zephyr-qwen2-7b-sft / trainer_state.json
tanliboy's picture
Model save
14d2744 verified
raw
history blame
34.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 956,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0010460251046025104,
"grad_norm": 4.858334608923654,
"learning_rate": 2.0833333333333333e-07,
"loss": 1.3092,
"step": 1
},
{
"epoch": 0.005230125523012552,
"grad_norm": 4.6974531972545845,
"learning_rate": 1.0416666666666667e-06,
"loss": 1.2901,
"step": 5
},
{
"epoch": 0.010460251046025104,
"grad_norm": 4.050379213449396,
"learning_rate": 2.0833333333333334e-06,
"loss": 1.2488,
"step": 10
},
{
"epoch": 0.015690376569037656,
"grad_norm": 1.260482829918926,
"learning_rate": 3.125e-06,
"loss": 1.1402,
"step": 15
},
{
"epoch": 0.02092050209205021,
"grad_norm": 0.7572315531712823,
"learning_rate": 4.166666666666667e-06,
"loss": 1.1238,
"step": 20
},
{
"epoch": 0.02615062761506276,
"grad_norm": 0.6345745967373482,
"learning_rate": 5.208333333333334e-06,
"loss": 1.0882,
"step": 25
},
{
"epoch": 0.03138075313807531,
"grad_norm": 0.5480254516819103,
"learning_rate": 6.25e-06,
"loss": 1.0972,
"step": 30
},
{
"epoch": 0.036610878661087864,
"grad_norm": 0.5577040974895372,
"learning_rate": 7.291666666666667e-06,
"loss": 1.0967,
"step": 35
},
{
"epoch": 0.04184100418410042,
"grad_norm": 0.5806448242568741,
"learning_rate": 8.333333333333334e-06,
"loss": 1.0806,
"step": 40
},
{
"epoch": 0.04707112970711297,
"grad_norm": 0.6520252066307457,
"learning_rate": 9.375000000000001e-06,
"loss": 1.0799,
"step": 45
},
{
"epoch": 0.05230125523012552,
"grad_norm": 0.5935927194187152,
"learning_rate": 1.0416666666666668e-05,
"loss": 1.0776,
"step": 50
},
{
"epoch": 0.05753138075313807,
"grad_norm": 0.539119622245633,
"learning_rate": 1.1458333333333333e-05,
"loss": 1.0899,
"step": 55
},
{
"epoch": 0.06276150627615062,
"grad_norm": 0.5602624433341231,
"learning_rate": 1.25e-05,
"loss": 1.0765,
"step": 60
},
{
"epoch": 0.06799163179916318,
"grad_norm": 0.6159503903121701,
"learning_rate": 1.3541666666666668e-05,
"loss": 1.0822,
"step": 65
},
{
"epoch": 0.07322175732217573,
"grad_norm": 0.6004379020843004,
"learning_rate": 1.4583333333333333e-05,
"loss": 1.0676,
"step": 70
},
{
"epoch": 0.07845188284518828,
"grad_norm": 0.5874082854599968,
"learning_rate": 1.5625e-05,
"loss": 1.094,
"step": 75
},
{
"epoch": 0.08368200836820083,
"grad_norm": 0.6751852785193894,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.0929,
"step": 80
},
{
"epoch": 0.08891213389121339,
"grad_norm": 0.5828293222590728,
"learning_rate": 1.7708333333333335e-05,
"loss": 1.1021,
"step": 85
},
{
"epoch": 0.09414225941422594,
"grad_norm": 0.6900402033657895,
"learning_rate": 1.8750000000000002e-05,
"loss": 1.0856,
"step": 90
},
{
"epoch": 0.09937238493723849,
"grad_norm": 0.5853837848957884,
"learning_rate": 1.979166666666667e-05,
"loss": 1.0862,
"step": 95
},
{
"epoch": 0.10460251046025104,
"grad_norm": 0.6701191850895604,
"learning_rate": 1.9998932457674904e-05,
"loss": 1.0933,
"step": 100
},
{
"epoch": 0.1098326359832636,
"grad_norm": 0.5533200148308522,
"learning_rate": 1.999459595761682e-05,
"loss": 1.1147,
"step": 105
},
{
"epoch": 0.11506276150627615,
"grad_norm": 0.7369714526144382,
"learning_rate": 1.9986925223989665e-05,
"loss": 1.093,
"step": 110
},
{
"epoch": 0.1202928870292887,
"grad_norm": 0.5547288913187312,
"learning_rate": 1.9975922815778082e-05,
"loss": 1.0831,
"step": 115
},
{
"epoch": 0.12552301255230125,
"grad_norm": 0.6897888287129971,
"learning_rate": 1.996159240342547e-05,
"loss": 1.0825,
"step": 120
},
{
"epoch": 0.1307531380753138,
"grad_norm": 0.6130228169149287,
"learning_rate": 1.9943938767609515e-05,
"loss": 1.083,
"step": 125
},
{
"epoch": 0.13598326359832635,
"grad_norm": 0.5733760917524161,
"learning_rate": 1.9922967797647357e-05,
"loss": 1.0905,
"step": 130
},
{
"epoch": 0.1412133891213389,
"grad_norm": 0.5893547730249119,
"learning_rate": 1.9898686489530852e-05,
"loss": 1.0974,
"step": 135
},
{
"epoch": 0.14644351464435146,
"grad_norm": 0.5801179495005143,
"learning_rate": 1.9871102943592717e-05,
"loss": 1.0731,
"step": 140
},
{
"epoch": 0.151673640167364,
"grad_norm": 0.5638361223609211,
"learning_rate": 1.9840226361804223e-05,
"loss": 1.1055,
"step": 145
},
{
"epoch": 0.15690376569037656,
"grad_norm": 0.5449934219233721,
"learning_rate": 1.9806067044705375e-05,
"loss": 1.0761,
"step": 150
},
{
"epoch": 0.16213389121338911,
"grad_norm": 0.6274931137807601,
"learning_rate": 1.9768636387968616e-05,
"loss": 1.1064,
"step": 155
},
{
"epoch": 0.16736401673640167,
"grad_norm": 0.6250809185735244,
"learning_rate": 1.9727946878597193e-05,
"loss": 1.0791,
"step": 160
},
{
"epoch": 0.17259414225941422,
"grad_norm": 0.5284412080420968,
"learning_rate": 1.968401209075946e-05,
"loss": 1.0839,
"step": 165
},
{
"epoch": 0.17782426778242677,
"grad_norm": 0.5821294853027748,
"learning_rate": 1.963684668126046e-05,
"loss": 1.1018,
"step": 170
},
{
"epoch": 0.18305439330543932,
"grad_norm": 0.6298467979748181,
"learning_rate": 1.9586466384652366e-05,
"loss": 1.0847,
"step": 175
},
{
"epoch": 0.18828451882845187,
"grad_norm": 0.6400162596224541,
"learning_rate": 1.9532888007985408e-05,
"loss": 1.0957,
"step": 180
},
{
"epoch": 0.19351464435146443,
"grad_norm": 0.6000021573219635,
"learning_rate": 1.9476129425200935e-05,
"loss": 1.0938,
"step": 185
},
{
"epoch": 0.19874476987447698,
"grad_norm": 0.6537737768236229,
"learning_rate": 1.9416209571168648e-05,
"loss": 1.089,
"step": 190
},
{
"epoch": 0.20397489539748953,
"grad_norm": 0.5787562041762513,
"learning_rate": 1.9353148435369816e-05,
"loss": 1.0906,
"step": 195
},
{
"epoch": 0.20920502092050208,
"grad_norm": 0.5695490038983394,
"learning_rate": 1.9286967055228744e-05,
"loss": 1.093,
"step": 200
},
{
"epoch": 0.21443514644351463,
"grad_norm": 0.6745539583613525,
"learning_rate": 1.9217687509094596e-05,
"loss": 1.0759,
"step": 205
},
{
"epoch": 0.2196652719665272,
"grad_norm": 0.6099357395998225,
"learning_rate": 1.9145332908875984e-05,
"loss": 1.0911,
"step": 210
},
{
"epoch": 0.22489539748953974,
"grad_norm": 0.6131129695854803,
"learning_rate": 1.9069927392330742e-05,
"loss": 1.1017,
"step": 215
},
{
"epoch": 0.2301255230125523,
"grad_norm": 0.700006659493977,
"learning_rate": 1.89914961150135e-05,
"loss": 1.082,
"step": 220
},
{
"epoch": 0.23535564853556484,
"grad_norm": 0.532783820430617,
"learning_rate": 1.891006524188368e-05,
"loss": 1.0797,
"step": 225
},
{
"epoch": 0.2405857740585774,
"grad_norm": 0.6213200850147771,
"learning_rate": 1.8825661938576784e-05,
"loss": 1.085,
"step": 230
},
{
"epoch": 0.24581589958158995,
"grad_norm": 0.6011279894568407,
"learning_rate": 1.8738314362341828e-05,
"loss": 1.085,
"step": 235
},
{
"epoch": 0.2510460251046025,
"grad_norm": 0.5489112313533119,
"learning_rate": 1.864805165264799e-05,
"loss": 1.0952,
"step": 240
},
{
"epoch": 0.25627615062761505,
"grad_norm": 0.6812709510575856,
"learning_rate": 1.8554903921463578e-05,
"loss": 1.0889,
"step": 245
},
{
"epoch": 0.2615062761506276,
"grad_norm": 0.5886412319389147,
"learning_rate": 1.8458902243210558e-05,
"loss": 1.0822,
"step": 250
},
{
"epoch": 0.26673640167364016,
"grad_norm": 0.6816262740472222,
"learning_rate": 1.8360078644398028e-05,
"loss": 1.0788,
"step": 255
},
{
"epoch": 0.2719665271966527,
"grad_norm": 0.5947592675988007,
"learning_rate": 1.8258466092938042e-05,
"loss": 1.101,
"step": 260
},
{
"epoch": 0.27719665271966526,
"grad_norm": 0.57966086068774,
"learning_rate": 1.8154098487147405e-05,
"loss": 1.0776,
"step": 265
},
{
"epoch": 0.2824267782426778,
"grad_norm": 0.7809661648951529,
"learning_rate": 1.8047010644439074e-05,
"loss": 1.0838,
"step": 270
},
{
"epoch": 0.28765690376569036,
"grad_norm": 0.5662869436802427,
"learning_rate": 1.7937238289706927e-05,
"loss": 1.0822,
"step": 275
},
{
"epoch": 0.2928870292887029,
"grad_norm": 0.6222425392268467,
"learning_rate": 1.7824818043407828e-05,
"loss": 1.0641,
"step": 280
},
{
"epoch": 0.29811715481171547,
"grad_norm": 0.5849161394935216,
"learning_rate": 1.7709787409344893e-05,
"loss": 1.075,
"step": 285
},
{
"epoch": 0.303347280334728,
"grad_norm": 0.6269611916504507,
"learning_rate": 1.75921847621561e-05,
"loss": 1.0863,
"step": 290
},
{
"epoch": 0.30857740585774057,
"grad_norm": 0.6490170961985677,
"learning_rate": 1.7472049334512354e-05,
"loss": 1.0753,
"step": 295
},
{
"epoch": 0.3138075313807531,
"grad_norm": 0.5516576130404307,
"learning_rate": 1.7349421204029343e-05,
"loss": 1.0838,
"step": 300
},
{
"epoch": 0.3190376569037657,
"grad_norm": 0.590710588774648,
"learning_rate": 1.7224341279897473e-05,
"loss": 1.0937,
"step": 305
},
{
"epoch": 0.32426778242677823,
"grad_norm": 0.6160629723045913,
"learning_rate": 1.7096851289234448e-05,
"loss": 1.0861,
"step": 310
},
{
"epoch": 0.3294979079497908,
"grad_norm": 0.6756970289943374,
"learning_rate": 1.6966993763164928e-05,
"loss": 1.0863,
"step": 315
},
{
"epoch": 0.33472803347280333,
"grad_norm": 0.5500515778659105,
"learning_rate": 1.6834812022632e-05,
"loss": 1.0916,
"step": 320
},
{
"epoch": 0.3399581589958159,
"grad_norm": 0.5597424586372212,
"learning_rate": 1.670035016394513e-05,
"loss": 1.0875,
"step": 325
},
{
"epoch": 0.34518828451882844,
"grad_norm": 0.5731613149047409,
"learning_rate": 1.656365304406953e-05,
"loss": 1.0812,
"step": 330
},
{
"epoch": 0.350418410041841,
"grad_norm": 0.5799594878521428,
"learning_rate": 1.6424766265661648e-05,
"loss": 1.084,
"step": 335
},
{
"epoch": 0.35564853556485354,
"grad_norm": 0.6429338013882139,
"learning_rate": 1.6283736161855995e-05,
"loss": 1.0772,
"step": 340
},
{
"epoch": 0.3608786610878661,
"grad_norm": 0.5360765914109888,
"learning_rate": 1.614060978080823e-05,
"loss": 1.0928,
"step": 345
},
{
"epoch": 0.36610878661087864,
"grad_norm": 0.5564731824128625,
"learning_rate": 1.5995434869999723e-05,
"loss": 1.079,
"step": 350
},
{
"epoch": 0.3713389121338912,
"grad_norm": 0.561749976298582,
"learning_rate": 1.584825986030883e-05,
"loss": 1.0876,
"step": 355
},
{
"epoch": 0.37656903765690375,
"grad_norm": 0.5636262441838287,
"learning_rate": 1.5699133849854164e-05,
"loss": 1.0889,
"step": 360
},
{
"epoch": 0.3817991631799163,
"grad_norm": 0.5703557438762126,
"learning_rate": 1.5548106587615295e-05,
"loss": 1.067,
"step": 365
},
{
"epoch": 0.38702928870292885,
"grad_norm": 0.5519913100989867,
"learning_rate": 1.5395228456836298e-05,
"loss": 1.084,
"step": 370
},
{
"epoch": 0.3922594142259414,
"grad_norm": 0.6641842447650281,
"learning_rate": 1.524055045821773e-05,
"loss": 1.0915,
"step": 375
},
{
"epoch": 0.39748953974895396,
"grad_norm": 0.6318777460752142,
"learning_rate": 1.5084124192902612e-05,
"loss": 1.0788,
"step": 380
},
{
"epoch": 0.4027196652719665,
"grad_norm": 0.5871983626150223,
"learning_rate": 1.4926001845262081e-05,
"loss": 1.0831,
"step": 385
},
{
"epoch": 0.40794979079497906,
"grad_norm": 0.5413531651813962,
"learning_rate": 1.4766236165486526e-05,
"loss": 1.0798,
"step": 390
},
{
"epoch": 0.4131799163179916,
"grad_norm": 0.5573908575490272,
"learning_rate": 1.4604880451987895e-05,
"loss": 1.0763,
"step": 395
},
{
"epoch": 0.41841004184100417,
"grad_norm": 0.5194023684186305,
"learning_rate": 1.4441988533619182e-05,
"loss": 1.0735,
"step": 400
},
{
"epoch": 0.4236401673640167,
"grad_norm": 0.5830453281999414,
"learning_rate": 1.4277614751716911e-05,
"loss": 1.0821,
"step": 405
},
{
"epoch": 0.42887029288702927,
"grad_norm": 0.5637920375973177,
"learning_rate": 1.4111813941972672e-05,
"loss": 1.0961,
"step": 410
},
{
"epoch": 0.4341004184100418,
"grad_norm": 0.5841576073237601,
"learning_rate": 1.3944641416139737e-05,
"loss": 1.0817,
"step": 415
},
{
"epoch": 0.4393305439330544,
"grad_norm": 0.5585707054156458,
"learning_rate": 1.3776152943580846e-05,
"loss": 1.0913,
"step": 420
},
{
"epoch": 0.4445606694560669,
"grad_norm": 0.5469930648842268,
"learning_rate": 1.3606404732663365e-05,
"loss": 1.0698,
"step": 425
},
{
"epoch": 0.4497907949790795,
"grad_norm": 0.6030829852829179,
"learning_rate": 1.3435453412007949e-05,
"loss": 1.0679,
"step": 430
},
{
"epoch": 0.45502092050209203,
"grad_norm": 0.5093855577577417,
"learning_rate": 1.3263356011597045e-05,
"loss": 1.0757,
"step": 435
},
{
"epoch": 0.4602510460251046,
"grad_norm": 0.5491996878865052,
"learning_rate": 1.3090169943749475e-05,
"loss": 1.0787,
"step": 440
},
{
"epoch": 0.46548117154811713,
"grad_norm": 0.512248753718038,
"learning_rate": 1.2915952983967477e-05,
"loss": 1.0877,
"step": 445
},
{
"epoch": 0.4707112970711297,
"grad_norm": 0.5483058294847694,
"learning_rate": 1.2740763251662585e-05,
"loss": 1.1004,
"step": 450
},
{
"epoch": 0.47594142259414224,
"grad_norm": 0.6156670796243979,
"learning_rate": 1.2564659190766788e-05,
"loss": 1.0764,
"step": 455
},
{
"epoch": 0.4811715481171548,
"grad_norm": 0.5548599105126784,
"learning_rate": 1.2387699550235419e-05,
"loss": 1.0858,
"step": 460
},
{
"epoch": 0.48640167364016734,
"grad_norm": 0.4944121544407168,
"learning_rate": 1.2209943364448298e-05,
"loss": 1.0584,
"step": 465
},
{
"epoch": 0.4916317991631799,
"grad_norm": 0.552437443100853,
"learning_rate": 1.2031449933515625e-05,
"loss": 1.0687,
"step": 470
},
{
"epoch": 0.49686192468619245,
"grad_norm": 0.5439607969799337,
"learning_rate": 1.1852278803495267e-05,
"loss": 1.0574,
"step": 475
},
{
"epoch": 0.502092050209205,
"grad_norm": 0.644493491235614,
"learning_rate": 1.1672489746527979e-05,
"loss": 1.0758,
"step": 480
},
{
"epoch": 0.5073221757322176,
"grad_norm": 0.6395625397020802,
"learning_rate": 1.1492142740897188e-05,
"loss": 1.0677,
"step": 485
},
{
"epoch": 0.5125523012552301,
"grad_norm": 0.556432612437615,
"learning_rate": 1.1311297951020028e-05,
"loss": 1.0652,
"step": 490
},
{
"epoch": 0.5177824267782427,
"grad_norm": 0.7228902841727272,
"learning_rate": 1.1130015707376262e-05,
"loss": 1.0715,
"step": 495
},
{
"epoch": 0.5230125523012552,
"grad_norm": 0.5377283657779242,
"learning_rate": 1.0948356486381829e-05,
"loss": 1.0769,
"step": 500
},
{
"epoch": 0.5282426778242678,
"grad_norm": 0.5540329931082308,
"learning_rate": 1.0766380890213675e-05,
"loss": 1.0806,
"step": 505
},
{
"epoch": 0.5334728033472803,
"grad_norm": 0.5757757293198247,
"learning_rate": 1.0584149626592662e-05,
"loss": 1.056,
"step": 510
},
{
"epoch": 0.5387029288702929,
"grad_norm": 0.6168745200076227,
"learning_rate": 1.0401723488531238e-05,
"loss": 1.0683,
"step": 515
},
{
"epoch": 0.5439330543933054,
"grad_norm": 0.541484533483186,
"learning_rate": 1.0219163334052682e-05,
"loss": 1.0818,
"step": 520
},
{
"epoch": 0.549163179916318,
"grad_norm": 0.6087775414515405,
"learning_rate": 1.003653006588863e-05,
"loss": 1.0757,
"step": 525
},
{
"epoch": 0.5543933054393305,
"grad_norm": 0.514881446809141,
"learning_rate": 9.853884611161709e-06,
"loss": 1.047,
"step": 530
},
{
"epoch": 0.5596234309623431,
"grad_norm": 0.5141201235730006,
"learning_rate": 9.671287901060035e-06,
"loss": 1.064,
"step": 535
},
{
"epoch": 0.5648535564853556,
"grad_norm": 0.558678168889332,
"learning_rate": 9.48880085051033e-06,
"loss": 1.0734,
"step": 540
},
{
"epoch": 0.5700836820083682,
"grad_norm": 0.5515688630887745,
"learning_rate": 9.306484337856497e-06,
"loss": 1.0776,
"step": 545
},
{
"epoch": 0.5753138075313807,
"grad_norm": 0.5198540078292905,
"learning_rate": 9.124399184550377e-06,
"loss": 1.0818,
"step": 550
},
{
"epoch": 0.5805439330543933,
"grad_norm": 0.5376863184344793,
"learning_rate": 8.94260613486149e-06,
"loss": 1.0757,
"step": 555
},
{
"epoch": 0.5857740585774058,
"grad_norm": 0.5211594345995101,
"learning_rate": 8.76116583561252e-06,
"loss": 1.059,
"step": 560
},
{
"epoch": 0.5910041841004184,
"grad_norm": 0.5365035781726012,
"learning_rate": 8.580138815947336e-06,
"loss": 1.0651,
"step": 565
},
{
"epoch": 0.5962343096234309,
"grad_norm": 0.6305280632754058,
"learning_rate": 8.399585467138215e-06,
"loss": 1.0763,
"step": 570
},
{
"epoch": 0.6014644351464435,
"grad_norm": 0.6210565261791966,
"learning_rate": 8.219566022439128e-06,
"loss": 1.0792,
"step": 575
},
{
"epoch": 0.606694560669456,
"grad_norm": 0.5614735395071686,
"learning_rate": 8.040140536991688e-06,
"loss": 1.0768,
"step": 580
},
{
"epoch": 0.6119246861924686,
"grad_norm": 0.6033992631021817,
"learning_rate": 7.861368867790541e-06,
"loss": 1.0793,
"step": 585
},
{
"epoch": 0.6171548117154811,
"grad_norm": 0.5261950818886735,
"learning_rate": 7.683310653714857e-06,
"loss": 1.0723,
"step": 590
},
{
"epoch": 0.6223849372384938,
"grad_norm": 0.6939327155318498,
"learning_rate": 7.506025295632577e-06,
"loss": 1.0647,
"step": 595
},
{
"epoch": 0.6276150627615062,
"grad_norm": 0.5456155215789399,
"learning_rate": 7.329571936584072e-06,
"loss": 1.0566,
"step": 600
},
{
"epoch": 0.6328451882845189,
"grad_norm": 0.5593291123422247,
"learning_rate": 7.154009442051813e-06,
"loss": 1.0562,
"step": 605
},
{
"epoch": 0.6380753138075314,
"grad_norm": 0.5765286225968658,
"learning_rate": 6.979396380322621e-06,
"loss": 1.0744,
"step": 610
},
{
"epoch": 0.643305439330544,
"grad_norm": 0.5457349758684551,
"learning_rate": 6.805791002949085e-06,
"loss": 1.0688,
"step": 615
},
{
"epoch": 0.6485355648535565,
"grad_norm": 0.5037932117786891,
"learning_rate": 6.63325122531663e-06,
"loss": 1.0541,
"step": 620
},
{
"epoch": 0.6537656903765691,
"grad_norm": 0.5198402167489076,
"learning_rate": 6.461834607322719e-06,
"loss": 1.0547,
"step": 625
},
{
"epoch": 0.6589958158995816,
"grad_norm": 0.532907357801191,
"learning_rate": 6.291598334174685e-06,
"loss": 1.0572,
"step": 630
},
{
"epoch": 0.6642259414225942,
"grad_norm": 0.49718822363891185,
"learning_rate": 6.122599197312525e-06,
"loss": 1.0858,
"step": 635
},
{
"epoch": 0.6694560669456067,
"grad_norm": 0.5022400515016125,
"learning_rate": 5.954893575463064e-06,
"loss": 1.0637,
"step": 640
},
{
"epoch": 0.6746861924686193,
"grad_norm": 0.5200472376048936,
"learning_rate": 5.788537415831825e-06,
"loss": 1.055,
"step": 645
},
{
"epoch": 0.6799163179916318,
"grad_norm": 0.5001640685326711,
"learning_rate": 5.623586215438813e-06,
"loss": 1.0641,
"step": 650
},
{
"epoch": 0.6851464435146444,
"grad_norm": 0.5054182621927257,
"learning_rate": 5.460095002604533e-06,
"loss": 1.0477,
"step": 655
},
{
"epoch": 0.6903765690376569,
"grad_norm": 0.540901612852172,
"learning_rate": 5.298118318592316e-06,
"loss": 1.0654,
"step": 660
},
{
"epoch": 0.6956066945606695,
"grad_norm": 0.5013385695399603,
"learning_rate": 5.137710199413182e-06,
"loss": 1.0686,
"step": 665
},
{
"epoch": 0.700836820083682,
"grad_norm": 0.5598314300382486,
"learning_rate": 4.978924157799208e-06,
"loss": 1.0601,
"step": 670
},
{
"epoch": 0.7060669456066946,
"grad_norm": 0.48540319643618013,
"learning_rate": 4.821813165351495e-06,
"loss": 1.0581,
"step": 675
},
{
"epoch": 0.7112970711297071,
"grad_norm": 0.5008016797878718,
"learning_rate": 4.666429634868651e-06,
"loss": 1.0851,
"step": 680
},
{
"epoch": 0.7165271966527197,
"grad_norm": 0.4863424416345438,
"learning_rate": 4.512825402861686e-06,
"loss": 1.0552,
"step": 685
},
{
"epoch": 0.7217573221757322,
"grad_norm": 0.5716252229591956,
"learning_rate": 4.361051712261173e-06,
"loss": 1.0596,
"step": 690
},
{
"epoch": 0.7269874476987448,
"grad_norm": 0.5290120655638039,
"learning_rate": 4.211159195322404e-06,
"loss": 1.05,
"step": 695
},
{
"epoch": 0.7322175732217573,
"grad_norm": 0.5201644477315278,
"learning_rate": 4.063197856734295e-06,
"loss": 1.0783,
"step": 700
},
{
"epoch": 0.7374476987447699,
"grad_norm": 0.5029601729726882,
"learning_rate": 3.917217056937649e-06,
"loss": 1.0636,
"step": 705
},
{
"epoch": 0.7426778242677824,
"grad_norm": 0.5165741001138463,
"learning_rate": 3.773265495658309e-06,
"loss": 1.0655,
"step": 710
},
{
"epoch": 0.747907949790795,
"grad_norm": 0.4952826724245214,
"learning_rate": 3.6313911956607827e-06,
"loss": 1.0854,
"step": 715
},
{
"epoch": 0.7531380753138075,
"grad_norm": 0.5146856714184609,
"learning_rate": 3.491641486727645e-06,
"loss": 1.0684,
"step": 720
},
{
"epoch": 0.7583682008368201,
"grad_norm": 0.5340447988495044,
"learning_rate": 3.354062989870166e-06,
"loss": 1.0505,
"step": 725
},
{
"epoch": 0.7635983263598326,
"grad_norm": 0.5104098596156186,
"learning_rate": 3.2187016017753714e-06,
"loss": 1.066,
"step": 730
},
{
"epoch": 0.7688284518828452,
"grad_norm": 0.5171571509002556,
"learning_rate": 3.085602479494735e-06,
"loss": 1.0602,
"step": 735
},
{
"epoch": 0.7740585774058577,
"grad_norm": 0.5015256263302731,
"learning_rate": 2.954810025379633e-06,
"loss": 1.071,
"step": 740
},
{
"epoch": 0.7792887029288703,
"grad_norm": 0.5031988802471105,
"learning_rate": 2.826367872268543e-06,
"loss": 1.0517,
"step": 745
},
{
"epoch": 0.7845188284518828,
"grad_norm": 0.5291480112168645,
"learning_rate": 2.700318868930977e-06,
"loss": 1.0683,
"step": 750
},
{
"epoch": 0.7897489539748954,
"grad_norm": 0.5301814683256604,
"learning_rate": 2.576705065772985e-06,
"loss": 1.0543,
"step": 755
},
{
"epoch": 0.7949790794979079,
"grad_norm": 0.4858932325034876,
"learning_rate": 2.455567700808974e-06,
"loss": 1.0437,
"step": 760
},
{
"epoch": 0.8002092050209205,
"grad_norm": 0.5226986538780068,
"learning_rate": 2.3369471859045558e-06,
"loss": 1.0637,
"step": 765
},
{
"epoch": 0.805439330543933,
"grad_norm": 0.5018373716025176,
"learning_rate": 2.2208830932950175e-06,
"loss": 1.0558,
"step": 770
},
{
"epoch": 0.8106694560669456,
"grad_norm": 0.5012264173655667,
"learning_rate": 2.107414142383857e-06,
"loss": 1.0636,
"step": 775
},
{
"epoch": 0.8158995815899581,
"grad_norm": 0.5329303605558987,
"learning_rate": 1.996578186825876e-06,
"loss": 1.0652,
"step": 780
},
{
"epoch": 0.8211297071129707,
"grad_norm": 0.5063762746528606,
"learning_rate": 1.8884122018990292e-06,
"loss": 1.0574,
"step": 785
},
{
"epoch": 0.8263598326359832,
"grad_norm": 0.507270043815261,
"learning_rate": 1.7829522721693738e-06,
"loss": 1.0476,
"step": 790
},
{
"epoch": 0.8315899581589958,
"grad_norm": 0.524242600565246,
"learning_rate": 1.680233579453101e-06,
"loss": 1.0604,
"step": 795
},
{
"epoch": 0.8368200836820083,
"grad_norm": 0.5006885648328792,
"learning_rate": 1.5802903910797584e-06,
"loss": 1.0623,
"step": 800
},
{
"epoch": 0.8420502092050209,
"grad_norm": 0.4983602387356114,
"learning_rate": 1.483156048460561e-06,
"loss": 1.0702,
"step": 805
},
{
"epoch": 0.8472803347280334,
"grad_norm": 0.5141728007092325,
"learning_rate": 1.3888629559655497e-06,
"loss": 1.065,
"step": 810
},
{
"epoch": 0.852510460251046,
"grad_norm": 0.4977890298090738,
"learning_rate": 1.2974425701133941e-06,
"loss": 1.0519,
"step": 815
},
{
"epoch": 0.8577405857740585,
"grad_norm": 0.5073113010910021,
"learning_rate": 1.2089253890773789e-06,
"loss": 1.0592,
"step": 820
},
{
"epoch": 0.8629707112970711,
"grad_norm": 0.4919709083678304,
"learning_rate": 1.1233409425110952e-06,
"loss": 1.0562,
"step": 825
},
{
"epoch": 0.8682008368200836,
"grad_norm": 0.4868416396718501,
"learning_rate": 1.0407177816972558e-06,
"loss": 1.0469,
"step": 830
},
{
"epoch": 0.8734309623430963,
"grad_norm": 0.5022480244107099,
"learning_rate": 9.610834700228654e-07,
"loss": 1.0775,
"step": 835
},
{
"epoch": 0.8786610878661087,
"grad_norm": 0.5053021620216933,
"learning_rate": 8.844645737839874e-07,
"loss": 1.0567,
"step": 840
},
{
"epoch": 0.8838912133891214,
"grad_norm": 0.4754078550874839,
"learning_rate": 8.108866533231252e-07,
"loss": 1.0428,
"step": 845
},
{
"epoch": 0.8891213389121339,
"grad_norm": 0.5132996856432627,
"learning_rate": 7.403742545021986e-07,
"loss": 1.0559,
"step": 850
},
{
"epoch": 0.8943514644351465,
"grad_norm": 0.4962332804391809,
"learning_rate": 6.729509005139634e-07,
"loss": 1.0415,
"step": 855
},
{
"epoch": 0.899581589958159,
"grad_norm": 0.5018234838855186,
"learning_rate": 6.086390840345758e-07,
"loss": 1.0499,
"step": 860
},
{
"epoch": 0.9048117154811716,
"grad_norm": 0.49057781568767944,
"learning_rate": 5.474602597199607e-07,
"loss": 1.0704,
"step": 865
},
{
"epoch": 0.9100418410041841,
"grad_norm": 0.4970194272797173,
"learning_rate": 4.894348370484648e-07,
"loss": 1.0691,
"step": 870
},
{
"epoch": 0.9152719665271967,
"grad_norm": 0.5221499540872924,
"learning_rate": 4.34582173512168e-07,
"loss": 1.0676,
"step": 875
},
{
"epoch": 0.9205020920502092,
"grad_norm": 0.48846308721575726,
"learning_rate": 3.8292056815916965e-07,
"loss": 1.059,
"step": 880
},
{
"epoch": 0.9257322175732218,
"grad_norm": 0.5300257475921499,
"learning_rate": 3.344672554889428e-07,
"loss": 1.0692,
"step": 885
},
{
"epoch": 0.9309623430962343,
"grad_norm": 0.4959679907039524,
"learning_rate": 2.8923839970285473e-07,
"loss": 1.0592,
"step": 890
},
{
"epoch": 0.9361924686192469,
"grad_norm": 0.49081091342104494,
"learning_rate": 2.472490893117274e-07,
"loss": 1.0458,
"step": 895
},
{
"epoch": 0.9414225941422594,
"grad_norm": 0.5007780005543897,
"learning_rate": 2.0851333210225032e-07,
"loss": 1.0608,
"step": 900
},
{
"epoch": 0.946652719665272,
"grad_norm": 0.48329063115650944,
"learning_rate": 1.730440504639408e-07,
"loss": 1.0579,
"step": 905
},
{
"epoch": 0.9518828451882845,
"grad_norm": 0.48895599668986894,
"learning_rate": 1.408530770781813e-07,
"loss": 1.0527,
"step": 910
},
{
"epoch": 0.9571129707112971,
"grad_norm": 0.4977013333272583,
"learning_rate": 1.1195115097079268e-07,
"loss": 1.0685,
"step": 915
},
{
"epoch": 0.9623430962343096,
"grad_norm": 0.48972695236839464,
"learning_rate": 8.634791392946429e-08,
"loss": 1.0677,
"step": 920
},
{
"epoch": 0.9675732217573222,
"grad_norm": 0.4907567001368972,
"learning_rate": 6.405190728721033e-08,
"loss": 1.0455,
"step": 925
},
{
"epoch": 0.9728033472803347,
"grad_norm": 0.4772333774679202,
"learning_rate": 4.5070569072952485e-08,
"loss": 1.061,
"step": 930
},
{
"epoch": 0.9780334728033473,
"grad_norm": 0.49589302731265184,
"learning_rate": 2.9410231530168087e-08,
"loss": 1.0499,
"step": 935
},
{
"epoch": 0.9832635983263598,
"grad_norm": 0.47194869812634765,
"learning_rate": 1.7076119004429958e-08,
"loss": 1.0763,
"step": 940
},
{
"epoch": 0.9884937238493724,
"grad_norm": 0.4944519541688841,
"learning_rate": 8.072346200544979e-09,
"loss": 1.0672,
"step": 945
},
{
"epoch": 0.9937238493723849,
"grad_norm": 0.4887018274547105,
"learning_rate": 2.401916809872118e-09,
"loss": 1.0628,
"step": 950
},
{
"epoch": 0.9989539748953975,
"grad_norm": 0.47672135170607305,
"learning_rate": 6.672250828620996e-11,
"loss": 1.0627,
"step": 955
},
{
"epoch": 1.0,
"eval_loss": 1.064562439918518,
"eval_runtime": 286.7813,
"eval_samples_per_second": 47.21,
"eval_steps_per_second": 0.739,
"step": 956
},
{
"epoch": 1.0,
"step": 956,
"total_flos": 500925122740224.0,
"train_loss": 1.0766765792000743,
"train_runtime": 10814.9645,
"train_samples_per_second": 11.312,
"train_steps_per_second": 0.088
}
],
"logging_steps": 5,
"max_steps": 956,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 500925122740224.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}