zephyr-7b-sft-full / trainer_state.json
yuzhe-123's picture
Model save
d474d22 verified
raw
history blame
39.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1090,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0009174311926605505,
"grad_norm": 8.567075219673832,
"learning_rate": 1.8348623853211012e-07,
"loss": 1.1366,
"step": 1
},
{
"epoch": 0.0045871559633027525,
"grad_norm": 7.502231595297741,
"learning_rate": 9.174311926605506e-07,
"loss": 1.1311,
"step": 5
},
{
"epoch": 0.009174311926605505,
"grad_norm": 4.041886291445951,
"learning_rate": 1.8348623853211011e-06,
"loss": 1.0455,
"step": 10
},
{
"epoch": 0.013761467889908258,
"grad_norm": 2.66931200242608,
"learning_rate": 2.7522935779816517e-06,
"loss": 1.0276,
"step": 15
},
{
"epoch": 0.01834862385321101,
"grad_norm": 2.655264853297466,
"learning_rate": 3.6697247706422022e-06,
"loss": 0.996,
"step": 20
},
{
"epoch": 0.022935779816513763,
"grad_norm": 2.0824902429417502,
"learning_rate": 4.587155963302753e-06,
"loss": 0.9943,
"step": 25
},
{
"epoch": 0.027522935779816515,
"grad_norm": 2.223113856453528,
"learning_rate": 5.504587155963303e-06,
"loss": 0.9914,
"step": 30
},
{
"epoch": 0.03211009174311927,
"grad_norm": 2.8245029253119083,
"learning_rate": 6.422018348623854e-06,
"loss": 0.9877,
"step": 35
},
{
"epoch": 0.03669724770642202,
"grad_norm": 1.9055889003234818,
"learning_rate": 7.3394495412844045e-06,
"loss": 0.9948,
"step": 40
},
{
"epoch": 0.04128440366972477,
"grad_norm": 2.2824223465177687,
"learning_rate": 8.256880733944956e-06,
"loss": 0.9779,
"step": 45
},
{
"epoch": 0.045871559633027525,
"grad_norm": 2.238766365680375,
"learning_rate": 9.174311926605506e-06,
"loss": 0.9842,
"step": 50
},
{
"epoch": 0.05045871559633028,
"grad_norm": 2.3168005986367777,
"learning_rate": 1.0091743119266055e-05,
"loss": 0.9935,
"step": 55
},
{
"epoch": 0.05504587155963303,
"grad_norm": 2.0671072791875775,
"learning_rate": 1.1009174311926607e-05,
"loss": 0.9973,
"step": 60
},
{
"epoch": 0.05963302752293578,
"grad_norm": 2.3870937792624987,
"learning_rate": 1.1926605504587156e-05,
"loss": 0.9812,
"step": 65
},
{
"epoch": 0.06422018348623854,
"grad_norm": 2.872503228286113,
"learning_rate": 1.2844036697247708e-05,
"loss": 0.9866,
"step": 70
},
{
"epoch": 0.06880733944954129,
"grad_norm": 2.918010569550636,
"learning_rate": 1.3761467889908258e-05,
"loss": 1.0001,
"step": 75
},
{
"epoch": 0.07339449541284404,
"grad_norm": 3.0830710791214773,
"learning_rate": 1.4678899082568809e-05,
"loss": 0.9886,
"step": 80
},
{
"epoch": 0.0779816513761468,
"grad_norm": 2.6063312280690574,
"learning_rate": 1.559633027522936e-05,
"loss": 0.9927,
"step": 85
},
{
"epoch": 0.08256880733944955,
"grad_norm": 2.209724380985353,
"learning_rate": 1.6513761467889912e-05,
"loss": 1.0107,
"step": 90
},
{
"epoch": 0.0871559633027523,
"grad_norm": 2.76948096722416,
"learning_rate": 1.743119266055046e-05,
"loss": 1.0185,
"step": 95
},
{
"epoch": 0.09174311926605505,
"grad_norm": 2.1716697491458707,
"learning_rate": 1.834862385321101e-05,
"loss": 0.9981,
"step": 100
},
{
"epoch": 0.0963302752293578,
"grad_norm": 2.1757317389304482,
"learning_rate": 1.9266055045871563e-05,
"loss": 1.014,
"step": 105
},
{
"epoch": 0.10091743119266056,
"grad_norm": 2.5856692970053055,
"learning_rate": 1.999994872196626e-05,
"loss": 1.0231,
"step": 110
},
{
"epoch": 0.10550458715596331,
"grad_norm": 2.0176395767985897,
"learning_rate": 1.9998154046002822e-05,
"loss": 1.005,
"step": 115
},
{
"epoch": 0.11009174311926606,
"grad_norm": 1.9369629064989813,
"learning_rate": 1.999379599421534e-05,
"loss": 1.0051,
"step": 120
},
{
"epoch": 0.11467889908256881,
"grad_norm": 1.9803383518176758,
"learning_rate": 1.9986875683942535e-05,
"loss": 1.0137,
"step": 125
},
{
"epoch": 0.11926605504587157,
"grad_norm": 1.9751164482121784,
"learning_rate": 1.9977394889447526e-05,
"loss": 1.0125,
"step": 130
},
{
"epoch": 0.12385321100917432,
"grad_norm": 1.766026099400838,
"learning_rate": 1.9965356041462954e-05,
"loss": 0.993,
"step": 135
},
{
"epoch": 0.12844036697247707,
"grad_norm": 1.8456571124434127,
"learning_rate": 1.9950762226567783e-05,
"loss": 1.0208,
"step": 140
},
{
"epoch": 0.13302752293577982,
"grad_norm": 2.2558886400833136,
"learning_rate": 1.9933617186395917e-05,
"loss": 1.0273,
"step": 145
},
{
"epoch": 0.13761467889908258,
"grad_norm": 2.0067810742214367,
"learning_rate": 1.9913925316676946e-05,
"loss": 1.0214,
"step": 150
},
{
"epoch": 0.14220183486238533,
"grad_norm": 2.0354769519738696,
"learning_rate": 1.9891691666109112e-05,
"loss": 1.0449,
"step": 155
},
{
"epoch": 0.14678899082568808,
"grad_norm": 1.9853898034162918,
"learning_rate": 1.9866921935064907e-05,
"loss": 1.0348,
"step": 160
},
{
"epoch": 0.15137614678899083,
"grad_norm": 1.8771707161324125,
"learning_rate": 1.9839622474129595e-05,
"loss": 1.0447,
"step": 165
},
{
"epoch": 0.1559633027522936,
"grad_norm": 1.7555490053031533,
"learning_rate": 1.9809800282473014e-05,
"loss": 1.0319,
"step": 170
},
{
"epoch": 0.16055045871559634,
"grad_norm": 3.813249317869119,
"learning_rate": 1.977746300605507e-05,
"loss": 1.0432,
"step": 175
},
{
"epoch": 0.1651376146788991,
"grad_norm": 2.932842350036964,
"learning_rate": 1.9742618935665478e-05,
"loss": 1.0337,
"step": 180
},
{
"epoch": 0.16972477064220184,
"grad_norm": 1.9490193595251735,
"learning_rate": 1.9705277004798072e-05,
"loss": 1.0354,
"step": 185
},
{
"epoch": 0.1743119266055046,
"grad_norm": 2.0315194096887073,
"learning_rate": 1.9665446787360444e-05,
"loss": 1.0381,
"step": 190
},
{
"epoch": 0.17889908256880735,
"grad_norm": 2.01041642690734,
"learning_rate": 1.9623138495219292e-05,
"loss": 1.0176,
"step": 195
},
{
"epoch": 0.1834862385321101,
"grad_norm": 1.8379234276601883,
"learning_rate": 1.957836297558229e-05,
"loss": 1.0179,
"step": 200
},
{
"epoch": 0.18807339449541285,
"grad_norm": 1.8383111439189317,
"learning_rate": 1.9531131708217005e-05,
"loss": 1.023,
"step": 205
},
{
"epoch": 0.1926605504587156,
"grad_norm": 1.7211278520137383,
"learning_rate": 1.948145680250766e-05,
"loss": 1.0201,
"step": 210
},
{
"epoch": 0.19724770642201836,
"grad_norm": 1.6488508269613613,
"learning_rate": 1.9429350994350483e-05,
"loss": 1.0337,
"step": 215
},
{
"epoch": 0.2018348623853211,
"grad_norm": 1.6195453176201335,
"learning_rate": 1.93748276428884e-05,
"loss": 1.041,
"step": 220
},
{
"epoch": 0.20642201834862386,
"grad_norm": 1.6631810770290123,
"learning_rate": 1.931790072708596e-05,
"loss": 1.0253,
"step": 225
},
{
"epoch": 0.21100917431192662,
"grad_norm": 1.682342166301177,
"learning_rate": 1.9258584842145342e-05,
"loss": 1.0446,
"step": 230
},
{
"epoch": 0.21559633027522937,
"grad_norm": 1.9113148172207652,
"learning_rate": 1.9196895195764363e-05,
"loss": 1.0249,
"step": 235
},
{
"epoch": 0.22018348623853212,
"grad_norm": 1.921572422455064,
"learning_rate": 1.913284760423745e-05,
"loss": 1.0289,
"step": 240
},
{
"epoch": 0.22477064220183487,
"grad_norm": 1.6160022720196747,
"learning_rate": 1.9066458488400586e-05,
"loss": 1.0281,
"step": 245
},
{
"epoch": 0.22935779816513763,
"grad_norm": 1.546114425823443,
"learning_rate": 1.8997744869421248e-05,
"loss": 1.0237,
"step": 250
},
{
"epoch": 0.23394495412844038,
"grad_norm": 1.6113177148502253,
"learning_rate": 1.8926724364434447e-05,
"loss": 1.0159,
"step": 255
},
{
"epoch": 0.23853211009174313,
"grad_norm": 1.756084810660972,
"learning_rate": 1.8853415182025953e-05,
"loss": 1.03,
"step": 260
},
{
"epoch": 0.24311926605504589,
"grad_norm": 1.7722580062413746,
"learning_rate": 1.8777836117563894e-05,
"loss": 1.0268,
"step": 265
},
{
"epoch": 0.24770642201834864,
"grad_norm": 1.6552997587829843,
"learning_rate": 1.8700006548379898e-05,
"loss": 1.0214,
"step": 270
},
{
"epoch": 0.25229357798165136,
"grad_norm": 1.7554082828396729,
"learning_rate": 1.861994642880105e-05,
"loss": 1.0411,
"step": 275
},
{
"epoch": 0.25688073394495414,
"grad_norm": 1.6101850875118928,
"learning_rate": 1.8537676285033886e-05,
"loss": 1.0289,
"step": 280
},
{
"epoch": 0.26146788990825687,
"grad_norm": 1.6754749206063948,
"learning_rate": 1.845321720990181e-05,
"loss": 1.03,
"step": 285
},
{
"epoch": 0.26605504587155965,
"grad_norm": 1.5989084659148887,
"learning_rate": 1.8366590857437182e-05,
"loss": 1.0332,
"step": 290
},
{
"epoch": 0.2706422018348624,
"grad_norm": 1.7373362181608558,
"learning_rate": 1.8277819437329577e-05,
"loss": 1.0196,
"step": 295
},
{
"epoch": 0.27522935779816515,
"grad_norm": 1.6415478980043552,
"learning_rate": 1.8186925709231534e-05,
"loss": 1.0331,
"step": 300
},
{
"epoch": 0.2798165137614679,
"grad_norm": 1.595584297931999,
"learning_rate": 1.809393297692334e-05,
"loss": 1.0257,
"step": 305
},
{
"epoch": 0.28440366972477066,
"grad_norm": 1.8203998942508204,
"learning_rate": 1.799886508233829e-05,
"loss": 1.044,
"step": 310
},
{
"epoch": 0.2889908256880734,
"grad_norm": 1.6518600023400687,
"learning_rate": 1.790174639944997e-05,
"loss": 1.0387,
"step": 315
},
{
"epoch": 0.29357798165137616,
"grad_norm": 1.7796790607246211,
"learning_rate": 1.780260182802314e-05,
"loss": 1.0261,
"step": 320
},
{
"epoch": 0.2981651376146789,
"grad_norm": 1.6534133312284494,
"learning_rate": 1.7701456787229805e-05,
"loss": 1.0338,
"step": 325
},
{
"epoch": 0.30275229357798167,
"grad_norm": 1.5254168104289345,
"learning_rate": 1.7598337209132142e-05,
"loss": 1.0155,
"step": 330
},
{
"epoch": 0.3073394495412844,
"grad_norm": 1.5916672455155843,
"learning_rate": 1.7493269532033882e-05,
"loss": 1.0181,
"step": 335
},
{
"epoch": 0.3119266055045872,
"grad_norm": 2.469796014086296,
"learning_rate": 1.738628069370195e-05,
"loss": 1.0291,
"step": 340
},
{
"epoch": 0.3165137614678899,
"grad_norm": 1.7002430789318252,
"learning_rate": 1.7277398124460022e-05,
"loss": 1.0261,
"step": 345
},
{
"epoch": 0.3211009174311927,
"grad_norm": 1.6452789420112204,
"learning_rate": 1.71666497401558e-05,
"loss": 0.9936,
"step": 350
},
{
"epoch": 0.3256880733944954,
"grad_norm": 1.6090081655393613,
"learning_rate": 1.7054063935003813e-05,
"loss": 1.0253,
"step": 355
},
{
"epoch": 0.3302752293577982,
"grad_norm": 1.6521899111347487,
"learning_rate": 1.6939669574305565e-05,
"loss": 1.0076,
"step": 360
},
{
"epoch": 0.3348623853211009,
"grad_norm": 1.5087732073233964,
"learning_rate": 1.6823495987048922e-05,
"loss": 1.0098,
"step": 365
},
{
"epoch": 0.3394495412844037,
"grad_norm": 1.4806406906585232,
"learning_rate": 1.6705572958388576e-05,
"loss": 1.01,
"step": 370
},
{
"epoch": 0.3440366972477064,
"grad_norm": 1.6035884549602302,
"learning_rate": 1.6585930722009602e-05,
"loss": 1.0335,
"step": 375
},
{
"epoch": 0.3486238532110092,
"grad_norm": 1.5314286198965614,
"learning_rate": 1.6464599952375998e-05,
"loss": 1.0227,
"step": 380
},
{
"epoch": 0.3532110091743119,
"grad_norm": 1.630708086376815,
"learning_rate": 1.63416117568662e-05,
"loss": 1.0186,
"step": 385
},
{
"epoch": 0.3577981651376147,
"grad_norm": 1.484014704538301,
"learning_rate": 1.621699766779763e-05,
"loss": 1.0141,
"step": 390
},
{
"epoch": 0.3623853211009174,
"grad_norm": 1.489313766253633,
"learning_rate": 1.6090789634342278e-05,
"loss": 1.0049,
"step": 395
},
{
"epoch": 0.3669724770642202,
"grad_norm": 1.5152327147526468,
"learning_rate": 1.5963020014335437e-05,
"loss": 1.0105,
"step": 400
},
{
"epoch": 0.37155963302752293,
"grad_norm": 1.4449410389905726,
"learning_rate": 1.583372156597961e-05,
"loss": 1.0031,
"step": 405
},
{
"epoch": 0.3761467889908257,
"grad_norm": 1.4506061967201511,
"learning_rate": 1.570292743944583e-05,
"loss": 1.002,
"step": 410
},
{
"epoch": 0.38073394495412843,
"grad_norm": 1.428745280078687,
"learning_rate": 1.557067116837444e-05,
"loss": 1.0223,
"step": 415
},
{
"epoch": 0.3853211009174312,
"grad_norm": 1.4567183232187022,
"learning_rate": 1.5436986661277578e-05,
"loss": 1.0133,
"step": 420
},
{
"epoch": 0.38990825688073394,
"grad_norm": 1.6991055564741833,
"learning_rate": 1.530190819284555e-05,
"loss": 1.0038,
"step": 425
},
{
"epoch": 0.3944954128440367,
"grad_norm": 1.5791854852416984,
"learning_rate": 1.5165470395159314e-05,
"loss": 1.0335,
"step": 430
},
{
"epoch": 0.39908256880733944,
"grad_norm": 1.6518586876063224,
"learning_rate": 1.5027708248811331e-05,
"loss": 1.0006,
"step": 435
},
{
"epoch": 0.4036697247706422,
"grad_norm": 1.37950132663011,
"learning_rate": 1.4888657073937077e-05,
"loss": 1.009,
"step": 440
},
{
"epoch": 0.40825688073394495,
"grad_norm": 1.4704035924378334,
"learning_rate": 1.4748352521159492e-05,
"loss": 0.9958,
"step": 445
},
{
"epoch": 0.41284403669724773,
"grad_norm": 1.4596704904359323,
"learning_rate": 1.4606830562448692e-05,
"loss": 1.005,
"step": 450
},
{
"epoch": 0.41743119266055045,
"grad_norm": 1.4454837075407818,
"learning_rate": 1.4464127481899312e-05,
"loss": 1.0073,
"step": 455
},
{
"epoch": 0.42201834862385323,
"grad_norm": 1.4689484755348563,
"learning_rate": 1.4320279866427798e-05,
"loss": 1.0064,
"step": 460
},
{
"epoch": 0.42660550458715596,
"grad_norm": 1.485956826048098,
"learning_rate": 1.4175324596392075e-05,
"loss": 1.0203,
"step": 465
},
{
"epoch": 0.43119266055045874,
"grad_norm": 1.573521513583627,
"learning_rate": 1.402929883613599e-05,
"loss": 1.0074,
"step": 470
},
{
"epoch": 0.43577981651376146,
"grad_norm": 1.4783455247528414,
"learning_rate": 1.3882240024460928e-05,
"loss": 0.9925,
"step": 475
},
{
"epoch": 0.44036697247706424,
"grad_norm": 1.4099865921237311,
"learning_rate": 1.3734185865027061e-05,
"loss": 1.0105,
"step": 480
},
{
"epoch": 0.44495412844036697,
"grad_norm": 1.4586347859412025,
"learning_rate": 1.358517431668672e-05,
"loss": 0.9992,
"step": 485
},
{
"epoch": 0.44954128440366975,
"grad_norm": 1.3606557347465276,
"learning_rate": 1.3435243583752294e-05,
"loss": 0.9936,
"step": 490
},
{
"epoch": 0.4541284403669725,
"grad_norm": 1.423923281806998,
"learning_rate": 1.3284432106201233e-05,
"loss": 0.9879,
"step": 495
},
{
"epoch": 0.45871559633027525,
"grad_norm": 1.5219817365897736,
"learning_rate": 1.313277854982062e-05,
"loss": 1.0079,
"step": 500
},
{
"epoch": 0.463302752293578,
"grad_norm": 1.6212946711670189,
"learning_rate": 1.2980321796293838e-05,
"loss": 0.987,
"step": 505
},
{
"epoch": 0.46788990825688076,
"grad_norm": 1.4932426551096094,
"learning_rate": 1.2827100933231904e-05,
"loss": 1.0052,
"step": 510
},
{
"epoch": 0.4724770642201835,
"grad_norm": 1.5771080120985452,
"learning_rate": 1.2673155244151985e-05,
"loss": 1.0052,
"step": 515
},
{
"epoch": 0.47706422018348627,
"grad_norm": 1.5047726515081563,
"learning_rate": 1.2518524198405699e-05,
"loss": 0.9992,
"step": 520
},
{
"epoch": 0.481651376146789,
"grad_norm": 1.5275724292638637,
"learning_rate": 1.2363247441059775e-05,
"loss": 0.9852,
"step": 525
},
{
"epoch": 0.48623853211009177,
"grad_norm": 1.4373779721892144,
"learning_rate": 1.2207364782731657e-05,
"loss": 1.0087,
"step": 530
},
{
"epoch": 0.4908256880733945,
"grad_norm": 1.4497248778716865,
"learning_rate": 1.2050916189382646e-05,
"loss": 0.9732,
"step": 535
},
{
"epoch": 0.4954128440366973,
"grad_norm": 1.4508227174593424,
"learning_rate": 1.189394177207125e-05,
"loss": 0.98,
"step": 540
},
{
"epoch": 0.5,
"grad_norm": 1.3723993793012572,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.9873,
"step": 545
},
{
"epoch": 0.5045871559633027,
"grad_norm": 1.400736917075807,
"learning_rate": 1.1578576573543541e-05,
"loss": 0.964,
"step": 550
},
{
"epoch": 0.5091743119266054,
"grad_norm": 1.4328329143842218,
"learning_rate": 1.1420266647205232e-05,
"loss": 0.9744,
"step": 555
},
{
"epoch": 0.5137614678899083,
"grad_norm": 1.4005342497275977,
"learning_rate": 1.1261592585930576e-05,
"loss": 0.9845,
"step": 560
},
{
"epoch": 0.518348623853211,
"grad_norm": 1.3870143214943034,
"learning_rate": 1.1102595071354471e-05,
"loss": 0.9849,
"step": 565
},
{
"epoch": 0.5229357798165137,
"grad_norm": 1.4037256110851137,
"learning_rate": 1.0943314868040365e-05,
"loss": 0.978,
"step": 570
},
{
"epoch": 0.5275229357798165,
"grad_norm": 1.4524357204964615,
"learning_rate": 1.0783792813028828e-05,
"loss": 0.9883,
"step": 575
},
{
"epoch": 0.5321100917431193,
"grad_norm": 1.4168006943303706,
"learning_rate": 1.0624069805367558e-05,
"loss": 0.9694,
"step": 580
},
{
"epoch": 0.536697247706422,
"grad_norm": 1.3868495069798417,
"learning_rate": 1.0464186795625481e-05,
"loss": 0.9722,
"step": 585
},
{
"epoch": 0.5412844036697247,
"grad_norm": 1.484113669184668,
"learning_rate": 1.0304184775393642e-05,
"loss": 0.9858,
"step": 590
},
{
"epoch": 0.5458715596330275,
"grad_norm": 1.4378478721351533,
"learning_rate": 1.0144104766775574e-05,
"loss": 0.9688,
"step": 595
},
{
"epoch": 0.5504587155963303,
"grad_norm": 1.3497763152773319,
"learning_rate": 9.983987811869863e-06,
"loss": 0.9619,
"step": 600
},
{
"epoch": 0.555045871559633,
"grad_norm": 1.3638902119410679,
"learning_rate": 9.823874962247565e-06,
"loss": 0.9869,
"step": 605
},
{
"epoch": 0.5596330275229358,
"grad_norm": 1.4727242463849317,
"learning_rate": 9.663807268427197e-06,
"loss": 0.9497,
"step": 610
},
{
"epoch": 0.5642201834862385,
"grad_norm": 1.3817944954956538,
"learning_rate": 9.503825769350016e-06,
"loss": 0.9759,
"step": 615
},
{
"epoch": 0.5688073394495413,
"grad_norm": 1.4097103167722767,
"learning_rate": 9.343971481858246e-06,
"loss": 0.9721,
"step": 620
},
{
"epoch": 0.573394495412844,
"grad_norm": 1.3297950318079086,
"learning_rate": 9.184285390178978e-06,
"loss": 0.9679,
"step": 625
},
{
"epoch": 0.5779816513761468,
"grad_norm": 1.3397331767173486,
"learning_rate": 9.024808435416435e-06,
"loss": 0.9555,
"step": 630
},
{
"epoch": 0.5825688073394495,
"grad_norm": 1.3793546519569164,
"learning_rate": 8.865581505055292e-06,
"loss": 0.9657,
"step": 635
},
{
"epoch": 0.5871559633027523,
"grad_norm": 1.3699122633364083,
"learning_rate": 8.706645422477739e-06,
"loss": 0.9611,
"step": 640
},
{
"epoch": 0.591743119266055,
"grad_norm": 1.652929217335104,
"learning_rate": 8.548040936496989e-06,
"loss": 0.9656,
"step": 645
},
{
"epoch": 0.5963302752293578,
"grad_norm": 1.4000463163388799,
"learning_rate": 8.389808710909881e-06,
"loss": 0.9689,
"step": 650
},
{
"epoch": 0.6009174311926605,
"grad_norm": 1.3744642974899226,
"learning_rate": 8.231989314071318e-06,
"loss": 0.9675,
"step": 655
},
{
"epoch": 0.6055045871559633,
"grad_norm": 1.367828783708881,
"learning_rate": 8.07462320849313e-06,
"loss": 0.9827,
"step": 660
},
{
"epoch": 0.6100917431192661,
"grad_norm": 1.3746153840246962,
"learning_rate": 7.917750740470116e-06,
"loss": 0.9612,
"step": 665
},
{
"epoch": 0.6146788990825688,
"grad_norm": 1.3634420428996064,
"learning_rate": 7.761412129735853e-06,
"loss": 0.9606,
"step": 670
},
{
"epoch": 0.6192660550458715,
"grad_norm": 1.365419016371242,
"learning_rate": 7.605647459150961e-06,
"loss": 0.9701,
"step": 675
},
{
"epoch": 0.6238532110091743,
"grad_norm": 1.2994526356261897,
"learning_rate": 7.4504966644264775e-06,
"loss": 0.9496,
"step": 680
},
{
"epoch": 0.6284403669724771,
"grad_norm": 1.3036192585622581,
"learning_rate": 7.295999523884921e-06,
"loss": 0.9603,
"step": 685
},
{
"epoch": 0.6330275229357798,
"grad_norm": 1.3656479207440784,
"learning_rate": 7.142195648261747e-06,
"loss": 0.971,
"step": 690
},
{
"epoch": 0.6376146788990825,
"grad_norm": 1.3493578112863471,
"learning_rate": 6.989124470549746e-06,
"loss": 0.9401,
"step": 695
},
{
"epoch": 0.6422018348623854,
"grad_norm": 1.321094455754546,
"learning_rate": 6.83682523588902e-06,
"loss": 0.9632,
"step": 700
},
{
"epoch": 0.6467889908256881,
"grad_norm": 1.3430854963278231,
"learning_rate": 6.685336991505122e-06,
"loss": 0.9648,
"step": 705
},
{
"epoch": 0.6513761467889908,
"grad_norm": 1.295245677177454,
"learning_rate": 6.5346985766979384e-06,
"loss": 0.9547,
"step": 710
},
{
"epoch": 0.6559633027522935,
"grad_norm": 1.3017382650713638,
"learning_rate": 6.384948612883872e-06,
"loss": 0.9597,
"step": 715
},
{
"epoch": 0.6605504587155964,
"grad_norm": 1.2936500217660096,
"learning_rate": 6.2361254936939e-06,
"loss": 0.9603,
"step": 720
},
{
"epoch": 0.6651376146788991,
"grad_norm": 1.325710868385086,
"learning_rate": 6.0882673751300235e-06,
"loss": 0.9476,
"step": 725
},
{
"epoch": 0.6697247706422018,
"grad_norm": 1.3419334755973713,
"learning_rate": 5.941412165782645e-06,
"loss": 0.9565,
"step": 730
},
{
"epoch": 0.6743119266055045,
"grad_norm": 1.437071920630301,
"learning_rate": 5.79559751711138e-06,
"loss": 0.9479,
"step": 735
},
{
"epoch": 0.6788990825688074,
"grad_norm": 1.3574467430871946,
"learning_rate": 5.650860813791786e-06,
"loss": 0.9485,
"step": 740
},
{
"epoch": 0.6834862385321101,
"grad_norm": 1.3767014536400128,
"learning_rate": 5.507239164130501e-06,
"loss": 0.9492,
"step": 745
},
{
"epoch": 0.6880733944954128,
"grad_norm": 1.3706065876046587,
"learning_rate": 5.364769390551225e-06,
"loss": 0.9678,
"step": 750
},
{
"epoch": 0.6926605504587156,
"grad_norm": 1.295329148997341,
"learning_rate": 5.223488020154028e-06,
"loss": 0.9597,
"step": 755
},
{
"epoch": 0.6972477064220184,
"grad_norm": 1.317633111030153,
"learning_rate": 5.083431275350312e-06,
"loss": 0.9255,
"step": 760
},
{
"epoch": 0.7018348623853211,
"grad_norm": 1.296366909205948,
"learning_rate": 4.9446350645759885e-06,
"loss": 0.9418,
"step": 765
},
{
"epoch": 0.7064220183486238,
"grad_norm": 1.3869599027709183,
"learning_rate": 4.807134973085036e-06,
"loss": 0.9666,
"step": 770
},
{
"epoch": 0.7110091743119266,
"grad_norm": 1.2574891887163746,
"learning_rate": 4.670966253826027e-06,
"loss": 0.935,
"step": 775
},
{
"epoch": 0.7155963302752294,
"grad_norm": 1.3824364498830946,
"learning_rate": 4.53616381840377e-06,
"loss": 0.9468,
"step": 780
},
{
"epoch": 0.7201834862385321,
"grad_norm": 1.3301006886667626,
"learning_rate": 4.402762228128531e-06,
"loss": 0.9393,
"step": 785
},
{
"epoch": 0.7247706422018348,
"grad_norm": 1.3916604005630384,
"learning_rate": 4.270795685155001e-06,
"loss": 0.9478,
"step": 790
},
{
"epoch": 0.7293577981651376,
"grad_norm": 1.3065920210924657,
"learning_rate": 4.140298023713416e-06,
"loss": 0.9301,
"step": 795
},
{
"epoch": 0.7339449541284404,
"grad_norm": 1.3321418112780183,
"learning_rate": 4.0113027014349374e-06,
"loss": 0.9148,
"step": 800
},
{
"epoch": 0.7385321100917431,
"grad_norm": 1.3066885947900655,
"learning_rate": 3.883842790773647e-06,
"loss": 0.9574,
"step": 805
},
{
"epoch": 0.7431192660550459,
"grad_norm": 1.2788509584031447,
"learning_rate": 3.757950970527249e-06,
"loss": 0.9261,
"step": 810
},
{
"epoch": 0.7477064220183486,
"grad_norm": 1.2817821352009064,
"learning_rate": 3.633659517458736e-06,
"loss": 0.9248,
"step": 815
},
{
"epoch": 0.7522935779816514,
"grad_norm": 1.2551311767393507,
"learning_rate": 3.511000298021098e-06,
"loss": 0.9232,
"step": 820
},
{
"epoch": 0.7568807339449541,
"grad_norm": 1.3288724774154381,
"learning_rate": 3.39000476018726e-06,
"loss": 0.9266,
"step": 825
},
{
"epoch": 0.7614678899082569,
"grad_norm": 1.2641453349887841,
"learning_rate": 3.2707039253872796e-06,
"loss": 0.9231,
"step": 830
},
{
"epoch": 0.7660550458715596,
"grad_norm": 1.2984387481989768,
"learning_rate": 3.153128380554941e-06,
"loss": 0.9315,
"step": 835
},
{
"epoch": 0.7706422018348624,
"grad_norm": 1.2705124222427846,
"learning_rate": 3.037308270285709e-06,
"loss": 0.9231,
"step": 840
},
{
"epoch": 0.7752293577981652,
"grad_norm": 1.267466264071829,
"learning_rate": 2.923273289108115e-06,
"loss": 0.9411,
"step": 845
},
{
"epoch": 0.7798165137614679,
"grad_norm": 1.356820160448841,
"learning_rate": 2.8110526738705345e-06,
"loss": 0.924,
"step": 850
},
{
"epoch": 0.7844036697247706,
"grad_norm": 1.3040569720039672,
"learning_rate": 2.700675196245288e-06,
"loss": 0.9388,
"step": 855
},
{
"epoch": 0.7889908256880734,
"grad_norm": 1.2420186482046902,
"learning_rate": 2.592169155352031e-06,
"loss": 0.9421,
"step": 860
},
{
"epoch": 0.7935779816513762,
"grad_norm": 1.2616233171009335,
"learning_rate": 2.485562370502279e-06,
"loss": 0.9432,
"step": 865
},
{
"epoch": 0.7981651376146789,
"grad_norm": 1.3424259509025755,
"learning_rate": 2.3808821740669608e-06,
"loss": 0.9352,
"step": 870
},
{
"epoch": 0.8027522935779816,
"grad_norm": 1.322358943195577,
"learning_rate": 2.2781554044688015e-06,
"loss": 0.9379,
"step": 875
},
{
"epoch": 0.8073394495412844,
"grad_norm": 1.2847587486081231,
"learning_rate": 2.1774083993013715e-06,
"loss": 0.935,
"step": 880
},
{
"epoch": 0.8119266055045872,
"grad_norm": 1.2963379429813555,
"learning_rate": 2.0786669885765044e-06,
"loss": 0.9271,
"step": 885
},
{
"epoch": 0.8165137614678899,
"grad_norm": 1.2767056632273088,
"learning_rate": 1.981956488101898e-06,
"loss": 0.9402,
"step": 890
},
{
"epoch": 0.8211009174311926,
"grad_norm": 1.2867778108368306,
"learning_rate": 1.8873016929904942e-06,
"loss": 0.9421,
"step": 895
},
{
"epoch": 0.8256880733944955,
"grad_norm": 1.2551888718092907,
"learning_rate": 1.7947268713034128e-06,
"loss": 0.9119,
"step": 900
},
{
"epoch": 0.8302752293577982,
"grad_norm": 1.2420897547436003,
"learning_rate": 1.704255757827963e-06,
"loss": 0.9146,
"step": 905
},
{
"epoch": 0.8348623853211009,
"grad_norm": 1.2938041997903864,
"learning_rate": 1.6159115479924259e-06,
"loss": 0.9308,
"step": 910
},
{
"epoch": 0.8394495412844036,
"grad_norm": 1.3003180857551158,
"learning_rate": 1.529716891919074e-06,
"loss": 0.9396,
"step": 915
},
{
"epoch": 0.8440366972477065,
"grad_norm": 1.2759086944542821,
"learning_rate": 1.4456938886170413e-06,
"loss": 0.9109,
"step": 920
},
{
"epoch": 0.8486238532110092,
"grad_norm": 1.33319189540699,
"learning_rate": 1.3638640803164516e-06,
"loss": 0.9237,
"step": 925
},
{
"epoch": 0.8532110091743119,
"grad_norm": 1.2809646406448443,
"learning_rate": 1.2842484469453365e-06,
"loss": 0.914,
"step": 930
},
{
"epoch": 0.8577981651376146,
"grad_norm": 1.2918537502920435,
"learning_rate": 1.2068674007506787e-06,
"loss": 0.9286,
"step": 935
},
{
"epoch": 0.8623853211009175,
"grad_norm": 1.3494320187543551,
"learning_rate": 1.1317407810650372e-06,
"loss": 0.9264,
"step": 940
},
{
"epoch": 0.8669724770642202,
"grad_norm": 1.247741786729038,
"learning_rate": 1.0588878492200261e-06,
"loss": 0.9321,
"step": 945
},
{
"epoch": 0.8715596330275229,
"grad_norm": 1.2685992470376304,
"learning_rate": 9.883272836080116e-07,
"loss": 0.9199,
"step": 950
},
{
"epoch": 0.8761467889908257,
"grad_norm": 1.2982688058402074,
"learning_rate": 9.200771748932513e-07,
"loss": 0.9326,
"step": 955
},
{
"epoch": 0.8807339449541285,
"grad_norm": 1.3089054904372284,
"learning_rate": 8.541550213737171e-07,
"loss": 0.9114,
"step": 960
},
{
"epoch": 0.8853211009174312,
"grad_norm": 1.252254149851376,
"learning_rate": 7.905777244947954e-07,
"loss": 0.918,
"step": 965
},
{
"epoch": 0.8899082568807339,
"grad_norm": 1.2933393496044958,
"learning_rate": 7.293615845160196e-07,
"loss": 0.9288,
"step": 970
},
{
"epoch": 0.8944954128440367,
"grad_norm": 1.2618705359183393,
"learning_rate": 6.705222963319191e-07,
"loss": 0.9376,
"step": 975
},
{
"epoch": 0.8990825688073395,
"grad_norm": 1.2592063070116135,
"learning_rate": 6.140749454480932e-07,
"loss": 0.9116,
"step": 980
},
{
"epoch": 0.9036697247706422,
"grad_norm": 1.2987200032177229,
"learning_rate": 5.600340041135133e-07,
"loss": 0.9181,
"step": 985
},
{
"epoch": 0.908256880733945,
"grad_norm": 1.300028729042509,
"learning_rate": 5.0841332761005e-07,
"loss": 0.9333,
"step": 990
},
{
"epoch": 0.9128440366972477,
"grad_norm": 1.178743058473968,
"learning_rate": 4.592261507001994e-07,
"loss": 0.9145,
"step": 995
},
{
"epoch": 0.9174311926605505,
"grad_norm": 1.2612257727695169,
"learning_rate": 4.124850842338779e-07,
"loss": 0.9198,
"step": 1000
},
{
"epoch": 0.9220183486238532,
"grad_norm": 1.2664257049420728,
"learning_rate": 3.6820211191520127e-07,
"loss": 0.9237,
"step": 1005
},
{
"epoch": 0.926605504587156,
"grad_norm": 1.2636645907476838,
"learning_rate": 3.263885872300343e-07,
"loss": 0.9228,
"step": 1010
},
{
"epoch": 0.9311926605504587,
"grad_norm": 1.242318963263546,
"learning_rate": 2.870552305351382e-07,
"loss": 0.9005,
"step": 1015
},
{
"epoch": 0.9357798165137615,
"grad_norm": 1.2809884089732457,
"learning_rate": 2.5021212630962246e-07,
"loss": 0.9276,
"step": 1020
},
{
"epoch": 0.9403669724770642,
"grad_norm": 1.2119069638655617,
"learning_rate": 2.158687205694443e-07,
"loss": 0.907,
"step": 1025
},
{
"epoch": 0.944954128440367,
"grad_norm": 1.2376811762313147,
"learning_rate": 1.840338184455881e-07,
"loss": 0.9269,
"step": 1030
},
{
"epoch": 0.9495412844036697,
"grad_norm": 1.2478362914541763,
"learning_rate": 1.5471558192656776e-07,
"loss": 0.9159,
"step": 1035
},
{
"epoch": 0.9541284403669725,
"grad_norm": 1.2217359410822013,
"learning_rate": 1.279215277658097e-07,
"loss": 0.932,
"step": 1040
},
{
"epoch": 0.9587155963302753,
"grad_norm": 1.2656920108994318,
"learning_rate": 1.0365852555447642e-07,
"loss": 0.925,
"step": 1045
},
{
"epoch": 0.963302752293578,
"grad_norm": 1.2272545903225942,
"learning_rate": 8.19327959602012e-08,
"loss": 0.9346,
"step": 1050
},
{
"epoch": 0.9678899082568807,
"grad_norm": 1.2652144722057859,
"learning_rate": 6.274990913221035e-08,
"loss": 0.9288,
"step": 1055
},
{
"epoch": 0.9724770642201835,
"grad_norm": 1.2861426339061892,
"learning_rate": 4.6114783273213395e-08,
"loss": 0.9134,
"step": 1060
},
{
"epoch": 0.9770642201834863,
"grad_norm": 1.2468482937718688,
"learning_rate": 3.203168337845508e-08,
"loss": 0.9062,
"step": 1065
},
{
"epoch": 0.981651376146789,
"grad_norm": 1.2763891416814128,
"learning_rate": 2.05042201422323e-08,
"loss": 0.9156,
"step": 1070
},
{
"epoch": 0.9862385321100917,
"grad_norm": 1.2950677825635273,
"learning_rate": 1.1535349032167908e-08,
"loss": 0.8987,
"step": 1075
},
{
"epoch": 0.9908256880733946,
"grad_norm": 1.2554986306370373,
"learning_rate": 5.127369531473525e-09,
"loss": 0.9281,
"step": 1080
},
{
"epoch": 0.9954128440366973,
"grad_norm": 1.2713525280166649,
"learning_rate": 1.2819245493955746e-09,
"loss": 0.9239,
"step": 1085
},
{
"epoch": 1.0,
"grad_norm": 1.4188473493340228,
"learning_rate": 0.0,
"loss": 0.9203,
"step": 1090
},
{
"epoch": 1.0,
"eval_loss": 0.9367018342018127,
"eval_runtime": 316.9529,
"eval_samples_per_second": 48.685,
"eval_steps_per_second": 0.764,
"step": 1090
},
{
"epoch": 1.0,
"step": 1090,
"total_flos": 456447649382400.0,
"train_loss": 0.9783222812031387,
"train_runtime": 12140.5299,
"train_samples_per_second": 11.484,
"train_steps_per_second": 0.09
}
],
"logging_steps": 5,
"max_steps": 1090,
"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": false
},
"attributes": {}
}
},
"total_flos": 456447649382400.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}