qwen_feedback / trainer_state.json
terry69's picture
Model save
c7c0b30 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": 40.966713680527505,
"learning_rate": 5.98802395209581e-08,
"loss": 2.0373,
"step": 1
},
{
"epoch": 0.0030102347983142685,
"grad_norm": 39.14616421730173,
"learning_rate": 2.9940119760479047e-07,
"loss": 2.0379,
"step": 5
},
{
"epoch": 0.006020469596628537,
"grad_norm": 27.910266092453746,
"learning_rate": 5.988023952095809e-07,
"loss": 1.9953,
"step": 10
},
{
"epoch": 0.009030704394942806,
"grad_norm": 15.201013735116844,
"learning_rate": 8.982035928143713e-07,
"loss": 1.7804,
"step": 15
},
{
"epoch": 0.012040939193257074,
"grad_norm": 5.295520564674036,
"learning_rate": 1.1976047904191619e-06,
"loss": 1.502,
"step": 20
},
{
"epoch": 0.015051173991571343,
"grad_norm": 4.314026946089969,
"learning_rate": 1.4970059880239521e-06,
"loss": 1.3422,
"step": 25
},
{
"epoch": 0.018061408789885613,
"grad_norm": 2.8393277241114694,
"learning_rate": 1.7964071856287426e-06,
"loss": 1.2716,
"step": 30
},
{
"epoch": 0.02107164358819988,
"grad_norm": 1.9881246281959908,
"learning_rate": 2.095808383233533e-06,
"loss": 1.142,
"step": 35
},
{
"epoch": 0.024081878386514148,
"grad_norm": 1.4576393482889902,
"learning_rate": 2.3952095808383237e-06,
"loss": 1.091,
"step": 40
},
{
"epoch": 0.027092113184828417,
"grad_norm": 1.3452736306984228,
"learning_rate": 2.694610778443114e-06,
"loss": 1.0808,
"step": 45
},
{
"epoch": 0.030102347983142687,
"grad_norm": 1.2853833849515133,
"learning_rate": 2.9940119760479042e-06,
"loss": 1.0534,
"step": 50
},
{
"epoch": 0.033112582781456956,
"grad_norm": 1.3461267648995063,
"learning_rate": 3.2934131736526947e-06,
"loss": 1.0335,
"step": 55
},
{
"epoch": 0.036122817579771226,
"grad_norm": 1.2720929793887874,
"learning_rate": 3.592814371257485e-06,
"loss": 1.0202,
"step": 60
},
{
"epoch": 0.03913305237808549,
"grad_norm": 1.2405264880245372,
"learning_rate": 3.892215568862276e-06,
"loss": 1.0059,
"step": 65
},
{
"epoch": 0.04214328717639976,
"grad_norm": 1.1810599448574401,
"learning_rate": 4.191616766467066e-06,
"loss": 1.0041,
"step": 70
},
{
"epoch": 0.04515352197471403,
"grad_norm": 1.2433212471597048,
"learning_rate": 4.4910179640718566e-06,
"loss": 0.9915,
"step": 75
},
{
"epoch": 0.048163756773028296,
"grad_norm": 1.2370160588056265,
"learning_rate": 4.7904191616766475e-06,
"loss": 0.9683,
"step": 80
},
{
"epoch": 0.051173991571342566,
"grad_norm": 1.2799056245464302,
"learning_rate": 5.0898203592814375e-06,
"loss": 0.9757,
"step": 85
},
{
"epoch": 0.054184226369656835,
"grad_norm": 1.2232343401620152,
"learning_rate": 5.389221556886228e-06,
"loss": 0.9664,
"step": 90
},
{
"epoch": 0.057194461167971104,
"grad_norm": 1.2749429574013502,
"learning_rate": 5.6886227544910184e-06,
"loss": 0.9531,
"step": 95
},
{
"epoch": 0.060204695966285374,
"grad_norm": 1.341905992560118,
"learning_rate": 5.9880239520958085e-06,
"loss": 0.936,
"step": 100
},
{
"epoch": 0.06321493076459964,
"grad_norm": 1.3312129649486055,
"learning_rate": 6.2874251497005985e-06,
"loss": 0.9288,
"step": 105
},
{
"epoch": 0.06622516556291391,
"grad_norm": 1.3057249774040953,
"learning_rate": 6.586826347305389e-06,
"loss": 0.9137,
"step": 110
},
{
"epoch": 0.06923540036122817,
"grad_norm": 1.3480198308811908,
"learning_rate": 6.88622754491018e-06,
"loss": 0.9087,
"step": 115
},
{
"epoch": 0.07224563515954245,
"grad_norm": 1.2851052893092942,
"learning_rate": 7.18562874251497e-06,
"loss": 0.9006,
"step": 120
},
{
"epoch": 0.07525586995785671,
"grad_norm": 1.3337166654009118,
"learning_rate": 7.485029940119761e-06,
"loss": 0.8995,
"step": 125
},
{
"epoch": 0.07826610475617098,
"grad_norm": 1.259701546947682,
"learning_rate": 7.784431137724551e-06,
"loss": 0.877,
"step": 130
},
{
"epoch": 0.08127633955448525,
"grad_norm": 1.3284638098388712,
"learning_rate": 8.083832335329342e-06,
"loss": 0.8847,
"step": 135
},
{
"epoch": 0.08428657435279951,
"grad_norm": 1.3016966435741442,
"learning_rate": 8.383233532934131e-06,
"loss": 0.8562,
"step": 140
},
{
"epoch": 0.08729680915111379,
"grad_norm": 1.272771944986212,
"learning_rate": 8.682634730538922e-06,
"loss": 0.8537,
"step": 145
},
{
"epoch": 0.09030704394942805,
"grad_norm": 1.2404717495109163,
"learning_rate": 8.982035928143713e-06,
"loss": 0.8402,
"step": 150
},
{
"epoch": 0.09331727874774233,
"grad_norm": 1.3171631151874839,
"learning_rate": 9.281437125748504e-06,
"loss": 0.8596,
"step": 155
},
{
"epoch": 0.09632751354605659,
"grad_norm": 1.2739485119594887,
"learning_rate": 9.580838323353295e-06,
"loss": 0.8548,
"step": 160
},
{
"epoch": 0.09933774834437085,
"grad_norm": 1.2415391381244691,
"learning_rate": 9.880239520958084e-06,
"loss": 0.8348,
"step": 165
},
{
"epoch": 0.10234798314268513,
"grad_norm": 1.2732863554156713,
"learning_rate": 9.999900509954779e-06,
"loss": 0.8326,
"step": 170
},
{
"epoch": 0.1053582179409994,
"grad_norm": 1.3293213562766069,
"learning_rate": 9.999292529572152e-06,
"loss": 0.839,
"step": 175
},
{
"epoch": 0.10836845273931367,
"grad_norm": 1.3011820572360617,
"learning_rate": 9.998131908181262e-06,
"loss": 0.8377,
"step": 180
},
{
"epoch": 0.11137868753762793,
"grad_norm": 1.2833557284441266,
"learning_rate": 9.996418774081658e-06,
"loss": 0.8228,
"step": 185
},
{
"epoch": 0.11438892233594221,
"grad_norm": 1.286488000985355,
"learning_rate": 9.994153316649769e-06,
"loss": 0.8318,
"step": 190
},
{
"epoch": 0.11739915713425647,
"grad_norm": 1.2384872269253941,
"learning_rate": 9.991335786317964e-06,
"loss": 0.8209,
"step": 195
},
{
"epoch": 0.12040939193257075,
"grad_norm": 1.2713798885562295,
"learning_rate": 9.987966494546873e-06,
"loss": 0.8069,
"step": 200
},
{
"epoch": 0.12341962673088501,
"grad_norm": 1.2482060631896188,
"learning_rate": 9.984045813790959e-06,
"loss": 0.8142,
"step": 205
},
{
"epoch": 0.12642986152919927,
"grad_norm": 1.3415986154597055,
"learning_rate": 9.979574177457337e-06,
"loss": 0.813,
"step": 210
},
{
"epoch": 0.12944009632751355,
"grad_norm": 1.2350908672442804,
"learning_rate": 9.974552079857873e-06,
"loss": 0.8109,
"step": 215
},
{
"epoch": 0.13245033112582782,
"grad_norm": 1.3964490955664748,
"learning_rate": 9.968980076154533e-06,
"loss": 0.8293,
"step": 220
},
{
"epoch": 0.13546056592414207,
"grad_norm": 1.2591771311022242,
"learning_rate": 9.962858782298023e-06,
"loss": 0.802,
"step": 225
},
{
"epoch": 0.13847080072245635,
"grad_norm": 1.288853392082994,
"learning_rate": 9.956188874959686e-06,
"loss": 0.8151,
"step": 230
},
{
"epoch": 0.14148103552077063,
"grad_norm": 1.197725597237524,
"learning_rate": 9.948971091456715e-06,
"loss": 0.8071,
"step": 235
},
{
"epoch": 0.1444912703190849,
"grad_norm": 1.2226857692711433,
"learning_rate": 9.941206229670634e-06,
"loss": 0.8053,
"step": 240
},
{
"epoch": 0.14750150511739915,
"grad_norm": 1.3119146876238303,
"learning_rate": 9.932895147959106e-06,
"loss": 0.7955,
"step": 245
},
{
"epoch": 0.15051173991571343,
"grad_norm": 1.3326937913135584,
"learning_rate": 9.924038765061042e-06,
"loss": 0.7833,
"step": 250
},
{
"epoch": 0.1535219747140277,
"grad_norm": 1.2647168837729537,
"learning_rate": 9.91463805999504e-06,
"loss": 0.7933,
"step": 255
},
{
"epoch": 0.15653220951234195,
"grad_norm": 1.2033595276379918,
"learning_rate": 9.904694071951167e-06,
"loss": 0.8004,
"step": 260
},
{
"epoch": 0.15954244431065623,
"grad_norm": 1.1708378372158956,
"learning_rate": 9.894207900176074e-06,
"loss": 0.7877,
"step": 265
},
{
"epoch": 0.1625526791089705,
"grad_norm": 1.277798553181957,
"learning_rate": 9.883180703851488e-06,
"loss": 0.7899,
"step": 270
},
{
"epoch": 0.16556291390728478,
"grad_norm": 1.207840548329369,
"learning_rate": 9.871613701966067e-06,
"loss": 0.7885,
"step": 275
},
{
"epoch": 0.16857314870559903,
"grad_norm": 1.156274795904894,
"learning_rate": 9.859508173180653e-06,
"loss": 0.7664,
"step": 280
},
{
"epoch": 0.1715833835039133,
"grad_norm": 1.164171992678941,
"learning_rate": 9.846865455686915e-06,
"loss": 0.7777,
"step": 285
},
{
"epoch": 0.17459361830222758,
"grad_norm": 1.1762216211439502,
"learning_rate": 9.833686947059436e-06,
"loss": 0.7727,
"step": 290
},
{
"epoch": 0.17760385310054183,
"grad_norm": 1.1617067559555396,
"learning_rate": 9.819974104101198e-06,
"loss": 0.7724,
"step": 295
},
{
"epoch": 0.1806140878988561,
"grad_norm": 1.1855967777933063,
"learning_rate": 9.80572844268256e-06,
"loss": 0.7645,
"step": 300
},
{
"epoch": 0.18362432269717038,
"grad_norm": 1.1944671571121344,
"learning_rate": 9.790951537573686e-06,
"loss": 0.7819,
"step": 305
},
{
"epoch": 0.18663455749548466,
"grad_norm": 1.1965936475386203,
"learning_rate": 9.775645022270448e-06,
"loss": 0.778,
"step": 310
},
{
"epoch": 0.1896447922937989,
"grad_norm": 1.256280786074695,
"learning_rate": 9.759810588813872e-06,
"loss": 0.774,
"step": 315
},
{
"epoch": 0.19265502709211318,
"grad_norm": 1.3027821404071966,
"learning_rate": 9.743449987603082e-06,
"loss": 0.7682,
"step": 320
},
{
"epoch": 0.19566526189042746,
"grad_norm": 1.2406739802662574,
"learning_rate": 9.726565027201813e-06,
"loss": 0.788,
"step": 325
},
{
"epoch": 0.1986754966887417,
"grad_norm": 1.1751549969003265,
"learning_rate": 9.70915757413847e-06,
"loss": 0.7815,
"step": 330
},
{
"epoch": 0.20168573148705599,
"grad_norm": 1.1964609554923433,
"learning_rate": 9.691229552699817e-06,
"loss": 0.7615,
"step": 335
},
{
"epoch": 0.20469596628537026,
"grad_norm": 1.2774212868740151,
"learning_rate": 9.672782944718234e-06,
"loss": 0.7677,
"step": 340
},
{
"epoch": 0.20770620108368454,
"grad_norm": 1.202397374963106,
"learning_rate": 9.65381978935266e-06,
"loss": 0.7671,
"step": 345
},
{
"epoch": 0.2107164358819988,
"grad_norm": 1.2159818292840359,
"learning_rate": 9.634342182863163e-06,
"loss": 0.7476,
"step": 350
},
{
"epoch": 0.21372667068031306,
"grad_norm": 1.194513717936444,
"learning_rate": 9.614352278379217e-06,
"loss": 0.7788,
"step": 355
},
{
"epoch": 0.21673690547862734,
"grad_norm": 1.2060005508903568,
"learning_rate": 9.593852285661684e-06,
"loss": 0.7709,
"step": 360
},
{
"epoch": 0.2197471402769416,
"grad_norm": 1.203716084337312,
"learning_rate": 9.572844470858537e-06,
"loss": 0.7584,
"step": 365
},
{
"epoch": 0.22275737507525586,
"grad_norm": 1.282647241669985,
"learning_rate": 9.551331156254358e-06,
"loss": 0.7696,
"step": 370
},
{
"epoch": 0.22576760987357014,
"grad_norm": 1.2009722082380498,
"learning_rate": 9.529314720013618e-06,
"loss": 0.7526,
"step": 375
},
{
"epoch": 0.22877784467188442,
"grad_norm": 1.2201198977296086,
"learning_rate": 9.506797595917787e-06,
"loss": 0.7511,
"step": 380
},
{
"epoch": 0.23178807947019867,
"grad_norm": 1.3415385517504728,
"learning_rate": 9.483782273096295e-06,
"loss": 0.7513,
"step": 385
},
{
"epoch": 0.23479831426851294,
"grad_norm": 1.26924665240572,
"learning_rate": 9.460271295751373e-06,
"loss": 0.7649,
"step": 390
},
{
"epoch": 0.23780854906682722,
"grad_norm": 1.2785361560349413,
"learning_rate": 9.436267262876808e-06,
"loss": 0.751,
"step": 395
},
{
"epoch": 0.2408187838651415,
"grad_norm": 1.1852007444354873,
"learning_rate": 9.411772827970642e-06,
"loss": 0.7548,
"step": 400
},
{
"epoch": 0.24382901866345574,
"grad_norm": 1.1910921737694808,
"learning_rate": 9.38679069874184e-06,
"loss": 0.7507,
"step": 405
},
{
"epoch": 0.24683925346177002,
"grad_norm": 1.3053941058405514,
"learning_rate": 9.36132363681097e-06,
"loss": 0.7557,
"step": 410
},
{
"epoch": 0.2498494882600843,
"grad_norm": 1.2465327608331673,
"learning_rate": 9.335374457404928e-06,
"loss": 0.7649,
"step": 415
},
{
"epoch": 0.25285972305839854,
"grad_norm": 1.2813369476453156,
"learning_rate": 9.308946029045726e-06,
"loss": 0.7325,
"step": 420
},
{
"epoch": 0.25586995785671285,
"grad_norm": 1.2539019797830948,
"learning_rate": 9.282041273233402e-06,
"loss": 0.726,
"step": 425
},
{
"epoch": 0.2588801926550271,
"grad_norm": 1.2685847502715055,
"learning_rate": 9.254663164123052e-06,
"loss": 0.7372,
"step": 430
},
{
"epoch": 0.26189042745334135,
"grad_norm": 1.2404250071973326,
"learning_rate": 9.226814728196072e-06,
"loss": 0.7333,
"step": 435
},
{
"epoch": 0.26490066225165565,
"grad_norm": 1.206906014688777,
"learning_rate": 9.198499043925591e-06,
"loss": 0.7305,
"step": 440
},
{
"epoch": 0.2679108970499699,
"grad_norm": 1.2568238232780307,
"learning_rate": 9.169719241436162e-06,
"loss": 0.7437,
"step": 445
},
{
"epoch": 0.27092113184828415,
"grad_norm": 1.201047028691787,
"learning_rate": 9.14047850215775e-06,
"loss": 0.7229,
"step": 450
},
{
"epoch": 0.27393136664659845,
"grad_norm": 1.1628890533112184,
"learning_rate": 9.110780058474052e-06,
"loss": 0.7252,
"step": 455
},
{
"epoch": 0.2769416014449127,
"grad_norm": 1.222552632587458,
"learning_rate": 9.080627193365155e-06,
"loss": 0.7348,
"step": 460
},
{
"epoch": 0.27995183624322695,
"grad_norm": 1.204712682838897,
"learning_rate": 9.050023240044649e-06,
"loss": 0.7394,
"step": 465
},
{
"epoch": 0.28296207104154125,
"grad_norm": 1.1699974198146679,
"learning_rate": 9.018971581591141e-06,
"loss": 0.7172,
"step": 470
},
{
"epoch": 0.2859723058398555,
"grad_norm": 1.2143354714842383,
"learning_rate": 8.987475650574289e-06,
"loss": 0.7459,
"step": 475
},
{
"epoch": 0.2889825406381698,
"grad_norm": 1.2344181327548198,
"learning_rate": 8.955538928675343e-06,
"loss": 0.7217,
"step": 480
},
{
"epoch": 0.29199277543648405,
"grad_norm": 1.2200821617384021,
"learning_rate": 8.923164946302274e-06,
"loss": 0.7355,
"step": 485
},
{
"epoch": 0.2950030102347983,
"grad_norm": 1.2730755552957604,
"learning_rate": 8.890357282199504e-06,
"loss": 0.7407,
"step": 490
},
{
"epoch": 0.2980132450331126,
"grad_norm": 1.326965523163219,
"learning_rate": 8.857119563052301e-06,
"loss": 0.7112,
"step": 495
},
{
"epoch": 0.30102347983142685,
"grad_norm": 1.2881202459959005,
"learning_rate": 8.823455463085873e-06,
"loss": 0.7299,
"step": 500
},
{
"epoch": 0.3040337146297411,
"grad_norm": 1.3023027317084608,
"learning_rate": 8.789368703659199e-06,
"loss": 0.7291,
"step": 505
},
{
"epoch": 0.3070439494280554,
"grad_norm": 1.192827763147033,
"learning_rate": 8.754863052853658e-06,
"loss": 0.7242,
"step": 510
},
{
"epoch": 0.31005418422636966,
"grad_norm": 1.172549917556519,
"learning_rate": 8.719942325056496e-06,
"loss": 0.707,
"step": 515
},
{
"epoch": 0.3130644190246839,
"grad_norm": 1.1750104521638358,
"learning_rate": 8.68461038053916e-06,
"loss": 0.7152,
"step": 520
},
{
"epoch": 0.3160746538229982,
"grad_norm": 1.3380189595946483,
"learning_rate": 8.648871125030576e-06,
"loss": 0.711,
"step": 525
},
{
"epoch": 0.31908488862131246,
"grad_norm": 1.226348488844478,
"learning_rate": 8.612728509285395e-06,
"loss": 0.7104,
"step": 530
},
{
"epoch": 0.3220951234196267,
"grad_norm": 1.245273272201369,
"learning_rate": 8.576186528647253e-06,
"loss": 0.7286,
"step": 535
},
{
"epoch": 0.325105358217941,
"grad_norm": 1.1790288496340813,
"learning_rate": 8.53924922260712e-06,
"loss": 0.7075,
"step": 540
},
{
"epoch": 0.32811559301625526,
"grad_norm": 1.1914412291760281,
"learning_rate": 8.501920674356755e-06,
"loss": 0.7008,
"step": 545
},
{
"epoch": 0.33112582781456956,
"grad_norm": 1.1806518773814458,
"learning_rate": 8.46420501033733e-06,
"loss": 0.6934,
"step": 550
},
{
"epoch": 0.3341360626128838,
"grad_norm": 1.2723595572649768,
"learning_rate": 8.42610639978329e-06,
"loss": 0.7019,
"step": 555
},
{
"epoch": 0.33714629741119806,
"grad_norm": 1.2225157243802531,
"learning_rate": 8.387629054261454e-06,
"loss": 0.7002,
"step": 560
},
{
"epoch": 0.34015653220951236,
"grad_norm": 1.2003267518231666,
"learning_rate": 8.348777227205462e-06,
"loss": 0.6984,
"step": 565
},
{
"epoch": 0.3431667670078266,
"grad_norm": 1.1755635510106648,
"learning_rate": 8.309555213445583e-06,
"loss": 0.7123,
"step": 570
},
{
"epoch": 0.34617700180614086,
"grad_norm": 1.2065493450135722,
"learning_rate": 8.269967348733947e-06,
"loss": 0.6855,
"step": 575
},
{
"epoch": 0.34918723660445516,
"grad_norm": 1.2455393618923039,
"learning_rate": 8.230018009265255e-06,
"loss": 0.6971,
"step": 580
},
{
"epoch": 0.3521974714027694,
"grad_norm": 1.3331603991829393,
"learning_rate": 8.189711611193012e-06,
"loss": 0.6985,
"step": 585
},
{
"epoch": 0.35520770620108366,
"grad_norm": 1.221254992942756,
"learning_rate": 8.149052610141357e-06,
"loss": 0.6843,
"step": 590
},
{
"epoch": 0.35821794099939797,
"grad_norm": 1.2577154855191386,
"learning_rate": 8.108045500712518e-06,
"loss": 0.7015,
"step": 595
},
{
"epoch": 0.3612281757977122,
"grad_norm": 1.2678823261096912,
"learning_rate": 8.066694815989961e-06,
"loss": 0.7054,
"step": 600
},
{
"epoch": 0.36423841059602646,
"grad_norm": 1.3068755574009954,
"learning_rate": 8.025005127037282e-06,
"loss": 0.6909,
"step": 605
},
{
"epoch": 0.36724864539434077,
"grad_norm": 1.25345913435044,
"learning_rate": 7.982981042392907e-06,
"loss": 0.6804,
"step": 610
},
{
"epoch": 0.370258880192655,
"grad_norm": 1.2171026231986553,
"learning_rate": 7.940627207560655e-06,
"loss": 0.6792,
"step": 615
},
{
"epoch": 0.3732691149909693,
"grad_norm": 1.1878268346505996,
"learning_rate": 7.897948304496189e-06,
"loss": 0.7002,
"step": 620
},
{
"epoch": 0.37627934978928357,
"grad_norm": 1.2021282337780814,
"learning_rate": 7.854949051089467e-06,
"loss": 0.7026,
"step": 625
},
{
"epoch": 0.3792895845875978,
"grad_norm": 1.219850402255314,
"learning_rate": 7.811634200643202e-06,
"loss": 0.7081,
"step": 630
},
{
"epoch": 0.3822998193859121,
"grad_norm": 1.25066752923423,
"learning_rate": 7.768008541347423e-06,
"loss": 0.6709,
"step": 635
},
{
"epoch": 0.38531005418422637,
"grad_norm": 1.233821739163781,
"learning_rate": 7.72407689575016e-06,
"loss": 0.6821,
"step": 640
},
{
"epoch": 0.3883202889825406,
"grad_norm": 1.2400488427811083,
"learning_rate": 7.67984412022434e-06,
"loss": 0.6797,
"step": 645
},
{
"epoch": 0.3913305237808549,
"grad_norm": 1.2744104516354757,
"learning_rate": 7.635315104430959e-06,
"loss": 0.6785,
"step": 650
},
{
"epoch": 0.39434075857916917,
"grad_norm": 1.264910637088745,
"learning_rate": 7.5904947707785434e-06,
"loss": 0.649,
"step": 655
},
{
"epoch": 0.3973509933774834,
"grad_norm": 1.2672765451192602,
"learning_rate": 7.545388073879018e-06,
"loss": 0.6906,
"step": 660
},
{
"epoch": 0.4003612281757977,
"grad_norm": 1.2685599345911318,
"learning_rate": 7.500000000000001e-06,
"loss": 0.6671,
"step": 665
},
{
"epoch": 0.40337146297411197,
"grad_norm": 1.2358020853032294,
"learning_rate": 7.454335566513603e-06,
"loss": 0.6775,
"step": 670
},
{
"epoch": 0.4063816977724263,
"grad_norm": 1.316177091586594,
"learning_rate": 7.408399821341787e-06,
"loss": 0.679,
"step": 675
},
{
"epoch": 0.4093919325707405,
"grad_norm": 1.1839753807419882,
"learning_rate": 7.362197842398355e-06,
"loss": 0.6712,
"step": 680
},
{
"epoch": 0.4124021673690548,
"grad_norm": 1.2764557732073194,
"learning_rate": 7.315734737027612e-06,
"loss": 0.6719,
"step": 685
},
{
"epoch": 0.4154124021673691,
"grad_norm": 1.2259459384980307,
"learning_rate": 7.2690156414397775e-06,
"loss": 0.657,
"step": 690
},
{
"epoch": 0.4184226369656833,
"grad_norm": 1.2344897988837857,
"learning_rate": 7.22204572014322e-06,
"loss": 0.6553,
"step": 695
},
{
"epoch": 0.4214328717639976,
"grad_norm": 1.2553454967603142,
"learning_rate": 7.174830165373542e-06,
"loss": 0.6602,
"step": 700
},
{
"epoch": 0.4244431065623119,
"grad_norm": 1.2370627727074957,
"learning_rate": 7.127374196519616e-06,
"loss": 0.6566,
"step": 705
},
{
"epoch": 0.4274533413606261,
"grad_norm": 1.23827568170562,
"learning_rate": 7.079683059546607e-06,
"loss": 0.6566,
"step": 710
},
{
"epoch": 0.4304635761589404,
"grad_norm": 1.2717726441044983,
"learning_rate": 7.031762026416074e-06,
"loss": 0.6523,
"step": 715
},
{
"epoch": 0.4334738109572547,
"grad_norm": 1.2414138030074853,
"learning_rate": 6.983616394503177e-06,
"loss": 0.668,
"step": 720
},
{
"epoch": 0.43648404575556893,
"grad_norm": 1.2863979258950011,
"learning_rate": 6.9352514860110876e-06,
"loss": 0.6723,
"step": 725
},
{
"epoch": 0.4394942805538832,
"grad_norm": 1.1998975466871244,
"learning_rate": 6.886672647382653e-06,
"loss": 0.6496,
"step": 730
},
{
"epoch": 0.4425045153521975,
"grad_norm": 1.2253729409328844,
"learning_rate": 6.837885248709386e-06,
"loss": 0.6837,
"step": 735
},
{
"epoch": 0.44551475015051173,
"grad_norm": 1.3977058259662651,
"learning_rate": 6.788894683137822e-06,
"loss": 0.6535,
"step": 740
},
{
"epoch": 0.44852498494882603,
"grad_norm": 1.2252881439154193,
"learning_rate": 6.739706366273346e-06,
"loss": 0.6668,
"step": 745
},
{
"epoch": 0.4515352197471403,
"grad_norm": 1.2685171799824646,
"learning_rate": 6.690325735581532e-06,
"loss": 0.6408,
"step": 750
},
{
"epoch": 0.45454545454545453,
"grad_norm": 1.2559367914781316,
"learning_rate": 6.640758249787067e-06,
"loss": 0.6551,
"step": 755
},
{
"epoch": 0.45755568934376883,
"grad_norm": 1.3414553027680045,
"learning_rate": 6.591009388270315e-06,
"loss": 0.6483,
"step": 760
},
{
"epoch": 0.4605659241420831,
"grad_norm": 1.2833745957368083,
"learning_rate": 6.54108465046161e-06,
"loss": 0.6591,
"step": 765
},
{
"epoch": 0.46357615894039733,
"grad_norm": 1.2801896169833944,
"learning_rate": 6.490989555233328e-06,
"loss": 0.6481,
"step": 770
},
{
"epoch": 0.46658639373871164,
"grad_norm": 1.2982217970316745,
"learning_rate": 6.440729640289809e-06,
"loss": 0.6445,
"step": 775
},
{
"epoch": 0.4695966285370259,
"grad_norm": 1.2399362763796928,
"learning_rate": 6.3903104615551956e-06,
"loss": 0.6428,
"step": 780
},
{
"epoch": 0.47260686333534013,
"grad_norm": 1.252116968840507,
"learning_rate": 6.3397375925592675e-06,
"loss": 0.621,
"step": 785
},
{
"epoch": 0.47561709813365444,
"grad_norm": 1.1634766201317972,
"learning_rate": 6.289016623821308e-06,
"loss": 0.6396,
"step": 790
},
{
"epoch": 0.4786273329319687,
"grad_norm": 1.2089828599308985,
"learning_rate": 6.2381531622321234e-06,
"loss": 0.6429,
"step": 795
},
{
"epoch": 0.481637567730283,
"grad_norm": 1.2611778575722072,
"learning_rate": 6.18715283043422e-06,
"loss": 0.6282,
"step": 800
},
{
"epoch": 0.48464780252859724,
"grad_norm": 1.2634822793060227,
"learning_rate": 6.136021266200271e-06,
"loss": 0.6472,
"step": 805
},
{
"epoch": 0.4876580373269115,
"grad_norm": 1.2731517375276795,
"learning_rate": 6.084764121809878e-06,
"loss": 0.6187,
"step": 810
},
{
"epoch": 0.4906682721252258,
"grad_norm": 1.2681414747392186,
"learning_rate": 6.033387063424765e-06,
"loss": 0.6318,
"step": 815
},
{
"epoch": 0.49367850692354004,
"grad_norm": 1.2152691698415932,
"learning_rate": 5.9818957704624046e-06,
"loss": 0.6442,
"step": 820
},
{
"epoch": 0.4966887417218543,
"grad_norm": 1.2163481834250993,
"learning_rate": 5.930295934968197e-06,
"loss": 0.633,
"step": 825
},
{
"epoch": 0.4996989765201686,
"grad_norm": 1.2644071741742902,
"learning_rate": 5.878593260986256e-06,
"loss": 0.6333,
"step": 830
},
{
"epoch": 0.5027092113184829,
"grad_norm": 1.2998551781398153,
"learning_rate": 5.8267934639288525e-06,
"loss": 0.6377,
"step": 835
},
{
"epoch": 0.5057194461167971,
"grad_norm": 1.2747330171688573,
"learning_rate": 5.77490226994462e-06,
"loss": 0.6385,
"step": 840
},
{
"epoch": 0.5087296809151114,
"grad_norm": 1.2147166700112426,
"learning_rate": 5.722925415285555e-06,
"loss": 0.6335,
"step": 845
},
{
"epoch": 0.5117399157134257,
"grad_norm": 1.2814771734916943,
"learning_rate": 5.670868645672916e-06,
"loss": 0.6316,
"step": 850
},
{
"epoch": 0.5147501505117399,
"grad_norm": 1.2118662457570313,
"learning_rate": 5.618737715662067e-06,
"loss": 0.6219,
"step": 855
},
{
"epoch": 0.5177603853100542,
"grad_norm": 1.2322127749127771,
"learning_rate": 5.566538388006351e-06,
"loss": 0.623,
"step": 860
},
{
"epoch": 0.5207706201083685,
"grad_norm": 1.1991462675329747,
"learning_rate": 5.514276433020044e-06,
"loss": 0.6107,
"step": 865
},
{
"epoch": 0.5237808549066827,
"grad_norm": 1.2149020805519655,
"learning_rate": 5.461957627940489e-06,
"loss": 0.6191,
"step": 870
},
{
"epoch": 0.526791089704997,
"grad_norm": 1.1877405672596075,
"learning_rate": 5.409587756289462e-06,
"loss": 0.6112,
"step": 875
},
{
"epoch": 0.5298013245033113,
"grad_norm": 1.1970060104579194,
"learning_rate": 5.357172607233831e-06,
"loss": 0.6284,
"step": 880
},
{
"epoch": 0.5328115593016255,
"grad_norm": 1.368716689528383,
"learning_rate": 5.304717974945596e-06,
"loss": 0.6163,
"step": 885
},
{
"epoch": 0.5358217940999398,
"grad_norm": 1.273293450405195,
"learning_rate": 5.252229657961394e-06,
"loss": 0.6214,
"step": 890
},
{
"epoch": 0.5388320288982541,
"grad_norm": 1.196878746141804,
"learning_rate": 5.199713458541495e-06,
"loss": 0.6247,
"step": 895
},
{
"epoch": 0.5418422636965683,
"grad_norm": 1.2084713164747305,
"learning_rate": 5.1471751820284e-06,
"loss": 0.6103,
"step": 900
},
{
"epoch": 0.5448524984948826,
"grad_norm": 1.2196639084600536,
"learning_rate": 5.094620636205096e-06,
"loss": 0.6221,
"step": 905
},
{
"epoch": 0.5478627332931969,
"grad_norm": 1.1988440703906922,
"learning_rate": 5.042055630653042e-06,
"loss": 0.6085,
"step": 910
},
{
"epoch": 0.5508729680915111,
"grad_norm": 1.26407878369998,
"learning_rate": 4.98948597610996e-06,
"loss": 0.6114,
"step": 915
},
{
"epoch": 0.5538832028898254,
"grad_norm": 1.2541252408565282,
"learning_rate": 4.936917483827483e-06,
"loss": 0.6098,
"step": 920
},
{
"epoch": 0.5568934376881397,
"grad_norm": 1.2461032064160447,
"learning_rate": 4.884355964928767e-06,
"loss": 0.6288,
"step": 925
},
{
"epoch": 0.5599036724864539,
"grad_norm": 1.1876409214621886,
"learning_rate": 4.831807229766101e-06,
"loss": 0.5967,
"step": 930
},
{
"epoch": 0.5629139072847682,
"grad_norm": 1.3653959531629782,
"learning_rate": 4.779277087278615e-06,
"loss": 0.5981,
"step": 935
},
{
"epoch": 0.5659241420830825,
"grad_norm": 1.259758029782513,
"learning_rate": 4.7267713443501274e-06,
"loss": 0.6208,
"step": 940
},
{
"epoch": 0.5689343768813967,
"grad_norm": 1.2512441621206507,
"learning_rate": 4.67429580516724e-06,
"loss": 0.6036,
"step": 945
},
{
"epoch": 0.571944611679711,
"grad_norm": 1.2148856832751866,
"learning_rate": 4.6218562705777185e-06,
"loss": 0.5918,
"step": 950
},
{
"epoch": 0.5749548464780253,
"grad_norm": 1.2618875581171025,
"learning_rate": 4.5694585374492314e-06,
"loss": 0.5956,
"step": 955
},
{
"epoch": 0.5779650812763396,
"grad_norm": 1.2405407377914555,
"learning_rate": 4.517108398028566e-06,
"loss": 0.6005,
"step": 960
},
{
"epoch": 0.5809753160746538,
"grad_norm": 1.2775566642146432,
"learning_rate": 4.464811639301314e-06,
"loss": 0.589,
"step": 965
},
{
"epoch": 0.5839855508729681,
"grad_norm": 1.3117899815699137,
"learning_rate": 4.412574042352156e-06,
"loss": 0.5979,
"step": 970
},
{
"epoch": 0.5869957856712824,
"grad_norm": 1.269362160884011,
"learning_rate": 4.360401381725806e-06,
"loss": 0.5887,
"step": 975
},
{
"epoch": 0.5900060204695966,
"grad_norm": 1.248523016364606,
"learning_rate": 4.308299424788667e-06,
"loss": 0.6022,
"step": 980
},
{
"epoch": 0.5930162552679109,
"grad_norm": 1.2366195530748443,
"learning_rate": 4.256273931091284e-06,
"loss": 0.6,
"step": 985
},
{
"epoch": 0.5960264900662252,
"grad_norm": 1.222474401781831,
"learning_rate": 4.204330651731662e-06,
"loss": 0.5871,
"step": 990
},
{
"epoch": 0.5990367248645394,
"grad_norm": 1.2424032915235261,
"learning_rate": 4.152475328719517e-06,
"loss": 0.5783,
"step": 995
},
{
"epoch": 0.6020469596628537,
"grad_norm": 1.2570025380680472,
"learning_rate": 4.1007136943415325e-06,
"loss": 0.5636,
"step": 1000
},
{
"epoch": 0.605057194461168,
"grad_norm": 1.2170323945106316,
"learning_rate": 4.049051470527692e-06,
"loss": 0.601,
"step": 1005
},
{
"epoch": 0.6080674292594822,
"grad_norm": 1.1958865931929052,
"learning_rate": 3.997494368218745e-06,
"loss": 0.5878,
"step": 1010
},
{
"epoch": 0.6110776640577965,
"grad_norm": 1.3367291697694699,
"learning_rate": 3.946048086734921e-06,
"loss": 0.5924,
"step": 1015
},
{
"epoch": 0.6140878988561108,
"grad_norm": 1.2820724684669875,
"learning_rate": 3.894718313145873e-06,
"loss": 0.5847,
"step": 1020
},
{
"epoch": 0.617098133654425,
"grad_norm": 1.2719469697058394,
"learning_rate": 3.843510721642036e-06,
"loss": 0.5744,
"step": 1025
},
{
"epoch": 0.6201083684527393,
"grad_norm": 1.2274652244842972,
"learning_rate": 3.7924309729073616e-06,
"loss": 0.5751,
"step": 1030
},
{
"epoch": 0.6231186032510536,
"grad_norm": 1.3030807337065518,
"learning_rate": 3.7414847134935716e-06,
"loss": 0.5899,
"step": 1035
},
{
"epoch": 0.6261288380493678,
"grad_norm": 1.1948808509249107,
"learning_rate": 3.6906775751959667e-06,
"loss": 0.5755,
"step": 1040
},
{
"epoch": 0.6291390728476821,
"grad_norm": 1.2576164970306511,
"learning_rate": 3.640015174430864e-06,
"loss": 0.5682,
"step": 1045
},
{
"epoch": 0.6321493076459964,
"grad_norm": 1.2585913656336474,
"learning_rate": 3.5895031116147355e-06,
"loss": 0.5791,
"step": 1050
},
{
"epoch": 0.6351595424443106,
"grad_norm": 1.303783858936202,
"learning_rate": 3.539146970545124e-06,
"loss": 0.5789,
"step": 1055
},
{
"epoch": 0.6381697772426249,
"grad_norm": 1.2996679794519685,
"learning_rate": 3.488952317783374e-06,
"loss": 0.5669,
"step": 1060
},
{
"epoch": 0.6411800120409392,
"grad_norm": 1.227591078819204,
"learning_rate": 3.438924702039301e-06,
"loss": 0.5698,
"step": 1065
},
{
"epoch": 0.6441902468392534,
"grad_norm": 1.2857752436654992,
"learning_rate": 3.389069653557805e-06,
"loss": 0.5739,
"step": 1070
},
{
"epoch": 0.6472004816375677,
"grad_norm": 1.2421495144623809,
"learning_rate": 3.3393926835075307e-06,
"loss": 0.5788,
"step": 1075
},
{
"epoch": 0.650210716435882,
"grad_norm": 1.3331817251764237,
"learning_rate": 3.289899283371657e-06,
"loss": 0.5606,
"step": 1080
},
{
"epoch": 0.6532209512341962,
"grad_norm": 1.2708137446575152,
"learning_rate": 3.240594924340835e-06,
"loss": 0.5615,
"step": 1085
},
{
"epoch": 0.6562311860325105,
"grad_norm": 1.2640775164953746,
"learning_rate": 3.1914850567083866e-06,
"loss": 0.565,
"step": 1090
},
{
"epoch": 0.6592414208308248,
"grad_norm": 1.2780803430813366,
"learning_rate": 3.1425751092678064e-06,
"loss": 0.5508,
"step": 1095
},
{
"epoch": 0.6622516556291391,
"grad_norm": 1.2424354197819152,
"learning_rate": 3.0938704887126425e-06,
"loss": 0.5541,
"step": 1100
},
{
"epoch": 0.6652618904274533,
"grad_norm": 1.2085302041635138,
"learning_rate": 3.045376579038821e-06,
"loss": 0.5507,
"step": 1105
},
{
"epoch": 0.6682721252257676,
"grad_norm": 1.2913791921928985,
"learning_rate": 2.9970987409494784e-06,
"loss": 0.5751,
"step": 1110
},
{
"epoch": 0.6712823600240819,
"grad_norm": 1.2151986171427258,
"learning_rate": 2.9490423112623646e-06,
"loss": 0.5545,
"step": 1115
},
{
"epoch": 0.6742925948223961,
"grad_norm": 1.285546926147462,
"learning_rate": 2.9012126023198973e-06,
"loss": 0.5559,
"step": 1120
},
{
"epoch": 0.6773028296207104,
"grad_norm": 1.2516039295058423,
"learning_rate": 2.853614901401909e-06,
"loss": 0.5492,
"step": 1125
},
{
"epoch": 0.6803130644190247,
"grad_norm": 1.2619394657656795,
"learning_rate": 2.806254470141174e-06,
"loss": 0.5535,
"step": 1130
},
{
"epoch": 0.6833232992173389,
"grad_norm": 1.263160410510985,
"learning_rate": 2.759136543941773e-06,
"loss": 0.5491,
"step": 1135
},
{
"epoch": 0.6863335340156532,
"grad_norm": 1.2808685326776417,
"learning_rate": 2.712266331400332e-06,
"loss": 0.5355,
"step": 1140
},
{
"epoch": 0.6893437688139675,
"grad_norm": 1.2224213775798936,
"learning_rate": 2.66564901373027e-06,
"loss": 0.5497,
"step": 1145
},
{
"epoch": 0.6923540036122817,
"grad_norm": 1.2417187357822892,
"learning_rate": 2.6192897441890337e-06,
"loss": 0.5395,
"step": 1150
},
{
"epoch": 0.695364238410596,
"grad_norm": 1.2539889085403395,
"learning_rate": 2.573193647508426e-06,
"loss": 0.5484,
"step": 1155
},
{
"epoch": 0.6983744732089103,
"grad_norm": 1.277361909713358,
"learning_rate": 2.5273658193281252e-06,
"loss": 0.5359,
"step": 1160
},
{
"epoch": 0.7013847080072245,
"grad_norm": 1.2511254070143052,
"learning_rate": 2.4818113256323745e-06,
"loss": 0.5549,
"step": 1165
},
{
"epoch": 0.7043949428055388,
"grad_norm": 1.3131375870134938,
"learning_rate": 2.4365352021899635e-06,
"loss": 0.5555,
"step": 1170
},
{
"epoch": 0.7074051776038531,
"grad_norm": 1.2825609178954607,
"learning_rate": 2.391542453997578e-06,
"loss": 0.5452,
"step": 1175
},
{
"epoch": 0.7104154124021673,
"grad_norm": 1.2877459534486142,
"learning_rate": 2.346838054726505e-06,
"loss": 0.5417,
"step": 1180
},
{
"epoch": 0.7134256472004816,
"grad_norm": 1.2529996046706589,
"learning_rate": 2.302426946172836e-06,
"loss": 0.54,
"step": 1185
},
{
"epoch": 0.7164358819987959,
"grad_norm": 1.2558399980367174,
"learning_rate": 2.258314037711184e-06,
"loss": 0.5484,
"step": 1190
},
{
"epoch": 0.7194461167971101,
"grad_norm": 1.272832955666527,
"learning_rate": 2.214504205751971e-06,
"loss": 0.5515,
"step": 1195
},
{
"epoch": 0.7224563515954244,
"grad_norm": 1.2469393251594179,
"learning_rate": 2.1710022932023805e-06,
"loss": 0.5519,
"step": 1200
},
{
"epoch": 0.7254665863937387,
"grad_norm": 1.243045028807727,
"learning_rate": 2.127813108931007e-06,
"loss": 0.5438,
"step": 1205
},
{
"epoch": 0.7284768211920529,
"grad_norm": 1.2114723202770783,
"learning_rate": 2.084941427236245e-06,
"loss": 0.5435,
"step": 1210
},
{
"epoch": 0.7314870559903672,
"grad_norm": 1.2939587849010044,
"learning_rate": 2.04239198731855e-06,
"loss": 0.5393,
"step": 1215
},
{
"epoch": 0.7344972907886815,
"grad_norm": 1.2378436974378162,
"learning_rate": 2.000169492756523e-06,
"loss": 0.5455,
"step": 1220
},
{
"epoch": 0.7375075255869958,
"grad_norm": 1.2231608996738805,
"learning_rate": 1.9582786109869713e-06,
"loss": 0.5421,
"step": 1225
},
{
"epoch": 0.74051776038531,
"grad_norm": 1.3275912626430932,
"learning_rate": 1.9167239727889527e-06,
"loss": 0.5361,
"step": 1230
},
{
"epoch": 0.7435279951836243,
"grad_norm": 1.3041604998744816,
"learning_rate": 1.875510171771865e-06,
"loss": 0.5331,
"step": 1235
},
{
"epoch": 0.7465382299819386,
"grad_norm": 1.279896389842292,
"learning_rate": 1.8346417638676533e-06,
"loss": 0.5286,
"step": 1240
},
{
"epoch": 0.7495484647802528,
"grad_norm": 1.309403322072463,
"learning_rate": 1.7941232668271863e-06,
"loss": 0.5479,
"step": 1245
},
{
"epoch": 0.7525586995785671,
"grad_norm": 1.2455386392310572,
"learning_rate": 1.753959159720836e-06,
"loss": 0.5353,
"step": 1250
},
{
"epoch": 0.7555689343768814,
"grad_norm": 1.2637146932290357,
"learning_rate": 1.7141538824433506e-06,
"loss": 0.5349,
"step": 1255
},
{
"epoch": 0.7585791691751956,
"grad_norm": 1.258655302158318,
"learning_rate": 1.6747118352230495e-06,
"loss": 0.538,
"step": 1260
},
{
"epoch": 0.7615894039735099,
"grad_norm": 1.2756271345276968,
"learning_rate": 1.6356373781354058e-06,
"loss": 0.5167,
"step": 1265
},
{
"epoch": 0.7645996387718242,
"grad_norm": 1.2413970049455438,
"learning_rate": 1.5969348306210692e-06,
"loss": 0.5309,
"step": 1270
},
{
"epoch": 0.7676098735701384,
"grad_norm": 1.3035393545030223,
"learning_rate": 1.5586084710083737e-06,
"loss": 0.5239,
"step": 1275
},
{
"epoch": 0.7706201083684527,
"grad_norm": 1.2294394327286127,
"learning_rate": 1.5206625360403943e-06,
"loss": 0.5305,
"step": 1280
},
{
"epoch": 0.773630343166767,
"grad_norm": 1.2174444367598383,
"learning_rate": 1.4831012204066114e-06,
"loss": 0.531,
"step": 1285
},
{
"epoch": 0.7766405779650812,
"grad_norm": 1.2366691168478938,
"learning_rate": 1.445928676279199e-06,
"loss": 0.518,
"step": 1290
},
{
"epoch": 0.7796508127633955,
"grad_norm": 1.264298324101515,
"learning_rate": 1.4091490128540374e-06,
"loss": 0.5068,
"step": 1295
},
{
"epoch": 0.7826610475617098,
"grad_norm": 1.2332308291363379,
"learning_rate": 1.3727662958964627e-06,
"loss": 0.5203,
"step": 1300
},
{
"epoch": 0.785671282360024,
"grad_norm": 1.2299329633898903,
"learning_rate": 1.3367845472918272e-06,
"loss": 0.5298,
"step": 1305
},
{
"epoch": 0.7886815171583383,
"grad_norm": 1.2504340191558994,
"learning_rate": 1.3012077446008969e-06,
"loss": 0.5248,
"step": 1310
},
{
"epoch": 0.7916917519566526,
"grad_norm": 1.3049347135811968,
"learning_rate": 1.266039820620159e-06,
"loss": 0.523,
"step": 1315
},
{
"epoch": 0.7947019867549668,
"grad_norm": 1.2179046011678003,
"learning_rate": 1.2312846629470826e-06,
"loss": 0.4974,
"step": 1320
},
{
"epoch": 0.7977122215532811,
"grad_norm": 1.3164892931071699,
"learning_rate": 1.1969461135503573e-06,
"loss": 0.526,
"step": 1325
},
{
"epoch": 0.8007224563515954,
"grad_norm": 1.271375375821384,
"learning_rate": 1.163027968345195e-06,
"loss": 0.5161,
"step": 1330
},
{
"epoch": 0.8037326911499096,
"grad_norm": 1.3139938511991043,
"learning_rate": 1.1295339767737125e-06,
"loss": 0.5113,
"step": 1335
},
{
"epoch": 0.8067429259482239,
"grad_norm": 1.286007032144441,
"learning_rate": 1.0964678413904529e-06,
"loss": 0.5217,
"step": 1340
},
{
"epoch": 0.8097531607465382,
"grad_norm": 1.3030418958577936,
"learning_rate": 1.0638332174530953e-06,
"loss": 0.5326,
"step": 1345
},
{
"epoch": 0.8127633955448526,
"grad_norm": 1.2573877263232616,
"learning_rate": 1.0316337125183817e-06,
"loss": 0.5172,
"step": 1350
},
{
"epoch": 0.8157736303431667,
"grad_norm": 1.3008639038029295,
"learning_rate": 9.998728860433277e-07,
"loss": 0.512,
"step": 1355
},
{
"epoch": 0.818783865141481,
"grad_norm": 1.2333298206728438,
"learning_rate": 9.685542489917494e-07,
"loss": 0.5047,
"step": 1360
},
{
"epoch": 0.8217940999397954,
"grad_norm": 1.289522737500026,
"learning_rate": 9.376812634461418e-07,
"loss": 0.5104,
"step": 1365
},
{
"epoch": 0.8248043347381095,
"grad_norm": 1.2659811011921933,
"learning_rate": 9.072573422249692e-07,
"loss": 0.5232,
"step": 1370
},
{
"epoch": 0.8278145695364238,
"grad_norm": 1.22881385464476,
"learning_rate": 8.772858485054042e-07,
"loss": 0.5199,
"step": 1375
},
{
"epoch": 0.8308248043347382,
"grad_norm": 1.2715826823700018,
"learning_rate": 8.477700954515372e-07,
"loss": 0.5202,
"step": 1380
},
{
"epoch": 0.8338350391330523,
"grad_norm": 1.2503742943128184,
"learning_rate": 8.187133458481416e-07,
"loss": 0.5285,
"step": 1385
},
{
"epoch": 0.8368452739313667,
"grad_norm": 1.2943224950483547,
"learning_rate": 7.901188117399817e-07,
"loss": 0.5146,
"step": 1390
},
{
"epoch": 0.839855508729681,
"grad_norm": 1.2465983369075755,
"learning_rate": 7.619896540767435e-07,
"loss": 0.5214,
"step": 1395
},
{
"epoch": 0.8428657435279951,
"grad_norm": 1.2931198121674863,
"learning_rate": 7.343289823636168e-07,
"loss": 0.5103,
"step": 1400
},
{
"epoch": 0.8458759783263095,
"grad_norm": 1.3126713125625864,
"learning_rate": 7.0713985431755e-07,
"loss": 0.5131,
"step": 1405
},
{
"epoch": 0.8488862131246238,
"grad_norm": 1.2514737845776198,
"learning_rate": 6.804252755292429e-07,
"loss": 0.4972,
"step": 1410
},
{
"epoch": 0.851896447922938,
"grad_norm": 1.2888999488065467,
"learning_rate": 6.541881991309013e-07,
"loss": 0.5086,
"step": 1415
},
{
"epoch": 0.8549066827212523,
"grad_norm": 1.2929120768720084,
"learning_rate": 6.284315254697726e-07,
"loss": 0.5131,
"step": 1420
},
{
"epoch": 0.8579169175195666,
"grad_norm": 1.2348017732169416,
"learning_rate": 6.031581017875482e-07,
"loss": 0.5147,
"step": 1425
},
{
"epoch": 0.8609271523178808,
"grad_norm": 1.2685705494636663,
"learning_rate": 5.783707219056078e-07,
"loss": 0.5092,
"step": 1430
},
{
"epoch": 0.863937387116195,
"grad_norm": 1.2976102586906246,
"learning_rate": 5.540721259161774e-07,
"loss": 0.5115,
"step": 1435
},
{
"epoch": 0.8669476219145094,
"grad_norm": 1.250337429227757,
"learning_rate": 5.302649998794368e-07,
"loss": 0.5169,
"step": 1440
},
{
"epoch": 0.8699578567128236,
"grad_norm": 1.2461623782975149,
"learning_rate": 5.0695197552659e-07,
"loss": 0.5078,
"step": 1445
},
{
"epoch": 0.8729680915111379,
"grad_norm": 1.2782693319086738,
"learning_rate": 4.841356299689359e-07,
"loss": 0.5157,
"step": 1450
},
{
"epoch": 0.8759783263094522,
"grad_norm": 1.2663091436447071,
"learning_rate": 4.618184854129981e-07,
"loss": 0.5094,
"step": 1455
},
{
"epoch": 0.8789885611077664,
"grad_norm": 1.2889286164793572,
"learning_rate": 4.4000300888169753e-07,
"loss": 0.5097,
"step": 1460
},
{
"epoch": 0.8819987959060807,
"grad_norm": 1.270161020673374,
"learning_rate": 4.1869161194164565e-07,
"loss": 0.5178,
"step": 1465
},
{
"epoch": 0.885009030704395,
"grad_norm": 1.3028978532412478,
"learning_rate": 3.9788665043656083e-07,
"loss": 0.5071,
"step": 1470
},
{
"epoch": 0.8880192655027093,
"grad_norm": 1.2451090536200642,
"learning_rate": 3.775904242268391e-07,
"loss": 0.5161,
"step": 1475
},
{
"epoch": 0.8910295003010235,
"grad_norm": 1.3526909015751019,
"learning_rate": 3.578051769353219e-07,
"loss": 0.5156,
"step": 1480
},
{
"epoch": 0.8940397350993378,
"grad_norm": 1.349791093552624,
"learning_rate": 3.385330956992816e-07,
"loss": 0.5163,
"step": 1485
},
{
"epoch": 0.8970499698976521,
"grad_norm": 1.274654975609038,
"learning_rate": 3.1977631092863613e-07,
"loss": 0.5085,
"step": 1490
},
{
"epoch": 0.9000602046959663,
"grad_norm": 1.2723546077606147,
"learning_rate": 3.015368960704584e-07,
"loss": 0.4992,
"step": 1495
},
{
"epoch": 0.9030704394942806,
"grad_norm": 1.2492243057949244,
"learning_rate": 2.8381686737975867e-07,
"loss": 0.5117,
"step": 1500
},
{
"epoch": 0.9060806742925949,
"grad_norm": 1.2506347579447048,
"learning_rate": 2.666181836966053e-07,
"loss": 0.5061,
"step": 1505
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.2651285126889307,
"learning_rate": 2.4994274622958726e-07,
"loss": 0.5117,
"step": 1510
},
{
"epoch": 0.9121011438892234,
"grad_norm": 1.2928036851193043,
"learning_rate": 2.3379239834564526e-07,
"loss": 0.5052,
"step": 1515
},
{
"epoch": 0.9151113786875377,
"grad_norm": 1.327963392569787,
"learning_rate": 2.1816892536629775e-07,
"loss": 0.505,
"step": 1520
},
{
"epoch": 0.9181216134858519,
"grad_norm": 1.2391568358607692,
"learning_rate": 2.0307405437029027e-07,
"loss": 0.4938,
"step": 1525
},
{
"epoch": 0.9211318482841662,
"grad_norm": 1.2967086784362822,
"learning_rate": 1.8850945400266994e-07,
"loss": 0.5108,
"step": 1530
},
{
"epoch": 0.9241420830824805,
"grad_norm": 1.29545735590143,
"learning_rate": 1.7447673429033361e-07,
"loss": 0.5032,
"step": 1535
},
{
"epoch": 0.9271523178807947,
"grad_norm": 1.2875158001193772,
"learning_rate": 1.6097744646404457e-07,
"loss": 0.5107,
"step": 1540
},
{
"epoch": 0.930162552679109,
"grad_norm": 1.266148830144139,
"learning_rate": 1.4801308278695636e-07,
"loss": 0.5074,
"step": 1545
},
{
"epoch": 0.9331727874774233,
"grad_norm": 1.3306638459879772,
"learning_rate": 1.3558507638965158e-07,
"loss": 0.507,
"step": 1550
},
{
"epoch": 0.9361830222757375,
"grad_norm": 1.333069277227204,
"learning_rate": 1.2369480111171784e-07,
"loss": 0.5015,
"step": 1555
},
{
"epoch": 0.9391932570740518,
"grad_norm": 1.3040224501040143,
"learning_rate": 1.1234357134987717e-07,
"loss": 0.498,
"step": 1560
},
{
"epoch": 0.9422034918723661,
"grad_norm": 1.2723009240100882,
"learning_rate": 1.0153264191269052e-07,
"loss": 0.522,
"step": 1565
},
{
"epoch": 0.9452137266706803,
"grad_norm": 1.2733119108501816,
"learning_rate": 9.126320788184374e-08,
"loss": 0.5084,
"step": 1570
},
{
"epoch": 0.9482239614689946,
"grad_norm": 1.2479278083881131,
"learning_rate": 8.153640448003875e-08,
"loss": 0.4962,
"step": 1575
},
{
"epoch": 0.9512341962673089,
"grad_norm": 1.2959385578328415,
"learning_rate": 7.235330694550402e-08,
"loss": 0.5114,
"step": 1580
},
{
"epoch": 0.9542444310656231,
"grad_norm": 1.231634916892265,
"learning_rate": 6.371493041313126e-08,
"loss": 0.4951,
"step": 1585
},
{
"epoch": 0.9572546658639374,
"grad_norm": 1.27585215530384,
"learning_rate": 5.562222980225907e-08,
"loss": 0.5071,
"step": 1590
},
{
"epoch": 0.9602649006622517,
"grad_norm": 1.225825933770062,
"learning_rate": 4.807609971111238e-08,
"loss": 0.4971,
"step": 1595
},
{
"epoch": 0.963275135460566,
"grad_norm": 1.2453682011592957,
"learning_rate": 4.107737431791159e-08,
"loss": 0.5054,
"step": 1600
},
{
"epoch": 0.9662853702588802,
"grad_norm": 1.2857076641073053,
"learning_rate": 3.462682728865685e-08,
"loss": 0.5043,
"step": 1605
},
{
"epoch": 0.9692956050571945,
"grad_norm": 1.2778563737126165,
"learning_rate": 2.8725171691605934e-08,
"loss": 0.5075,
"step": 1610
},
{
"epoch": 0.9723058398555088,
"grad_norm": 1.2554905316985445,
"learning_rate": 2.3373059918448958e-08,
"loss": 0.4938,
"step": 1615
},
{
"epoch": 0.975316074653823,
"grad_norm": 1.3192898914729378,
"learning_rate": 1.8571083612188845e-08,
"loss": 0.5088,
"step": 1620
},
{
"epoch": 0.9783263094521373,
"grad_norm": 1.2731885499582947,
"learning_rate": 1.431977360173975e-08,
"loss": 0.4969,
"step": 1625
},
{
"epoch": 0.9813365442504516,
"grad_norm": 1.304095478961953,
"learning_rate": 1.0619599843249006e-08,
"loss": 0.5139,
"step": 1630
},
{
"epoch": 0.9843467790487658,
"grad_norm": 1.27242623092078,
"learning_rate": 7.470971368142011e-09,
"loss": 0.5156,
"step": 1635
},
{
"epoch": 0.9873570138470801,
"grad_norm": 1.239577403155516,
"learning_rate": 4.874236237911723e-09,
"loss": 0.5042,
"step": 1640
},
{
"epoch": 0.9903672486453944,
"grad_norm": 1.2559422923667112,
"learning_rate": 2.8296815056377824e-09,
"loss": 0.5124,
"step": 1645
},
{
"epoch": 0.9933774834437086,
"grad_norm": 1.2978012230245286,
"learning_rate": 1.3375331842574446e-09,
"loss": 0.5151,
"step": 1650
},
{
"epoch": 0.9963877182420229,
"grad_norm": 1.2286869775237814,
"learning_rate": 3.9795622158111945e-10,
"loss": 0.5093,
"step": 1655
},
{
"epoch": 0.9993979530403372,
"grad_norm": 1.2619068290294162,
"learning_rate": 1.1054482056405136e-11,
"loss": 0.5114,
"step": 1660
},
{
"epoch": 1.0,
"eval_runtime": 3.8298,
"eval_samples_per_second": 2.611,
"eval_steps_per_second": 0.783,
"step": 1661
},
{
"epoch": 1.0,
"step": 1661,
"total_flos": 430711668473856.0,
"train_loss": 0.669738974236064,
"train_runtime": 16716.2181,
"train_samples_per_second": 1.59,
"train_steps_per_second": 0.099
}
],
"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
}