0.45454545454545453,
"grad_norm": 1.2834186580973637,
"learning_rate": 6.640758249787067e-06,
"loss": 0.6288,
"step": 755
},
{
"epoch": 0.45755568934376883,
"grad_norm": 1.3482463281404071,
"learning_rate": 6.591009388270315e-06,
"loss": 0.638,
"step": 760
},
{
"epoch": 0.4605659241420831,
"grad_norm": 1.2245377747283543,
"learning_rate": 6.54108465046161e-06,
"loss": 0.6577,
"step": 765
},
{
"epoch": 0.46357615894039733,
"grad_norm": 1.2532400498054517,
"learning_rate": 6.490989555233328e-06,
"loss": 0.6398,
"step": 770
},
{
"epoch": 0.46658639373871164,
"grad_norm": 1.2133221218062722,
"learning_rate": 6.440729640289809e-06,
"loss": 0.6493,
"step": 775
},
{
"epoch": 0.4695966285370259,
"grad_norm": 1.2934721949521188,
"learning_rate": 6.3903104615551956e-06,
"loss": 0.6337,
"step": 780
},
{
"epoch": 0.47260686333534013,
"grad_norm": 1.208460074456768,
"learning_rate": 6.3397375925592675e-06,
"loss": 0.6426,
"step": 785
},
{
"epoch": 0.47561709813365444,
"grad_norm": 1.2643602971960362,
"learning_rate": 6.289016623821308e-06,
"loss": 0.6414,
"step": 790
},
{
"epoch": 0.4786273329319687,
"grad_norm": 1.199529441446071,
"learning_rate": 6.2381531622321234e-06,
"loss": 0.6487,
"step": 795
},
{
"epoch": 0.481637567730283,
"grad_norm": 1.2923378006389945,
"learning_rate": 6.18715283043422e-06,
"loss": 0.6242,
"step": 800
},
{
"epoch": 0.48464780252859724,
"grad_norm": 1.2138746939708784,
"learning_rate": 6.136021266200271e-06,
"loss": 0.6326,
"step": 805
},
{
"epoch": 0.4876580373269115,
"grad_norm": 1.261846103287385,
"learning_rate": 6.084764121809878e-06,
"loss": 0.6309,
"step": 810
},
{
"epoch": 0.4906682721252258,
"grad_norm": 1.2939948005814752,
"learning_rate": 6.033387063424765e-06,
"loss": 0.6252,
"step": 815
},
{
"epoch": 0.49367850692354004,
"grad_norm": 1.3038410279687773,
"learning_rate": 5.9818957704624046e-06,
"loss": 0.6335,
"step": 820
},
{
"epoch": 0.4966887417218543,
"grad_norm": 1.201247170646155,
"learning_rate": 5.930295934968197e-06,
"loss": 0.6262,
"step": 825
},
{
"epoch": 0.4996989765201686,
"grad_norm": 1.194976486878025,
"learning_rate": 5.878593260986256e-06,
"loss": 0.6341,
"step": 830
},
{
"epoch": 0.5027092113184829,
"grad_norm": 1.285380478781876,
"learning_rate": 5.8267934639288525e-06,
"loss": 0.6423,
"step": 835
},
{
"epoch": 0.5057194461167971,
"grad_norm": 1.2142854289892278,
"learning_rate": 5.77490226994462e-06,
"loss": 0.6217,
"step": 840
},
{
"epoch": 0.5087296809151114,
"grad_norm": 1.2387931529444611,
"learning_rate": 5.722925415285555e-06,
"loss": 0.6325,
"step": 845
},
{
"epoch": 0.5117399157134257,
"grad_norm": 1.166501906073654,
"learning_rate": 5.670868645672916e-06,
"loss": 0.6265,
"step": 850
},
{
"epoch": 0.5147501505117399,
"grad_norm": 1.2315216626331218,
"learning_rate": 5.618737715662067e-06,
"loss": 0.6163,
"step": 855
},
{
"epoch": 0.5177603853100542,
"grad_norm": 1.2402640957841304,
"learning_rate": 5.566538388006351e-06,
"loss": 0.6376,
"step": 860
},
{
"epoch": 0.5207706201083685,
"grad_norm": 1.2844662700519471,
"learning_rate": 5.514276433020044e-06,
"loss": 0.6146,
"step": 865
},
{
"epoch": 0.5237808549066827,
"grad_norm": 1.2752483334016254,
"learning_rate": 5.461957627940489e-06,
"loss": 0.6394,
"step": 870
},
{
"epoch": 0.526791089704997,
"grad_norm": 1.227395524806422,
"learning_rate": 5.409587756289462e-06,
"loss": 0.6139,
"step": 875
},
{
"epoch": 0.5298013245033113,
"grad_norm": 1.2049284439880659,
"learning_rate": 5.357172607233831e-06,
"loss": 0.6144,
"step": 880
},
{
"epoch": 0.5328115593016255,
"grad_norm": 1.245827654247563,
"learning_rate": 5.304717974945596e-06,
"loss": 0.6152,
"step": 885
},
{
"epoch": 0.5358217940999398,
"grad_norm": 1.3299253926984027,
"learning_rate": 5.252229657961394e-06,
"loss": 0.6121,
"step": 890
},
{
"epoch": 0.5388320288982541,
"grad_norm": 1.209710810071064,
"learning_rate": 5.199713458541495e-06,
"loss": 0.5971,
"step": 895
},
{
"epoch": 0.5418422636965683,
"grad_norm": 1.254649968737726,
"learning_rate": 5.1471751820284e-06,
"loss": 0.6022,
"step": 900
},
{
"epoch": 0.5448524984948826,
"grad_norm": 1.267331556965335,
"learning_rate": 5.094620636205096e-06,
"loss": 0.6163,
"step": 905
},
{
"epoch": 0.5478627332931969,
"grad_norm": 1.2114393038072488,
"learning_rate": 5.042055630653042e-06,
"loss": 0.6162,
"step": 910
},
{
"epoch": 0.5508729680915111,
"grad_norm": 1.2119812194103736,
"learning_rate": 4.98948597610996e-06,
"loss": 0.6103,
"step": 915
},
{
"epoch": 0.5538832028898254,
"grad_norm": 1.2118940977952493,
"learning_rate": 4.936917483827483e-06,
"loss": 0.609,
"step": 920
},
{
"epoch": 0.5568934376881397,
"grad_norm": 1.216555361613017,
"learning_rate": 4.884355964928767e-06,
"loss": 0.5869,
"step": 925
},
{
"epoch": 0.5599036724864539,
"grad_norm": 1.2125436521321353,
"learning_rate": 4.831807229766101e-06,
"loss": 0.5959,
"step": 930
},
{
"epoch": 0.5629139072847682,
"grad_norm": 1.2273004433080854,
"learning_rate": 4.779277087278615e-06,
"loss": 0.5956,
"step": 935
},
{
"epoch": 0.5659241420830825,
"grad_norm": 1.327408954158384,
"learning_rate": 4.7267713443501274e-06,
"loss": 0.6143,
"step": 940
},
{
"epoch": 0.5689343768813967,
"grad_norm": 1.2414363124005465,
"learning_rate": 4.67429580516724e-06,
"loss": 0.5887,
"step": 945
},
{
"epoch": 0.571944611679711,
"grad_norm": 1.203442995074465,
"learning_rate": 4.6218562705777185e-06,
"loss": 0.6004,
"step": 950
},
{
"epoch": 0.5749548464780253,
"grad_norm": 1.2174481353824513,
"learning_rate": 4.5694585374492314e-06,
"loss": 0.5853,
"step": 955
},
{
"epoch": 0.5779650812763396,
"grad_norm": 1.2904799934518778,
"learning_rate": 4.517108398028566e-06,
"loss": 0.6017,
"step": 960
},
{
"epoch": 0.5809753160746538,
"grad_norm": 1.2287011726696084,
"learning_rate": 4.464811639301314e-06,
"loss": 0.606,
"step": 965
},
{
"epoch": 0.5839855508729681,
"grad_norm": 1.2571252783339275,
"learning_rate": 4.412574042352156e-06,
"loss": 0.6011,
"step": 970
},
{
"epoch": 0.5869957856712824,
"grad_norm": 1.2246675913495295,
"learning_rate": 4.360401381725806e-06,
"loss": 0.5845,
"step": 975
},
{
"epoch": 0.5900060204695966,
"grad_norm": 1.269602392062893,
"learning_rate": 4.308299424788667e-06,
"loss": 0.578,
"step": 980
},
{
"epoch": 0.5930162552679109,
"grad_norm": 1.2266667172046457,
"learning_rate": 4.256273931091284e-06,
"loss": 0.596,
"step": 985
},
{
"epoch": 0.5960264900662252,
"grad_norm": 1.205718345046694,
"learning_rate": 4.204330651731662e-06,
"loss": 0.5936,
"step": 990
},
{
"epoch": 0.5990367248645394,
"grad_norm": 1.1952203148817158,
"learning_rate": 4.152475328719517e-06,
"loss": 0.5738,
"step": 995
},
{
"epoch": 0.6020469596628537,
"grad_norm": 1.276847431530527,
"learning_rate": 4.1007136943415325e-06,
"loss": 0.5686,
"step": 1000
},
{
"epoch": 0.605057194461168,
"grad_norm": 1.2563772516846479,
"learning_rate": 4.049051470527692e-06,
"loss": 0.5825,
"step": 1005
},
{
"epoch": 0.6080674292594822,
"grad_norm": 1.3085315216721447,
"learning_rate": 3.997494368218745e-06,
"loss": 0.5796,
"step": 1010
},
{
"epoch": 0.6110776640577965,
"grad_norm": 1.252832406830381,
"learning_rate": 3.946048086734921e-06,
"loss": 0.5912,
"step": 1015
},
{
"epoch": 0.6140878988561108,
"grad_norm": 1.241503492216098,
"learning_rate": 3.894718313145873e-06,
"loss": 0.5811,
"step": 1020
},
{
"epoch": 0.617098133654425,
"grad_norm": 1.1913066136989139,
"learning_rate": 3.843510721642036e-06,
"loss": 0.5718,
"step": 1025
},
{
"epoch": 0.6201083684527393,
"grad_norm": 1.3275708487415967,
"learning_rate": 3.7924309729073616e-06,
"loss": 0.5609,
"step": 1030
},
{
"epoch": 0.6231186032510536,
"grad_norm": 1.256537476019095,
"learning_rate": 3.7414847134935716e-06,
"loss": 0.5825,
"step": 1035
},
{
"epoch": 0.6261288380493678,
"grad_norm": 1.221832356712663,
"learning_rate": 3.6906775751959667e-06,
"loss": 0.554,
"step": 1040
},
{
"epoch": 0.6291390728476821,
"grad_norm": 1.2248781750755409,
"learning_rate": 3.640015174430864e-06,
"loss": 0.5638,
"step": 1045
},
{
"epoch": 0.6321493076459964,
"grad_norm": 1.2047806028877057,
"learning_rate": 3.5895031116147355e-06,
"loss": 0.5733,
"step": 1050
},
{
"epoch": 0.6351595424443106,
"grad_norm": 1.275159952276474,
"learning_rate": 3.539146970545124e-06,
"loss": 0.5799,
"step": 1055
},
{
"epoch": 0.6381697772426249,
"grad_norm": 1.284257273469691,
"learning_rate": 3.488952317783374e-06,
"loss": 0.5726,
"step": 1060
},
{
"epoch": 0.6411800120409392,
"grad_norm": 1.2383109080935772,
"learning_rate": 3.438924702039301e-06,
"loss": 0.5708,
"step": 1065
},
{
"epoch": 0.6441902468392534,
"grad_norm": 1.2190673552207432,
"learning_rate": 3.389069653557805e-06,
"loss": 0.5538,
"step": 1070
},
{
"epoch": 0.6472004816375677,
"grad_norm": 1.2506993933070911,
"learning_rate": 3.3393926835075307e-06,
"loss": 0.5797,
"step": 1075
},
{
"epoch": 0.650210716435882,
"grad_norm": 1.2356328454342058,
"learning_rate": 3.289899283371657e-06,
"loss": 0.5585,
"step": 1080
},
{
"epoch": 0.6532209512341962,
"grad_norm": 1.3063384197021242,
"learning_rate": 3.240594924340835e-06,
"loss": 0.5796,
"step": 1085
},
{
"epoch": 0.6562311860325105,
"grad_norm": 1.2373220459625451,
"learning_rate": 3.1914850567083866e-06,
"loss": 0.5622,
"step": 1090
},
{
"epoch": 0.6592414208308248,
"grad_norm": 1.3531129912740778,
"learning_rate": 3.1425751092678064e-06,
"loss": 0.5596,
"step": 1095
},
{
"epoch": 0.6622516556291391,
"grad_norm": 1.2389296621775976,
"learning_rate": 3.0938704887126425e-06,
"loss": 0.5535,
"step": 1100
},
{
"epoch": 0.6652618904274533,
"grad_norm": 1.2515989438388835,
"learning_rate": 3.045376579038821e-06,
"loss": 0.5681,
"step": 1105
},
{
"epoch": 0.6682721252257676,
"grad_norm": 1.2185663400088427,
"learning_rate": 2.9970987409494784e-06,
"loss": 0.571,
"step": 1110
},
{
"epoch": 0.6712823600240819,
"grad_norm": 1.269873255758116,
"learning_rate": 2.9490423112623646e-06,
"loss": 0.5658,
"step": 1115
},
{
"epoch": 0.6742925948223961,
"grad_norm": 1.2244029668023797,
"learning_rate": 2.9012126023198973e-06,
"loss": 0.538,
"step": 1120
},
{
"epoch": 0.6773028296207104,
"grad_norm": 1.2816311543199834,
"learning_rate": 2.853614901401909e-06,
"loss": 0.5484,
"step": 1125
},
{
"epoch": 0.6803130644190247,
"grad_norm": 1.276081217254366,
"learning_rate": 2.806254470141174e-06,
"loss": 0.5691,
"step": 1130
},
{
"epoch": 0.6833232992173389,
"grad_norm": 1.266389306582438,
"learning_rate": 2.759136543941773e-06,
"loss": 0.5567,
"step": 1135
},
{
"epoch": 0.6863335340156532,
"grad_norm": 1.2924072931517483,
"learning_rate": 2.712266331400332e-06,
"loss": 0.5439,
"step": 1140
},
{
"epoch": 0.6893437688139675,
"grad_norm": 1.254669923456662,
"learning_rate": 2.66564901373027e-06,
"loss": 0.5613,
"step": 1145
},
{
"epoch": 0.6923540036122817,
"grad_norm": 1.3474144476855063,
"learning_rate": 2.6192897441890337e-06,
"loss": 0.5489,
"step": 1150
},
{
"epoch": 0.695364238410596,
"grad_norm": 1.269185993074116,
"learning_rate": 2.573193647508426e-06,
"loss": 0.5483,
"step": 1155
},
{
"epoch": 0.6983744732089103,
"grad_norm": 1.3562380015815203,
"learning_rate": 2.5273658193281252e-06,
"loss": 0.5567,
"step": 1160
},
{
"epoch": 0.7013847080072245,
"grad_norm": 1.2580827828033503,
"learning_rate": 2.4818113256323745e-06,
"loss": 0.5417,
"step": 1165
},
{
"epoch": 0.7043949428055388,
"grad_norm": 1.2738218692009522,
"learning_rate": 2.4365352021899635e-06,
"loss": 0.5418,
"step": 1170
},
{
"epoch": 0.7074051776038531,
"grad_norm": 1.257678286567073,
"learning_rate": 2.391542453997578e-06,
"loss": 0.5502,
"step": 1175
},
{
"epoch": 0.7104154124021673,
"grad_norm": 1.275010857155547,
"learning_rate": 2.346838054726505e-06,
"loss": 0.5281,
"step": 1180
},
{
"epoch": 0.7134256472004816,
"grad_norm": 1.3071543695724546,
"learning_rate": 2.302426946172836e-06,
"loss": 0.5324,
"step": 1185
},
{
"epoch": 0.7164358819987959,
"grad_norm": 1.2139842130794491,
"learning_rate": 2.258314037711184e-06,
"loss": 0.554,
"step": 1190
},
{
"epoch": 0.7194461167971101,
"grad_norm": 1.2099665519481047,
"learning_rate": 2.214504205751971e-06,
"loss": 0.5429,
"step": 1195
},
{
"epoch": 0.7224563515954244,
"grad_norm": 1.3437644441417467,
"learning_rate": 2.1710022932023805e-06,
"loss": 0.5472,
"step": 1200
},
{
"epoch": 0.7254665863937387,
"grad_norm": 1.26435415137662,
"learning_rate": 2.127813108931007e-06,
"loss": 0.5487,
"step": 1205
},
{
"epoch": 0.7284768211920529,
"grad_norm": 1.2092596990063904,
"learning_rate": 2.084941427236245e-06,
"loss": 0.533,
"step": 1210
},
{
"epoch": 0.7314870559903672,
"grad_norm": 1.2313416526959684,
"learning_rate": 2.04239198731855e-06,
"loss": 0.5349,
"step": 1215
},
{
"epoch": 0.7344972907886815,
"grad_norm": 1.3072779429716,
"learning_rate": 2.000169492756523e-06,
"loss": 0.5402,
"step": 1220
},
{
"epoch": 0.7375075255869958,
"grad_norm": 1.2795138536801312,
"learning_rate": 1.9582786109869713e-06,
"loss": 0.5478,
"step": 1225
},
{
"epoch": 0.74051776038531,
"grad_norm": 1.2961377151032474,
"learning_rate": 1.9167239727889527e-06,
"loss": 0.543,
"step": 1230
},
{
"epoch": 0.7435279951836243,
"grad_norm": 1.2566416288826483,
"learning_rate": 1.875510171771865e-06,
"loss": 0.5277,
"step": 1235
},
{
"epoch": 0.7465382299819386,
"grad_norm": 1.2674802415781152,
"learning_rate": 1.8346417638676533e-06,
"loss": 0.5401,
"step": 1240
},
{
"epoch": 0.7495484647802528,
"grad_norm": 1.2584995829737369,
"learning_rate": 1.7941232668271863e-06,
"loss": 0.5315,
"step": 1245
},
{
"epoch": 0.7525586995785671,
"grad_norm": 1.2445698965082652,
"learning_rate": 1.753959159720836e-06,
"loss": 0.5294,
"step": 1250
},
{
"epoch": 0.7555689343768814,
"grad_norm": 1.2674536046077867,
"learning_rate": 1.7141538824433506e-06,
"loss": 0.5278,
"step": 1255
},
{
"epoch": 0.7585791691751956,
"grad_norm": 1.294324607883709,
"learning_rate": 1.6747118352230495e-06,
"loss": 0.5392,
"step": 1260
},
{
"epoch": 0.7615894039735099,
"grad_norm": 1.2340338478309871,
"learning_rate": 1.6356373781354058e-06,
"loss": 0.5335,
"step": 1265
},
{
"epoch": 0.7645996387718242,
"grad_norm": 1.3398026544287736,
"learning_rate": 1.5969348306210692e-06,
"loss": 0.5322,
"step": 1270
},
{
"epoch": 0.7676098735701384,
"grad_norm": 1.3532898551242525,
"learning_rate": 1.5586084710083737e-06,
"loss": 0.5198,
"step": 1275
},
{
"epoch": 0.7706201083684527,
"grad_norm": 1.315480840691035,
"learning_rate": 1.5206625360403943e-06,
"loss": 0.5319,
"step": 1280
},
{
"epoch": 0.773630343166767,
"grad_norm": 1.255766242972646,
"learning_rate": 1.4831012204066114e-06,
"loss": 0.5101,
"step": 1285
},
{
"epoch": 0.7766405779650812,
"grad_norm": 1.3117533667822703,
"learning_rate": 1.445928676279199e-06,
"loss": 0.5369,
"step": 1290
},
{
"epoch": 0.7796508127633955,
"grad_norm": 1.288985322714688,
"learning_rate": 1.4091490128540374e-06,
"loss": 0.526,
"step": 1295
},
{
"epoch": 0.7826610475617098,
"grad_norm": 1.2093920386965087,
"learning_rate": 1.3727662958964627e-06,
"loss": 0.5152,
"step": 1300
},
{
"epoch": 0.785671282360024,
"grad_norm": 1.288741507825481,
"learning_rate": 1.3367845472918272e-06,
"loss": 0.5242,
"step": 1305
},
{
"epoch": 0.7886815171583383,
"grad_norm": 1.289266020619647,
"learning_rate": 1.3012077446008969e-06,
"loss": 0.519,
"step": 1310
},
{
"epoch": 0.7916917519566526,
"grad_norm": 1.305675104370065,
"learning_rate": 1.266039820620159e-06,
"loss": 0.5241,
"step": 1315
},
{
"epoch": 0.7947019867549668,
"grad_norm": 1.2633206609620018,
"learning_rate": 1.2312846629470826e-06,
"loss": 0.5298,
"step": 1320
},
{
"epoch": 0.7977122215532811,
"grad_norm": 1.2569813999396267,
"learning_rate": 1.1969461135503573e-06,
"loss": 0.5284,
"step": 1325
},
{
"epoch": 0.8007224563515954,
"grad_norm": 1.2500080013485295,
"learning_rate": 1.163027968345195e-06,
"loss": 0.5013,
"step": 1330
},
{
"epoch": 0.8037326911499096,
"grad_norm": 1.2227749344044816,
"learning_rate": 1.1295339767737125e-06,
"loss": 0.5181,
"step": 1335
},
{
"epoch": 0.8067429259482239,
"grad_norm": 1.2039813893211253,
"learning_rate": 1.0964678413904529e-06,
"loss": 0.5077,
"step": 1340
},
{
"epoch": 0.8097531607465382,
"grad_norm": 1.2432336763434526,
"learning_rate": 1.0638332174530953e-06,
"loss": 0.5183,
"step": 1345
},
{
"epoch": 0.8127633955448526,
"grad_norm": 1.3224425175232468,
"learning_rate": 1.0316337125183817e-06,
"loss": 0.5085,
"step": 1350
},
{
"epoch": 0.8157736303431667,
"grad_norm": 1.2490625646083682,
"learning_rate": 9.998728860433277e-07,
"loss": 0.512,
"step": 1355
},
{
"epoch": 0.818783865141481,
"grad_norm": 1.2834749344502785,
"learning_rate": 9.685542489917494e-07,
"loss": 0.5148,
"step": 1360
},
{
"epoch": 0.8217940999397954,
"grad_norm": 1.2717678687726444,
"learning_rate": 9.376812634461418e-07,
"loss": 0.5165,
"step": 1365
},
{
"epoch": 0.8248043347381095,
"grad_norm": 1.273693819691485,
"learning_rate": 9.072573422249692e-07,
"loss": 0.5138,
"step": 1370
},
{
"epoch": 0.8278145695364238,
"grad_norm": 1.2541602658022404,
"learning_rate": 8.772858485054042e-07,
"loss": 0.5341,
"step": 1375
},
{
"epoch": 0.8308248043347382,
"grad_norm": 1.311251853744712,
"learning_rate": 8.477700954515372e-07,
"loss": 0.5104,
"step": 1380
},
{
"epoch": 0.8338350391330523,
"grad_norm": 1.2486954380131434,
"learning_rate": 8.187133458481416e-07,
"loss": 0.5156,
"step": 1385
},
{
"epoch": 0.8368452739313667,
"grad_norm": 1.2747215761089814,
"learning_rate": 7.901188117399817e-07,
"loss": 0.5211,
"step": 1390
},
{
"epoch": 0.839855508729681,
"grad_norm": 1.234543288759014,
"learning_rate": 7.619896540767435e-07,
"loss": 0.5157,
"step": 1395
},
{
"epoch": 0.8428657435279951,
"grad_norm": 1.3263990218789858,
"learning_rate": 7.343289823636168e-07,
"loss": 0.5147,
"step": 1400
},
{
"epoch": 0.8458759783263095,
"grad_norm": 1.2622046736064572,
"learning_rate": 7.0713985431755e-07,
"loss": 0.5222,
"step": 1405
},
{
"epoch": 0.8488862131246238,
"grad_norm": 1.2345168269397355,
"learning_rate": 6.804252755292429e-07,
"loss": 0.5226,
"step": 1410
},
{
"epoch": 0.851896447922938,
"grad_norm": 1.244419436243612,
"learning_rate": 6.541881991309013e-07,
"loss": 0.5211,
"step": 1415
},
{
"epoch": 0.8549066827212523,
"grad_norm": 1.277021130264081,
"learning_rate": 6.284315254697726e-07,
"loss": 0.514,
"step": 1420
},
{
"epoch": 0.8579169175195666,
"grad_norm": 1.3651795136190366,
"learning_rate": 6.031581017875482e-07,
"loss": 0.5196,
"step": 1425
},
{
"epoch": 0.8609271523178808,
"grad_norm": 1.2924139249953397,
"learning_rate": 5.783707219056078e-07,
"loss": 0.4952,
"step": 1430
},
{
"epoch": 0.863937387116195,
"grad_norm": 1.29441777957266,
"learning_rate": 5.540721259161774e-07,
"loss": 0.5093,
"step": 1435
},
{
"epoch": 0.8669476219145094,
"grad_norm": 1.2349372001584507,
"learning_rate": 5.302649998794368e-07,
"loss": 0.5189,
"step": 1440
},
{
"epoch": 0.8699578567128236,
"grad_norm": 1.3010846373559104,
"learning_rate": 5.0695197552659e-07,
"loss": 0.5125,
"step": 1445
},
{
"epoch": 0.8729680915111379,
"grad_norm": 1.2313741385914343,
"learning_rate": 4.841356299689359e-07,
"loss": 0.5019,
"step": 1450
},
{
"epoch": 0.8759783263094522,
"grad_norm": 1.2736789351010163,
"learning_rate": 4.618184854129981e-07,
"loss": 0.5162,
"step": 1455
},
{
"epoch": 0.8789885611077664,
"grad_norm": 1.3450829294599627,
"learning_rate": 4.4000300888169753e-07,
"loss": 0.5122,
"step": 1460
},
{
"epoch": 0.8819987959060807,
"grad_norm": 1.3345961379400417,
"learning_rate": 4.1869161194164565e-07,
"loss": 0.5195,
"step": 1465
},
{
"epoch": 0.885009030704395,
"grad_norm": 1.2706980513616788,
"learning_rate": 3.9788665043656083e-07,
"loss": 0.5272,
"step": 1470
},
{
"epoch": 0.8880192655027093,
"grad_norm": 1.298097370684443,
"learning_rate": 3.775904242268391e-07,
"loss": 0.5189,
"step": 1475
},
{
"epoch": 0.8910295003010235,
"grad_norm": 1.264558600518237,
"learning_rate": 3.578051769353219e-07,
"loss": 0.5105,
"step": 1480
},
{
"epoch": 0.8940397350993378,
"grad_norm": 1.2417670245253374,
"learning_rate": 3.385330956992816e-07,
"loss": 0.4906,
"step": 1485
},
{
"epoch": 0.8970499698976521,
"grad_norm": 1.3169348811520523,
"learning_rate": 3.1977631092863613e-07,
"loss": 0.5162,
"step": 1490
},
{
"epoch": 0.9000602046959663,
"grad_norm": 1.3331828477675507,
"learning_rate": 3.015368960704584e-07,
"loss": 0.5118,
"step": 1495
},
{
"epoch": 0.9030704394942806,
"grad_norm": 1.2820985050226046,
"learning_rate": 2.8381686737975867e-07,
"loss": 0.5034,
"step": 1500
},
{
"epoch": 0.9060806742925949,
"grad_norm": 1.369084535263281,
"learning_rate": 2.666181836966053e-07,
"loss": 0.5111,
"step": 1505
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.2852930092164696,
"learning_rate": 2.4994274622958726e-07,
"loss": 0.4986,
"step": 1510
},
{
"epoch": 0.9121011438892234,
"grad_norm": 1.2402160274810259,
"learning_rate": 2.3379239834564526e-07,
"loss": 0.5118,
"step": 1515
},
{
"epoch": 0.9151113786875377,
"grad_norm": 1.237567686573788,
"learning_rate": 2.1816892536629775e-07,
"loss": 0.5035,
"step": 1520
},
{
"epoch": 0.9181216134858519,
"grad_norm": 1.2495841691480962,
"learning_rate": 2.0307405437029027e-07,
"loss": 0.5111,
"step": 1525
},
{
"epoch": 0.9211318482841662,
"grad_norm": 1.2363443537540169,
"learning_rate": 1.8850945400266994e-07,
"loss": 0.4958,
"step": 1530
},
{
"epoch": 0.9241420830824805,
"grad_norm": 1.2925244771883408,
"learning_rate": 1.7447673429033361e-07,
"loss": 0.5073,
"step": 1535
},
{
"epoch": 0.9271523178807947,
"grad_norm": 1.2918239018472708,
"learning_rate": 1.6097744646404457e-07,
"loss": 0.505,
"step": 1540
},
{
"epoch": 0.930162552679109,
"grad_norm": 1.3462621823153225,
"learning_rate": 1.4801308278695636e-07,
"loss": 0.5208,
"step": 1545
},
{
"epoch": 0.9331727874774233,
"grad_norm": 1.3040362281207198,
"learning_rate": 1.3558507638965158e-07,
"loss": 0.5068,
"step": 1550
},
{
"epoch": 0.9361830222757375,
"grad_norm": 1.2501483756470055,
"learning_rate": 1.2369480111171784e-07,
"loss": 0.4984,
"step": 1555
},
{
"epoch": 0.9391932570740518,
"grad_norm": 1.24010506497212,
"learning_rate": 1.1234357134987717e-07,
"loss": 0.5036,
"step": 1560
},
{
"epoch": 0.9422034918723661,
"grad_norm": 1.2731162265485154,
"learning_rate": 1.0153264191269052e-07,
"loss": 0.5159,
"step": 1565
},
{
"epoch": 0.9452137266706803,
"grad_norm": 1.241141289633183,
"learning_rate": 9.126320788184374e-08,
"loss": 0.5077,
"step": 1570
},
{
"epoch": 0.9482239614689946,
"grad_norm": 1.2598787839937213,
"learning_rate": 8.153640448003875e-08,
"loss": 0.4974,
"step": 1575
},
{
"epoch": 0.9512341962673089,
"grad_norm": 1.2922592226571687,
"learning_rate": 7.235330694550402e-08,
"loss": 0.4975,
"step": 1580
},
{
"epoch": 0.9542444310656231,
"grad_norm": 1.2596424466646803,
"learning_rate": 6.371493041313126e-08,
"loss": 0.52,
"step": 1585
},
{
"epoch": 0.9572546658639374,
"grad_norm": 1.2829293637589985,
"learning_rate": 5.562222980225907e-08,
"loss": 0.5118,
"step": 1590
},
{
"epoch": 0.9602649006622517,
"grad_norm": 1.2885001691506446,
"learning_rate": 4.807609971111238e-08,
"loss": 0.5082,
"step": 1595
},
{
"epoch": 0.963275135460566,
"grad_norm": 1.251029961869992,
"learning_rate": 4.107737431791159e-08,
"loss": 0.5135,
"step": 1600
},
{
"epoch": 0.9662853702588802,
"grad_norm": 1.2394267183031216,
"learning_rate": 3.462682728865685e-08,
"loss": 0.5104,
"step": 1605
},
{
"epoch": 0.9692956050571945,
"grad_norm": 1.2966890851556148,
"learning_rate": 2.8725171691605934e-08,
"loss": 0.5123,
"step": 1610
},
{
"epoch": 0.9723058398555088,
"grad_norm": 1.2442957383652233,
"learning_rate": 2.3373059918448958e-08,
"loss": 0.5047,
"step": 1615
},
{
"epoch": 0.975316074653823,
"grad_norm": 1.2908864465886667,
"learning_rate": 1.8571083612188845e-08,
"loss": 0.5063,
"step": 1620
},
{
"epoch": 0.9783263094521373,
"grad_norm": 1.2475359893632505,
"learning_rate": 1.431977360173975e-08,
"loss": 0.5029,
"step": 1625
},
{
"epoch": 0.9813365442504516,
"grad_norm": 1.2712485645750238,
"learning_rate": 1.0619599843249006e-08,
"loss": 0.5147,
"step": 1630
},
{
"epoch": 0.9843467790487658,
"grad_norm": 1.2710371635621938,
"learning_rate": 7.470971368142011e-09,
"loss": 0.4974,
"step": 1635
},
{
"epoch": 0.9873570138470801,
"grad_norm": 1.3076474088761778,
"learning_rate": 4.874236237911723e-09,
"loss": 0.504,
"step": 1640
},
{
"epoch": 0.9903672486453944,
"grad_norm": 1.2813980219783268,
"learning_rate": 2.8296815056377824e-09,
"loss": 0.5086,
"step": 1645
},
{
"epoch": 0.9933774834437086,
"grad_norm": 1.2247806264161647,
"learning_rate": 1.3375331842574446e-09,
"loss": 0.4984,
"step": 1650
},
{
"epoch": 0.9963877182420229,
"grad_norm": 1.2500205386368362,
"learning_rate": 3.9795622158111945e-10,
"loss": 0.5102,
"step": 1655
},
{
"epoch": 0.9993979530403372,
"grad_norm": 1.3276901052559071,
"learning_rate": 1.1054482056405136e-11,
"loss": 0.5088,
"step": 1660
},
{
"epoch": 1.0,
"eval_runtime": 3.3023,
"eval_samples_per_second": 3.028,
"eval_steps_per_second": 0.908,
"step": 1661
},
{
"epoch": 1.0,
"step": 1661,
"total_flos": 430711668473856.0,
"train_loss": 0.6697713777575416,
"train_runtime": 21206.7621,
"train_samples_per_second": 1.253,
"train_steps_per_second": 0.078
}
],
"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": 430711668473856.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}