864 lines
21 KiB
JSON
864 lines
21 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_steps": 1000,
|
||
|
|
"global_step": 1164,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.01718213058419244,
|
||
|
|
"grad_norm": 2.712413787841797,
|
||
|
|
"learning_rate": 1.5384615384615387e-06,
|
||
|
|
"loss": 0.9951,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03436426116838488,
|
||
|
|
"grad_norm": 1.447428584098816,
|
||
|
|
"learning_rate": 3.247863247863248e-06,
|
||
|
|
"loss": 0.8983,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05154639175257732,
|
||
|
|
"grad_norm": 0.9766889214515686,
|
||
|
|
"learning_rate": 4.957264957264958e-06,
|
||
|
|
"loss": 0.809,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06872852233676977,
|
||
|
|
"grad_norm": 0.8826571702957153,
|
||
|
|
"learning_rate": 6.666666666666667e-06,
|
||
|
|
"loss": 0.775,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0859106529209622,
|
||
|
|
"grad_norm": 0.8237293362617493,
|
||
|
|
"learning_rate": 8.376068376068377e-06,
|
||
|
|
"loss": 0.7402,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10309278350515463,
|
||
|
|
"grad_norm": 0.7995133399963379,
|
||
|
|
"learning_rate": 1.0085470085470086e-05,
|
||
|
|
"loss": 0.7305,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12027491408934708,
|
||
|
|
"grad_norm": 0.8882474303245544,
|
||
|
|
"learning_rate": 1.1794871794871796e-05,
|
||
|
|
"loss": 0.7005,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13745704467353953,
|
||
|
|
"grad_norm": 0.8641324043273926,
|
||
|
|
"learning_rate": 1.3504273504273506e-05,
|
||
|
|
"loss": 0.7067,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15463917525773196,
|
||
|
|
"grad_norm": 0.8109490275382996,
|
||
|
|
"learning_rate": 1.5213675213675214e-05,
|
||
|
|
"loss": 0.6664,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1718213058419244,
|
||
|
|
"grad_norm": 0.8584981560707092,
|
||
|
|
"learning_rate": 1.6923076923076924e-05,
|
||
|
|
"loss": 0.684,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18900343642611683,
|
||
|
|
"grad_norm": 0.9134799838066101,
|
||
|
|
"learning_rate": 1.8632478632478636e-05,
|
||
|
|
"loss": 0.6805,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20618556701030927,
|
||
|
|
"grad_norm": 0.8744814395904541,
|
||
|
|
"learning_rate": 1.999981993260809e-05,
|
||
|
|
"loss": 0.6771,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22336769759450173,
|
||
|
|
"grad_norm": 0.871300995349884,
|
||
|
|
"learning_rate": 1.9993518254774517e-05,
|
||
|
|
"loss": 0.6922,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24054982817869416,
|
||
|
|
"grad_norm": 0.8517248034477234,
|
||
|
|
"learning_rate": 1.9978219691148676e-05,
|
||
|
|
"loss": 0.6622,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25773195876288657,
|
||
|
|
"grad_norm": 0.8589627742767334,
|
||
|
|
"learning_rate": 1.9953938014600757e-05,
|
||
|
|
"loss": 0.655,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27491408934707906,
|
||
|
|
"grad_norm": 0.8162297606468201,
|
||
|
|
"learning_rate": 1.9920695085247012e-05,
|
||
|
|
"loss": 0.6625,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2920962199312715,
|
||
|
|
"grad_norm": 0.8791314959526062,
|
||
|
|
"learning_rate": 1.9878520830769675e-05,
|
||
|
|
"loss": 0.6634,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30927835051546393,
|
||
|
|
"grad_norm": 0.8827916383743286,
|
||
|
|
"learning_rate": 1.9827453219473925e-05,
|
||
|
|
"loss": 0.6611,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32646048109965636,
|
||
|
|
"grad_norm": 0.8067660927772522,
|
||
|
|
"learning_rate": 1.9767538226106078e-05,
|
||
|
|
"loss": 0.6578,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3436426116838488,
|
||
|
|
"grad_norm": 0.8449408411979675,
|
||
|
|
"learning_rate": 1.9698829790463792e-05,
|
||
|
|
"loss": 0.6586,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36082474226804123,
|
||
|
|
"grad_norm": 0.8589709401130676,
|
||
|
|
"learning_rate": 1.962138976883558e-05,
|
||
|
|
"loss": 0.6454,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37800687285223367,
|
||
|
|
"grad_norm": 0.7897211909294128,
|
||
|
|
"learning_rate": 1.9535287878313315e-05,
|
||
|
|
"loss": 0.6488,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3951890034364261,
|
||
|
|
"grad_norm": 0.830895185470581,
|
||
|
|
"learning_rate": 1.9440601634027892e-05,
|
||
|
|
"loss": 0.6343,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41237113402061853,
|
||
|
|
"grad_norm": 0.781804084777832,
|
||
|
|
"learning_rate": 1.9337416279364486e-05,
|
||
|
|
"loss": 0.6463,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42955326460481097,
|
||
|
|
"grad_norm": 0.8396164774894714,
|
||
|
|
"learning_rate": 1.922582470922034e-05,
|
||
|
|
"loss": 0.6406,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44673539518900346,
|
||
|
|
"grad_norm": 0.834781289100647,
|
||
|
|
"learning_rate": 1.910592738637407e-05,
|
||
|
|
"loss": 0.6175,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4639175257731959,
|
||
|
|
"grad_norm": 0.80084627866745,
|
||
|
|
"learning_rate": 1.897783225104181e-05,
|
||
|
|
"loss": 0.6458,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48109965635738833,
|
||
|
|
"grad_norm": 0.8287182450294495,
|
||
|
|
"learning_rate": 1.8841654623701673e-05,
|
||
|
|
"loss": 0.6376,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.49828178694158076,
|
||
|
|
"grad_norm": 0.8581538200378418,
|
||
|
|
"learning_rate": 1.869751710127392e-05,
|
||
|
|
"loss": 0.6172,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5154639175257731,
|
||
|
|
"grad_norm": 0.8504770398139954,
|
||
|
|
"learning_rate": 1.8545549446750392e-05,
|
||
|
|
"loss": 0.6275,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5326460481099656,
|
||
|
|
"grad_norm": 0.8199458718299866,
|
||
|
|
"learning_rate": 1.8385888472372474e-05,
|
||
|
|
"loss": 0.6259,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5498281786941581,
|
||
|
|
"grad_norm": 0.8343937397003174,
|
||
|
|
"learning_rate": 1.8218677916462882e-05,
|
||
|
|
"loss": 0.6256,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5670103092783505,
|
||
|
|
"grad_norm": 0.8714265823364258,
|
||
|
|
"learning_rate": 1.8044068314022057e-05,
|
||
|
|
"loss": 0.6222,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.584192439862543,
|
||
|
|
"grad_norm": 0.7972601056098938,
|
||
|
|
"learning_rate": 1.786221686120567e-05,
|
||
|
|
"loss": 0.6297,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6013745704467354,
|
||
|
|
"grad_norm": 0.8275905251502991,
|
||
|
|
"learning_rate": 1.7673287273805342e-05,
|
||
|
|
"loss": 0.6356,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6185567010309279,
|
||
|
|
"grad_norm": 0.873673141002655,
|
||
|
|
"learning_rate": 1.7477449639859876e-05,
|
||
|
|
"loss": 0.6213,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6357388316151202,
|
||
|
|
"grad_norm": 0.8335562348365784,
|
||
|
|
"learning_rate": 1.7274880266529716e-05,
|
||
|
|
"loss": 0.6159,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6529209621993127,
|
||
|
|
"grad_norm": 0.78938227891922,
|
||
|
|
"learning_rate": 1.7065761521372562e-05,
|
||
|
|
"loss": 0.6091,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6701030927835051,
|
||
|
|
"grad_norm": 0.7613410353660583,
|
||
|
|
"learning_rate": 1.6850281668162908e-05,
|
||
|
|
"loss": 0.619,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6872852233676976,
|
||
|
|
"grad_norm": 0.9851221442222595,
|
||
|
|
"learning_rate": 1.6628634697403447e-05,
|
||
|
|
"loss": 0.6205,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7044673539518901,
|
||
|
|
"grad_norm": 0.7949865460395813,
|
||
|
|
"learning_rate": 1.6401020151680815e-05,
|
||
|
|
"loss": 0.6054,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7216494845360825,
|
||
|
|
"grad_norm": 0.7933169007301331,
|
||
|
|
"learning_rate": 1.6167642946022956e-05,
|
||
|
|
"loss": 0.6119,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.738831615120275,
|
||
|
|
"grad_norm": 0.8381890058517456,
|
||
|
|
"learning_rate": 1.592871318341986e-05,
|
||
|
|
"loss": 0.6263,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7560137457044673,
|
||
|
|
"grad_norm": 0.8074691891670227,
|
||
|
|
"learning_rate": 1.5684445965673693e-05,
|
||
|
|
"loss": 0.6196,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7731958762886598,
|
||
|
|
"grad_norm": 0.7967894077301025,
|
||
|
|
"learning_rate": 1.5435061199748625e-05,
|
||
|
|
"loss": 0.6176,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7903780068728522,
|
||
|
|
"grad_norm": 0.8072052001953125,
|
||
|
|
"learning_rate": 1.5180783399794749e-05,
|
||
|
|
"loss": 0.6206,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8075601374570447,
|
||
|
|
"grad_norm": 0.7594694495201111,
|
||
|
|
"learning_rate": 1.4921841485024236e-05,
|
||
|
|
"loss": 0.6106,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8247422680412371,
|
||
|
|
"grad_norm": 0.7110355496406555,
|
||
|
|
"learning_rate": 1.4658468573621752e-05,
|
||
|
|
"loss": 0.6086,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8419243986254296,
|
||
|
|
"grad_norm": 0.7381004691123962,
|
||
|
|
"learning_rate": 1.4390901772874668e-05,
|
||
|
|
"loss": 0.6157,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8591065292096219,
|
||
|
|
"grad_norm": 0.791502058506012,
|
||
|
|
"learning_rate": 1.4119381965711976e-05,
|
||
|
|
"loss": 0.6033,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8762886597938144,
|
||
|
|
"grad_norm": 0.7640668749809265,
|
||
|
|
"learning_rate": 1.3844153593844098e-05,
|
||
|
|
"loss": 0.5957,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8934707903780069,
|
||
|
|
"grad_norm": 0.783951461315155,
|
||
|
|
"learning_rate": 1.356546443769885e-05,
|
||
|
|
"loss": 0.5949,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9106529209621993,
|
||
|
|
"grad_norm": 0.7650246620178223,
|
||
|
|
"learning_rate": 1.328356539335161e-05,
|
||
|
|
"loss": 0.6193,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9278350515463918,
|
||
|
|
"grad_norm": 0.7882317304611206,
|
||
|
|
"learning_rate": 1.2998710246650594e-05,
|
||
|
|
"loss": 0.6009,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9450171821305842,
|
||
|
|
"grad_norm": 0.7401001453399658,
|
||
|
|
"learning_rate": 1.2711155444740529e-05,
|
||
|
|
"loss": 0.6033,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9621993127147767,
|
||
|
|
"grad_norm": 0.73549884557724,
|
||
|
|
"learning_rate": 1.242115986519044e-05,
|
||
|
|
"loss": 0.6276,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.979381443298969,
|
||
|
|
"grad_norm": 0.7159228324890137,
|
||
|
|
"learning_rate": 1.2128984582933397e-05,
|
||
|
|
"loss": 0.5978,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9965635738831615,
|
||
|
|
"grad_norm": 0.7005369663238525,
|
||
|
|
"learning_rate": 1.1834892635228024e-05,
|
||
|
|
"loss": 0.5901,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.013745704467354,
|
||
|
|
"grad_norm": 0.8413263559341431,
|
||
|
|
"learning_rate": 1.1539148784853404e-05,
|
||
|
|
"loss": 0.5547,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0309278350515463,
|
||
|
|
"grad_norm": 0.8096709251403809,
|
||
|
|
"learning_rate": 1.1242019281750525e-05,
|
||
|
|
"loss": 0.5384,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0481099656357389,
|
||
|
|
"grad_norm": 0.755794882774353,
|
||
|
|
"learning_rate": 1.0943771623324884e-05,
|
||
|
|
"loss": 0.5263,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0652920962199313,
|
||
|
|
"grad_norm": 0.8318384289741516,
|
||
|
|
"learning_rate": 1.064467431362603e-05,
|
||
|
|
"loss": 0.5314,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0824742268041236,
|
||
|
|
"grad_norm": 0.7728995680809021,
|
||
|
|
"learning_rate": 1.0344996621620862e-05,
|
||
|
|
"loss": 0.5425,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0996563573883162,
|
||
|
|
"grad_norm": 0.7516904473304749,
|
||
|
|
"learning_rate": 1.004500833877828e-05,
|
||
|
|
"loss": 0.5297,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1168384879725086,
|
||
|
|
"grad_norm": 0.7600984573364258,
|
||
|
|
"learning_rate": 9.744979536183458e-06,
|
||
|
|
"loss": 0.5368,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.134020618556701,
|
||
|
|
"grad_norm": 0.7411098480224609,
|
||
|
|
"learning_rate": 9.445180321400373e-06,
|
||
|
|
"loss": 0.5415,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1512027491408934,
|
||
|
|
"grad_norm": 0.7564770579338074,
|
||
|
|
"learning_rate": 9.145880595301495e-06,
|
||
|
|
"loss": 0.5312,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.168384879725086,
|
||
|
|
"grad_norm": 0.7903248071670532,
|
||
|
|
"learning_rate": 8.847349809083557e-06,
|
||
|
|
"loss": 0.5354,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1855670103092784,
|
||
|
|
"grad_norm": 0.7636477947235107,
|
||
|
|
"learning_rate": 8.54985672168817e-06,
|
||
|
|
"loss": 0.545,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2027491408934707,
|
||
|
|
"grad_norm": 0.7316245436668396,
|
||
|
|
"learning_rate": 8.253669157845632e-06,
|
||
|
|
"loss": 0.5239,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2199312714776633,
|
||
|
|
"grad_norm": 0.7267956137657166,
|
||
|
|
"learning_rate": 7.959053766959785e-06,
|
||
|
|
"loss": 0.5212,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2371134020618557,
|
||
|
|
"grad_norm": 0.7424056529998779,
|
||
|
|
"learning_rate": 7.666275783051012e-06,
|
||
|
|
"loss": 0.5314,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.254295532646048,
|
||
|
|
"grad_norm": 0.7225111722946167,
|
||
|
|
"learning_rate": 7.375598785973429e-06,
|
||
|
|
"loss": 0.5298,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2714776632302405,
|
||
|
|
"grad_norm": 0.7220715880393982,
|
||
|
|
"learning_rate": 7.087284464121305e-06,
|
||
|
|
"loss": 0.5283,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2886597938144329,
|
||
|
|
"grad_norm": 0.7413302063941956,
|
||
|
|
"learning_rate": 6.801592378838282e-06,
|
||
|
|
"loss": 0.5312,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3058419243986255,
|
||
|
|
"grad_norm": 0.7235661149024963,
|
||
|
|
"learning_rate": 6.518779730741555e-06,
|
||
|
|
"loss": 0.5414,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3230240549828178,
|
||
|
|
"grad_norm": 0.749245822429657,
|
||
|
|
"learning_rate": 6.239101128171317e-06,
|
||
|
|
"loss": 0.5143,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3402061855670104,
|
||
|
|
"grad_norm": 0.7819613218307495,
|
||
|
|
"learning_rate": 5.9628083579739715e-06,
|
||
|
|
"loss": 0.5381,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3573883161512028,
|
||
|
|
"grad_norm": 0.7491976618766785,
|
||
|
|
"learning_rate": 5.690150158825462e-06,
|
||
|
|
"loss": 0.5366,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3745704467353952,
|
||
|
|
"grad_norm": 0.7514867186546326,
|
||
|
|
"learning_rate": 5.421371997298781e-06,
|
||
|
|
"loss": 0.5306,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3917525773195876,
|
||
|
|
"grad_norm": 0.7224606275558472,
|
||
|
|
"learning_rate": 5.156715846877234e-06,
|
||
|
|
"loss": 0.5248,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.40893470790378,
|
||
|
|
"grad_norm": 0.7374058365821838,
|
||
|
|
"learning_rate": 4.896419970112499e-06,
|
||
|
|
"loss": 0.5282,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4261168384879725,
|
||
|
|
"grad_norm": 0.7273322939872742,
|
||
|
|
"learning_rate": 4.640718704123472e-06,
|
||
|
|
"loss": 0.5181,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.443298969072165,
|
||
|
|
"grad_norm": 0.7601420879364014,
|
||
|
|
"learning_rate": 4.389842249629095e-06,
|
||
|
|
"loss": 0.5302,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4604810996563573,
|
||
|
|
"grad_norm": 0.7110586762428284,
|
||
|
|
"learning_rate": 4.144016463705081e-06,
|
||
|
|
"loss": 0.5305,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.47766323024055,
|
||
|
|
"grad_norm": 0.6905546188354492,
|
||
|
|
"learning_rate": 3.903462656451074e-06,
|
||
|
|
"loss": 0.5216,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4948453608247423,
|
||
|
|
"grad_norm": 0.7205499410629272,
|
||
|
|
"learning_rate": 3.668397391751336e-06,
|
||
|
|
"loss": 0.5188,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5120274914089347,
|
||
|
|
"grad_norm": 0.7174751162528992,
|
||
|
|
"learning_rate": 3.4390322923083385e-06,
|
||
|
|
"loss": 0.5255,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.529209621993127,
|
||
|
|
"grad_norm": 0.7230581641197205,
|
||
|
|
"learning_rate": 3.2155738491247303e-06,
|
||
|
|
"loss": 0.5114,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5463917525773194,
|
||
|
|
"grad_norm": 0.7528461813926697,
|
||
|
|
"learning_rate": 2.99822323560525e-06,
|
||
|
|
"loss": 0.5227,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.563573883161512,
|
||
|
|
"grad_norm": 0.7461559772491455,
|
||
|
|
"learning_rate": 2.787176126445923e-06,
|
||
|
|
"loss": 0.5234,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5807560137457046,
|
||
|
|
"grad_norm": 0.7162157893180847,
|
||
|
|
"learning_rate": 2.582622521473588e-06,
|
||
|
|
"loss": 0.5344,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.597938144329897,
|
||
|
|
"grad_norm": 0.7246734499931335,
|
||
|
|
"learning_rate": 2.384746574594343e-06,
|
||
|
|
"loss": 0.5283,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6151202749140894,
|
||
|
|
"grad_norm": 0.7509973049163818,
|
||
|
|
"learning_rate": 2.1937264280049365e-06,
|
||
|
|
"loss": 0.5232,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6323024054982818,
|
||
|
|
"grad_norm": 0.716250479221344,
|
||
|
|
"learning_rate": 2.0097340518163e-06,
|
||
|
|
"loss": 0.5252,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6494845360824741,
|
||
|
|
"grad_norm": 0.7006597518920898,
|
||
|
|
"learning_rate": 1.8329350892336616e-06,
|
||
|
|
"loss": 0.5137,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6666666666666665,
|
||
|
|
"grad_norm": 0.696034848690033,
|
||
|
|
"learning_rate": 1.6634887074325844e-06,
|
||
|
|
"loss": 0.5284,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6838487972508591,
|
||
|
|
"grad_norm": 0.767427384853363,
|
||
|
|
"learning_rate": 1.5015474542651964e-06,
|
||
|
|
"loss": 0.5255,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7010309278350515,
|
||
|
|
"grad_norm": 0.7569447159767151,
|
||
|
|
"learning_rate": 1.3472571209256157e-06,
|
||
|
|
"loss": 0.5297,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.718213058419244,
|
||
|
|
"grad_norm": 0.7058409452438354,
|
||
|
|
"learning_rate": 1.200756610698205e-06,
|
||
|
|
"loss": 0.5103,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.718213058419244,
|
||
|
|
"eval_loss": 0.5760038495063782,
|
||
|
|
"eval_runtime": 10.0494,
|
||
|
|
"eval_samples_per_second": 10.747,
|
||
|
|
"eval_steps_per_second": 5.373,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7353951890034365,
|
||
|
|
"grad_norm": 0.7059236168861389,
|
||
|
|
"learning_rate": 1.0621778139068128e-06,
|
||
|
|
"loss": 0.5118,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7525773195876289,
|
||
|
|
"grad_norm": 0.7298533916473389,
|
||
|
|
"learning_rate": 9.316454891775983e-07,
|
||
|
|
"loss": 0.516,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7697594501718212,
|
||
|
|
"grad_norm": 0.6977025270462036,
|
||
|
|
"learning_rate": 8.092771511223185e-07,
|
||
|
|
"loss": 0.5148,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7869415807560136,
|
||
|
|
"grad_norm": 0.6981027126312256,
|
||
|
|
"learning_rate": 6.951829645431985e-07,
|
||
|
|
"loss": 0.5145,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8041237113402062,
|
||
|
|
"grad_norm": 0.7186697125434875,
|
||
|
|
"learning_rate": 5.894656452546343e-07,
|
||
|
|
"loss": 0.5217,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8213058419243986,
|
||
|
|
"grad_norm": 0.6785486936569214,
|
||
|
|
"learning_rate": 4.92220367611006e-07,
|
||
|
|
"loss": 0.5186,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8384879725085912,
|
||
|
|
"grad_norm": 0.6929068565368652,
|
||
|
|
"learning_rate": 4.035346788238681e-07,
|
||
|
|
"loss": 0.5275,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8556701030927836,
|
||
|
|
"grad_norm": 0.6964775919914246,
|
||
|
|
"learning_rate": 3.2348842014562564e-07,
|
||
|
|
"loss": 0.5199,
|
||
|
|
"step": 1080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.872852233676976,
|
||
|
|
"grad_norm": 0.717461884021759,
|
||
|
|
"learning_rate": 2.5215365499069446e-07,
|
||
|
|
"loss": 0.5291,
|
||
|
|
"step": 1090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8900343642611683,
|
||
|
|
"grad_norm": 0.6823316216468811,
|
||
|
|
"learning_rate": 1.895946040588048e-07,
|
||
|
|
"loss": 0.5278,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9072164948453607,
|
||
|
|
"grad_norm": 0.7319381237030029,
|
||
|
|
"learning_rate": 1.3586758751890638e-07,
|
||
|
|
"loss": 0.5164,
|
||
|
|
"step": 1110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9243986254295533,
|
||
|
|
"grad_norm": 0.6861236691474915,
|
||
|
|
"learning_rate": 9.10209743056889e-08,
|
||
|
|
"loss": 0.5357,
|
||
|
|
"step": 1120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9415807560137457,
|
||
|
|
"grad_norm": 0.6966984272003174,
|
||
|
|
"learning_rate": 5.509513857437365e-08,
|
||
|
|
"loss": 0.5264,
|
||
|
|
"step": 1130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9587628865979383,
|
||
|
|
"grad_norm": 0.6996172666549683,
|
||
|
|
"learning_rate": 2.812242335299642e-08,
|
||
|
|
"loss": 0.5197,
|
||
|
|
"step": 1140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9759450171821307,
|
||
|
|
"grad_norm": 0.7039883136749268,
|
||
|
|
"learning_rate": 1.0127111424872437e-08,
|
||
|
|
"loss": 0.5238,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.993127147766323,
|
||
|
|
"grad_norm": 0.7203370332717896,
|
||
|
|
"learning_rate": 1.125403467490127e-09,
|
||
|
|
"loss": 0.5161,
|
||
|
|
"step": 1160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"step": 1164,
|
||
|
|
"total_flos": 9.78689173969961e+17,
|
||
|
|
"train_loss": 0.5901876624507183,
|
||
|
|
"train_runtime": 6155.6604,
|
||
|
|
"train_samples_per_second": 3.026,
|
||
|
|
"train_steps_per_second": 0.189
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 1164,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 2,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 9.78689173969961e+17,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|