8295 lines
234 KiB
JSON
8295 lines
234 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 3.0,
|
|
"eval_steps": 500,
|
|
"global_step": 4125,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0007272727272727272,
|
|
"grad_norm": 22.59613609313965,
|
|
"learning_rate": 8.064516129032259e-08,
|
|
"loss": 2.312417507171631,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1,
|
|
"token_acc": 0.4763092269326683,
|
|
"train_speed(iter/s)": 0.040402
|
|
},
|
|
{
|
|
"epoch": 0.0036363636363636364,
|
|
"grad_norm": 16.399948120117188,
|
|
"learning_rate": 4.032258064516129e-07,
|
|
"loss": 2.0219309329986572,
|
|
"memory(GiB)": 64.44,
|
|
"step": 5,
|
|
"token_acc": 0.5876224213377788,
|
|
"train_speed(iter/s)": 0.148345
|
|
},
|
|
{
|
|
"epoch": 0.007272727272727273,
|
|
"grad_norm": 25.756488800048828,
|
|
"learning_rate": 8.064516129032258e-07,
|
|
"loss": 2.0684181213378907,
|
|
"memory(GiB)": 64.44,
|
|
"step": 10,
|
|
"token_acc": 0.5817155756207675,
|
|
"train_speed(iter/s)": 0.226534
|
|
},
|
|
{
|
|
"epoch": 0.01090909090909091,
|
|
"grad_norm": 14.12072467803955,
|
|
"learning_rate": 1.2096774193548388e-06,
|
|
"loss": 1.9113319396972657,
|
|
"memory(GiB)": 64.44,
|
|
"step": 15,
|
|
"token_acc": 0.5714888232813159,
|
|
"train_speed(iter/s)": 0.275262
|
|
},
|
|
{
|
|
"epoch": 0.014545454545454545,
|
|
"grad_norm": 9.777295112609863,
|
|
"learning_rate": 1.6129032258064516e-06,
|
|
"loss": 1.6752943038940429,
|
|
"memory(GiB)": 64.44,
|
|
"step": 20,
|
|
"token_acc": 0.6,
|
|
"train_speed(iter/s)": 0.308005
|
|
},
|
|
{
|
|
"epoch": 0.01818181818181818,
|
|
"grad_norm": 5.061804294586182,
|
|
"learning_rate": 2.0161290322580646e-06,
|
|
"loss": 1.5955514907836914,
|
|
"memory(GiB)": 64.44,
|
|
"step": 25,
|
|
"token_acc": 0.6175671083927965,
|
|
"train_speed(iter/s)": 0.328805
|
|
},
|
|
{
|
|
"epoch": 0.02181818181818182,
|
|
"grad_norm": 5.698326110839844,
|
|
"learning_rate": 2.4193548387096776e-06,
|
|
"loss": 1.599413776397705,
|
|
"memory(GiB)": 64.44,
|
|
"step": 30,
|
|
"token_acc": 0.6125984251968504,
|
|
"train_speed(iter/s)": 0.349264
|
|
},
|
|
{
|
|
"epoch": 0.025454545454545455,
|
|
"grad_norm": 4.843620777130127,
|
|
"learning_rate": 2.822580645161291e-06,
|
|
"loss": 1.6385580062866212,
|
|
"memory(GiB)": 64.44,
|
|
"step": 35,
|
|
"token_acc": 0.6209428129829985,
|
|
"train_speed(iter/s)": 0.363951
|
|
},
|
|
{
|
|
"epoch": 0.02909090909090909,
|
|
"grad_norm": 5.110261917114258,
|
|
"learning_rate": 3.225806451612903e-06,
|
|
"loss": 1.573188877105713,
|
|
"memory(GiB)": 64.44,
|
|
"step": 40,
|
|
"token_acc": 0.6010611419909045,
|
|
"train_speed(iter/s)": 0.376755
|
|
},
|
|
{
|
|
"epoch": 0.03272727272727273,
|
|
"grad_norm": 5.215554237365723,
|
|
"learning_rate": 3.6290322580645166e-06,
|
|
"loss": 1.5826381683349608,
|
|
"memory(GiB)": 64.44,
|
|
"step": 45,
|
|
"token_acc": 0.6275129884797831,
|
|
"train_speed(iter/s)": 0.387234
|
|
},
|
|
{
|
|
"epoch": 0.03636363636363636,
|
|
"grad_norm": 5.08247709274292,
|
|
"learning_rate": 4.032258064516129e-06,
|
|
"loss": 1.5879148483276366,
|
|
"memory(GiB)": 64.44,
|
|
"step": 50,
|
|
"token_acc": 0.6302843966179862,
|
|
"train_speed(iter/s)": 0.394999
|
|
},
|
|
{
|
|
"epoch": 0.04,
|
|
"grad_norm": 5.114762306213379,
|
|
"learning_rate": 4.435483870967742e-06,
|
|
"loss": 1.5051856994628907,
|
|
"memory(GiB)": 64.44,
|
|
"step": 55,
|
|
"token_acc": 0.6086616227498043,
|
|
"train_speed(iter/s)": 0.401855
|
|
},
|
|
{
|
|
"epoch": 0.04363636363636364,
|
|
"grad_norm": 4.984416961669922,
|
|
"learning_rate": 4.838709677419355e-06,
|
|
"loss": 1.565500259399414,
|
|
"memory(GiB)": 64.44,
|
|
"step": 60,
|
|
"token_acc": 0.6135120350109409,
|
|
"train_speed(iter/s)": 0.408094
|
|
},
|
|
{
|
|
"epoch": 0.04727272727272727,
|
|
"grad_norm": 4.8500075340271,
|
|
"learning_rate": 5.241935483870968e-06,
|
|
"loss": 1.4776456832885743,
|
|
"memory(GiB)": 64.44,
|
|
"step": 65,
|
|
"token_acc": 0.6234521101844832,
|
|
"train_speed(iter/s)": 0.412103
|
|
},
|
|
{
|
|
"epoch": 0.05090909090909091,
|
|
"grad_norm": 4.941230297088623,
|
|
"learning_rate": 5.645161290322582e-06,
|
|
"loss": 1.5210559844970704,
|
|
"memory(GiB)": 64.44,
|
|
"step": 70,
|
|
"token_acc": 0.6208510638297873,
|
|
"train_speed(iter/s)": 0.41635
|
|
},
|
|
{
|
|
"epoch": 0.05454545454545454,
|
|
"grad_norm": 4.713766098022461,
|
|
"learning_rate": 6.048387096774194e-06,
|
|
"loss": 1.5052054405212403,
|
|
"memory(GiB)": 64.44,
|
|
"step": 75,
|
|
"token_acc": 0.646598030438675,
|
|
"train_speed(iter/s)": 0.420387
|
|
},
|
|
{
|
|
"epoch": 0.05818181818181818,
|
|
"grad_norm": 4.036681175231934,
|
|
"learning_rate": 6.451612903225806e-06,
|
|
"loss": 1.4002185821533204,
|
|
"memory(GiB)": 64.44,
|
|
"step": 80,
|
|
"token_acc": 0.67150771073514,
|
|
"train_speed(iter/s)": 0.41958
|
|
},
|
|
{
|
|
"epoch": 0.06181818181818182,
|
|
"grad_norm": 4.7215728759765625,
|
|
"learning_rate": 6.854838709677419e-06,
|
|
"loss": 1.549690341949463,
|
|
"memory(GiB)": 64.44,
|
|
"step": 85,
|
|
"token_acc": 0.6353428916911383,
|
|
"train_speed(iter/s)": 0.421693
|
|
},
|
|
{
|
|
"epoch": 0.06545454545454546,
|
|
"grad_norm": 4.367216110229492,
|
|
"learning_rate": 7.258064516129033e-06,
|
|
"loss": 1.485302448272705,
|
|
"memory(GiB)": 64.44,
|
|
"step": 90,
|
|
"token_acc": 0.6345565749235474,
|
|
"train_speed(iter/s)": 0.424745
|
|
},
|
|
{
|
|
"epoch": 0.06909090909090909,
|
|
"grad_norm": 4.972701072692871,
|
|
"learning_rate": 7.661290322580646e-06,
|
|
"loss": 1.5724483489990235,
|
|
"memory(GiB)": 64.44,
|
|
"step": 95,
|
|
"token_acc": 0.6130210571551353,
|
|
"train_speed(iter/s)": 0.42788
|
|
},
|
|
{
|
|
"epoch": 0.07272727272727272,
|
|
"grad_norm": 4.559494972229004,
|
|
"learning_rate": 8.064516129032258e-06,
|
|
"loss": 1.581673526763916,
|
|
"memory(GiB)": 64.44,
|
|
"step": 100,
|
|
"token_acc": 0.6047289761320545,
|
|
"train_speed(iter/s)": 0.430426
|
|
},
|
|
{
|
|
"epoch": 0.07636363636363637,
|
|
"grad_norm": 4.930873394012451,
|
|
"learning_rate": 8.467741935483872e-06,
|
|
"loss": 1.4878676414489747,
|
|
"memory(GiB)": 64.44,
|
|
"step": 105,
|
|
"token_acc": 0.5901130569718466,
|
|
"train_speed(iter/s)": 0.430305
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"grad_norm": 5.295114994049072,
|
|
"learning_rate": 8.870967741935484e-06,
|
|
"loss": 1.541814422607422,
|
|
"memory(GiB)": 64.44,
|
|
"step": 110,
|
|
"token_acc": 0.6382845188284518,
|
|
"train_speed(iter/s)": 0.432117
|
|
},
|
|
{
|
|
"epoch": 0.08363636363636363,
|
|
"grad_norm": 5.135087013244629,
|
|
"learning_rate": 9.274193548387097e-06,
|
|
"loss": 1.5753223419189453,
|
|
"memory(GiB)": 64.44,
|
|
"step": 115,
|
|
"token_acc": 0.6122940940136601,
|
|
"train_speed(iter/s)": 0.434155
|
|
},
|
|
{
|
|
"epoch": 0.08727272727272728,
|
|
"grad_norm": 4.172939300537109,
|
|
"learning_rate": 9.67741935483871e-06,
|
|
"loss": 1.552765464782715,
|
|
"memory(GiB)": 64.44,
|
|
"step": 120,
|
|
"token_acc": 0.6434352517985612,
|
|
"train_speed(iter/s)": 0.435597
|
|
},
|
|
{
|
|
"epoch": 0.09090909090909091,
|
|
"grad_norm": 4.920270919799805,
|
|
"learning_rate": 9.999998458645166e-06,
|
|
"loss": 1.5966070175170899,
|
|
"memory(GiB)": 64.44,
|
|
"step": 125,
|
|
"token_acc": 0.6312905844155844,
|
|
"train_speed(iter/s)": 0.437896
|
|
},
|
|
{
|
|
"epoch": 0.09454545454545454,
|
|
"grad_norm": 5.000640869140625,
|
|
"learning_rate": 9.999944511325734e-06,
|
|
"loss": 1.4817375183105468,
|
|
"memory(GiB)": 64.44,
|
|
"step": 130,
|
|
"token_acc": 0.6199952617863066,
|
|
"train_speed(iter/s)": 0.439898
|
|
},
|
|
{
|
|
"epoch": 0.09818181818181818,
|
|
"grad_norm": 4.871465682983398,
|
|
"learning_rate": 9.999813497214877e-06,
|
|
"loss": 1.46666259765625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 135,
|
|
"token_acc": 0.644193127446716,
|
|
"train_speed(iter/s)": 0.440049
|
|
},
|
|
{
|
|
"epoch": 0.10181818181818182,
|
|
"grad_norm": 4.193850517272949,
|
|
"learning_rate": 9.99960541833198e-06,
|
|
"loss": 1.5098167419433595,
|
|
"memory(GiB)": 64.44,
|
|
"step": 140,
|
|
"token_acc": 0.6258551545175749,
|
|
"train_speed(iter/s)": 0.440616
|
|
},
|
|
{
|
|
"epoch": 0.10545454545454545,
|
|
"grad_norm": 4.325166702270508,
|
|
"learning_rate": 9.999320277884277e-06,
|
|
"loss": 1.4981783866882323,
|
|
"memory(GiB)": 64.44,
|
|
"step": 145,
|
|
"token_acc": 0.6175044091710759,
|
|
"train_speed(iter/s)": 0.442088
|
|
},
|
|
{
|
|
"epoch": 0.10909090909090909,
|
|
"grad_norm": 4.8125410079956055,
|
|
"learning_rate": 9.998958080266787e-06,
|
|
"loss": 1.407313346862793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 150,
|
|
"token_acc": 0.6471645367412141,
|
|
"train_speed(iter/s)": 0.442736
|
|
},
|
|
{
|
|
"epoch": 0.11272727272727273,
|
|
"grad_norm": 4.802966117858887,
|
|
"learning_rate": 9.998518831062255e-06,
|
|
"loss": 1.5133844375610352,
|
|
"memory(GiB)": 64.44,
|
|
"step": 155,
|
|
"token_acc": 0.6264990328820116,
|
|
"train_speed(iter/s)": 0.443566
|
|
},
|
|
{
|
|
"epoch": 0.11636363636363636,
|
|
"grad_norm": 4.249505043029785,
|
|
"learning_rate": 9.99800253704106e-06,
|
|
"loss": 1.5245509147644043,
|
|
"memory(GiB)": 64.44,
|
|
"step": 160,
|
|
"token_acc": 0.6210456942003515,
|
|
"train_speed(iter/s)": 0.444249
|
|
},
|
|
{
|
|
"epoch": 0.12,
|
|
"grad_norm": 4.372662544250488,
|
|
"learning_rate": 9.997409206161116e-06,
|
|
"loss": 1.422879409790039,
|
|
"memory(GiB)": 64.44,
|
|
"step": 165,
|
|
"token_acc": 0.6574539363484088,
|
|
"train_speed(iter/s)": 0.445374
|
|
},
|
|
{
|
|
"epoch": 0.12363636363636364,
|
|
"grad_norm": 5.232819557189941,
|
|
"learning_rate": 9.996738847567745e-06,
|
|
"loss": 1.6023344039916991,
|
|
"memory(GiB)": 64.44,
|
|
"step": 170,
|
|
"token_acc": 0.6377105743860362,
|
|
"train_speed(iter/s)": 0.446738
|
|
},
|
|
{
|
|
"epoch": 0.12727272727272726,
|
|
"grad_norm": 4.823113918304443,
|
|
"learning_rate": 9.995991471593538e-06,
|
|
"loss": 1.5789835929870606,
|
|
"memory(GiB)": 64.44,
|
|
"step": 175,
|
|
"token_acc": 0.6327939590075512,
|
|
"train_speed(iter/s)": 0.447515
|
|
},
|
|
{
|
|
"epoch": 0.13090909090909092,
|
|
"grad_norm": 4.816397666931152,
|
|
"learning_rate": 9.9951670897582e-06,
|
|
"loss": 1.4952729225158692,
|
|
"memory(GiB)": 64.44,
|
|
"step": 180,
|
|
"token_acc": 0.622705212070057,
|
|
"train_speed(iter/s)": 0.448454
|
|
},
|
|
{
|
|
"epoch": 0.13454545454545455,
|
|
"grad_norm": 4.636293411254883,
|
|
"learning_rate": 9.99426571476836e-06,
|
|
"loss": 1.6462493896484376,
|
|
"memory(GiB)": 64.44,
|
|
"step": 185,
|
|
"token_acc": 0.6159734626180148,
|
|
"train_speed(iter/s)": 0.449812
|
|
},
|
|
{
|
|
"epoch": 0.13818181818181818,
|
|
"grad_norm": 4.711672782897949,
|
|
"learning_rate": 9.993287360517394e-06,
|
|
"loss": 1.5444934844970704,
|
|
"memory(GiB)": 64.44,
|
|
"step": 190,
|
|
"token_acc": 0.6438658428949691,
|
|
"train_speed(iter/s)": 0.450862
|
|
},
|
|
{
|
|
"epoch": 0.14181818181818182,
|
|
"grad_norm": 4.75740385055542,
|
|
"learning_rate": 9.992232042085188e-06,
|
|
"loss": 1.5141837120056152,
|
|
"memory(GiB)": 64.44,
|
|
"step": 195,
|
|
"token_acc": 0.6383887135697048,
|
|
"train_speed(iter/s)": 0.451385
|
|
},
|
|
{
|
|
"epoch": 0.14545454545454545,
|
|
"grad_norm": 4.457281589508057,
|
|
"learning_rate": 9.991099775737925e-06,
|
|
"loss": 1.5545944213867187,
|
|
"memory(GiB)": 64.44,
|
|
"step": 200,
|
|
"token_acc": 0.6363013698630137,
|
|
"train_speed(iter/s)": 0.452181
|
|
},
|
|
{
|
|
"epoch": 0.14909090909090908,
|
|
"grad_norm": 4.13571310043335,
|
|
"learning_rate": 9.989890578927827e-06,
|
|
"loss": 1.5171931266784668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 205,
|
|
"token_acc": 0.6591174021648626,
|
|
"train_speed(iter/s)": 0.452587
|
|
},
|
|
{
|
|
"epoch": 0.15272727272727274,
|
|
"grad_norm": 4.348774433135986,
|
|
"learning_rate": 9.988604470292886e-06,
|
|
"loss": 1.4988608360290527,
|
|
"memory(GiB)": 64.44,
|
|
"step": 210,
|
|
"token_acc": 0.6556742815033162,
|
|
"train_speed(iter/s)": 0.452838
|
|
},
|
|
{
|
|
"epoch": 0.15636363636363637,
|
|
"grad_norm": 4.25404167175293,
|
|
"learning_rate": 9.987241469656573e-06,
|
|
"loss": 1.5212635040283202,
|
|
"memory(GiB)": 64.44,
|
|
"step": 215,
|
|
"token_acc": 0.6363988383349467,
|
|
"train_speed(iter/s)": 0.453261
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"grad_norm": 4.63402795791626,
|
|
"learning_rate": 9.985801598027536e-06,
|
|
"loss": 1.5325578689575194,
|
|
"memory(GiB)": 64.44,
|
|
"step": 220,
|
|
"token_acc": 0.6563048383053397,
|
|
"train_speed(iter/s)": 0.453851
|
|
},
|
|
{
|
|
"epoch": 0.16363636363636364,
|
|
"grad_norm": 4.500482559204102,
|
|
"learning_rate": 9.984284877599281e-06,
|
|
"loss": 1.6204137802124023,
|
|
"memory(GiB)": 64.44,
|
|
"step": 225,
|
|
"token_acc": 0.60273095796885,
|
|
"train_speed(iter/s)": 0.45439
|
|
},
|
|
{
|
|
"epoch": 0.16727272727272727,
|
|
"grad_norm": 4.2809929847717285,
|
|
"learning_rate": 9.982691331749826e-06,
|
|
"loss": 1.5170854568481444,
|
|
"memory(GiB)": 64.44,
|
|
"step": 230,
|
|
"token_acc": 0.6392360292383872,
|
|
"train_speed(iter/s)": 0.45458
|
|
},
|
|
{
|
|
"epoch": 0.1709090909090909,
|
|
"grad_norm": 5.02462911605835,
|
|
"learning_rate": 9.98102098504133e-06,
|
|
"loss": 1.591007900238037,
|
|
"memory(GiB)": 64.44,
|
|
"step": 235,
|
|
"token_acc": 0.6016401590457257,
|
|
"train_speed(iter/s)": 0.455408
|
|
},
|
|
{
|
|
"epoch": 0.17454545454545456,
|
|
"grad_norm": 5.187936782836914,
|
|
"learning_rate": 9.979273863219736e-06,
|
|
"loss": 1.5465394973754882,
|
|
"memory(GiB)": 64.44,
|
|
"step": 240,
|
|
"token_acc": 0.596551724137931,
|
|
"train_speed(iter/s)": 0.455611
|
|
},
|
|
{
|
|
"epoch": 0.1781818181818182,
|
|
"grad_norm": 4.898781776428223,
|
|
"learning_rate": 9.977449993214357e-06,
|
|
"loss": 1.604863739013672,
|
|
"memory(GiB)": 64.44,
|
|
"step": 245,
|
|
"token_acc": 0.6101733063245555,
|
|
"train_speed(iter/s)": 0.456036
|
|
},
|
|
{
|
|
"epoch": 0.18181818181818182,
|
|
"grad_norm": 4.115358829498291,
|
|
"learning_rate": 9.975549403137464e-06,
|
|
"loss": 1.4797225952148438,
|
|
"memory(GiB)": 64.44,
|
|
"step": 250,
|
|
"token_acc": 0.6420706973150473,
|
|
"train_speed(iter/s)": 0.456135
|
|
},
|
|
{
|
|
"epoch": 0.18545454545454546,
|
|
"grad_norm": 4.969156742095947,
|
|
"learning_rate": 9.973572122283862e-06,
|
|
"loss": 1.5260903358459472,
|
|
"memory(GiB)": 64.44,
|
|
"step": 255,
|
|
"token_acc": 0.632973944294699,
|
|
"train_speed(iter/s)": 0.456308
|
|
},
|
|
{
|
|
"epoch": 0.1890909090909091,
|
|
"grad_norm": 4.20519495010376,
|
|
"learning_rate": 9.971518181130424e-06,
|
|
"loss": 1.470450210571289,
|
|
"memory(GiB)": 64.44,
|
|
"step": 260,
|
|
"token_acc": 0.6416337285902504,
|
|
"train_speed(iter/s)": 0.456191
|
|
},
|
|
{
|
|
"epoch": 0.19272727272727272,
|
|
"grad_norm": 4.706243991851807,
|
|
"learning_rate": 9.969387611335632e-06,
|
|
"loss": 1.5608243942260742,
|
|
"memory(GiB)": 64.44,
|
|
"step": 265,
|
|
"token_acc": 0.589919545669664,
|
|
"train_speed(iter/s)": 0.456389
|
|
},
|
|
{
|
|
"epoch": 0.19636363636363635,
|
|
"grad_norm": 4.728813648223877,
|
|
"learning_rate": 9.96718044573909e-06,
|
|
"loss": 1.5417936325073243,
|
|
"memory(GiB)": 64.44,
|
|
"step": 270,
|
|
"token_acc": 0.6227795193312434,
|
|
"train_speed(iter/s)": 0.457028
|
|
},
|
|
{
|
|
"epoch": 0.2,
|
|
"grad_norm": 4.621528148651123,
|
|
"learning_rate": 9.964896718361007e-06,
|
|
"loss": 1.589951229095459,
|
|
"memory(GiB)": 64.44,
|
|
"step": 275,
|
|
"token_acc": 0.6673909107369184,
|
|
"train_speed(iter/s)": 0.457304
|
|
},
|
|
{
|
|
"epoch": 0.20363636363636364,
|
|
"grad_norm": 4.198657989501953,
|
|
"learning_rate": 9.96253646440168e-06,
|
|
"loss": 1.494888687133789,
|
|
"memory(GiB)": 64.44,
|
|
"step": 280,
|
|
"token_acc": 0.6309142318567389,
|
|
"train_speed(iter/s)": 0.457335
|
|
},
|
|
{
|
|
"epoch": 0.20727272727272728,
|
|
"grad_norm": 4.37053108215332,
|
|
"learning_rate": 9.960099720240957e-06,
|
|
"loss": 1.4807355880737305,
|
|
"memory(GiB)": 64.44,
|
|
"step": 285,
|
|
"token_acc": 0.6721584984358707,
|
|
"train_speed(iter/s)": 0.457916
|
|
},
|
|
{
|
|
"epoch": 0.2109090909090909,
|
|
"grad_norm": 4.357964992523193,
|
|
"learning_rate": 9.957586523437662e-06,
|
|
"loss": 1.5848698616027832,
|
|
"memory(GiB)": 64.44,
|
|
"step": 290,
|
|
"token_acc": 0.6120053020261315,
|
|
"train_speed(iter/s)": 0.458254
|
|
},
|
|
{
|
|
"epoch": 0.21454545454545454,
|
|
"grad_norm": 3.9031291007995605,
|
|
"learning_rate": 9.954996912729029e-06,
|
|
"loss": 1.5905929565429688,
|
|
"memory(GiB)": 64.44,
|
|
"step": 295,
|
|
"token_acc": 0.6347525394424033,
|
|
"train_speed(iter/s)": 0.458959
|
|
},
|
|
{
|
|
"epoch": 0.21818181818181817,
|
|
"grad_norm": 4.532529830932617,
|
|
"learning_rate": 9.9523309280301e-06,
|
|
"loss": 1.4738840103149413,
|
|
"memory(GiB)": 64.44,
|
|
"step": 300,
|
|
"token_acc": 0.636405109489051,
|
|
"train_speed(iter/s)": 0.458664
|
|
},
|
|
{
|
|
"epoch": 0.22181818181818183,
|
|
"grad_norm": 4.186645030975342,
|
|
"learning_rate": 9.949588610433105e-06,
|
|
"loss": 1.5529683113098145,
|
|
"memory(GiB)": 64.44,
|
|
"step": 305,
|
|
"token_acc": 0.584009009009009,
|
|
"train_speed(iter/s)": 0.458812
|
|
},
|
|
{
|
|
"epoch": 0.22545454545454546,
|
|
"grad_norm": 4.240685939788818,
|
|
"learning_rate": 9.94677000220684e-06,
|
|
"loss": 1.5312615394592286,
|
|
"memory(GiB)": 64.44,
|
|
"step": 310,
|
|
"token_acc": 0.6109083100708611,
|
|
"train_speed(iter/s)": 0.45898
|
|
},
|
|
{
|
|
"epoch": 0.2290909090909091,
|
|
"grad_norm": 4.726016521453857,
|
|
"learning_rate": 9.943875146796005e-06,
|
|
"loss": 1.5821155548095702,
|
|
"memory(GiB)": 64.44,
|
|
"step": 315,
|
|
"token_acc": 0.6087295401402962,
|
|
"train_speed(iter/s)": 0.459411
|
|
},
|
|
{
|
|
"epoch": 0.23272727272727273,
|
|
"grad_norm": 3.997502088546753,
|
|
"learning_rate": 9.940904088820538e-06,
|
|
"loss": 1.4851863861083985,
|
|
"memory(GiB)": 64.44,
|
|
"step": 320,
|
|
"token_acc": 0.6362646930779278,
|
|
"train_speed(iter/s)": 0.459939
|
|
},
|
|
{
|
|
"epoch": 0.23636363636363636,
|
|
"grad_norm": 4.03656530380249,
|
|
"learning_rate": 9.93785687407493e-06,
|
|
"loss": 1.5195300102233886,
|
|
"memory(GiB)": 64.44,
|
|
"step": 325,
|
|
"token_acc": 0.6184857423795477,
|
|
"train_speed(iter/s)": 0.459636
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"grad_norm": 4.356072425842285,
|
|
"learning_rate": 9.93473354952751e-06,
|
|
"loss": 1.471431827545166,
|
|
"memory(GiB)": 64.44,
|
|
"step": 330,
|
|
"token_acc": 0.6255809254394827,
|
|
"train_speed(iter/s)": 0.459145
|
|
},
|
|
{
|
|
"epoch": 0.24363636363636362,
|
|
"grad_norm": 4.420890808105469,
|
|
"learning_rate": 9.931534163319739e-06,
|
|
"loss": 1.6016006469726562,
|
|
"memory(GiB)": 64.44,
|
|
"step": 335,
|
|
"token_acc": 0.6178384050367262,
|
|
"train_speed(iter/s)": 0.45933
|
|
},
|
|
{
|
|
"epoch": 0.24727272727272728,
|
|
"grad_norm": 4.427598476409912,
|
|
"learning_rate": 9.928258764765446e-06,
|
|
"loss": 1.5727561950683593,
|
|
"memory(GiB)": 64.44,
|
|
"step": 340,
|
|
"token_acc": 0.6156725729972548,
|
|
"train_speed(iter/s)": 0.459915
|
|
},
|
|
{
|
|
"epoch": 0.2509090909090909,
|
|
"grad_norm": 4.281289577484131,
|
|
"learning_rate": 9.924907404350084e-06,
|
|
"loss": 1.4837968826293946,
|
|
"memory(GiB)": 64.44,
|
|
"step": 345,
|
|
"token_acc": 0.6198141011108592,
|
|
"train_speed(iter/s)": 0.460137
|
|
},
|
|
{
|
|
"epoch": 0.2545454545454545,
|
|
"grad_norm": 4.332330226898193,
|
|
"learning_rate": 9.921480133729943e-06,
|
|
"loss": 1.4289453506469727,
|
|
"memory(GiB)": 64.44,
|
|
"step": 350,
|
|
"token_acc": 0.6394877267876201,
|
|
"train_speed(iter/s)": 0.460435
|
|
},
|
|
{
|
|
"epoch": 0.2581818181818182,
|
|
"grad_norm": 4.192288875579834,
|
|
"learning_rate": 9.917977005731365e-06,
|
|
"loss": 1.4650839805603026,
|
|
"memory(GiB)": 64.44,
|
|
"step": 355,
|
|
"token_acc": 0.6584795321637427,
|
|
"train_speed(iter/s)": 0.460398
|
|
},
|
|
{
|
|
"epoch": 0.26181818181818184,
|
|
"grad_norm": 4.511104583740234,
|
|
"learning_rate": 9.914398074349908e-06,
|
|
"loss": 1.5645034790039063,
|
|
"memory(GiB)": 64.44,
|
|
"step": 360,
|
|
"token_acc": 0.6630366492146597,
|
|
"train_speed(iter/s)": 0.460742
|
|
},
|
|
{
|
|
"epoch": 0.26545454545454544,
|
|
"grad_norm": 4.1415510177612305,
|
|
"learning_rate": 9.910743394749545e-06,
|
|
"loss": 1.5121275901794433,
|
|
"memory(GiB)": 64.44,
|
|
"step": 365,
|
|
"token_acc": 0.6393345216874629,
|
|
"train_speed(iter/s)": 0.460505
|
|
},
|
|
{
|
|
"epoch": 0.2690909090909091,
|
|
"grad_norm": 4.683187007904053,
|
|
"learning_rate": 9.90701302326178e-06,
|
|
"loss": 1.5751832962036132,
|
|
"memory(GiB)": 64.44,
|
|
"step": 370,
|
|
"token_acc": 0.6360116166505324,
|
|
"train_speed(iter/s)": 0.460985
|
|
},
|
|
{
|
|
"epoch": 0.2727272727272727,
|
|
"grad_norm": 4.6351470947265625,
|
|
"learning_rate": 9.903207017384805e-06,
|
|
"loss": 1.5309032440185546,
|
|
"memory(GiB)": 64.44,
|
|
"step": 375,
|
|
"token_acc": 0.5923818946326985,
|
|
"train_speed(iter/s)": 0.461128
|
|
},
|
|
{
|
|
"epoch": 0.27636363636363637,
|
|
"grad_norm": 4.136954307556152,
|
|
"learning_rate": 9.899325435782605e-06,
|
|
"loss": 1.5937362670898438,
|
|
"memory(GiB)": 64.44,
|
|
"step": 380,
|
|
"token_acc": 0.6055309734513274,
|
|
"train_speed(iter/s)": 0.461166
|
|
},
|
|
{
|
|
"epoch": 0.28,
|
|
"grad_norm": 5.078017711639404,
|
|
"learning_rate": 9.895368338284052e-06,
|
|
"loss": 1.465833854675293,
|
|
"memory(GiB)": 64.44,
|
|
"step": 385,
|
|
"token_acc": 0.639680262348842,
|
|
"train_speed(iter/s)": 0.461334
|
|
},
|
|
{
|
|
"epoch": 0.28363636363636363,
|
|
"grad_norm": 4.52872371673584,
|
|
"learning_rate": 9.891335785881983e-06,
|
|
"loss": 1.5299781799316405,
|
|
"memory(GiB)": 64.44,
|
|
"step": 390,
|
|
"token_acc": 0.628995633187773,
|
|
"train_speed(iter/s)": 0.461287
|
|
},
|
|
{
|
|
"epoch": 0.2872727272727273,
|
|
"grad_norm": 4.213586807250977,
|
|
"learning_rate": 9.887227840732264e-06,
|
|
"loss": 1.509080696105957,
|
|
"memory(GiB)": 64.44,
|
|
"step": 395,
|
|
"token_acc": 0.637125748502994,
|
|
"train_speed(iter/s)": 0.461288
|
|
},
|
|
{
|
|
"epoch": 0.2909090909090909,
|
|
"grad_norm": 7.419249534606934,
|
|
"learning_rate": 9.883044566152824e-06,
|
|
"loss": 1.566788387298584,
|
|
"memory(GiB)": 64.44,
|
|
"step": 400,
|
|
"token_acc": 0.6288452964779313,
|
|
"train_speed(iter/s)": 0.461127
|
|
},
|
|
{
|
|
"epoch": 0.29454545454545455,
|
|
"grad_norm": 4.467213153839111,
|
|
"learning_rate": 9.878786026622696e-06,
|
|
"loss": 1.5006021499633788,
|
|
"memory(GiB)": 64.44,
|
|
"step": 405,
|
|
"token_acc": 0.6567649106521235,
|
|
"train_speed(iter/s)": 0.461494
|
|
},
|
|
{
|
|
"epoch": 0.29818181818181816,
|
|
"grad_norm": 5.979769706726074,
|
|
"learning_rate": 9.874452287780998e-06,
|
|
"loss": 1.5232136726379395,
|
|
"memory(GiB)": 64.44,
|
|
"step": 410,
|
|
"token_acc": 0.6336889751050759,
|
|
"train_speed(iter/s)": 0.460984
|
|
},
|
|
{
|
|
"epoch": 0.3018181818181818,
|
|
"grad_norm": 4.591410160064697,
|
|
"learning_rate": 9.870043416425942e-06,
|
|
"loss": 1.4568933486938476,
|
|
"memory(GiB)": 64.44,
|
|
"step": 415,
|
|
"token_acc": 0.6160337552742616,
|
|
"train_speed(iter/s)": 0.461261
|
|
},
|
|
{
|
|
"epoch": 0.3054545454545455,
|
|
"grad_norm": 4.580557346343994,
|
|
"learning_rate": 9.865559480513797e-06,
|
|
"loss": 1.594901180267334,
|
|
"memory(GiB)": 64.44,
|
|
"step": 420,
|
|
"token_acc": 0.6259556218534402,
|
|
"train_speed(iter/s)": 0.461368
|
|
},
|
|
{
|
|
"epoch": 0.3090909090909091,
|
|
"grad_norm": 4.55140495300293,
|
|
"learning_rate": 9.861000549157839e-06,
|
|
"loss": 1.4992039680480957,
|
|
"memory(GiB)": 64.44,
|
|
"step": 425,
|
|
"token_acc": 0.63933626159102,
|
|
"train_speed(iter/s)": 0.461652
|
|
},
|
|
{
|
|
"epoch": 0.31272727272727274,
|
|
"grad_norm": 4.030912399291992,
|
|
"learning_rate": 9.856366692627296e-06,
|
|
"loss": 1.5157204627990724,
|
|
"memory(GiB)": 64.44,
|
|
"step": 430,
|
|
"token_acc": 0.6167369901547117,
|
|
"train_speed(iter/s)": 0.461884
|
|
},
|
|
{
|
|
"epoch": 0.31636363636363635,
|
|
"grad_norm": 4.710945129394531,
|
|
"learning_rate": 9.851657982346244e-06,
|
|
"loss": 1.524489402770996,
|
|
"memory(GiB)": 64.44,
|
|
"step": 435,
|
|
"token_acc": 0.6184729635646322,
|
|
"train_speed(iter/s)": 0.461909
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 4.01018762588501,
|
|
"learning_rate": 9.846874490892528e-06,
|
|
"loss": 1.5859844207763671,
|
|
"memory(GiB)": 64.44,
|
|
"step": 440,
|
|
"token_acc": 0.630187144019528,
|
|
"train_speed(iter/s)": 0.462173
|
|
},
|
|
{
|
|
"epoch": 0.3236363636363636,
|
|
"grad_norm": 4.102193832397461,
|
|
"learning_rate": 9.842016291996638e-06,
|
|
"loss": 1.6166738510131835,
|
|
"memory(GiB)": 64.44,
|
|
"step": 445,
|
|
"token_acc": 0.6040579710144928,
|
|
"train_speed(iter/s)": 0.462381
|
|
},
|
|
{
|
|
"epoch": 0.32727272727272727,
|
|
"grad_norm": 4.053755760192871,
|
|
"learning_rate": 9.837083460540565e-06,
|
|
"loss": 1.5673515319824218,
|
|
"memory(GiB)": 64.44,
|
|
"step": 450,
|
|
"token_acc": 0.6129250168880883,
|
|
"train_speed(iter/s)": 0.462643
|
|
},
|
|
{
|
|
"epoch": 0.33090909090909093,
|
|
"grad_norm": 4.424112796783447,
|
|
"learning_rate": 9.832076072556649e-06,
|
|
"loss": 1.4549053192138672,
|
|
"memory(GiB)": 64.44,
|
|
"step": 455,
|
|
"token_acc": 0.6802023121387283,
|
|
"train_speed(iter/s)": 0.462184
|
|
},
|
|
{
|
|
"epoch": 0.33454545454545453,
|
|
"grad_norm": 4.756387710571289,
|
|
"learning_rate": 9.82699420522641e-06,
|
|
"loss": 1.5000262260437012,
|
|
"memory(GiB)": 64.44,
|
|
"step": 460,
|
|
"token_acc": 0.6323158136332339,
|
|
"train_speed(iter/s)": 0.462438
|
|
},
|
|
{
|
|
"epoch": 0.3381818181818182,
|
|
"grad_norm": 4.306171894073486,
|
|
"learning_rate": 9.821837936879364e-06,
|
|
"loss": 1.4740945816040039,
|
|
"memory(GiB)": 64.44,
|
|
"step": 465,
|
|
"token_acc": 0.6459806231997905,
|
|
"train_speed(iter/s)": 0.462268
|
|
},
|
|
{
|
|
"epoch": 0.3418181818181818,
|
|
"grad_norm": 3.8091623783111572,
|
|
"learning_rate": 9.816607346991801e-06,
|
|
"loss": 1.4778839111328126,
|
|
"memory(GiB)": 64.44,
|
|
"step": 470,
|
|
"token_acc": 0.6258960573476703,
|
|
"train_speed(iter/s)": 0.46241
|
|
},
|
|
{
|
|
"epoch": 0.34545454545454546,
|
|
"grad_norm": 4.151661396026611,
|
|
"learning_rate": 9.811302516185572e-06,
|
|
"loss": 1.5408092498779298,
|
|
"memory(GiB)": 64.44,
|
|
"step": 475,
|
|
"token_acc": 0.6218589899975604,
|
|
"train_speed(iter/s)": 0.462383
|
|
},
|
|
{
|
|
"epoch": 0.3490909090909091,
|
|
"grad_norm": 4.382880210876465,
|
|
"learning_rate": 9.805923526226843e-06,
|
|
"loss": 1.5352145195007325,
|
|
"memory(GiB)": 64.44,
|
|
"step": 480,
|
|
"token_acc": 0.6172420674516065,
|
|
"train_speed(iter/s)": 0.462555
|
|
},
|
|
{
|
|
"epoch": 0.3527272727272727,
|
|
"grad_norm": 4.075268745422363,
|
|
"learning_rate": 9.800470460024835e-06,
|
|
"loss": 1.4970654487609862,
|
|
"memory(GiB)": 64.44,
|
|
"step": 485,
|
|
"token_acc": 0.6273666092943201,
|
|
"train_speed(iter/s)": 0.462386
|
|
},
|
|
{
|
|
"epoch": 0.3563636363636364,
|
|
"grad_norm": 4.728382110595703,
|
|
"learning_rate": 9.794943401630543e-06,
|
|
"loss": 1.4647459030151366,
|
|
"memory(GiB)": 64.44,
|
|
"step": 490,
|
|
"token_acc": 0.6437184827888621,
|
|
"train_speed(iter/s)": 0.462727
|
|
},
|
|
{
|
|
"epoch": 0.36,
|
|
"grad_norm": 3.9103758335113525,
|
|
"learning_rate": 9.78934243623544e-06,
|
|
"loss": 1.5026863098144532,
|
|
"memory(GiB)": 64.44,
|
|
"step": 495,
|
|
"token_acc": 0.641298958849777,
|
|
"train_speed(iter/s)": 0.462645
|
|
},
|
|
{
|
|
"epoch": 0.36363636363636365,
|
|
"grad_norm": 4.6131086349487305,
|
|
"learning_rate": 9.783667650170177e-06,
|
|
"loss": 1.5507336616516114,
|
|
"memory(GiB)": 64.44,
|
|
"step": 500,
|
|
"token_acc": 0.5890446521287642,
|
|
"train_speed(iter/s)": 0.462845
|
|
},
|
|
{
|
|
"epoch": 0.36727272727272725,
|
|
"grad_norm": 4.001507759094238,
|
|
"learning_rate": 9.777919130903228e-06,
|
|
"loss": 1.5329437255859375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 505,
|
|
"token_acc": 0.5871156661786238,
|
|
"train_speed(iter/s)": 0.462924
|
|
},
|
|
{
|
|
"epoch": 0.3709090909090909,
|
|
"grad_norm": 4.340017318725586,
|
|
"learning_rate": 9.77209696703957e-06,
|
|
"loss": 1.4759044647216797,
|
|
"memory(GiB)": 64.44,
|
|
"step": 510,
|
|
"token_acc": 0.6089973142345568,
|
|
"train_speed(iter/s)": 0.463076
|
|
},
|
|
{
|
|
"epoch": 0.37454545454545457,
|
|
"grad_norm": 4.634973049163818,
|
|
"learning_rate": 9.766201248319295e-06,
|
|
"loss": 1.5061256408691406,
|
|
"memory(GiB)": 64.44,
|
|
"step": 515,
|
|
"token_acc": 0.603759207518415,
|
|
"train_speed(iter/s)": 0.462996
|
|
},
|
|
{
|
|
"epoch": 0.3781818181818182,
|
|
"grad_norm": 3.9917287826538086,
|
|
"learning_rate": 9.760232065616233e-06,
|
|
"loss": 1.4734699249267578,
|
|
"memory(GiB)": 64.44,
|
|
"step": 520,
|
|
"token_acc": 0.6146010186757216,
|
|
"train_speed(iter/s)": 0.462922
|
|
},
|
|
{
|
|
"epoch": 0.38181818181818183,
|
|
"grad_norm": 4.241364479064941,
|
|
"learning_rate": 9.754189510936561e-06,
|
|
"loss": 1.5880233764648437,
|
|
"memory(GiB)": 64.44,
|
|
"step": 525,
|
|
"token_acc": 0.5949421345906558,
|
|
"train_speed(iter/s)": 0.463191
|
|
},
|
|
{
|
|
"epoch": 0.38545454545454544,
|
|
"grad_norm": 4.494994163513184,
|
|
"learning_rate": 9.748073677417371e-06,
|
|
"loss": 1.4729201316833496,
|
|
"memory(GiB)": 64.44,
|
|
"step": 530,
|
|
"token_acc": 0.6479982498359221,
|
|
"train_speed(iter/s)": 0.463331
|
|
},
|
|
{
|
|
"epoch": 0.3890909090909091,
|
|
"grad_norm": 4.426039218902588,
|
|
"learning_rate": 9.741884659325242e-06,
|
|
"loss": 1.5634471893310546,
|
|
"memory(GiB)": 64.44,
|
|
"step": 535,
|
|
"token_acc": 0.6027777777777777,
|
|
"train_speed(iter/s)": 0.463485
|
|
},
|
|
{
|
|
"epoch": 0.3927272727272727,
|
|
"grad_norm": 4.643458366394043,
|
|
"learning_rate": 9.735622552054786e-06,
|
|
"loss": 1.549774742126465,
|
|
"memory(GiB)": 64.44,
|
|
"step": 540,
|
|
"token_acc": 0.6494584005531229,
|
|
"train_speed(iter/s)": 0.463578
|
|
},
|
|
{
|
|
"epoch": 0.39636363636363636,
|
|
"grad_norm": 4.178001880645752,
|
|
"learning_rate": 9.729287452127178e-06,
|
|
"loss": 1.5748735427856446,
|
|
"memory(GiB)": 64.44,
|
|
"step": 545,
|
|
"token_acc": 0.5911589008363202,
|
|
"train_speed(iter/s)": 0.463746
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"grad_norm": 3.9572534561157227,
|
|
"learning_rate": 9.722879457188666e-06,
|
|
"loss": 1.5722023010253907,
|
|
"memory(GiB)": 64.44,
|
|
"step": 550,
|
|
"token_acc": 0.6157480314960629,
|
|
"train_speed(iter/s)": 0.46385
|
|
},
|
|
{
|
|
"epoch": 0.4036363636363636,
|
|
"grad_norm": 3.675825834274292,
|
|
"learning_rate": 9.716398666009067e-06,
|
|
"loss": 1.5456720352172852,
|
|
"memory(GiB)": 64.44,
|
|
"step": 555,
|
|
"token_acc": 0.6297201243891604,
|
|
"train_speed(iter/s)": 0.463989
|
|
},
|
|
{
|
|
"epoch": 0.4072727272727273,
|
|
"grad_norm": 3.31160306930542,
|
|
"learning_rate": 9.70984517848025e-06,
|
|
"loss": 1.6060224533081056,
|
|
"memory(GiB)": 64.44,
|
|
"step": 560,
|
|
"token_acc": 0.6269613259668508,
|
|
"train_speed(iter/s)": 0.46403
|
|
},
|
|
{
|
|
"epoch": 0.4109090909090909,
|
|
"grad_norm": 4.05873441696167,
|
|
"learning_rate": 9.703219095614583e-06,
|
|
"loss": 1.5519798278808594,
|
|
"memory(GiB)": 64.44,
|
|
"step": 565,
|
|
"token_acc": 0.6137970571373973,
|
|
"train_speed(iter/s)": 0.463591
|
|
},
|
|
{
|
|
"epoch": 0.41454545454545455,
|
|
"grad_norm": 4.47987699508667,
|
|
"learning_rate": 9.69652051954339e-06,
|
|
"loss": 1.5114023208618164,
|
|
"memory(GiB)": 64.44,
|
|
"step": 570,
|
|
"token_acc": 0.6189915174363808,
|
|
"train_speed(iter/s)": 0.463785
|
|
},
|
|
{
|
|
"epoch": 0.41818181818181815,
|
|
"grad_norm": 3.981717586517334,
|
|
"learning_rate": 9.68974955351537e-06,
|
|
"loss": 1.4660893440246583,
|
|
"memory(GiB)": 64.44,
|
|
"step": 575,
|
|
"token_acc": 0.6405854716299241,
|
|
"train_speed(iter/s)": 0.463635
|
|
},
|
|
{
|
|
"epoch": 0.4218181818181818,
|
|
"grad_norm": 4.990572452545166,
|
|
"learning_rate": 9.682906301895007e-06,
|
|
"loss": 1.4636313438415527,
|
|
"memory(GiB)": 64.44,
|
|
"step": 580,
|
|
"token_acc": 0.6241696474195196,
|
|
"train_speed(iter/s)": 0.463595
|
|
},
|
|
{
|
|
"epoch": 0.4254545454545455,
|
|
"grad_norm": 4.313155651092529,
|
|
"learning_rate": 9.675990870160958e-06,
|
|
"loss": 1.5075316429138184,
|
|
"memory(GiB)": 64.44,
|
|
"step": 585,
|
|
"token_acc": 0.597252882021094,
|
|
"train_speed(iter/s)": 0.463552
|
|
},
|
|
{
|
|
"epoch": 0.4290909090909091,
|
|
"grad_norm": 3.7366018295288086,
|
|
"learning_rate": 9.669003364904438e-06,
|
|
"loss": 1.5338315963745117,
|
|
"memory(GiB)": 64.44,
|
|
"step": 590,
|
|
"token_acc": 0.6448745802883665,
|
|
"train_speed(iter/s)": 0.46378
|
|
},
|
|
{
|
|
"epoch": 0.43272727272727274,
|
|
"grad_norm": 3.7582857608795166,
|
|
"learning_rate": 9.66194389382756e-06,
|
|
"loss": 1.5849173545837403,
|
|
"memory(GiB)": 64.44,
|
|
"step": 595,
|
|
"token_acc": 0.6113561190738699,
|
|
"train_speed(iter/s)": 0.4639
|
|
},
|
|
{
|
|
"epoch": 0.43636363636363634,
|
|
"grad_norm": 3.267991065979004,
|
|
"learning_rate": 9.65481256574169e-06,
|
|
"loss": 1.4916378021240235,
|
|
"memory(GiB)": 64.44,
|
|
"step": 600,
|
|
"token_acc": 0.6791225416036308,
|
|
"train_speed(iter/s)": 0.463656
|
|
},
|
|
{
|
|
"epoch": 0.44,
|
|
"grad_norm": 3.9313509464263916,
|
|
"learning_rate": 9.647609490565762e-06,
|
|
"loss": 1.486070442199707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 605,
|
|
"token_acc": 0.6800168634064081,
|
|
"train_speed(iter/s)": 0.463546
|
|
},
|
|
{
|
|
"epoch": 0.44363636363636366,
|
|
"grad_norm": 4.258699417114258,
|
|
"learning_rate": 9.640334779324589e-06,
|
|
"loss": 1.5047850608825684,
|
|
"memory(GiB)": 64.44,
|
|
"step": 610,
|
|
"token_acc": 0.6266666666666667,
|
|
"train_speed(iter/s)": 0.463839
|
|
},
|
|
{
|
|
"epoch": 0.44727272727272727,
|
|
"grad_norm": 4.216398239135742,
|
|
"learning_rate": 9.632988544147143e-06,
|
|
"loss": 1.4978628158569336,
|
|
"memory(GiB)": 64.44,
|
|
"step": 615,
|
|
"token_acc": 0.6422018348623854,
|
|
"train_speed(iter/s)": 0.463915
|
|
},
|
|
{
|
|
"epoch": 0.4509090909090909,
|
|
"grad_norm": 4.547543048858643,
|
|
"learning_rate": 9.625570898264837e-06,
|
|
"loss": 1.4514757156372071,
|
|
"memory(GiB)": 64.44,
|
|
"step": 620,
|
|
"token_acc": 0.643418013856813,
|
|
"train_speed(iter/s)": 0.464248
|
|
},
|
|
{
|
|
"epoch": 0.45454545454545453,
|
|
"grad_norm": 4.483821392059326,
|
|
"learning_rate": 9.618081956009772e-06,
|
|
"loss": 1.6247974395751954,
|
|
"memory(GiB)": 64.44,
|
|
"step": 625,
|
|
"token_acc": 0.6099777822662088,
|
|
"train_speed(iter/s)": 0.464408
|
|
},
|
|
{
|
|
"epoch": 0.4581818181818182,
|
|
"grad_norm": 4.023407936096191,
|
|
"learning_rate": 9.610521832812982e-06,
|
|
"loss": 1.5059638977050782,
|
|
"memory(GiB)": 64.44,
|
|
"step": 630,
|
|
"token_acc": 0.6152820968458463,
|
|
"train_speed(iter/s)": 0.464375
|
|
},
|
|
{
|
|
"epoch": 0.4618181818181818,
|
|
"grad_norm": 3.8227059841156006,
|
|
"learning_rate": 9.602890645202647e-06,
|
|
"loss": 1.5907604217529296,
|
|
"memory(GiB)": 64.44,
|
|
"step": 635,
|
|
"token_acc": 0.6613013698630137,
|
|
"train_speed(iter/s)": 0.46445
|
|
},
|
|
{
|
|
"epoch": 0.46545454545454545,
|
|
"grad_norm": 4.068626403808594,
|
|
"learning_rate": 9.595188510802299e-06,
|
|
"loss": 1.5536697387695313,
|
|
"memory(GiB)": 64.44,
|
|
"step": 640,
|
|
"token_acc": 0.6237579042457091,
|
|
"train_speed(iter/s)": 0.464352
|
|
},
|
|
{
|
|
"epoch": 0.4690909090909091,
|
|
"grad_norm": 4.049459934234619,
|
|
"learning_rate": 9.587415548329013e-06,
|
|
"loss": 1.4787757873535157,
|
|
"memory(GiB)": 64.44,
|
|
"step": 645,
|
|
"token_acc": 0.6037911168927709,
|
|
"train_speed(iter/s)": 0.464418
|
|
},
|
|
{
|
|
"epoch": 0.4727272727272727,
|
|
"grad_norm": 3.7583351135253906,
|
|
"learning_rate": 9.579571877591576e-06,
|
|
"loss": 1.5354381561279298,
|
|
"memory(GiB)": 64.44,
|
|
"step": 650,
|
|
"token_acc": 0.6248664815210425,
|
|
"train_speed(iter/s)": 0.464484
|
|
},
|
|
{
|
|
"epoch": 0.4763636363636364,
|
|
"grad_norm": 4.056765079498291,
|
|
"learning_rate": 9.571657619488635e-06,
|
|
"loss": 1.472429370880127,
|
|
"memory(GiB)": 64.44,
|
|
"step": 655,
|
|
"token_acc": 0.6824114356743319,
|
|
"train_speed(iter/s)": 0.464459
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"grad_norm": 3.8453173637390137,
|
|
"learning_rate": 9.563672896006842e-06,
|
|
"loss": 1.5348154067993165,
|
|
"memory(GiB)": 64.44,
|
|
"step": 660,
|
|
"token_acc": 0.6112652016215063,
|
|
"train_speed(iter/s)": 0.464415
|
|
},
|
|
{
|
|
"epoch": 0.48363636363636364,
|
|
"grad_norm": 4.265034198760986,
|
|
"learning_rate": 9.555617830218964e-06,
|
|
"loss": 1.5670215606689453,
|
|
"memory(GiB)": 64.44,
|
|
"step": 665,
|
|
"token_acc": 0.6162583518930957,
|
|
"train_speed(iter/s)": 0.464479
|
|
},
|
|
{
|
|
"epoch": 0.48727272727272725,
|
|
"grad_norm": 4.171985149383545,
|
|
"learning_rate": 9.547492546281996e-06,
|
|
"loss": 1.5287425994873047,
|
|
"memory(GiB)": 64.44,
|
|
"step": 670,
|
|
"token_acc": 0.6656884875846502,
|
|
"train_speed(iter/s)": 0.464691
|
|
},
|
|
{
|
|
"epoch": 0.4909090909090909,
|
|
"grad_norm": 3.559645414352417,
|
|
"learning_rate": 9.539297169435238e-06,
|
|
"loss": 1.5655836105346679,
|
|
"memory(GiB)": 64.44,
|
|
"step": 675,
|
|
"token_acc": 0.6459044368600683,
|
|
"train_speed(iter/s)": 0.464634
|
|
},
|
|
{
|
|
"epoch": 0.49454545454545457,
|
|
"grad_norm": 3.3797030448913574,
|
|
"learning_rate": 9.531031825998374e-06,
|
|
"loss": 1.4685256004333496,
|
|
"memory(GiB)": 64.44,
|
|
"step": 680,
|
|
"token_acc": 0.6592842240410849,
|
|
"train_speed(iter/s)": 0.464582
|
|
},
|
|
{
|
|
"epoch": 0.49818181818181817,
|
|
"grad_norm": 4.35471248626709,
|
|
"learning_rate": 9.522696643369515e-06,
|
|
"loss": 1.6089235305786134,
|
|
"memory(GiB)": 64.44,
|
|
"step": 685,
|
|
"token_acc": 0.6074537540805223,
|
|
"train_speed(iter/s)": 0.464786
|
|
},
|
|
{
|
|
"epoch": 0.5018181818181818,
|
|
"grad_norm": 4.220783233642578,
|
|
"learning_rate": 9.514291750023248e-06,
|
|
"loss": 1.4950236320495605,
|
|
"memory(GiB)": 64.44,
|
|
"step": 690,
|
|
"token_acc": 0.6523521795425119,
|
|
"train_speed(iter/s)": 0.46478
|
|
},
|
|
{
|
|
"epoch": 0.5054545454545455,
|
|
"grad_norm": 3.488475799560547,
|
|
"learning_rate": 9.505817275508636e-06,
|
|
"loss": 1.484272575378418,
|
|
"memory(GiB)": 64.44,
|
|
"step": 695,
|
|
"token_acc": 0.6576594090202177,
|
|
"train_speed(iter/s)": 0.464879
|
|
},
|
|
{
|
|
"epoch": 0.509090909090909,
|
|
"grad_norm": 3.5033652782440186,
|
|
"learning_rate": 9.497273350447247e-06,
|
|
"loss": 1.4467235565185548,
|
|
"memory(GiB)": 64.44,
|
|
"step": 700,
|
|
"token_acc": 0.6930147058823529,
|
|
"train_speed(iter/s)": 0.464836
|
|
},
|
|
{
|
|
"epoch": 0.5127272727272727,
|
|
"grad_norm": 3.9645371437072754,
|
|
"learning_rate": 9.488660106531115e-06,
|
|
"loss": 1.613375473022461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 705,
|
|
"token_acc": 0.5843076123668486,
|
|
"train_speed(iter/s)": 0.465038
|
|
},
|
|
{
|
|
"epoch": 0.5163636363636364,
|
|
"grad_norm": 3.6118175983428955,
|
|
"learning_rate": 9.479977676520734e-06,
|
|
"loss": 1.4993645668029785,
|
|
"memory(GiB)": 64.44,
|
|
"step": 710,
|
|
"token_acc": 0.595,
|
|
"train_speed(iter/s)": 0.464939
|
|
},
|
|
{
|
|
"epoch": 0.52,
|
|
"grad_norm": 3.7442569732666016,
|
|
"learning_rate": 9.471226194242987e-06,
|
|
"loss": 1.5396413803100586,
|
|
"memory(GiB)": 64.44,
|
|
"step": 715,
|
|
"token_acc": 0.6280975981700343,
|
|
"train_speed(iter/s)": 0.464999
|
|
},
|
|
{
|
|
"epoch": 0.5236363636363637,
|
|
"grad_norm": 4.1397576332092285,
|
|
"learning_rate": 9.462405794589109e-06,
|
|
"loss": 1.5331635475158691,
|
|
"memory(GiB)": 64.44,
|
|
"step": 720,
|
|
"token_acc": 0.6184054885261415,
|
|
"train_speed(iter/s)": 0.465084
|
|
},
|
|
{
|
|
"epoch": 0.5272727272727272,
|
|
"grad_norm": 3.7472028732299805,
|
|
"learning_rate": 9.453516613512585e-06,
|
|
"loss": 1.5267699241638184,
|
|
"memory(GiB)": 64.44,
|
|
"step": 725,
|
|
"token_acc": 0.6468453940737755,
|
|
"train_speed(iter/s)": 0.465192
|
|
},
|
|
{
|
|
"epoch": 0.5309090909090909,
|
|
"grad_norm": 3.989107131958008,
|
|
"learning_rate": 9.44455878802707e-06,
|
|
"loss": 1.585801124572754,
|
|
"memory(GiB)": 64.44,
|
|
"step": 730,
|
|
"token_acc": 0.6121025641025641,
|
|
"train_speed(iter/s)": 0.465308
|
|
},
|
|
{
|
|
"epoch": 0.5345454545454545,
|
|
"grad_norm": 3.7448976039886475,
|
|
"learning_rate": 9.435532456204268e-06,
|
|
"loss": 1.5152654647827148,
|
|
"memory(GiB)": 64.44,
|
|
"step": 735,
|
|
"token_acc": 0.6513828238719068,
|
|
"train_speed(iter/s)": 0.465448
|
|
},
|
|
{
|
|
"epoch": 0.5381818181818182,
|
|
"grad_norm": 4.547407150268555,
|
|
"learning_rate": 9.426437757171811e-06,
|
|
"loss": 1.5738471031188965,
|
|
"memory(GiB)": 64.44,
|
|
"step": 740,
|
|
"token_acc": 0.6231003039513677,
|
|
"train_speed(iter/s)": 0.46574
|
|
},
|
|
{
|
|
"epoch": 0.5418181818181819,
|
|
"grad_norm": 3.8383874893188477,
|
|
"learning_rate": 9.417274831111109e-06,
|
|
"loss": 1.4214591979980469,
|
|
"memory(GiB)": 64.44,
|
|
"step": 745,
|
|
"token_acc": 0.676269621421976,
|
|
"train_speed(iter/s)": 0.465722
|
|
},
|
|
{
|
|
"epoch": 0.5454545454545454,
|
|
"grad_norm": 3.543907880783081,
|
|
"learning_rate": 9.408043819255194e-06,
|
|
"loss": 1.430149745941162,
|
|
"memory(GiB)": 64.44,
|
|
"step": 750,
|
|
"token_acc": 0.6681834091704585,
|
|
"train_speed(iter/s)": 0.465558
|
|
},
|
|
{
|
|
"epoch": 0.5490909090909091,
|
|
"grad_norm": 3.295700788497925,
|
|
"learning_rate": 9.398744863886534e-06,
|
|
"loss": 1.4326640129089356,
|
|
"memory(GiB)": 64.44,
|
|
"step": 755,
|
|
"token_acc": 0.6369306369306369,
|
|
"train_speed(iter/s)": 0.46536
|
|
},
|
|
{
|
|
"epoch": 0.5527272727272727,
|
|
"grad_norm": 3.635788917541504,
|
|
"learning_rate": 9.389378108334852e-06,
|
|
"loss": 1.523570442199707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 760,
|
|
"token_acc": 0.6374558303886926,
|
|
"train_speed(iter/s)": 0.465517
|
|
},
|
|
{
|
|
"epoch": 0.5563636363636364,
|
|
"grad_norm": 4.037724494934082,
|
|
"learning_rate": 9.379943696974912e-06,
|
|
"loss": 1.4817398071289063,
|
|
"memory(GiB)": 64.44,
|
|
"step": 765,
|
|
"token_acc": 0.6240034475328593,
|
|
"train_speed(iter/s)": 0.465524
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"grad_norm": 3.898163318634033,
|
|
"learning_rate": 9.370441775224287e-06,
|
|
"loss": 1.4279541015625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 770,
|
|
"token_acc": 0.6828789958815454,
|
|
"train_speed(iter/s)": 0.465562
|
|
},
|
|
{
|
|
"epoch": 0.5636363636363636,
|
|
"grad_norm": 4.19639253616333,
|
|
"learning_rate": 9.360872489541129e-06,
|
|
"loss": 1.5240070343017578,
|
|
"memory(GiB)": 64.44,
|
|
"step": 775,
|
|
"token_acc": 0.615297833935018,
|
|
"train_speed(iter/s)": 0.465613
|
|
},
|
|
{
|
|
"epoch": 0.5672727272727273,
|
|
"grad_norm": 3.475050449371338,
|
|
"learning_rate": 9.351235987421901e-06,
|
|
"loss": 1.4884565353393555,
|
|
"memory(GiB)": 64.44,
|
|
"step": 780,
|
|
"token_acc": 0.6517972350230414,
|
|
"train_speed(iter/s)": 0.465728
|
|
},
|
|
{
|
|
"epoch": 0.5709090909090909,
|
|
"grad_norm": 4.0276265144348145,
|
|
"learning_rate": 9.341532417399113e-06,
|
|
"loss": 1.5572041511535644,
|
|
"memory(GiB)": 64.44,
|
|
"step": 785,
|
|
"token_acc": 0.6271844660194175,
|
|
"train_speed(iter/s)": 0.465581
|
|
},
|
|
{
|
|
"epoch": 0.5745454545454546,
|
|
"grad_norm": 3.860581398010254,
|
|
"learning_rate": 9.331761929039024e-06,
|
|
"loss": 1.4943767547607423,
|
|
"memory(GiB)": 64.44,
|
|
"step": 790,
|
|
"token_acc": 0.6216508210890234,
|
|
"train_speed(iter/s)": 0.465606
|
|
},
|
|
{
|
|
"epoch": 0.5781818181818181,
|
|
"grad_norm": 3.591012477874756,
|
|
"learning_rate": 9.321924672939349e-06,
|
|
"loss": 1.4894336700439452,
|
|
"memory(GiB)": 64.44,
|
|
"step": 795,
|
|
"token_acc": 0.599320882852292,
|
|
"train_speed(iter/s)": 0.465616
|
|
},
|
|
{
|
|
"epoch": 0.5818181818181818,
|
|
"grad_norm": 3.522542953491211,
|
|
"learning_rate": 9.312020800726918e-06,
|
|
"loss": 1.5618131637573243,
|
|
"memory(GiB)": 64.44,
|
|
"step": 800,
|
|
"token_acc": 0.6190360272638754,
|
|
"train_speed(iter/s)": 0.465756
|
|
},
|
|
{
|
|
"epoch": 0.5854545454545454,
|
|
"grad_norm": 3.7886240482330322,
|
|
"learning_rate": 9.302050465055357e-06,
|
|
"loss": 1.5633182525634766,
|
|
"memory(GiB)": 64.44,
|
|
"step": 805,
|
|
"token_acc": 0.6189818014282423,
|
|
"train_speed(iter/s)": 0.46599
|
|
},
|
|
{
|
|
"epoch": 0.5890909090909091,
|
|
"grad_norm": 3.3036091327667236,
|
|
"learning_rate": 9.292013819602726e-06,
|
|
"loss": 1.5002209663391113,
|
|
"memory(GiB)": 64.44,
|
|
"step": 810,
|
|
"token_acc": 0.607620320855615,
|
|
"train_speed(iter/s)": 0.465952
|
|
},
|
|
{
|
|
"epoch": 0.5927272727272728,
|
|
"grad_norm": 3.9183084964752197,
|
|
"learning_rate": 9.281911019069158e-06,
|
|
"loss": 1.5537723541259765,
|
|
"memory(GiB)": 64.44,
|
|
"step": 815,
|
|
"token_acc": 0.6308488126028685,
|
|
"train_speed(iter/s)": 0.466075
|
|
},
|
|
{
|
|
"epoch": 0.5963636363636363,
|
|
"grad_norm": 3.89871883392334,
|
|
"learning_rate": 9.271742219174464e-06,
|
|
"loss": 1.4676464080810547,
|
|
"memory(GiB)": 64.44,
|
|
"step": 820,
|
|
"token_acc": 0.6469534050179212,
|
|
"train_speed(iter/s)": 0.4662
|
|
},
|
|
{
|
|
"epoch": 0.6,
|
|
"grad_norm": 3.631673574447632,
|
|
"learning_rate": 9.261507576655738e-06,
|
|
"loss": 1.498595905303955,
|
|
"memory(GiB)": 64.44,
|
|
"step": 825,
|
|
"token_acc": 0.6465152098575279,
|
|
"train_speed(iter/s)": 0.466232
|
|
},
|
|
{
|
|
"epoch": 0.6036363636363636,
|
|
"grad_norm": 4.689718723297119,
|
|
"learning_rate": 9.251207249264943e-06,
|
|
"loss": 1.5046562194824218,
|
|
"memory(GiB)": 64.44,
|
|
"step": 830,
|
|
"token_acc": 0.6152659984579799,
|
|
"train_speed(iter/s)": 0.466249
|
|
},
|
|
{
|
|
"epoch": 0.6072727272727273,
|
|
"grad_norm": 4.173959255218506,
|
|
"learning_rate": 9.24084139576648e-06,
|
|
"loss": 1.5240567207336426,
|
|
"memory(GiB)": 64.44,
|
|
"step": 835,
|
|
"token_acc": 0.5942028985507246,
|
|
"train_speed(iter/s)": 0.466409
|
|
},
|
|
{
|
|
"epoch": 0.610909090909091,
|
|
"grad_norm": 3.9074172973632812,
|
|
"learning_rate": 9.230410175934733e-06,
|
|
"loss": 1.6020416259765624,
|
|
"memory(GiB)": 64.44,
|
|
"step": 840,
|
|
"token_acc": 0.6180246913580247,
|
|
"train_speed(iter/s)": 0.466524
|
|
},
|
|
{
|
|
"epoch": 0.6145454545454545,
|
|
"grad_norm": 3.913738489151001,
|
|
"learning_rate": 9.219913750551617e-06,
|
|
"loss": 1.395237636566162,
|
|
"memory(GiB)": 64.44,
|
|
"step": 845,
|
|
"token_acc": 0.665514261019879,
|
|
"train_speed(iter/s)": 0.466525
|
|
},
|
|
{
|
|
"epoch": 0.6181818181818182,
|
|
"grad_norm": 3.6115214824676514,
|
|
"learning_rate": 9.20935228140409e-06,
|
|
"loss": 1.5188591957092286,
|
|
"memory(GiB)": 64.44,
|
|
"step": 850,
|
|
"token_acc": 0.635885447106955,
|
|
"train_speed(iter/s)": 0.466588
|
|
},
|
|
{
|
|
"epoch": 0.6218181818181818,
|
|
"grad_norm": 4.478565692901611,
|
|
"learning_rate": 9.19872593128167e-06,
|
|
"loss": 1.522629451751709,
|
|
"memory(GiB)": 64.44,
|
|
"step": 855,
|
|
"token_acc": 0.6044009779951101,
|
|
"train_speed(iter/s)": 0.466658
|
|
},
|
|
{
|
|
"epoch": 0.6254545454545455,
|
|
"grad_norm": 3.820162296295166,
|
|
"learning_rate": 9.188034863973913e-06,
|
|
"loss": 1.4850282669067383,
|
|
"memory(GiB)": 64.44,
|
|
"step": 860,
|
|
"token_acc": 0.5907387580299786,
|
|
"train_speed(iter/s)": 0.466569
|
|
},
|
|
{
|
|
"epoch": 0.6290909090909091,
|
|
"grad_norm": 3.6879067420959473,
|
|
"learning_rate": 9.177279244267902e-06,
|
|
"loss": 1.5320751190185546,
|
|
"memory(GiB)": 64.44,
|
|
"step": 865,
|
|
"token_acc": 0.6051044083526682,
|
|
"train_speed(iter/s)": 0.466644
|
|
},
|
|
{
|
|
"epoch": 0.6327272727272727,
|
|
"grad_norm": 4.055009365081787,
|
|
"learning_rate": 9.166459237945696e-06,
|
|
"loss": 1.5284514427185059,
|
|
"memory(GiB)": 64.44,
|
|
"step": 870,
|
|
"token_acc": 0.6201756177251379,
|
|
"train_speed(iter/s)": 0.46667
|
|
},
|
|
{
|
|
"epoch": 0.6363636363636364,
|
|
"grad_norm": 4.183087348937988,
|
|
"learning_rate": 9.155575011781779e-06,
|
|
"loss": 1.554479217529297,
|
|
"memory(GiB)": 64.44,
|
|
"step": 875,
|
|
"token_acc": 0.6525652565256526,
|
|
"train_speed(iter/s)": 0.466644
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 3.3368186950683594,
|
|
"learning_rate": 9.14462673354049e-06,
|
|
"loss": 1.497207260131836,
|
|
"memory(GiB)": 64.44,
|
|
"step": 880,
|
|
"token_acc": 0.5918765866041789,
|
|
"train_speed(iter/s)": 0.466658
|
|
},
|
|
{
|
|
"epoch": 0.6436363636363637,
|
|
"grad_norm": 3.610478401184082,
|
|
"learning_rate": 9.133614571973436e-06,
|
|
"loss": 1.5291955947875977,
|
|
"memory(GiB)": 64.44,
|
|
"step": 885,
|
|
"token_acc": 0.628099173553719,
|
|
"train_speed(iter/s)": 0.466633
|
|
},
|
|
{
|
|
"epoch": 0.6472727272727272,
|
|
"grad_norm": 3.1254570484161377,
|
|
"learning_rate": 9.122538696816896e-06,
|
|
"loss": 1.5357439041137695,
|
|
"memory(GiB)": 64.44,
|
|
"step": 890,
|
|
"token_acc": 0.6493339063171465,
|
|
"train_speed(iter/s)": 0.466674
|
|
},
|
|
{
|
|
"epoch": 0.6509090909090909,
|
|
"grad_norm": 3.9878220558166504,
|
|
"learning_rate": 9.111399278789193e-06,
|
|
"loss": 1.4707233428955078,
|
|
"memory(GiB)": 64.44,
|
|
"step": 895,
|
|
"token_acc": 0.6466725043782837,
|
|
"train_speed(iter/s)": 0.466691
|
|
},
|
|
{
|
|
"epoch": 0.6545454545454545,
|
|
"grad_norm": 4.023503303527832,
|
|
"learning_rate": 9.100196489588076e-06,
|
|
"loss": 1.5020041465759277,
|
|
"memory(GiB)": 64.44,
|
|
"step": 900,
|
|
"token_acc": 0.6528338136407301,
|
|
"train_speed(iter/s)": 0.466756
|
|
},
|
|
{
|
|
"epoch": 0.6581818181818182,
|
|
"grad_norm": 3.8978965282440186,
|
|
"learning_rate": 9.088930501888064e-06,
|
|
"loss": 1.533686637878418,
|
|
"memory(GiB)": 64.44,
|
|
"step": 905,
|
|
"token_acc": 0.6118152524167562,
|
|
"train_speed(iter/s)": 0.466876
|
|
},
|
|
{
|
|
"epoch": 0.6618181818181819,
|
|
"grad_norm": 3.952993154525757,
|
|
"learning_rate": 9.077601489337788e-06,
|
|
"loss": 1.5572603225708008,
|
|
"memory(GiB)": 64.44,
|
|
"step": 910,
|
|
"token_acc": 0.6244158878504673,
|
|
"train_speed(iter/s)": 0.466842
|
|
},
|
|
{
|
|
"epoch": 0.6654545454545454,
|
|
"grad_norm": 3.9382662773132324,
|
|
"learning_rate": 9.066209626557318e-06,
|
|
"loss": 1.5762770652770997,
|
|
"memory(GiB)": 64.44,
|
|
"step": 915,
|
|
"token_acc": 0.5991866244916403,
|
|
"train_speed(iter/s)": 0.466835
|
|
},
|
|
{
|
|
"epoch": 0.6690909090909091,
|
|
"grad_norm": 4.0595808029174805,
|
|
"learning_rate": 9.054755089135463e-06,
|
|
"loss": 1.486503791809082,
|
|
"memory(GiB)": 64.44,
|
|
"step": 920,
|
|
"token_acc": 0.6291459369817579,
|
|
"train_speed(iter/s)": 0.466902
|
|
},
|
|
{
|
|
"epoch": 0.6727272727272727,
|
|
"grad_norm": 3.5437912940979004,
|
|
"learning_rate": 9.043238053627073e-06,
|
|
"loss": 1.4707436561584473,
|
|
"memory(GiB)": 64.44,
|
|
"step": 925,
|
|
"token_acc": 0.6164813919768004,
|
|
"train_speed(iter/s)": 0.466968
|
|
},
|
|
{
|
|
"epoch": 0.6763636363636364,
|
|
"grad_norm": 3.699765920639038,
|
|
"learning_rate": 9.031658697550311e-06,
|
|
"loss": 1.5214021682739258,
|
|
"memory(GiB)": 64.44,
|
|
"step": 930,
|
|
"token_acc": 0.6087981146897093,
|
|
"train_speed(iter/s)": 0.467078
|
|
},
|
|
{
|
|
"epoch": 0.68,
|
|
"grad_norm": 3.693669557571411,
|
|
"learning_rate": 9.020017199383926e-06,
|
|
"loss": 1.4087844848632813,
|
|
"memory(GiB)": 64.44,
|
|
"step": 935,
|
|
"token_acc": 0.6391267123287672,
|
|
"train_speed(iter/s)": 0.467101
|
|
},
|
|
{
|
|
"epoch": 0.6836363636363636,
|
|
"grad_norm": 3.6385085582733154,
|
|
"learning_rate": 9.008313738564485e-06,
|
|
"loss": 1.3953245162963868,
|
|
"memory(GiB)": 64.44,
|
|
"step": 940,
|
|
"token_acc": 0.6613488034807832,
|
|
"train_speed(iter/s)": 0.46707
|
|
},
|
|
{
|
|
"epoch": 0.6872727272727273,
|
|
"grad_norm": 3.7062394618988037,
|
|
"learning_rate": 8.996548495483627e-06,
|
|
"loss": 1.57774019241333,
|
|
"memory(GiB)": 64.44,
|
|
"step": 945,
|
|
"token_acc": 0.6061627347135291,
|
|
"train_speed(iter/s)": 0.466985
|
|
},
|
|
{
|
|
"epoch": 0.6909090909090909,
|
|
"grad_norm": 4.307070732116699,
|
|
"learning_rate": 8.984721651485276e-06,
|
|
"loss": 1.5156424522399903,
|
|
"memory(GiB)": 64.44,
|
|
"step": 950,
|
|
"token_acc": 0.6389528795811519,
|
|
"train_speed(iter/s)": 0.467075
|
|
},
|
|
{
|
|
"epoch": 0.6945454545454546,
|
|
"grad_norm": 4.296825408935547,
|
|
"learning_rate": 8.972833388862835e-06,
|
|
"loss": 1.465583324432373,
|
|
"memory(GiB)": 64.44,
|
|
"step": 955,
|
|
"token_acc": 0.6539109912339852,
|
|
"train_speed(iter/s)": 0.467103
|
|
},
|
|
{
|
|
"epoch": 0.6981818181818182,
|
|
"grad_norm": 3.43802809715271,
|
|
"learning_rate": 8.960883890856387e-06,
|
|
"loss": 1.4816664695739745,
|
|
"memory(GiB)": 64.44,
|
|
"step": 960,
|
|
"token_acc": 0.625,
|
|
"train_speed(iter/s)": 0.467064
|
|
},
|
|
{
|
|
"epoch": 0.7018181818181818,
|
|
"grad_norm": 3.5750675201416016,
|
|
"learning_rate": 8.948873341649873e-06,
|
|
"loss": 1.4105497360229493,
|
|
"memory(GiB)": 64.44,
|
|
"step": 965,
|
|
"token_acc": 0.6611288604898828,
|
|
"train_speed(iter/s)": 0.466926
|
|
},
|
|
{
|
|
"epoch": 0.7054545454545454,
|
|
"grad_norm": 3.355591058731079,
|
|
"learning_rate": 8.936801926368245e-06,
|
|
"loss": 1.4713809013366699,
|
|
"memory(GiB)": 64.44,
|
|
"step": 970,
|
|
"token_acc": 0.6177150192554557,
|
|
"train_speed(iter/s)": 0.466928
|
|
},
|
|
{
|
|
"epoch": 0.7090909090909091,
|
|
"grad_norm": 3.269821882247925,
|
|
"learning_rate": 8.924669831074617e-06,
|
|
"loss": 1.5545562744140624,
|
|
"memory(GiB)": 64.44,
|
|
"step": 975,
|
|
"token_acc": 0.6235870384325546,
|
|
"train_speed(iter/s)": 0.466981
|
|
},
|
|
{
|
|
"epoch": 0.7127272727272728,
|
|
"grad_norm": 3.8328733444213867,
|
|
"learning_rate": 8.912477242767395e-06,
|
|
"loss": 1.4566053390502929,
|
|
"memory(GiB)": 64.44,
|
|
"step": 980,
|
|
"token_acc": 0.6365413860525744,
|
|
"train_speed(iter/s)": 0.466921
|
|
},
|
|
{
|
|
"epoch": 0.7163636363636363,
|
|
"grad_norm": 3.6056768894195557,
|
|
"learning_rate": 8.900224349377396e-06,
|
|
"loss": 1.5423126220703125,
|
|
"memory(GiB)": 64.44,
|
|
"step": 985,
|
|
"token_acc": 0.6146384479717814,
|
|
"train_speed(iter/s)": 0.466988
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"grad_norm": 3.902472972869873,
|
|
"learning_rate": 8.887911339764954e-06,
|
|
"loss": 1.5114011764526367,
|
|
"memory(GiB)": 64.44,
|
|
"step": 990,
|
|
"token_acc": 0.623652942801879,
|
|
"train_speed(iter/s)": 0.467217
|
|
},
|
|
{
|
|
"epoch": 0.7236363636363636,
|
|
"grad_norm": 3.429377794265747,
|
|
"learning_rate": 8.875538403717003e-06,
|
|
"loss": 1.498643398284912,
|
|
"memory(GiB)": 64.44,
|
|
"step": 995,
|
|
"token_acc": 0.6255022203425671,
|
|
"train_speed(iter/s)": 0.467293
|
|
},
|
|
{
|
|
"epoch": 0.7272727272727273,
|
|
"grad_norm": 4.113056659698486,
|
|
"learning_rate": 8.863105731944153e-06,
|
|
"loss": 1.5103962898254395,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1000,
|
|
"token_acc": 0.6358121770388677,
|
|
"train_speed(iter/s)": 0.467345
|
|
},
|
|
{
|
|
"epoch": 0.730909090909091,
|
|
"grad_norm": 3.749250888824463,
|
|
"learning_rate": 8.85061351607776e-06,
|
|
"loss": 1.4655189514160156,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1005,
|
|
"token_acc": 0.6587071876507478,
|
|
"train_speed(iter/s)": 0.467499
|
|
},
|
|
{
|
|
"epoch": 0.7345454545454545,
|
|
"grad_norm": 3.5996060371398926,
|
|
"learning_rate": 8.838061948666956e-06,
|
|
"loss": 1.523210334777832,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1010,
|
|
"token_acc": 0.6319569120287253,
|
|
"train_speed(iter/s)": 0.467546
|
|
},
|
|
{
|
|
"epoch": 0.7381818181818182,
|
|
"grad_norm": 3.7059760093688965,
|
|
"learning_rate": 8.825451223175697e-06,
|
|
"loss": 1.4700775146484375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1015,
|
|
"token_acc": 0.6333166583291646,
|
|
"train_speed(iter/s)": 0.467627
|
|
},
|
|
{
|
|
"epoch": 0.7418181818181818,
|
|
"grad_norm": 4.267899990081787,
|
|
"learning_rate": 8.812781533979766e-06,
|
|
"loss": 1.494734764099121,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1020,
|
|
"token_acc": 0.6311154598825832,
|
|
"train_speed(iter/s)": 0.467767
|
|
},
|
|
{
|
|
"epoch": 0.7454545454545455,
|
|
"grad_norm": 3.691187620162964,
|
|
"learning_rate": 8.800053076363792e-06,
|
|
"loss": 1.4671652793884278,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1025,
|
|
"token_acc": 0.6183923110528615,
|
|
"train_speed(iter/s)": 0.467891
|
|
},
|
|
{
|
|
"epoch": 0.7490909090909091,
|
|
"grad_norm": 3.3653814792633057,
|
|
"learning_rate": 8.78726604651823e-06,
|
|
"loss": 1.4787571907043457,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1030,
|
|
"token_acc": 0.6354618722876627,
|
|
"train_speed(iter/s)": 0.467916
|
|
},
|
|
{
|
|
"epoch": 0.7527272727272727,
|
|
"grad_norm": 3.4715375900268555,
|
|
"learning_rate": 8.774420641536337e-06,
|
|
"loss": 1.4089265823364259,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1035,
|
|
"token_acc": 0.6256326825741142,
|
|
"train_speed(iter/s)": 0.467973
|
|
},
|
|
{
|
|
"epoch": 0.7563636363636363,
|
|
"grad_norm": 3.5086214542388916,
|
|
"learning_rate": 8.761517059411144e-06,
|
|
"loss": 1.4492218017578125,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1040,
|
|
"token_acc": 0.6436498150431565,
|
|
"train_speed(iter/s)": 0.467852
|
|
},
|
|
{
|
|
"epoch": 0.76,
|
|
"grad_norm": 3.7965283393859863,
|
|
"learning_rate": 8.74855549903239e-06,
|
|
"loss": 1.4732137680053712,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1045,
|
|
"token_acc": 0.6451612903225806,
|
|
"train_speed(iter/s)": 0.467928
|
|
},
|
|
{
|
|
"epoch": 0.7636363636363637,
|
|
"grad_norm": 4.0382232666015625,
|
|
"learning_rate": 8.735536160183464e-06,
|
|
"loss": 1.5476658821105957,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1050,
|
|
"token_acc": 0.6137471146447807,
|
|
"train_speed(iter/s)": 0.468098
|
|
},
|
|
{
|
|
"epoch": 0.7672727272727272,
|
|
"grad_norm": 3.921508312225342,
|
|
"learning_rate": 8.722459243538333e-06,
|
|
"loss": 1.4485240936279298,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1055,
|
|
"token_acc": 0.6001534919416731,
|
|
"train_speed(iter/s)": 0.468144
|
|
},
|
|
{
|
|
"epoch": 0.7709090909090909,
|
|
"grad_norm": 3.7814698219299316,
|
|
"learning_rate": 8.709324950658432e-06,
|
|
"loss": 1.545788288116455,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1060,
|
|
"token_acc": 0.6441542542338299,
|
|
"train_speed(iter/s)": 0.468211
|
|
},
|
|
{
|
|
"epoch": 0.7745454545454545,
|
|
"grad_norm": 3.402313232421875,
|
|
"learning_rate": 8.69613348398957e-06,
|
|
"loss": 1.428826904296875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1065,
|
|
"token_acc": 0.6991242702251876,
|
|
"train_speed(iter/s)": 0.468195
|
|
},
|
|
{
|
|
"epoch": 0.7781818181818182,
|
|
"grad_norm": 3.5627617835998535,
|
|
"learning_rate": 8.682885046858803e-06,
|
|
"loss": 1.4530531883239746,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1070,
|
|
"token_acc": 0.6563146997929606,
|
|
"train_speed(iter/s)": 0.468226
|
|
},
|
|
{
|
|
"epoch": 0.7818181818181819,
|
|
"grad_norm": 3.311058521270752,
|
|
"learning_rate": 8.66957984347131e-06,
|
|
"loss": 1.5207694053649903,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1075,
|
|
"token_acc": 0.666245487364621,
|
|
"train_speed(iter/s)": 0.46826
|
|
},
|
|
{
|
|
"epoch": 0.7854545454545454,
|
|
"grad_norm": 3.9101316928863525,
|
|
"learning_rate": 8.65621807890723e-06,
|
|
"loss": 1.4987771034240722,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1080,
|
|
"token_acc": 0.6131436314363143,
|
|
"train_speed(iter/s)": 0.468199
|
|
},
|
|
{
|
|
"epoch": 0.7890909090909091,
|
|
"grad_norm": 4.084334850311279,
|
|
"learning_rate": 8.642799959118517e-06,
|
|
"loss": 1.4973735809326172,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1085,
|
|
"token_acc": 0.6244131455399061,
|
|
"train_speed(iter/s)": 0.468307
|
|
},
|
|
{
|
|
"epoch": 0.7927272727272727,
|
|
"grad_norm": 3.905789852142334,
|
|
"learning_rate": 8.62932569092575e-06,
|
|
"loss": 1.5306825637817383,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1090,
|
|
"token_acc": 0.6427216047709406,
|
|
"train_speed(iter/s)": 0.468383
|
|
},
|
|
{
|
|
"epoch": 0.7963636363636364,
|
|
"grad_norm": 3.959667682647705,
|
|
"learning_rate": 8.61579548201496e-06,
|
|
"loss": 1.5060791969299316,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1095,
|
|
"token_acc": 0.6090833678971381,
|
|
"train_speed(iter/s)": 0.468461
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"grad_norm": 3.306290626525879,
|
|
"learning_rate": 8.602209540934419e-06,
|
|
"loss": 1.4021387100219727,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1100,
|
|
"token_acc": 0.6274509803921569,
|
|
"train_speed(iter/s)": 0.468377
|
|
},
|
|
{
|
|
"epoch": 0.8036363636363636,
|
|
"grad_norm": 3.746054172515869,
|
|
"learning_rate": 8.588568077091427e-06,
|
|
"loss": 1.468693733215332,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1105,
|
|
"token_acc": 0.6555926872419894,
|
|
"train_speed(iter/s)": 0.468391
|
|
},
|
|
{
|
|
"epoch": 0.8072727272727273,
|
|
"grad_norm": 3.7214505672454834,
|
|
"learning_rate": 8.574871300749087e-06,
|
|
"loss": 1.4309005737304688,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1110,
|
|
"token_acc": 0.65995670995671,
|
|
"train_speed(iter/s)": 0.468396
|
|
},
|
|
{
|
|
"epoch": 0.8109090909090909,
|
|
"grad_norm": 3.7076361179351807,
|
|
"learning_rate": 8.561119423023064e-06,
|
|
"loss": 1.505380344390869,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1115,
|
|
"token_acc": 0.6441098317094774,
|
|
"train_speed(iter/s)": 0.468426
|
|
},
|
|
{
|
|
"epoch": 0.8145454545454546,
|
|
"grad_norm": 3.645853042602539,
|
|
"learning_rate": 8.547312655878328e-06,
|
|
"loss": 1.491251277923584,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1120,
|
|
"token_acc": 0.6742215632281297,
|
|
"train_speed(iter/s)": 0.468413
|
|
},
|
|
{
|
|
"epoch": 0.8181818181818182,
|
|
"grad_norm": 3.672046661376953,
|
|
"learning_rate": 8.533451212125891e-06,
|
|
"loss": 1.489974594116211,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1125,
|
|
"token_acc": 0.6065996390822377,
|
|
"train_speed(iter/s)": 0.468418
|
|
},
|
|
{
|
|
"epoch": 0.8218181818181818,
|
|
"grad_norm": 3.495356321334839,
|
|
"learning_rate": 8.519535305419523e-06,
|
|
"loss": 1.3926969528198243,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1130,
|
|
"token_acc": 0.6418237615081105,
|
|
"train_speed(iter/s)": 0.468431
|
|
},
|
|
{
|
|
"epoch": 0.8254545454545454,
|
|
"grad_norm": 3.9424312114715576,
|
|
"learning_rate": 8.505565150252457e-06,
|
|
"loss": 1.4863405227661133,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1135,
|
|
"token_acc": 0.6287078934137758,
|
|
"train_speed(iter/s)": 0.468529
|
|
},
|
|
{
|
|
"epoch": 0.8290909090909091,
|
|
"grad_norm": 3.981257438659668,
|
|
"learning_rate": 8.491540961954094e-06,
|
|
"loss": 1.5392318725585938,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1140,
|
|
"token_acc": 0.6313190688925464,
|
|
"train_speed(iter/s)": 0.468624
|
|
},
|
|
{
|
|
"epoch": 0.8327272727272728,
|
|
"grad_norm": 3.6310641765594482,
|
|
"learning_rate": 8.477462956686669e-06,
|
|
"loss": 1.413975715637207,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1145,
|
|
"token_acc": 0.6229952093313893,
|
|
"train_speed(iter/s)": 0.468526
|
|
},
|
|
{
|
|
"epoch": 0.8363636363636363,
|
|
"grad_norm": 3.8132894039154053,
|
|
"learning_rate": 8.46333135144193e-06,
|
|
"loss": 1.5765387535095214,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1150,
|
|
"token_acc": 0.6065617497999467,
|
|
"train_speed(iter/s)": 0.468666
|
|
},
|
|
{
|
|
"epoch": 0.84,
|
|
"grad_norm": 3.9573326110839844,
|
|
"learning_rate": 8.44914636403779e-06,
|
|
"loss": 1.4768566131591796,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1155,
|
|
"token_acc": 0.6754981388219838,
|
|
"train_speed(iter/s)": 0.468646
|
|
},
|
|
{
|
|
"epoch": 0.8436363636363636,
|
|
"grad_norm": 3.644594669342041,
|
|
"learning_rate": 8.434908213114968e-06,
|
|
"loss": 1.5181507110595702,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1160,
|
|
"token_acc": 0.6126656848306333,
|
|
"train_speed(iter/s)": 0.468718
|
|
},
|
|
{
|
|
"epoch": 0.8472727272727273,
|
|
"grad_norm": 3.4235942363739014,
|
|
"learning_rate": 8.420617118133619e-06,
|
|
"loss": 1.4143997192382813,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1165,
|
|
"token_acc": 0.6732447817836812,
|
|
"train_speed(iter/s)": 0.468727
|
|
},
|
|
{
|
|
"epoch": 0.850909090909091,
|
|
"grad_norm": 3.506685256958008,
|
|
"learning_rate": 8.406273299369955e-06,
|
|
"loss": 1.5011482238769531,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1170,
|
|
"token_acc": 0.5982627578718784,
|
|
"train_speed(iter/s)": 0.468796
|
|
},
|
|
{
|
|
"epoch": 0.8545454545454545,
|
|
"grad_norm": 3.982557773590088,
|
|
"learning_rate": 8.391876977912852e-06,
|
|
"loss": 1.5165749549865724,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1175,
|
|
"token_acc": 0.6035743298131601,
|
|
"train_speed(iter/s)": 0.468896
|
|
},
|
|
{
|
|
"epoch": 0.8581818181818182,
|
|
"grad_norm": 3.354018211364746,
|
|
"learning_rate": 8.377428375660429e-06,
|
|
"loss": 1.415550136566162,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1180,
|
|
"token_acc": 0.6934698673829108,
|
|
"train_speed(iter/s)": 0.468771
|
|
},
|
|
{
|
|
"epoch": 0.8618181818181818,
|
|
"grad_norm": 3.8999528884887695,
|
|
"learning_rate": 8.36292771531664e-06,
|
|
"loss": 1.4483419418334962,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1185,
|
|
"token_acc": 0.6305418719211823,
|
|
"train_speed(iter/s)": 0.468762
|
|
},
|
|
{
|
|
"epoch": 0.8654545454545455,
|
|
"grad_norm": 3.604797840118408,
|
|
"learning_rate": 8.348375220387844e-06,
|
|
"loss": 1.5333918571472167,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1190,
|
|
"token_acc": 0.638868919242751,
|
|
"train_speed(iter/s)": 0.468837
|
|
},
|
|
{
|
|
"epoch": 0.8690909090909091,
|
|
"grad_norm": 4.271510601043701,
|
|
"learning_rate": 8.33377111517934e-06,
|
|
"loss": 1.5774791717529297,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1195,
|
|
"token_acc": 0.6245067952652346,
|
|
"train_speed(iter/s)": 0.468963
|
|
},
|
|
{
|
|
"epoch": 0.8727272727272727,
|
|
"grad_norm": 3.8670032024383545,
|
|
"learning_rate": 8.319115624791942e-06,
|
|
"loss": 1.5047279357910157,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1200,
|
|
"token_acc": 0.6603565231109008,
|
|
"train_speed(iter/s)": 0.468686
|
|
},
|
|
{
|
|
"epoch": 0.8763636363636363,
|
|
"grad_norm": 3.6218271255493164,
|
|
"learning_rate": 8.304408975118474e-06,
|
|
"loss": 1.5120525360107422,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1205,
|
|
"token_acc": 0.6243169398907104,
|
|
"train_speed(iter/s)": 0.468828
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"grad_norm": 3.629819631576538,
|
|
"learning_rate": 8.289651392840314e-06,
|
|
"loss": 1.516437816619873,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1210,
|
|
"token_acc": 0.6489115331171839,
|
|
"train_speed(iter/s)": 0.468904
|
|
},
|
|
{
|
|
"epoch": 0.8836363636363637,
|
|
"grad_norm": 3.4970250129699707,
|
|
"learning_rate": 8.27484310542389e-06,
|
|
"loss": 1.4951990127563477,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1215,
|
|
"token_acc": 0.622252435984591,
|
|
"train_speed(iter/s)": 0.46888
|
|
},
|
|
{
|
|
"epoch": 0.8872727272727273,
|
|
"grad_norm": 3.804673910140991,
|
|
"learning_rate": 8.259984341117175e-06,
|
|
"loss": 1.4283718109130858,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1220,
|
|
"token_acc": 0.6575117370892019,
|
|
"train_speed(iter/s)": 0.468919
|
|
},
|
|
{
|
|
"epoch": 0.8909090909090909,
|
|
"grad_norm": 3.604832172393799,
|
|
"learning_rate": 8.245075328946168e-06,
|
|
"loss": 1.4775312423706055,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1225,
|
|
"token_acc": 0.6120430107526882,
|
|
"train_speed(iter/s)": 0.469006
|
|
},
|
|
{
|
|
"epoch": 0.8945454545454545,
|
|
"grad_norm": 3.676591157913208,
|
|
"learning_rate": 8.230116298711368e-06,
|
|
"loss": 1.5129100799560546,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1230,
|
|
"token_acc": 0.6314168377823408,
|
|
"train_speed(iter/s)": 0.469052
|
|
},
|
|
{
|
|
"epoch": 0.8981818181818182,
|
|
"grad_norm": 3.6111202239990234,
|
|
"learning_rate": 8.21510748098422e-06,
|
|
"loss": 1.4718162536621093,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1235,
|
|
"token_acc": 0.6146128680479825,
|
|
"train_speed(iter/s)": 0.469126
|
|
},
|
|
{
|
|
"epoch": 0.9018181818181819,
|
|
"grad_norm": 3.0030624866485596,
|
|
"learning_rate": 8.200049107103583e-06,
|
|
"loss": 1.3085548400878906,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1240,
|
|
"token_acc": 0.6549242424242424,
|
|
"train_speed(iter/s)": 0.469076
|
|
},
|
|
{
|
|
"epoch": 0.9054545454545454,
|
|
"grad_norm": 3.4156432151794434,
|
|
"learning_rate": 8.18494140917214e-06,
|
|
"loss": 1.479377555847168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1245,
|
|
"token_acc": 0.6434093161546085,
|
|
"train_speed(iter/s)": 0.469066
|
|
},
|
|
{
|
|
"epoch": 0.9090909090909091,
|
|
"grad_norm": 3.6555395126342773,
|
|
"learning_rate": 8.16978462005284e-06,
|
|
"loss": 1.4368497848510742,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1250,
|
|
"token_acc": 0.634508348794063,
|
|
"train_speed(iter/s)": 0.469134
|
|
},
|
|
{
|
|
"epoch": 0.9127272727272727,
|
|
"grad_norm": 3.6969878673553467,
|
|
"learning_rate": 8.154578973365295e-06,
|
|
"loss": 1.5231738090515137,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1255,
|
|
"token_acc": 0.6483041693670339,
|
|
"train_speed(iter/s)": 0.4692
|
|
},
|
|
{
|
|
"epoch": 0.9163636363636364,
|
|
"grad_norm": 4.165112018585205,
|
|
"learning_rate": 8.139324703482185e-06,
|
|
"loss": 1.381281852722168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1260,
|
|
"token_acc": 0.721323395746228,
|
|
"train_speed(iter/s)": 0.469217
|
|
},
|
|
{
|
|
"epoch": 0.92,
|
|
"grad_norm": 3.383225440979004,
|
|
"learning_rate": 8.12402204552565e-06,
|
|
"loss": 1.4503177642822265,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1265,
|
|
"token_acc": 0.6532917802882743,
|
|
"train_speed(iter/s)": 0.469129
|
|
},
|
|
{
|
|
"epoch": 0.9236363636363636,
|
|
"grad_norm": 3.8582053184509277,
|
|
"learning_rate": 8.108671235363654e-06,
|
|
"loss": 1.4586336135864257,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1270,
|
|
"token_acc": 0.6378548895899053,
|
|
"train_speed(iter/s)": 0.469026
|
|
},
|
|
{
|
|
"epoch": 0.9272727272727272,
|
|
"grad_norm": 3.600813865661621,
|
|
"learning_rate": 8.093272509606362e-06,
|
|
"loss": 1.3761531829833984,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1275,
|
|
"token_acc": 0.6523583728475169,
|
|
"train_speed(iter/s)": 0.468929
|
|
},
|
|
{
|
|
"epoch": 0.9309090909090909,
|
|
"grad_norm": 3.1909162998199463,
|
|
"learning_rate": 8.077826105602484e-06,
|
|
"loss": 1.4242467880249023,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1280,
|
|
"token_acc": 0.6917508096780339,
|
|
"train_speed(iter/s)": 0.468915
|
|
},
|
|
{
|
|
"epoch": 0.9345454545454546,
|
|
"grad_norm": 3.405683755874634,
|
|
"learning_rate": 8.062332261435623e-06,
|
|
"loss": 1.5797532081604004,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1285,
|
|
"token_acc": 0.5903821382567626,
|
|
"train_speed(iter/s)": 0.468972
|
|
},
|
|
{
|
|
"epoch": 0.9381818181818182,
|
|
"grad_norm": 3.5222887992858887,
|
|
"learning_rate": 8.046791215920598e-06,
|
|
"loss": 1.4545100212097168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1290,
|
|
"token_acc": 0.6477043673012318,
|
|
"train_speed(iter/s)": 0.46901
|
|
},
|
|
{
|
|
"epoch": 0.9418181818181818,
|
|
"grad_norm": 3.5817978382110596,
|
|
"learning_rate": 8.031203208599773e-06,
|
|
"loss": 1.534127426147461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1295,
|
|
"token_acc": 0.6471281792150918,
|
|
"train_speed(iter/s)": 0.469015
|
|
},
|
|
{
|
|
"epoch": 0.9454545454545454,
|
|
"grad_norm": 3.8275201320648193,
|
|
"learning_rate": 8.015568479739354e-06,
|
|
"loss": 1.5303773880004883,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1300,
|
|
"token_acc": 0.6103666245259166,
|
|
"train_speed(iter/s)": 0.469087
|
|
},
|
|
{
|
|
"epoch": 0.9490909090909091,
|
|
"grad_norm": 3.656036376953125,
|
|
"learning_rate": 7.999887270325694e-06,
|
|
"loss": 1.4996587753295898,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1305,
|
|
"token_acc": 0.6306327340649078,
|
|
"train_speed(iter/s)": 0.469097
|
|
},
|
|
{
|
|
"epoch": 0.9527272727272728,
|
|
"grad_norm": 3.534893035888672,
|
|
"learning_rate": 7.984159822061572e-06,
|
|
"loss": 1.3629323959350585,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1310,
|
|
"token_acc": 0.6900225363654989,
|
|
"train_speed(iter/s)": 0.469074
|
|
},
|
|
{
|
|
"epoch": 0.9563636363636364,
|
|
"grad_norm": 3.609980583190918,
|
|
"learning_rate": 7.968386377362478e-06,
|
|
"loss": 1.430015754699707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1315,
|
|
"token_acc": 0.6487435828154553,
|
|
"train_speed(iter/s)": 0.469086
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 3.791590452194214,
|
|
"learning_rate": 7.95256717935286e-06,
|
|
"loss": 1.461299514770508,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1320,
|
|
"token_acc": 0.6620127981384526,
|
|
"train_speed(iter/s)": 0.469157
|
|
},
|
|
{
|
|
"epoch": 0.9636363636363636,
|
|
"grad_norm": 3.705054998397827,
|
|
"learning_rate": 7.936702471862396e-06,
|
|
"loss": 1.4849932670593262,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1325,
|
|
"token_acc": 0.6486560614371915,
|
|
"train_speed(iter/s)": 0.469233
|
|
},
|
|
{
|
|
"epoch": 0.9672727272727273,
|
|
"grad_norm": 3.7390530109405518,
|
|
"learning_rate": 7.920792499422213e-06,
|
|
"loss": 1.525192642211914,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1330,
|
|
"token_acc": 0.6327765674170903,
|
|
"train_speed(iter/s)": 0.469275
|
|
},
|
|
{
|
|
"epoch": 0.9709090909090909,
|
|
"grad_norm": 3.7763898372650146,
|
|
"learning_rate": 7.904837507261144e-06,
|
|
"loss": 1.5329397201538086,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1335,
|
|
"token_acc": 0.6480872069107363,
|
|
"train_speed(iter/s)": 0.469341
|
|
},
|
|
{
|
|
"epoch": 0.9745454545454545,
|
|
"grad_norm": 3.635175943374634,
|
|
"learning_rate": 7.888837741301928e-06,
|
|
"loss": 1.4299137115478515,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1340,
|
|
"token_acc": 0.6698178464376296,
|
|
"train_speed(iter/s)": 0.469276
|
|
},
|
|
{
|
|
"epoch": 0.9781818181818182,
|
|
"grad_norm": 3.436739206314087,
|
|
"learning_rate": 7.872793448157426e-06,
|
|
"loss": 1.4139694213867187,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1345,
|
|
"token_acc": 0.6410256410256411,
|
|
"train_speed(iter/s)": 0.469339
|
|
},
|
|
{
|
|
"epoch": 0.9818181818181818,
|
|
"grad_norm": 3.4722468852996826,
|
|
"learning_rate": 7.856704875126819e-06,
|
|
"loss": 1.458078384399414,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1350,
|
|
"token_acc": 0.6286181139122315,
|
|
"train_speed(iter/s)": 0.469383
|
|
},
|
|
{
|
|
"epoch": 0.9854545454545455,
|
|
"grad_norm": 3.40093994140625,
|
|
"learning_rate": 7.840572270191803e-06,
|
|
"loss": 1.4649463653564454,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1355,
|
|
"token_acc": 0.6814643383126446,
|
|
"train_speed(iter/s)": 0.469484
|
|
},
|
|
{
|
|
"epoch": 0.9890909090909091,
|
|
"grad_norm": 3.787477493286133,
|
|
"learning_rate": 7.824395882012756e-06,
|
|
"loss": 1.3931927680969238,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1360,
|
|
"token_acc": 0.6691426146010186,
|
|
"train_speed(iter/s)": 0.469454
|
|
},
|
|
{
|
|
"epoch": 0.9927272727272727,
|
|
"grad_norm": 3.734520435333252,
|
|
"learning_rate": 7.808175959924913e-06,
|
|
"loss": 1.4553655624389648,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1365,
|
|
"token_acc": 0.6351840034035312,
|
|
"train_speed(iter/s)": 0.469387
|
|
},
|
|
{
|
|
"epoch": 0.9963636363636363,
|
|
"grad_norm": 3.851562023162842,
|
|
"learning_rate": 7.791912753934516e-06,
|
|
"loss": 1.4150389671325683,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1370,
|
|
"token_acc": 0.7106060606060606,
|
|
"train_speed(iter/s)": 0.469419
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"grad_norm": 3.3316128253936768,
|
|
"learning_rate": 7.775606514714971e-06,
|
|
"loss": 1.5171707153320313,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1375,
|
|
"token_acc": 0.6147117296222664,
|
|
"train_speed(iter/s)": 0.469455
|
|
},
|
|
{
|
|
"epoch": 1.0036363636363637,
|
|
"grad_norm": 3.1916940212249756,
|
|
"learning_rate": 7.759257493602974e-06,
|
|
"loss": 1.0122819900512696,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1380,
|
|
"token_acc": 0.7368881118881119,
|
|
"train_speed(iter/s)": 0.45697
|
|
},
|
|
{
|
|
"epoch": 1.0072727272727273,
|
|
"grad_norm": 3.803614854812622,
|
|
"learning_rate": 7.74286594259464e-06,
|
|
"loss": 0.9860285758972168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1385,
|
|
"token_acc": 0.7332010909992561,
|
|
"train_speed(iter/s)": 0.457105
|
|
},
|
|
{
|
|
"epoch": 1.010909090909091,
|
|
"grad_norm": 3.667937755584717,
|
|
"learning_rate": 7.726432114341622e-06,
|
|
"loss": 1.0007999420166016,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1390,
|
|
"token_acc": 0.7341772151898734,
|
|
"train_speed(iter/s)": 0.457171
|
|
},
|
|
{
|
|
"epoch": 1.0145454545454546,
|
|
"grad_norm": 4.310483455657959,
|
|
"learning_rate": 7.709956262147212e-06,
|
|
"loss": 1.0946765899658204,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1395,
|
|
"token_acc": 0.7179426404361223,
|
|
"train_speed(iter/s)": 0.457209
|
|
},
|
|
{
|
|
"epoch": 1.018181818181818,
|
|
"grad_norm": 3.2957613468170166,
|
|
"learning_rate": 7.693438639962447e-06,
|
|
"loss": 1.0460298538208008,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1400,
|
|
"token_acc": 0.7339094837389128,
|
|
"train_speed(iter/s)": 0.45725
|
|
},
|
|
{
|
|
"epoch": 1.0218181818181817,
|
|
"grad_norm": 4.106925964355469,
|
|
"learning_rate": 7.676879502382177e-06,
|
|
"loss": 1.0074480056762696,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1405,
|
|
"token_acc": 0.7509936871638999,
|
|
"train_speed(iter/s)": 0.457319
|
|
},
|
|
{
|
|
"epoch": 1.0254545454545454,
|
|
"grad_norm": 3.797349452972412,
|
|
"learning_rate": 7.660279104641154e-06,
|
|
"loss": 0.9518960952758789,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1410,
|
|
"token_acc": 0.7552194290583724,
|
|
"train_speed(iter/s)": 0.457307
|
|
},
|
|
{
|
|
"epoch": 1.029090909090909,
|
|
"grad_norm": 3.7703957557678223,
|
|
"learning_rate": 7.643637702610098e-06,
|
|
"loss": 1.0913700103759765,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1415,
|
|
"token_acc": 0.7454581752844879,
|
|
"train_speed(iter/s)": 0.457406
|
|
},
|
|
{
|
|
"epoch": 1.0327272727272727,
|
|
"grad_norm": 3.6022231578826904,
|
|
"learning_rate": 7.6269555527917485e-06,
|
|
"loss": 1.0375457763671876,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1420,
|
|
"token_acc": 0.742365352581899,
|
|
"train_speed(iter/s)": 0.457452
|
|
},
|
|
{
|
|
"epoch": 1.0363636363636364,
|
|
"grad_norm": 4.11036491394043,
|
|
"learning_rate": 7.610232912316906e-06,
|
|
"loss": 1.0097059249877929,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1425,
|
|
"token_acc": 0.7417205169628432,
|
|
"train_speed(iter/s)": 0.457507
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"grad_norm": 3.9686105251312256,
|
|
"learning_rate": 7.593470038940487e-06,
|
|
"loss": 1.0012422561645509,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1430,
|
|
"token_acc": 0.7476245654692931,
|
|
"train_speed(iter/s)": 0.457603
|
|
},
|
|
{
|
|
"epoch": 1.0436363636363637,
|
|
"grad_norm": 3.909630298614502,
|
|
"learning_rate": 7.576667191037525e-06,
|
|
"loss": 0.9417753219604492,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1435,
|
|
"token_acc": 0.7406999779881136,
|
|
"train_speed(iter/s)": 0.457645
|
|
},
|
|
{
|
|
"epoch": 1.0472727272727274,
|
|
"grad_norm": 3.4653851985931396,
|
|
"learning_rate": 7.5598246275992145e-06,
|
|
"loss": 1.0239160537719727,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1440,
|
|
"token_acc": 0.7312540055543687,
|
|
"train_speed(iter/s)": 0.457633
|
|
},
|
|
{
|
|
"epoch": 1.050909090909091,
|
|
"grad_norm": 3.6743690967559814,
|
|
"learning_rate": 7.5429426082289e-06,
|
|
"loss": 0.9335263252258301,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1445,
|
|
"token_acc": 0.7262321144674085,
|
|
"train_speed(iter/s)": 0.45757
|
|
},
|
|
{
|
|
"epoch": 1.0545454545454545,
|
|
"grad_norm": 3.187088966369629,
|
|
"learning_rate": 7.526021393138081e-06,
|
|
"loss": 0.9450448036193848,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1450,
|
|
"token_acc": 0.756238399670035,
|
|
"train_speed(iter/s)": 0.457549
|
|
},
|
|
{
|
|
"epoch": 1.0581818181818181,
|
|
"grad_norm": 3.3623409271240234,
|
|
"learning_rate": 7.509061243142406e-06,
|
|
"loss": 0.9245813369750977,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1455,
|
|
"token_acc": 0.7525731584258325,
|
|
"train_speed(iter/s)": 0.457493
|
|
},
|
|
{
|
|
"epoch": 1.0618181818181818,
|
|
"grad_norm": 3.6314806938171387,
|
|
"learning_rate": 7.49206241965764e-06,
|
|
"loss": 1.0117255210876466,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1460,
|
|
"token_acc": 0.7401989467524869,
|
|
"train_speed(iter/s)": 0.457535
|
|
},
|
|
{
|
|
"epoch": 1.0654545454545454,
|
|
"grad_norm": 3.7548723220825195,
|
|
"learning_rate": 7.47502518469565e-06,
|
|
"loss": 1.0736151695251466,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1465,
|
|
"token_acc": 0.72635998539613,
|
|
"train_speed(iter/s)": 0.457658
|
|
},
|
|
{
|
|
"epoch": 1.069090909090909,
|
|
"grad_norm": 3.9179763793945312,
|
|
"learning_rate": 7.457949800860358e-06,
|
|
"loss": 1.0174610137939453,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1470,
|
|
"token_acc": 0.7323055360896986,
|
|
"train_speed(iter/s)": 0.457707
|
|
},
|
|
{
|
|
"epoch": 1.0727272727272728,
|
|
"grad_norm": 3.4640631675720215,
|
|
"learning_rate": 7.440836531343692e-06,
|
|
"loss": 0.9873522758483887,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1475,
|
|
"token_acc": 0.7806191117092867,
|
|
"train_speed(iter/s)": 0.4577
|
|
},
|
|
{
|
|
"epoch": 1.0763636363636364,
|
|
"grad_norm": 3.791652202606201,
|
|
"learning_rate": 7.4236856399215344e-06,
|
|
"loss": 0.911875057220459,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1480,
|
|
"token_acc": 0.7158859470468432,
|
|
"train_speed(iter/s)": 0.457776
|
|
},
|
|
{
|
|
"epoch": 1.08,
|
|
"grad_norm": 3.838895082473755,
|
|
"learning_rate": 7.406497390949652e-06,
|
|
"loss": 1.0216828346252442,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1485,
|
|
"token_acc": 0.7165523156089194,
|
|
"train_speed(iter/s)": 0.457843
|
|
},
|
|
{
|
|
"epoch": 1.0836363636363637,
|
|
"grad_norm": 4.171408176422119,
|
|
"learning_rate": 7.3892720493596284e-06,
|
|
"loss": 0.974361801147461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1490,
|
|
"token_acc": 0.7775261324041812,
|
|
"train_speed(iter/s)": 0.457853
|
|
},
|
|
{
|
|
"epoch": 1.0872727272727274,
|
|
"grad_norm": 3.92596697807312,
|
|
"learning_rate": 7.372009880654767e-06,
|
|
"loss": 0.9939946174621582,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1495,
|
|
"token_acc": 0.7887160436498724,
|
|
"train_speed(iter/s)": 0.457899
|
|
},
|
|
{
|
|
"epoch": 1.0909090909090908,
|
|
"grad_norm": 3.189763069152832,
|
|
"learning_rate": 7.3547111509060145e-06,
|
|
"loss": 1.0516817092895507,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1500,
|
|
"token_acc": 0.7391479911164951,
|
|
"train_speed(iter/s)": 0.45792
|
|
},
|
|
{
|
|
"epoch": 1.0945454545454545,
|
|
"grad_norm": 3.6382617950439453,
|
|
"learning_rate": 7.337376126747849e-06,
|
|
"loss": 0.994780158996582,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1505,
|
|
"token_acc": 0.7239606616003577,
|
|
"train_speed(iter/s)": 0.45794
|
|
},
|
|
{
|
|
"epoch": 1.0981818181818181,
|
|
"grad_norm": 3.5191988945007324,
|
|
"learning_rate": 7.320005075374173e-06,
|
|
"loss": 0.9538056373596191,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1510,
|
|
"token_acc": 0.7650323774283071,
|
|
"train_speed(iter/s)": 0.457938
|
|
},
|
|
{
|
|
"epoch": 1.1018181818181818,
|
|
"grad_norm": 3.722946882247925,
|
|
"learning_rate": 7.302598264534197e-06,
|
|
"loss": 0.9156055450439453,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1515,
|
|
"token_acc": 0.7966154323862754,
|
|
"train_speed(iter/s)": 0.457866
|
|
},
|
|
{
|
|
"epoch": 1.1054545454545455,
|
|
"grad_norm": 4.167759418487549,
|
|
"learning_rate": 7.285155962528312e-06,
|
|
"loss": 1.0325860977172852,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1520,
|
|
"token_acc": 0.7377584330794341,
|
|
"train_speed(iter/s)": 0.457908
|
|
},
|
|
{
|
|
"epoch": 1.1090909090909091,
|
|
"grad_norm": 3.5554070472717285,
|
|
"learning_rate": 7.267678438203949e-06,
|
|
"loss": 0.9901115417480468,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1525,
|
|
"token_acc": 0.7337373737373737,
|
|
"train_speed(iter/s)": 0.457998
|
|
},
|
|
{
|
|
"epoch": 1.1127272727272728,
|
|
"grad_norm": 3.368116855621338,
|
|
"learning_rate": 7.250165960951445e-06,
|
|
"loss": 1.0267613410949707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1530,
|
|
"token_acc": 0.7259731691224983,
|
|
"train_speed(iter/s)": 0.457928
|
|
},
|
|
{
|
|
"epoch": 1.1163636363636364,
|
|
"grad_norm": 3.9949262142181396,
|
|
"learning_rate": 7.23261880069988e-06,
|
|
"loss": 0.9667708396911621,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1535,
|
|
"token_acc": 0.7625673249551167,
|
|
"train_speed(iter/s)": 0.457977
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"grad_norm": 3.527697801589966,
|
|
"learning_rate": 7.2150372279129245e-06,
|
|
"loss": 1.0338319778442382,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1540,
|
|
"token_acc": 0.7121308689186319,
|
|
"train_speed(iter/s)": 0.458058
|
|
},
|
|
{
|
|
"epoch": 1.1236363636363635,
|
|
"grad_norm": 3.553553342819214,
|
|
"learning_rate": 7.197421513584666e-06,
|
|
"loss": 0.9568606376647949,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1545,
|
|
"token_acc": 0.761969904240766,
|
|
"train_speed(iter/s)": 0.458148
|
|
},
|
|
{
|
|
"epoch": 1.1272727272727272,
|
|
"grad_norm": 3.8653059005737305,
|
|
"learning_rate": 7.179771929235433e-06,
|
|
"loss": 1.0294866561889648,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1550,
|
|
"token_acc": 0.702284064058808,
|
|
"train_speed(iter/s)": 0.458233
|
|
},
|
|
{
|
|
"epoch": 1.1309090909090909,
|
|
"grad_norm": 3.653136730194092,
|
|
"learning_rate": 7.162088746907613e-06,
|
|
"loss": 1.0458011627197266,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1555,
|
|
"token_acc": 0.7369407982855612,
|
|
"train_speed(iter/s)": 0.458366
|
|
},
|
|
{
|
|
"epoch": 1.1345454545454545,
|
|
"grad_norm": 3.158511161804199,
|
|
"learning_rate": 7.144372239161454e-06,
|
|
"loss": 1.0241386413574218,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1560,
|
|
"token_acc": 0.7366296670030272,
|
|
"train_speed(iter/s)": 0.458393
|
|
},
|
|
{
|
|
"epoch": 1.1381818181818182,
|
|
"grad_norm": 3.6491730213165283,
|
|
"learning_rate": 7.126622679070872e-06,
|
|
"loss": 0.9555923461914062,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1565,
|
|
"token_acc": 0.739185110663984,
|
|
"train_speed(iter/s)": 0.458359
|
|
},
|
|
{
|
|
"epoch": 1.1418181818181818,
|
|
"grad_norm": 3.801968812942505,
|
|
"learning_rate": 7.10884034021923e-06,
|
|
"loss": 1.0447402000427246,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1570,
|
|
"token_acc": 0.7483831851253031,
|
|
"train_speed(iter/s)": 0.458391
|
|
},
|
|
{
|
|
"epoch": 1.1454545454545455,
|
|
"grad_norm": 3.7026801109313965,
|
|
"learning_rate": 7.091025496695127e-06,
|
|
"loss": 0.9554638862609863,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1575,
|
|
"token_acc": 0.737876254180602,
|
|
"train_speed(iter/s)": 0.458447
|
|
},
|
|
{
|
|
"epoch": 1.1490909090909092,
|
|
"grad_norm": 4.218850135803223,
|
|
"learning_rate": 7.073178423088178e-06,
|
|
"loss": 0.9791478157043457,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1580,
|
|
"token_acc": 0.7147259520140876,
|
|
"train_speed(iter/s)": 0.458542
|
|
},
|
|
{
|
|
"epoch": 1.1527272727272728,
|
|
"grad_norm": 3.852479934692383,
|
|
"learning_rate": 7.055299394484778e-06,
|
|
"loss": 1.109478759765625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1585,
|
|
"token_acc": 0.7238214434710054,
|
|
"train_speed(iter/s)": 0.458621
|
|
},
|
|
{
|
|
"epoch": 1.1563636363636363,
|
|
"grad_norm": 3.196315288543701,
|
|
"learning_rate": 7.037388686463856e-06,
|
|
"loss": 0.9640076637268067,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1590,
|
|
"token_acc": 0.7785083213478529,
|
|
"train_speed(iter/s)": 0.458629
|
|
},
|
|
{
|
|
"epoch": 1.16,
|
|
"grad_norm": 3.4316623210906982,
|
|
"learning_rate": 7.0194465750926385e-06,
|
|
"loss": 1.0183427810668946,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1595,
|
|
"token_acc": 0.7284430410493078,
|
|
"train_speed(iter/s)": 0.458714
|
|
},
|
|
{
|
|
"epoch": 1.1636363636363636,
|
|
"grad_norm": 3.760284423828125,
|
|
"learning_rate": 7.001473336922382e-06,
|
|
"loss": 0.9476146697998047,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1600,
|
|
"token_acc": 0.7497454693545103,
|
|
"train_speed(iter/s)": 0.458727
|
|
},
|
|
{
|
|
"epoch": 1.1672727272727272,
|
|
"grad_norm": 3.6320013999938965,
|
|
"learning_rate": 6.983469248984125e-06,
|
|
"loss": 0.9717130661010742,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1605,
|
|
"token_acc": 0.7795737122557727,
|
|
"train_speed(iter/s)": 0.458757
|
|
},
|
|
{
|
|
"epoch": 1.170909090909091,
|
|
"grad_norm": 3.9939889907836914,
|
|
"learning_rate": 6.965434588784401e-06,
|
|
"loss": 1.0135496139526368,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1610,
|
|
"token_acc": 0.716100659662839,
|
|
"train_speed(iter/s)": 0.458863
|
|
},
|
|
{
|
|
"epoch": 1.1745454545454546,
|
|
"grad_norm": 3.6845226287841797,
|
|
"learning_rate": 6.947369634300974e-06,
|
|
"loss": 1.0554584503173827,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1615,
|
|
"token_acc": 0.7110694183864915,
|
|
"train_speed(iter/s)": 0.458836
|
|
},
|
|
{
|
|
"epoch": 1.1781818181818182,
|
|
"grad_norm": 3.7763588428497314,
|
|
"learning_rate": 6.9292746639785534e-06,
|
|
"loss": 0.9863467216491699,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1620,
|
|
"token_acc": 0.7300349556332347,
|
|
"train_speed(iter/s)": 0.458949
|
|
},
|
|
{
|
|
"epoch": 1.1818181818181819,
|
|
"grad_norm": 3.6125032901763916,
|
|
"learning_rate": 6.911149956724494e-06,
|
|
"loss": 1.057917594909668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1625,
|
|
"token_acc": 0.6973138239790347,
|
|
"train_speed(iter/s)": 0.459007
|
|
},
|
|
{
|
|
"epoch": 1.1854545454545455,
|
|
"grad_norm": 3.32602858543396,
|
|
"learning_rate": 6.8929957919044995e-06,
|
|
"loss": 1.1337769508361817,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1630,
|
|
"token_acc": 0.7133373566686784,
|
|
"train_speed(iter/s)": 0.459083
|
|
},
|
|
{
|
|
"epoch": 1.189090909090909,
|
|
"grad_norm": 3.699726104736328,
|
|
"learning_rate": 6.874812449338328e-06,
|
|
"loss": 0.9991239547729492,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1635,
|
|
"token_acc": 0.7829614604462475,
|
|
"train_speed(iter/s)": 0.459088
|
|
},
|
|
{
|
|
"epoch": 1.1927272727272726,
|
|
"grad_norm": 3.956362724304199,
|
|
"learning_rate": 6.856600209295459e-06,
|
|
"loss": 0.9848535537719727,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1640,
|
|
"token_acc": 0.7056864144658577,
|
|
"train_speed(iter/s)": 0.459203
|
|
},
|
|
{
|
|
"epoch": 1.1963636363636363,
|
|
"grad_norm": 4.043876647949219,
|
|
"learning_rate": 6.838359352490789e-06,
|
|
"loss": 1.076396656036377,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1645,
|
|
"token_acc": 0.7244575521381925,
|
|
"train_speed(iter/s)": 0.45924
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"grad_norm": 3.5813117027282715,
|
|
"learning_rate": 6.820090160080301e-06,
|
|
"loss": 1.1189224243164062,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1650,
|
|
"token_acc": 0.6809587217043941,
|
|
"train_speed(iter/s)": 0.459232
|
|
},
|
|
{
|
|
"epoch": 1.2036363636363636,
|
|
"grad_norm": 3.6248674392700195,
|
|
"learning_rate": 6.80179291365673e-06,
|
|
"loss": 1.0456843376159668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1655,
|
|
"token_acc": 0.7274615704196095,
|
|
"train_speed(iter/s)": 0.459302
|
|
},
|
|
{
|
|
"epoch": 1.2072727272727273,
|
|
"grad_norm": 4.285236835479736,
|
|
"learning_rate": 6.783467895245216e-06,
|
|
"loss": 1.062920093536377,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1660,
|
|
"token_acc": 0.76859151928404,
|
|
"train_speed(iter/s)": 0.459361
|
|
},
|
|
{
|
|
"epoch": 1.210909090909091,
|
|
"grad_norm": 4.260605335235596,
|
|
"learning_rate": 6.76511538729897e-06,
|
|
"loss": 1.0169434547424316,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1665,
|
|
"token_acc": 0.7671654929577465,
|
|
"train_speed(iter/s)": 0.459392
|
|
},
|
|
{
|
|
"epoch": 1.2145454545454546,
|
|
"grad_norm": 3.5965423583984375,
|
|
"learning_rate": 6.746735672694915e-06,
|
|
"loss": 1.0033367156982422,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1670,
|
|
"token_acc": 0.7534195240771969,
|
|
"train_speed(iter/s)": 0.459393
|
|
},
|
|
{
|
|
"epoch": 1.2181818181818183,
|
|
"grad_norm": 3.6282148361206055,
|
|
"learning_rate": 6.728329034729316e-06,
|
|
"loss": 1.0619688034057617,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1675,
|
|
"token_acc": 0.7052186177715092,
|
|
"train_speed(iter/s)": 0.459452
|
|
},
|
|
{
|
|
"epoch": 1.221818181818182,
|
|
"grad_norm": 3.6731245517730713,
|
|
"learning_rate": 6.709895757113432e-06,
|
|
"loss": 0.9825057983398438,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1680,
|
|
"token_acc": 0.7467287646928366,
|
|
"train_speed(iter/s)": 0.459543
|
|
},
|
|
{
|
|
"epoch": 1.2254545454545456,
|
|
"grad_norm": 3.2094638347625732,
|
|
"learning_rate": 6.691436123969126e-06,
|
|
"loss": 1.0286219596862793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1685,
|
|
"token_acc": 0.706904435299497,
|
|
"train_speed(iter/s)": 0.459495
|
|
},
|
|
{
|
|
"epoch": 1.229090909090909,
|
|
"grad_norm": 3.4451398849487305,
|
|
"learning_rate": 6.6729504198244935e-06,
|
|
"loss": 1.028489589691162,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1690,
|
|
"token_acc": 0.7215537259869116,
|
|
"train_speed(iter/s)": 0.459581
|
|
},
|
|
{
|
|
"epoch": 1.2327272727272727,
|
|
"grad_norm": 3.6798651218414307,
|
|
"learning_rate": 6.654438929609481e-06,
|
|
"loss": 1.0026846885681153,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1695,
|
|
"token_acc": 0.7153620761825031,
|
|
"train_speed(iter/s)": 0.459647
|
|
},
|
|
{
|
|
"epoch": 1.2363636363636363,
|
|
"grad_norm": 3.6463124752044678,
|
|
"learning_rate": 6.635901938651485e-06,
|
|
"loss": 1.0815553665161133,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1700,
|
|
"token_acc": 0.7078901805589909,
|
|
"train_speed(iter/s)": 0.459704
|
|
},
|
|
{
|
|
"epoch": 1.24,
|
|
"grad_norm": 3.657982110977173,
|
|
"learning_rate": 6.6173397326709575e-06,
|
|
"loss": 1.0929712295532226,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1705,
|
|
"token_acc": 0.7036560247167868,
|
|
"train_speed(iter/s)": 0.459738
|
|
},
|
|
{
|
|
"epoch": 1.2436363636363637,
|
|
"grad_norm": 3.6304128170013428,
|
|
"learning_rate": 6.59875259777701e-06,
|
|
"loss": 1.0496492385864258,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1710,
|
|
"token_acc": 0.7248382057576434,
|
|
"train_speed(iter/s)": 0.459763
|
|
},
|
|
{
|
|
"epoch": 1.2472727272727273,
|
|
"grad_norm": 3.586535930633545,
|
|
"learning_rate": 6.580140820462983e-06,
|
|
"loss": 1.0112442016601562,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1715,
|
|
"token_acc": 0.7460452457901003,
|
|
"train_speed(iter/s)": 0.459713
|
|
},
|
|
{
|
|
"epoch": 1.250909090909091,
|
|
"grad_norm": 3.4018948078155518,
|
|
"learning_rate": 6.561504687602061e-06,
|
|
"loss": 1.006149959564209,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1720,
|
|
"token_acc": 0.7438073394495412,
|
|
"train_speed(iter/s)": 0.4598
|
|
},
|
|
{
|
|
"epoch": 1.2545454545454544,
|
|
"grad_norm": 3.9034860134124756,
|
|
"learning_rate": 6.542844486442821e-06,
|
|
"loss": 1.0216346740722657,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1725,
|
|
"token_acc": 0.7470520231213873,
|
|
"train_speed(iter/s)": 0.459865
|
|
},
|
|
{
|
|
"epoch": 1.2581818181818183,
|
|
"grad_norm": 3.6191561222076416,
|
|
"learning_rate": 6.52416050460482e-06,
|
|
"loss": 1.0998344421386719,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1730,
|
|
"token_acc": 0.678407350689127,
|
|
"train_speed(iter/s)": 0.459933
|
|
},
|
|
{
|
|
"epoch": 1.2618181818181817,
|
|
"grad_norm": 3.7334041595458984,
|
|
"learning_rate": 6.505453030074161e-06,
|
|
"loss": 1.0943750381469726,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1735,
|
|
"token_acc": 0.7359395168725797,
|
|
"train_speed(iter/s)": 0.45994
|
|
},
|
|
{
|
|
"epoch": 1.2654545454545454,
|
|
"grad_norm": 3.683525562286377,
|
|
"learning_rate": 6.486722351199053e-06,
|
|
"loss": 0.999393081665039,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1740,
|
|
"token_acc": 0.7622227008399741,
|
|
"train_speed(iter/s)": 0.459977
|
|
},
|
|
{
|
|
"epoch": 1.269090909090909,
|
|
"grad_norm": 3.8208823204040527,
|
|
"learning_rate": 6.4679687566853635e-06,
|
|
"loss": 0.9925840377807618,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1745,
|
|
"token_acc": 0.7400342906129447,
|
|
"train_speed(iter/s)": 0.459979
|
|
},
|
|
{
|
|
"epoch": 1.2727272727272727,
|
|
"grad_norm": 3.513150691986084,
|
|
"learning_rate": 6.449192535592171e-06,
|
|
"loss": 1.0545754432678223,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1750,
|
|
"token_acc": 0.7310440651052005,
|
|
"train_speed(iter/s)": 0.46002
|
|
},
|
|
{
|
|
"epoch": 1.2763636363636364,
|
|
"grad_norm": 3.7287533283233643,
|
|
"learning_rate": 6.43039397732731e-06,
|
|
"loss": 1.0397346496582032,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1755,
|
|
"token_acc": 0.7179717298631366,
|
|
"train_speed(iter/s)": 0.460084
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"grad_norm": 3.5868330001831055,
|
|
"learning_rate": 6.411573371642913e-06,
|
|
"loss": 0.9834535598754883,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1760,
|
|
"token_acc": 0.7371490280777537,
|
|
"train_speed(iter/s)": 0.460105
|
|
},
|
|
{
|
|
"epoch": 1.2836363636363637,
|
|
"grad_norm": 4.112300395965576,
|
|
"learning_rate": 6.392731008630933e-06,
|
|
"loss": 1.0012317657470704,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1765,
|
|
"token_acc": 0.7569250590508911,
|
|
"train_speed(iter/s)": 0.460219
|
|
},
|
|
{
|
|
"epoch": 1.2872727272727273,
|
|
"grad_norm": 3.2801835536956787,
|
|
"learning_rate": 6.373867178718692e-06,
|
|
"loss": 1.0033071517944336,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1770,
|
|
"token_acc": 0.7949538693278102,
|
|
"train_speed(iter/s)": 0.460253
|
|
},
|
|
{
|
|
"epoch": 1.290909090909091,
|
|
"grad_norm": 3.687485694885254,
|
|
"learning_rate": 6.3549821726643765e-06,
|
|
"loss": 0.9472185134887695,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1775,
|
|
"token_acc": 0.7774779636625292,
|
|
"train_speed(iter/s)": 0.460162
|
|
},
|
|
{
|
|
"epoch": 1.2945454545454544,
|
|
"grad_norm": 3.748479127883911,
|
|
"learning_rate": 6.33607628155259e-06,
|
|
"loss": 0.9887983322143554,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1780,
|
|
"token_acc": 0.7205817540766858,
|
|
"train_speed(iter/s)": 0.460233
|
|
},
|
|
{
|
|
"epoch": 1.298181818181818,
|
|
"grad_norm": 3.500528335571289,
|
|
"learning_rate": 6.317149796789835e-06,
|
|
"loss": 1.0536639213562011,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1785,
|
|
"token_acc": 0.7099773242630385,
|
|
"train_speed(iter/s)": 0.460317
|
|
},
|
|
{
|
|
"epoch": 1.3018181818181818,
|
|
"grad_norm": 3.8819491863250732,
|
|
"learning_rate": 6.298203010100041e-06,
|
|
"loss": 1.042212677001953,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1790,
|
|
"token_acc": 0.7374644186555726,
|
|
"train_speed(iter/s)": 0.460343
|
|
},
|
|
{
|
|
"epoch": 1.3054545454545454,
|
|
"grad_norm": 3.779406785964966,
|
|
"learning_rate": 6.279236213520062e-06,
|
|
"loss": 1.0398897171020507,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1795,
|
|
"token_acc": 0.7061352090660414,
|
|
"train_speed(iter/s)": 0.460419
|
|
},
|
|
{
|
|
"epoch": 1.309090909090909,
|
|
"grad_norm": 3.7476646900177,
|
|
"learning_rate": 6.260249699395172e-06,
|
|
"loss": 0.9859634399414062,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1800,
|
|
"token_acc": 0.7423469387755102,
|
|
"train_speed(iter/s)": 0.460492
|
|
},
|
|
{
|
|
"epoch": 1.3127272727272727,
|
|
"grad_norm": 3.867429494857788,
|
|
"learning_rate": 6.241243760374562e-06,
|
|
"loss": 0.9892477035522461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1805,
|
|
"token_acc": 0.7313289927230946,
|
|
"train_speed(iter/s)": 0.460545
|
|
},
|
|
{
|
|
"epoch": 1.3163636363636364,
|
|
"grad_norm": 3.5159947872161865,
|
|
"learning_rate": 6.222218689406834e-06,
|
|
"loss": 1.0133953094482422,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1810,
|
|
"token_acc": 0.7424869037772264,
|
|
"train_speed(iter/s)": 0.460663
|
|
},
|
|
{
|
|
"epoch": 1.32,
|
|
"grad_norm": 4.202282905578613,
|
|
"learning_rate": 6.203174779735476e-06,
|
|
"loss": 1.1178067207336426,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1815,
|
|
"token_acc": 0.7210046786505787,
|
|
"train_speed(iter/s)": 0.460744
|
|
},
|
|
{
|
|
"epoch": 1.3236363636363637,
|
|
"grad_norm": 3.0678062438964844,
|
|
"learning_rate": 6.184112324894348e-06,
|
|
"loss": 0.9732847213745117,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1820,
|
|
"token_acc": 0.7407130856822849,
|
|
"train_speed(iter/s)": 0.46078
|
|
},
|
|
{
|
|
"epoch": 1.3272727272727272,
|
|
"grad_norm": 3.376113176345825,
|
|
"learning_rate": 6.165031618703159e-06,
|
|
"loss": 0.9729133605957031,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1825,
|
|
"token_acc": 0.797998366013072,
|
|
"train_speed(iter/s)": 0.460756
|
|
},
|
|
{
|
|
"epoch": 1.330909090909091,
|
|
"grad_norm": 3.703918695449829,
|
|
"learning_rate": 6.145932955262931e-06,
|
|
"loss": 0.8784601211547851,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1830,
|
|
"token_acc": 0.751122754491018,
|
|
"train_speed(iter/s)": 0.46079
|
|
},
|
|
{
|
|
"epoch": 1.3345454545454545,
|
|
"grad_norm": 3.489109516143799,
|
|
"learning_rate": 6.126816628951476e-06,
|
|
"loss": 1.028179931640625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1835,
|
|
"token_acc": 0.7180817936474985,
|
|
"train_speed(iter/s)": 0.460839
|
|
},
|
|
{
|
|
"epoch": 1.3381818181818181,
|
|
"grad_norm": 3.857232093811035,
|
|
"learning_rate": 6.107682934418847e-06,
|
|
"loss": 0.9387928009033203,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1840,
|
|
"token_acc": 0.7655922181956895,
|
|
"train_speed(iter/s)": 0.46081
|
|
},
|
|
{
|
|
"epoch": 1.3418181818181818,
|
|
"grad_norm": 3.459693193435669,
|
|
"learning_rate": 6.088532166582809e-06,
|
|
"loss": 1.0774751663208009,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1845,
|
|
"token_acc": 0.7526041666666666,
|
|
"train_speed(iter/s)": 0.46086
|
|
},
|
|
{
|
|
"epoch": 1.3454545454545455,
|
|
"grad_norm": 3.9030678272247314,
|
|
"learning_rate": 6.069364620624284e-06,
|
|
"loss": 1.0519866943359375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1850,
|
|
"token_acc": 0.7056686046511628,
|
|
"train_speed(iter/s)": 0.460949
|
|
},
|
|
{
|
|
"epoch": 1.3490909090909091,
|
|
"grad_norm": 3.601980686187744,
|
|
"learning_rate": 6.050180591982802e-06,
|
|
"loss": 0.9877336502075196,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1855,
|
|
"token_acc": 0.7090541451359245,
|
|
"train_speed(iter/s)": 0.461034
|
|
},
|
|
{
|
|
"epoch": 1.3527272727272728,
|
|
"grad_norm": 3.5950915813446045,
|
|
"learning_rate": 6.0309803763519535e-06,
|
|
"loss": 0.9389093399047852,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1860,
|
|
"token_acc": 0.7330003841721091,
|
|
"train_speed(iter/s)": 0.460922
|
|
},
|
|
{
|
|
"epoch": 1.3563636363636364,
|
|
"grad_norm": 3.591996908187866,
|
|
"learning_rate": 6.011764269674823e-06,
|
|
"loss": 0.9682686805725098,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1865,
|
|
"token_acc": 0.7354409317803661,
|
|
"train_speed(iter/s)": 0.460944
|
|
},
|
|
{
|
|
"epoch": 1.3599999999999999,
|
|
"grad_norm": 3.609264850616455,
|
|
"learning_rate": 5.992532568139438e-06,
|
|
"loss": 1.024522590637207,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1870,
|
|
"token_acc": 0.7284263959390863,
|
|
"train_speed(iter/s)": 0.461044
|
|
},
|
|
{
|
|
"epoch": 1.3636363636363638,
|
|
"grad_norm": 3.9778711795806885,
|
|
"learning_rate": 5.973285568174192e-06,
|
|
"loss": 1.0148784637451171,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1875,
|
|
"token_acc": 0.7046184243022968,
|
|
"train_speed(iter/s)": 0.461087
|
|
},
|
|
{
|
|
"epoch": 1.3672727272727272,
|
|
"grad_norm": 3.59853196144104,
|
|
"learning_rate": 5.954023566443283e-06,
|
|
"loss": 0.9733685493469239,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1880,
|
|
"token_acc": 0.766418046207022,
|
|
"train_speed(iter/s)": 0.461114
|
|
},
|
|
{
|
|
"epoch": 1.3709090909090909,
|
|
"grad_norm": 3.281967878341675,
|
|
"learning_rate": 5.934746859842141e-06,
|
|
"loss": 1.0568039894104004,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1885,
|
|
"token_acc": 0.7198658512527125,
|
|
"train_speed(iter/s)": 0.461138
|
|
},
|
|
{
|
|
"epoch": 1.3745454545454545,
|
|
"grad_norm": 3.522768974304199,
|
|
"learning_rate": 5.915455745492849e-06,
|
|
"loss": 1.0536276817321777,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1890,
|
|
"token_acc": 0.7116860337645253,
|
|
"train_speed(iter/s)": 0.461185
|
|
},
|
|
{
|
|
"epoch": 1.3781818181818182,
|
|
"grad_norm": 3.975327730178833,
|
|
"learning_rate": 5.896150520739563e-06,
|
|
"loss": 0.8941743850708008,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1895,
|
|
"token_acc": 0.7662427745664739,
|
|
"train_speed(iter/s)": 0.461165
|
|
},
|
|
{
|
|
"epoch": 1.3818181818181818,
|
|
"grad_norm": 3.616748332977295,
|
|
"learning_rate": 5.876831483143933e-06,
|
|
"loss": 0.9861072540283203,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1900,
|
|
"token_acc": 0.7113821138211383,
|
|
"train_speed(iter/s)": 0.461216
|
|
},
|
|
{
|
|
"epoch": 1.3854545454545455,
|
|
"grad_norm": 3.880399703979492,
|
|
"learning_rate": 5.857498930480511e-06,
|
|
"loss": 0.9869282722473145,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1905,
|
|
"token_acc": 0.7427873945849978,
|
|
"train_speed(iter/s)": 0.461175
|
|
},
|
|
{
|
|
"epoch": 1.3890909090909092,
|
|
"grad_norm": 4.036713600158691,
|
|
"learning_rate": 5.838153160732164e-06,
|
|
"loss": 1.035546875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1910,
|
|
"token_acc": 0.7298916801106246,
|
|
"train_speed(iter/s)": 0.461216
|
|
},
|
|
{
|
|
"epoch": 1.3927272727272726,
|
|
"grad_norm": 3.1300320625305176,
|
|
"learning_rate": 5.818794472085481e-06,
|
|
"loss": 0.9794349670410156,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1915,
|
|
"token_acc": 0.7693474962063733,
|
|
"train_speed(iter/s)": 0.461234
|
|
},
|
|
{
|
|
"epoch": 1.3963636363636365,
|
|
"grad_norm": 3.984023094177246,
|
|
"learning_rate": 5.799423162926181e-06,
|
|
"loss": 1.0017403602600097,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1920,
|
|
"token_acc": 0.71779679613781,
|
|
"train_speed(iter/s)": 0.461278
|
|
},
|
|
{
|
|
"epoch": 1.4,
|
|
"grad_norm": 4.028362274169922,
|
|
"learning_rate": 5.780039531834503e-06,
|
|
"loss": 0.9450706481933594,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1925,
|
|
"token_acc": 0.7490964428381206,
|
|
"train_speed(iter/s)": 0.461318
|
|
},
|
|
{
|
|
"epoch": 1.4036363636363636,
|
|
"grad_norm": 4.613126277923584,
|
|
"learning_rate": 5.760643877580612e-06,
|
|
"loss": 1.0643499374389649,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1930,
|
|
"token_acc": 0.711301283664737,
|
|
"train_speed(iter/s)": 0.461372
|
|
},
|
|
{
|
|
"epoch": 1.4072727272727272,
|
|
"grad_norm": 4.1977949142456055,
|
|
"learning_rate": 5.741236499119997e-06,
|
|
"loss": 1.0503165245056152,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1935,
|
|
"token_acc": 0.7509952229299363,
|
|
"train_speed(iter/s)": 0.461437
|
|
},
|
|
{
|
|
"epoch": 1.410909090909091,
|
|
"grad_norm": 3.390101909637451,
|
|
"learning_rate": 5.721817695588856e-06,
|
|
"loss": 0.9490262985229492,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1940,
|
|
"token_acc": 0.716691431401684,
|
|
"train_speed(iter/s)": 0.46149
|
|
},
|
|
{
|
|
"epoch": 1.4145454545454546,
|
|
"grad_norm": 3.2861225605010986,
|
|
"learning_rate": 5.702387766299484e-06,
|
|
"loss": 0.9019038200378418,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1945,
|
|
"token_acc": 0.7688659793814433,
|
|
"train_speed(iter/s)": 0.461429
|
|
},
|
|
{
|
|
"epoch": 1.4181818181818182,
|
|
"grad_norm": 3.669062376022339,
|
|
"learning_rate": 5.682947010735669e-06,
|
|
"loss": 1.0124034881591797,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1950,
|
|
"token_acc": 0.7542513479883866,
|
|
"train_speed(iter/s)": 0.461492
|
|
},
|
|
{
|
|
"epoch": 1.4218181818181819,
|
|
"grad_norm": 3.639533519744873,
|
|
"learning_rate": 5.663495728548064e-06,
|
|
"loss": 1.0447866439819335,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1955,
|
|
"token_acc": 0.7088954950834667,
|
|
"train_speed(iter/s)": 0.461568
|
|
},
|
|
{
|
|
"epoch": 1.4254545454545455,
|
|
"grad_norm": 3.64686918258667,
|
|
"learning_rate": 5.64403421954958e-06,
|
|
"loss": 0.9409385681152344,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1960,
|
|
"token_acc": 0.7370488583353794,
|
|
"train_speed(iter/s)": 0.461624
|
|
},
|
|
{
|
|
"epoch": 1.4290909090909092,
|
|
"grad_norm": 3.4092624187469482,
|
|
"learning_rate": 5.624562783710756e-06,
|
|
"loss": 0.9381342887878418,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1965,
|
|
"token_acc": 0.7217558167426241,
|
|
"train_speed(iter/s)": 0.461614
|
|
},
|
|
{
|
|
"epoch": 1.4327272727272726,
|
|
"grad_norm": 4.09271240234375,
|
|
"learning_rate": 5.6050817211551426e-06,
|
|
"loss": 1.0289035797119142,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1970,
|
|
"token_acc": 0.7752707581227437,
|
|
"train_speed(iter/s)": 0.461635
|
|
},
|
|
{
|
|
"epoch": 1.4363636363636363,
|
|
"grad_norm": 3.552288055419922,
|
|
"learning_rate": 5.5855913321546655e-06,
|
|
"loss": 0.9234642028808594,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1975,
|
|
"token_acc": 0.722303973092285,
|
|
"train_speed(iter/s)": 0.461673
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"grad_norm": 3.510248899459839,
|
|
"learning_rate": 5.566091917125007e-06,
|
|
"loss": 0.9744023323059082,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1980,
|
|
"token_acc": 0.7038145100972326,
|
|
"train_speed(iter/s)": 0.461799
|
|
},
|
|
{
|
|
"epoch": 1.4436363636363636,
|
|
"grad_norm": 4.241003513336182,
|
|
"learning_rate": 5.546583776620978e-06,
|
|
"loss": 0.9870964050292969,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1985,
|
|
"token_acc": 0.722917158366769,
|
|
"train_speed(iter/s)": 0.461849
|
|
},
|
|
{
|
|
"epoch": 1.4472727272727273,
|
|
"grad_norm": 4.006800174713135,
|
|
"learning_rate": 5.527067211331868e-06,
|
|
"loss": 1.1195402145385742,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1990,
|
|
"token_acc": 0.6974145486415425,
|
|
"train_speed(iter/s)": 0.461926
|
|
},
|
|
{
|
|
"epoch": 1.450909090909091,
|
|
"grad_norm": 3.797157049179077,
|
|
"learning_rate": 5.507542522076832e-06,
|
|
"loss": 1.0271331787109375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 1995,
|
|
"token_acc": 0.7195024077046549,
|
|
"train_speed(iter/s)": 0.461974
|
|
},
|
|
{
|
|
"epoch": 1.4545454545454546,
|
|
"grad_norm": 3.211519241333008,
|
|
"learning_rate": 5.488010009800241e-06,
|
|
"loss": 0.9283449172973632,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2000,
|
|
"token_acc": 0.7473524962178517,
|
|
"train_speed(iter/s)": 0.462008
|
|
},
|
|
{
|
|
"epoch": 1.4581818181818182,
|
|
"grad_norm": 3.639495849609375,
|
|
"learning_rate": 5.468469975567045e-06,
|
|
"loss": 1.0161470413208007,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2005,
|
|
"token_acc": 0.7487437185929648,
|
|
"train_speed(iter/s)": 0.46202
|
|
},
|
|
{
|
|
"epoch": 1.461818181818182,
|
|
"grad_norm": 3.784331798553467,
|
|
"learning_rate": 5.448922720558135e-06,
|
|
"loss": 1.0102046966552733,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2010,
|
|
"token_acc": 0.7574331550802139,
|
|
"train_speed(iter/s)": 0.462075
|
|
},
|
|
{
|
|
"epoch": 1.4654545454545453,
|
|
"grad_norm": 3.8442893028259277,
|
|
"learning_rate": 5.429368546065702e-06,
|
|
"loss": 1.0593183517456055,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2015,
|
|
"token_acc": 0.7385492000876617,
|
|
"train_speed(iter/s)": 0.462128
|
|
},
|
|
{
|
|
"epoch": 1.4690909090909092,
|
|
"grad_norm": 3.8028147220611572,
|
|
"learning_rate": 5.409807753488584e-06,
|
|
"loss": 0.9948017120361328,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2020,
|
|
"token_acc": 0.7277379733879222,
|
|
"train_speed(iter/s)": 0.462169
|
|
},
|
|
{
|
|
"epoch": 1.4727272727272727,
|
|
"grad_norm": 3.7433605194091797,
|
|
"learning_rate": 5.390240644327633e-06,
|
|
"loss": 1.015998649597168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2025,
|
|
"token_acc": 0.7299979744784282,
|
|
"train_speed(iter/s)": 0.462178
|
|
},
|
|
{
|
|
"epoch": 1.4763636363636363,
|
|
"grad_norm": 3.3795201778411865,
|
|
"learning_rate": 5.3706675201810635e-06,
|
|
"loss": 0.9729107856750489,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2030,
|
|
"token_acc": 0.7344326318073334,
|
|
"train_speed(iter/s)": 0.462188
|
|
},
|
|
{
|
|
"epoch": 1.48,
|
|
"grad_norm": 4.034217834472656,
|
|
"learning_rate": 5.351088682739793e-06,
|
|
"loss": 0.9540383338928222,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2035,
|
|
"token_acc": 0.7512517068730087,
|
|
"train_speed(iter/s)": 0.462252
|
|
},
|
|
{
|
|
"epoch": 1.4836363636363636,
|
|
"grad_norm": 4.032739639282227,
|
|
"learning_rate": 5.331504433782813e-06,
|
|
"loss": 1.0312958717346192,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2040,
|
|
"token_acc": 0.7813462384513858,
|
|
"train_speed(iter/s)": 0.462294
|
|
},
|
|
{
|
|
"epoch": 1.4872727272727273,
|
|
"grad_norm": 3.369480848312378,
|
|
"learning_rate": 5.311915075172513e-06,
|
|
"loss": 0.9297122955322266,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2045,
|
|
"token_acc": 0.7658692785835374,
|
|
"train_speed(iter/s)": 0.462269
|
|
},
|
|
{
|
|
"epoch": 1.490909090909091,
|
|
"grad_norm": 3.4090447425842285,
|
|
"learning_rate": 5.292320908850053e-06,
|
|
"loss": 0.984104061126709,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2050,
|
|
"token_acc": 0.7230137647911132,
|
|
"train_speed(iter/s)": 0.462301
|
|
},
|
|
{
|
|
"epoch": 1.4945454545454546,
|
|
"grad_norm": 3.4420931339263916,
|
|
"learning_rate": 5.272722236830685e-06,
|
|
"loss": 1.0333056449890137,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2055,
|
|
"token_acc": 0.7212076941806671,
|
|
"train_speed(iter/s)": 0.462332
|
|
},
|
|
{
|
|
"epoch": 1.498181818181818,
|
|
"grad_norm": 3.902665615081787,
|
|
"learning_rate": 5.2531193611991205e-06,
|
|
"loss": 0.9698718070983887,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2060,
|
|
"token_acc": 0.7210183042289081,
|
|
"train_speed(iter/s)": 0.462353
|
|
},
|
|
{
|
|
"epoch": 1.501818181818182,
|
|
"grad_norm": 4.051348686218262,
|
|
"learning_rate": 5.233512584104854e-06,
|
|
"loss": 0.9682827949523926,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2065,
|
|
"token_acc": 0.7305626897034788,
|
|
"train_speed(iter/s)": 0.462433
|
|
},
|
|
{
|
|
"epoch": 1.5054545454545454,
|
|
"grad_norm": 3.7025203704833984,
|
|
"learning_rate": 5.2139022077575204e-06,
|
|
"loss": 0.9961311340332031,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2070,
|
|
"token_acc": 0.7380658798661265,
|
|
"train_speed(iter/s)": 0.462478
|
|
},
|
|
{
|
|
"epoch": 1.509090909090909,
|
|
"grad_norm": 3.5820093154907227,
|
|
"learning_rate": 5.194288534422233e-06,
|
|
"loss": 0.9701545715332032,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2075,
|
|
"token_acc": 0.7684444444444445,
|
|
"train_speed(iter/s)": 0.46251
|
|
},
|
|
{
|
|
"epoch": 1.5127272727272727,
|
|
"grad_norm": 3.9114859104156494,
|
|
"learning_rate": 5.174671866414917e-06,
|
|
"loss": 1.0580403327941894,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2080,
|
|
"token_acc": 0.7120789530143746,
|
|
"train_speed(iter/s)": 0.462532
|
|
},
|
|
{
|
|
"epoch": 1.5163636363636364,
|
|
"grad_norm": 3.454298734664917,
|
|
"learning_rate": 5.155052506097667e-06,
|
|
"loss": 0.9580793380737305,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2085,
|
|
"token_acc": 0.779483828814113,
|
|
"train_speed(iter/s)": 0.462514
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"grad_norm": 3.2281603813171387,
|
|
"learning_rate": 5.1354307558740625e-06,
|
|
"loss": 0.9244537353515625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2090,
|
|
"token_acc": 0.7337402652073248,
|
|
"train_speed(iter/s)": 0.462538
|
|
},
|
|
{
|
|
"epoch": 1.5236363636363637,
|
|
"grad_norm": 4.469293594360352,
|
|
"learning_rate": 5.115806918184528e-06,
|
|
"loss": 0.966168212890625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2095,
|
|
"token_acc": 0.7608447488584474,
|
|
"train_speed(iter/s)": 0.462543
|
|
},
|
|
{
|
|
"epoch": 1.5272727272727273,
|
|
"grad_norm": 3.593106508255005,
|
|
"learning_rate": 5.096181295501663e-06,
|
|
"loss": 1.0489200592041015,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2100,
|
|
"token_acc": 0.7567863397548161,
|
|
"train_speed(iter/s)": 0.46258
|
|
},
|
|
{
|
|
"epoch": 1.5309090909090908,
|
|
"grad_norm": 3.4803240299224854,
|
|
"learning_rate": 5.076554190325579e-06,
|
|
"loss": 0.953731918334961,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2105,
|
|
"token_acc": 0.7805164319248826,
|
|
"train_speed(iter/s)": 0.462557
|
|
},
|
|
{
|
|
"epoch": 1.5345454545454547,
|
|
"grad_norm": 3.9683210849761963,
|
|
"learning_rate": 5.056925905179234e-06,
|
|
"loss": 1.0079194068908692,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2110,
|
|
"token_acc": 0.7628815628815628,
|
|
"train_speed(iter/s)": 0.462599
|
|
},
|
|
{
|
|
"epoch": 1.538181818181818,
|
|
"grad_norm": 3.7638962268829346,
|
|
"learning_rate": 5.0372967426037815e-06,
|
|
"loss": 0.9692018508911133,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2115,
|
|
"token_acc": 0.7856516400631468,
|
|
"train_speed(iter/s)": 0.462596
|
|
},
|
|
{
|
|
"epoch": 1.541818181818182,
|
|
"grad_norm": 3.9223406314849854,
|
|
"learning_rate": 5.017667005153891e-06,
|
|
"loss": 0.9786901473999023,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2120,
|
|
"token_acc": 0.7256913470115968,
|
|
"train_speed(iter/s)": 0.462545
|
|
},
|
|
{
|
|
"epoch": 1.5454545454545454,
|
|
"grad_norm": 3.5813755989074707,
|
|
"learning_rate": 4.998036995393099e-06,
|
|
"loss": 0.9308863639831543,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2125,
|
|
"token_acc": 0.7481103552532123,
|
|
"train_speed(iter/s)": 0.462501
|
|
},
|
|
{
|
|
"epoch": 1.549090909090909,
|
|
"grad_norm": 3.1303036212921143,
|
|
"learning_rate": 4.978407015889136e-06,
|
|
"loss": 0.9750682830810546,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2130,
|
|
"token_acc": 0.7527090409186479,
|
|
"train_speed(iter/s)": 0.462434
|
|
},
|
|
{
|
|
"epoch": 1.5527272727272727,
|
|
"grad_norm": 3.235821008682251,
|
|
"learning_rate": 4.958777369209268e-06,
|
|
"loss": 0.9330739974975586,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2135,
|
|
"token_acc": 0.7765167148163433,
|
|
"train_speed(iter/s)": 0.462449
|
|
},
|
|
{
|
|
"epoch": 1.5563636363636364,
|
|
"grad_norm": 3.6677234172821045,
|
|
"learning_rate": 4.939148357915628e-06,
|
|
"loss": 1.0006421089172364,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2140,
|
|
"token_acc": 0.7976786527082385,
|
|
"train_speed(iter/s)": 0.46251
|
|
},
|
|
{
|
|
"epoch": 1.56,
|
|
"grad_norm": 3.8058927059173584,
|
|
"learning_rate": 4.919520284560558e-06,
|
|
"loss": 0.9706598281860351,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2145,
|
|
"token_acc": 0.8033300128077415,
|
|
"train_speed(iter/s)": 0.462417
|
|
},
|
|
{
|
|
"epoch": 1.5636363636363635,
|
|
"grad_norm": 3.2341036796569824,
|
|
"learning_rate": 4.899893451681944e-06,
|
|
"loss": 0.9487602233886718,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2150,
|
|
"token_acc": 0.7283561951877671,
|
|
"train_speed(iter/s)": 0.462497
|
|
},
|
|
{
|
|
"epoch": 1.5672727272727274,
|
|
"grad_norm": 3.6466586589813232,
|
|
"learning_rate": 4.880268161798552e-06,
|
|
"loss": 1.0922494888305665,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2155,
|
|
"token_acc": 0.7327859675312558,
|
|
"train_speed(iter/s)": 0.462473
|
|
},
|
|
{
|
|
"epoch": 1.5709090909090908,
|
|
"grad_norm": 3.687445640563965,
|
|
"learning_rate": 4.860644717405358e-06,
|
|
"loss": 1.0499517440795898,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2160,
|
|
"token_acc": 0.7352428393524284,
|
|
"train_speed(iter/s)": 0.462559
|
|
},
|
|
{
|
|
"epoch": 1.5745454545454547,
|
|
"grad_norm": 3.905825614929199,
|
|
"learning_rate": 4.841023420968901e-06,
|
|
"loss": 0.9939208984375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2165,
|
|
"token_acc": 0.7321428571428571,
|
|
"train_speed(iter/s)": 0.462625
|
|
},
|
|
{
|
|
"epoch": 1.5781818181818181,
|
|
"grad_norm": 3.7845633029937744,
|
|
"learning_rate": 4.821404574922611e-06,
|
|
"loss": 0.9922620773315429,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2170,
|
|
"token_acc": 0.7227589305773984,
|
|
"train_speed(iter/s)": 0.462659
|
|
},
|
|
{
|
|
"epoch": 1.5818181818181818,
|
|
"grad_norm": 2.901113271713257,
|
|
"learning_rate": 4.801788481662145e-06,
|
|
"loss": 0.94874267578125,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2175,
|
|
"token_acc": 0.8261739799846035,
|
|
"train_speed(iter/s)": 0.462611
|
|
},
|
|
{
|
|
"epoch": 1.5854545454545454,
|
|
"grad_norm": 3.699376106262207,
|
|
"learning_rate": 4.782175443540733e-06,
|
|
"loss": 0.981483268737793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2180,
|
|
"token_acc": 0.7473048727899957,
|
|
"train_speed(iter/s)": 0.462622
|
|
},
|
|
{
|
|
"epoch": 1.589090909090909,
|
|
"grad_norm": 4.233500957489014,
|
|
"learning_rate": 4.762565762864513e-06,
|
|
"loss": 1.017900562286377,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2185,
|
|
"token_acc": 0.7384370015948963,
|
|
"train_speed(iter/s)": 0.462701
|
|
},
|
|
{
|
|
"epoch": 1.5927272727272728,
|
|
"grad_norm": 4.156431674957275,
|
|
"learning_rate": 4.742959741887872e-06,
|
|
"loss": 0.9760846138000489,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2190,
|
|
"token_acc": 0.7447447447447447,
|
|
"train_speed(iter/s)": 0.462745
|
|
},
|
|
{
|
|
"epoch": 1.5963636363636362,
|
|
"grad_norm": 3.670698881149292,
|
|
"learning_rate": 4.723357682808793e-06,
|
|
"loss": 1.0618730545043946,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2195,
|
|
"token_acc": 0.7577669902912622,
|
|
"train_speed(iter/s)": 0.462832
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"grad_norm": 4.159302711486816,
|
|
"learning_rate": 4.703759887764185e-06,
|
|
"loss": 0.9977631568908691,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2200,
|
|
"token_acc": 0.7650755902836759,
|
|
"train_speed(iter/s)": 0.462839
|
|
},
|
|
{
|
|
"epoch": 1.6036363636363635,
|
|
"grad_norm": 3.6215317249298096,
|
|
"learning_rate": 4.684166658825242e-06,
|
|
"loss": 1.054893684387207,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2205,
|
|
"token_acc": 0.7607733519683205,
|
|
"train_speed(iter/s)": 0.462843
|
|
},
|
|
{
|
|
"epoch": 1.6072727272727274,
|
|
"grad_norm": 3.5773353576660156,
|
|
"learning_rate": 4.664578297992767e-06,
|
|
"loss": 0.9340826034545898,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2210,
|
|
"token_acc": 0.7217975445911513,
|
|
"train_speed(iter/s)": 0.462864
|
|
},
|
|
{
|
|
"epoch": 1.6109090909090908,
|
|
"grad_norm": 3.176886558532715,
|
|
"learning_rate": 4.644995107192538e-06,
|
|
"loss": 0.9768821716308593,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2215,
|
|
"token_acc": 0.7468906829587606,
|
|
"train_speed(iter/s)": 0.462857
|
|
},
|
|
{
|
|
"epoch": 1.6145454545454545,
|
|
"grad_norm": 3.9905526638031006,
|
|
"learning_rate": 4.625417388270643e-06,
|
|
"loss": 1.0507758140563965,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2220,
|
|
"token_acc": 0.6920529801324503,
|
|
"train_speed(iter/s)": 0.462889
|
|
},
|
|
{
|
|
"epoch": 1.6181818181818182,
|
|
"grad_norm": 3.2976949214935303,
|
|
"learning_rate": 4.6058454429888215e-06,
|
|
"loss": 0.995277214050293,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2225,
|
|
"token_acc": 0.7551693796744391,
|
|
"train_speed(iter/s)": 0.462954
|
|
},
|
|
{
|
|
"epoch": 1.6218181818181818,
|
|
"grad_norm": 4.139730930328369,
|
|
"learning_rate": 4.586279573019829e-06,
|
|
"loss": 1.0577568054199218,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2230,
|
|
"token_acc": 0.704527402700556,
|
|
"train_speed(iter/s)": 0.463008
|
|
},
|
|
{
|
|
"epoch": 1.6254545454545455,
|
|
"grad_norm": 3.9827053546905518,
|
|
"learning_rate": 4.566720079942779e-06,
|
|
"loss": 0.8773839950561524,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2235,
|
|
"token_acc": 0.7590271281571562,
|
|
"train_speed(iter/s)": 0.462998
|
|
},
|
|
{
|
|
"epoch": 1.6290909090909091,
|
|
"grad_norm": 3.42809796333313,
|
|
"learning_rate": 4.547167265238488e-06,
|
|
"loss": 0.9219103813171386,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2240,
|
|
"token_acc": 0.7726116123062688,
|
|
"train_speed(iter/s)": 0.463011
|
|
},
|
|
{
|
|
"epoch": 1.6327272727272728,
|
|
"grad_norm": 3.682155132293701,
|
|
"learning_rate": 4.52762143028484e-06,
|
|
"loss": 0.9474153518676758,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2245,
|
|
"token_acc": 0.7516981132075472,
|
|
"train_speed(iter/s)": 0.463065
|
|
},
|
|
{
|
|
"epoch": 1.6363636363636362,
|
|
"grad_norm": 4.005671977996826,
|
|
"learning_rate": 4.508082876352136e-06,
|
|
"loss": 1.0329013824462892,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2250,
|
|
"token_acc": 0.7462128296240883,
|
|
"train_speed(iter/s)": 0.463106
|
|
},
|
|
{
|
|
"epoch": 1.6400000000000001,
|
|
"grad_norm": 3.734011650085449,
|
|
"learning_rate": 4.488551904598452e-06,
|
|
"loss": 0.9806623458862305,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2255,
|
|
"token_acc": 0.7730970891339721,
|
|
"train_speed(iter/s)": 0.463046
|
|
},
|
|
{
|
|
"epoch": 1.6436363636363636,
|
|
"grad_norm": 3.6952412128448486,
|
|
"learning_rate": 4.469028816064992e-06,
|
|
"loss": 1.0010573387145996,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2260,
|
|
"token_acc": 0.7182100132919805,
|
|
"train_speed(iter/s)": 0.463111
|
|
},
|
|
{
|
|
"epoch": 1.6472727272727272,
|
|
"grad_norm": 4.075192451477051,
|
|
"learning_rate": 4.4495139116714545e-06,
|
|
"loss": 0.878773307800293,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2265,
|
|
"token_acc": 0.75185508511567,
|
|
"train_speed(iter/s)": 0.463138
|
|
},
|
|
{
|
|
"epoch": 1.6509090909090909,
|
|
"grad_norm": 3.910977840423584,
|
|
"learning_rate": 4.430007492211392e-06,
|
|
"loss": 0.9923901557922363,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2270,
|
|
"token_acc": 0.7322286263208453,
|
|
"train_speed(iter/s)": 0.463156
|
|
},
|
|
{
|
|
"epoch": 1.6545454545454545,
|
|
"grad_norm": 4.175351619720459,
|
|
"learning_rate": 4.410509858347571e-06,
|
|
"loss": 1.054445457458496,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2275,
|
|
"token_acc": 0.7337883959044369,
|
|
"train_speed(iter/s)": 0.463189
|
|
},
|
|
{
|
|
"epoch": 1.6581818181818182,
|
|
"grad_norm": 3.577937602996826,
|
|
"learning_rate": 4.391021310607345e-06,
|
|
"loss": 0.9427365303039551,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2280,
|
|
"token_acc": 0.7608598298253471,
|
|
"train_speed(iter/s)": 0.463206
|
|
},
|
|
{
|
|
"epoch": 1.6618181818181819,
|
|
"grad_norm": 3.752776861190796,
|
|
"learning_rate": 4.371542149378017e-06,
|
|
"loss": 1.076515007019043,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2285,
|
|
"token_acc": 0.7492560999801626,
|
|
"train_speed(iter/s)": 0.463236
|
|
},
|
|
{
|
|
"epoch": 1.6654545454545455,
|
|
"grad_norm": 3.7193381786346436,
|
|
"learning_rate": 4.352072674902208e-06,
|
|
"loss": 0.9631393432617188,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2290,
|
|
"token_acc": 0.7573891625615764,
|
|
"train_speed(iter/s)": 0.463228
|
|
},
|
|
{
|
|
"epoch": 1.669090909090909,
|
|
"grad_norm": 3.5725324153900146,
|
|
"learning_rate": 4.332613187273234e-06,
|
|
"loss": 0.9510932922363281,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2295,
|
|
"token_acc": 0.7744831906452286,
|
|
"train_speed(iter/s)": 0.463264
|
|
},
|
|
{
|
|
"epoch": 1.6727272727272728,
|
|
"grad_norm": 3.7124252319335938,
|
|
"learning_rate": 4.313163986430482e-06,
|
|
"loss": 0.9919276237487793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2300,
|
|
"token_acc": 0.7588126159554731,
|
|
"train_speed(iter/s)": 0.463335
|
|
},
|
|
{
|
|
"epoch": 1.6763636363636363,
|
|
"grad_norm": 3.576068878173828,
|
|
"learning_rate": 4.293725372154775e-06,
|
|
"loss": 1.0437440872192383,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2305,
|
|
"token_acc": 0.7543238993710691,
|
|
"train_speed(iter/s)": 0.463326
|
|
},
|
|
{
|
|
"epoch": 1.6800000000000002,
|
|
"grad_norm": 3.792983055114746,
|
|
"learning_rate": 4.2742976440637666e-06,
|
|
"loss": 1.03680477142334,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2310,
|
|
"token_acc": 0.7257747171667487,
|
|
"train_speed(iter/s)": 0.463336
|
|
},
|
|
{
|
|
"epoch": 1.6836363636363636,
|
|
"grad_norm": 3.860466241836548,
|
|
"learning_rate": 4.254881101607315e-06,
|
|
"loss": 0.9429449081420899,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2315,
|
|
"token_acc": 0.7263529411764705,
|
|
"train_speed(iter/s)": 0.463353
|
|
},
|
|
{
|
|
"epoch": 1.6872727272727273,
|
|
"grad_norm": 4.08974027633667,
|
|
"learning_rate": 4.235476044062866e-06,
|
|
"loss": 0.9691776275634766,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2320,
|
|
"token_acc": 0.7541046200840015,
|
|
"train_speed(iter/s)": 0.463384
|
|
},
|
|
{
|
|
"epoch": 1.690909090909091,
|
|
"grad_norm": 3.7761855125427246,
|
|
"learning_rate": 4.216082770530843e-06,
|
|
"loss": 0.9898856163024903,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2325,
|
|
"token_acc": 0.7277238890142116,
|
|
"train_speed(iter/s)": 0.463395
|
|
},
|
|
{
|
|
"epoch": 1.6945454545454546,
|
|
"grad_norm": 3.6072325706481934,
|
|
"learning_rate": 4.196701579930038e-06,
|
|
"loss": 1.007903480529785,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2330,
|
|
"token_acc": 0.7275,
|
|
"train_speed(iter/s)": 0.463335
|
|
},
|
|
{
|
|
"epoch": 1.6981818181818182,
|
|
"grad_norm": 4.309933662414551,
|
|
"learning_rate": 4.177332770993e-06,
|
|
"loss": 0.9898582458496094,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2335,
|
|
"token_acc": 0.7726732673267327,
|
|
"train_speed(iter/s)": 0.463339
|
|
},
|
|
{
|
|
"epoch": 1.7018181818181817,
|
|
"grad_norm": 3.4883134365081787,
|
|
"learning_rate": 4.157976642261434e-06,
|
|
"loss": 0.9960001945495606,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2340,
|
|
"token_acc": 0.764170228582048,
|
|
"train_speed(iter/s)": 0.46331
|
|
},
|
|
{
|
|
"epoch": 1.7054545454545456,
|
|
"grad_norm": 3.0042526721954346,
|
|
"learning_rate": 4.138633492081597e-06,
|
|
"loss": 0.9730648994445801,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2345,
|
|
"token_acc": 0.7559236254888428,
|
|
"train_speed(iter/s)": 0.463311
|
|
},
|
|
{
|
|
"epoch": 1.709090909090909,
|
|
"grad_norm": 3.365316152572632,
|
|
"learning_rate": 4.119303618599705e-06,
|
|
"loss": 0.97021484375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2350,
|
|
"token_acc": 0.7311364912937652,
|
|
"train_speed(iter/s)": 0.463306
|
|
},
|
|
{
|
|
"epoch": 1.7127272727272729,
|
|
"grad_norm": 3.245842456817627,
|
|
"learning_rate": 4.099987319757326e-06,
|
|
"loss": 0.9768846511840821,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2355,
|
|
"token_acc": 0.7251812953238309,
|
|
"train_speed(iter/s)": 0.463285
|
|
},
|
|
{
|
|
"epoch": 1.7163636363636363,
|
|
"grad_norm": 3.2012290954589844,
|
|
"learning_rate": 4.0806848932868e-06,
|
|
"loss": 0.9911705017089844,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2360,
|
|
"token_acc": 0.7289512555391433,
|
|
"train_speed(iter/s)": 0.463339
|
|
},
|
|
{
|
|
"epoch": 1.72,
|
|
"grad_norm": 3.820871353149414,
|
|
"learning_rate": 4.0613966367066456e-06,
|
|
"loss": 1.0202957153320313,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2365,
|
|
"token_acc": 0.7401774397972116,
|
|
"train_speed(iter/s)": 0.463404
|
|
},
|
|
{
|
|
"epoch": 1.7236363636363636,
|
|
"grad_norm": 4.134464740753174,
|
|
"learning_rate": 4.042122847316973e-06,
|
|
"loss": 0.9151566505432129,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2370,
|
|
"token_acc": 0.7081224391419619,
|
|
"train_speed(iter/s)": 0.463391
|
|
},
|
|
{
|
|
"epoch": 1.7272727272727273,
|
|
"grad_norm": 3.890723705291748,
|
|
"learning_rate": 4.022863822194895e-06,
|
|
"loss": 1.0222078323364259,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2375,
|
|
"token_acc": 0.7584531281635959,
|
|
"train_speed(iter/s)": 0.463429
|
|
},
|
|
{
|
|
"epoch": 1.730909090909091,
|
|
"grad_norm": 3.866518497467041,
|
|
"learning_rate": 4.003619858189964e-06,
|
|
"loss": 0.9343522071838379,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2380,
|
|
"token_acc": 0.7731607629427792,
|
|
"train_speed(iter/s)": 0.463458
|
|
},
|
|
{
|
|
"epoch": 1.7345454545454544,
|
|
"grad_norm": 3.6248366832733154,
|
|
"learning_rate": 3.984391251919586e-06,
|
|
"loss": 0.9574322700500488,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2385,
|
|
"token_acc": 0.7667273094804933,
|
|
"train_speed(iter/s)": 0.463492
|
|
},
|
|
{
|
|
"epoch": 1.7381818181818183,
|
|
"grad_norm": 3.5135788917541504,
|
|
"learning_rate": 3.965178299764442e-06,
|
|
"loss": 0.9733609199523926,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2390,
|
|
"token_acc": 0.7768259693417493,
|
|
"train_speed(iter/s)": 0.463541
|
|
},
|
|
{
|
|
"epoch": 1.7418181818181817,
|
|
"grad_norm": 3.7573583126068115,
|
|
"learning_rate": 3.9459812978639366e-06,
|
|
"loss": 0.9856127738952637,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2395,
|
|
"token_acc": 0.6941059437950758,
|
|
"train_speed(iter/s)": 0.463547
|
|
},
|
|
{
|
|
"epoch": 1.7454545454545456,
|
|
"grad_norm": 3.8713910579681396,
|
|
"learning_rate": 3.926800542111624e-06,
|
|
"loss": 1.0458408355712892,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2400,
|
|
"token_acc": 0.7367864693446089,
|
|
"train_speed(iter/s)": 0.463544
|
|
},
|
|
{
|
|
"epoch": 1.749090909090909,
|
|
"grad_norm": 4.335873603820801,
|
|
"learning_rate": 3.907636328150643e-06,
|
|
"loss": 1.0263343811035157,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2405,
|
|
"token_acc": 0.7440373739857389,
|
|
"train_speed(iter/s)": 0.463594
|
|
},
|
|
{
|
|
"epoch": 1.7527272727272727,
|
|
"grad_norm": 3.356822967529297,
|
|
"learning_rate": 3.888488951369168e-06,
|
|
"loss": 0.9377895355224609,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2410,
|
|
"token_acc": 0.773037542662116,
|
|
"train_speed(iter/s)": 0.463587
|
|
},
|
|
{
|
|
"epoch": 1.7563636363636363,
|
|
"grad_norm": 3.5357346534729004,
|
|
"learning_rate": 3.869358706895855e-06,
|
|
"loss": 0.977473258972168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2415,
|
|
"token_acc": 0.7483623221142987,
|
|
"train_speed(iter/s)": 0.46363
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"grad_norm": 3.981171131134033,
|
|
"learning_rate": 3.850245889595288e-06,
|
|
"loss": 1.0373311042785645,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2420,
|
|
"token_acc": 0.7319004524886877,
|
|
"train_speed(iter/s)": 0.463625
|
|
},
|
|
{
|
|
"epoch": 1.7636363636363637,
|
|
"grad_norm": 3.7682418823242188,
|
|
"learning_rate": 3.831150794063434e-06,
|
|
"loss": 1.030501365661621,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2425,
|
|
"token_acc": 0.6902185444515171,
|
|
"train_speed(iter/s)": 0.463685
|
|
},
|
|
{
|
|
"epoch": 1.767272727272727,
|
|
"grad_norm": 4.16378116607666,
|
|
"learning_rate": 3.8120737146231114e-06,
|
|
"loss": 0.9865205764770508,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2430,
|
|
"token_acc": 0.7317015768416097,
|
|
"train_speed(iter/s)": 0.463737
|
|
},
|
|
{
|
|
"epoch": 1.770909090909091,
|
|
"grad_norm": 3.2445712089538574,
|
|
"learning_rate": 3.7930149453194443e-06,
|
|
"loss": 1.0164010047912597,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2435,
|
|
"token_acc": 0.7629260182876143,
|
|
"train_speed(iter/s)": 0.463707
|
|
},
|
|
{
|
|
"epoch": 1.7745454545454544,
|
|
"grad_norm": 4.097912788391113,
|
|
"learning_rate": 3.773974779915329e-06,
|
|
"loss": 1.02340087890625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2440,
|
|
"token_acc": 0.7449856733524355,
|
|
"train_speed(iter/s)": 0.46375
|
|
},
|
|
{
|
|
"epoch": 1.7781818181818183,
|
|
"grad_norm": 3.395263195037842,
|
|
"learning_rate": 3.754953511886917e-06,
|
|
"loss": 1.0150327682495117,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2445,
|
|
"token_acc": 0.7850148999574287,
|
|
"train_speed(iter/s)": 0.463806
|
|
},
|
|
{
|
|
"epoch": 1.7818181818181817,
|
|
"grad_norm": 4.005068302154541,
|
|
"learning_rate": 3.7359514344190815e-06,
|
|
"loss": 0.9756942749023437,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2450,
|
|
"token_acc": 0.7594743858312702,
|
|
"train_speed(iter/s)": 0.463838
|
|
},
|
|
{
|
|
"epoch": 1.7854545454545454,
|
|
"grad_norm": 4.276913642883301,
|
|
"learning_rate": 3.7169688404008975e-06,
|
|
"loss": 1.0148825645446777,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2455,
|
|
"token_acc": 0.7754668930390493,
|
|
"train_speed(iter/s)": 0.463893
|
|
},
|
|
{
|
|
"epoch": 1.789090909090909,
|
|
"grad_norm": 3.7833950519561768,
|
|
"learning_rate": 3.6980060224211368e-06,
|
|
"loss": 1.0185523986816407,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2460,
|
|
"token_acc": 0.7297499066815976,
|
|
"train_speed(iter/s)": 0.463928
|
|
},
|
|
{
|
|
"epoch": 1.7927272727272727,
|
|
"grad_norm": 3.901315212249756,
|
|
"learning_rate": 3.679063272763752e-06,
|
|
"loss": 1.0209418296813966,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2465,
|
|
"token_acc": 0.6788369042413642,
|
|
"train_speed(iter/s)": 0.463978
|
|
},
|
|
{
|
|
"epoch": 1.7963636363636364,
|
|
"grad_norm": 4.074141979217529,
|
|
"learning_rate": 3.660140883403369e-06,
|
|
"loss": 0.9935298919677734,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2470,
|
|
"token_acc": 0.723399355135882,
|
|
"train_speed(iter/s)": 0.46403
|
|
},
|
|
{
|
|
"epoch": 1.8,
|
|
"grad_norm": 3.7663679122924805,
|
|
"learning_rate": 3.641239146000792e-06,
|
|
"loss": 0.9881417274475097,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2475,
|
|
"token_acc": 0.7416889568207872,
|
|
"train_speed(iter/s)": 0.464076
|
|
},
|
|
{
|
|
"epoch": 1.8036363636363637,
|
|
"grad_norm": 3.5808398723602295,
|
|
"learning_rate": 3.6223583518985055e-06,
|
|
"loss": 0.9529662132263184,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2480,
|
|
"token_acc": 0.7698131062219068,
|
|
"train_speed(iter/s)": 0.464133
|
|
},
|
|
{
|
|
"epoch": 1.8072727272727271,
|
|
"grad_norm": 3.6277318000793457,
|
|
"learning_rate": 3.6034987921161834e-06,
|
|
"loss": 0.9884878158569336,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2485,
|
|
"token_acc": 0.740924092409241,
|
|
"train_speed(iter/s)": 0.464191
|
|
},
|
|
{
|
|
"epoch": 1.810909090909091,
|
|
"grad_norm": 4.147091388702393,
|
|
"learning_rate": 3.5846607573462023e-06,
|
|
"loss": 1.0621925354003907,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2490,
|
|
"token_acc": 0.6963470319634704,
|
|
"train_speed(iter/s)": 0.464255
|
|
},
|
|
{
|
|
"epoch": 1.8145454545454545,
|
|
"grad_norm": 3.680851697921753,
|
|
"learning_rate": 3.5658445379491647e-06,
|
|
"loss": 0.9677799224853516,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2495,
|
|
"token_acc": 0.781350977073396,
|
|
"train_speed(iter/s)": 0.464275
|
|
},
|
|
{
|
|
"epoch": 1.8181818181818183,
|
|
"grad_norm": 3.780642032623291,
|
|
"learning_rate": 3.5470504239494224e-06,
|
|
"loss": 1.0042539596557618,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2500,
|
|
"token_acc": 0.7281123014407093,
|
|
"train_speed(iter/s)": 0.464297
|
|
},
|
|
{
|
|
"epoch": 1.8218181818181818,
|
|
"grad_norm": 4.483243942260742,
|
|
"learning_rate": 3.528278705030599e-06,
|
|
"loss": 1.005680465698242,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2505,
|
|
"token_acc": 0.7122031507171408,
|
|
"train_speed(iter/s)": 0.46437
|
|
},
|
|
{
|
|
"epoch": 1.8254545454545454,
|
|
"grad_norm": 3.688572883605957,
|
|
"learning_rate": 3.509529670531138e-06,
|
|
"loss": 1.040778923034668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2510,
|
|
"token_acc": 0.7308886301901436,
|
|
"train_speed(iter/s)": 0.464352
|
|
},
|
|
{
|
|
"epoch": 1.829090909090909,
|
|
"grad_norm": 3.509716033935547,
|
|
"learning_rate": 3.4908036094398324e-06,
|
|
"loss": 1.0052711486816406,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2515,
|
|
"token_acc": 0.7249230769230769,
|
|
"train_speed(iter/s)": 0.464382
|
|
},
|
|
{
|
|
"epoch": 1.8327272727272728,
|
|
"grad_norm": 3.991241693496704,
|
|
"learning_rate": 3.4721008103913715e-06,
|
|
"loss": 1.0515169143676757,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2520,
|
|
"token_acc": 0.6967725794345759,
|
|
"train_speed(iter/s)": 0.464478
|
|
},
|
|
{
|
|
"epoch": 1.8363636363636364,
|
|
"grad_norm": 3.670917510986328,
|
|
"learning_rate": 3.4534215616619e-06,
|
|
"loss": 1.023714828491211,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2525,
|
|
"token_acc": 0.7610474631751227,
|
|
"train_speed(iter/s)": 0.464493
|
|
},
|
|
{
|
|
"epoch": 1.8399999999999999,
|
|
"grad_norm": 4.095003604888916,
|
|
"learning_rate": 3.434766151164565e-06,
|
|
"loss": 1.0314668655395507,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2530,
|
|
"token_acc": 0.7246666666666667,
|
|
"train_speed(iter/s)": 0.464568
|
|
},
|
|
{
|
|
"epoch": 1.8436363636363637,
|
|
"grad_norm": 3.6983370780944824,
|
|
"learning_rate": 3.416134866445086e-06,
|
|
"loss": 1.0155969619750977,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2535,
|
|
"token_acc": 0.720302596606011,
|
|
"train_speed(iter/s)": 0.464616
|
|
},
|
|
{
|
|
"epoch": 1.8472727272727272,
|
|
"grad_norm": 3.8502554893493652,
|
|
"learning_rate": 3.397527994677313e-06,
|
|
"loss": 0.9803200721740722,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2540,
|
|
"token_acc": 0.7569339384770549,
|
|
"train_speed(iter/s)": 0.464657
|
|
},
|
|
{
|
|
"epoch": 1.850909090909091,
|
|
"grad_norm": 3.3400015830993652,
|
|
"learning_rate": 3.378945822658811e-06,
|
|
"loss": 0.9980621337890625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2545,
|
|
"token_acc": 0.7765086941697921,
|
|
"train_speed(iter/s)": 0.464654
|
|
},
|
|
{
|
|
"epoch": 1.8545454545454545,
|
|
"grad_norm": 3.298196315765381,
|
|
"learning_rate": 3.3603886368064363e-06,
|
|
"loss": 1.0117303848266601,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2550,
|
|
"token_acc": 0.7541046938381302,
|
|
"train_speed(iter/s)": 0.464707
|
|
},
|
|
{
|
|
"epoch": 1.8581818181818182,
|
|
"grad_norm": 3.4078967571258545,
|
|
"learning_rate": 3.341856723151914e-06,
|
|
"loss": 1.0041580200195312,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2555,
|
|
"token_acc": 0.7841938405797102,
|
|
"train_speed(iter/s)": 0.464743
|
|
},
|
|
{
|
|
"epoch": 1.8618181818181818,
|
|
"grad_norm": 4.263998031616211,
|
|
"learning_rate": 3.32335036733744e-06,
|
|
"loss": 0.9891136169433594,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2560,
|
|
"token_acc": 0.7680060652009097,
|
|
"train_speed(iter/s)": 0.464712
|
|
},
|
|
{
|
|
"epoch": 1.8654545454545455,
|
|
"grad_norm": 3.5387911796569824,
|
|
"learning_rate": 3.3048698546112758e-06,
|
|
"loss": 0.9509899139404296,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2565,
|
|
"token_acc": 0.7269250382457929,
|
|
"train_speed(iter/s)": 0.464716
|
|
},
|
|
{
|
|
"epoch": 1.8690909090909091,
|
|
"grad_norm": 3.82328200340271,
|
|
"learning_rate": 3.2864154698233415e-06,
|
|
"loss": 0.9231161117553711,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2570,
|
|
"token_acc": 0.7634560906515581,
|
|
"train_speed(iter/s)": 0.464755
|
|
},
|
|
{
|
|
"epoch": 1.8727272727272726,
|
|
"grad_norm": 3.4116268157958984,
|
|
"learning_rate": 3.2679874974208404e-06,
|
|
"loss": 0.9480420112609863,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2575,
|
|
"token_acc": 0.7342569269521411,
|
|
"train_speed(iter/s)": 0.464772
|
|
},
|
|
{
|
|
"epoch": 1.8763636363636365,
|
|
"grad_norm": 3.618499517440796,
|
|
"learning_rate": 3.2495862214438665e-06,
|
|
"loss": 0.9754660606384278,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2580,
|
|
"token_acc": 0.7174500270124257,
|
|
"train_speed(iter/s)": 0.464804
|
|
},
|
|
{
|
|
"epoch": 1.88,
|
|
"grad_norm": 3.999831199645996,
|
|
"learning_rate": 3.2312119255210295e-06,
|
|
"loss": 0.9918521881103516,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2585,
|
|
"token_acc": 0.7208473154362416,
|
|
"train_speed(iter/s)": 0.464847
|
|
},
|
|
{
|
|
"epoch": 1.8836363636363638,
|
|
"grad_norm": 3.6423439979553223,
|
|
"learning_rate": 3.212864892865073e-06,
|
|
"loss": 0.9359735488891602,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2590,
|
|
"token_acc": 0.7750716332378224,
|
|
"train_speed(iter/s)": 0.464839
|
|
},
|
|
{
|
|
"epoch": 1.8872727272727272,
|
|
"grad_norm": 3.45302677154541,
|
|
"learning_rate": 3.1945454062685262e-06,
|
|
"loss": 0.980652904510498,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2595,
|
|
"token_acc": 0.7254253308128544,
|
|
"train_speed(iter/s)": 0.464846
|
|
},
|
|
{
|
|
"epoch": 1.8909090909090909,
|
|
"grad_norm": 3.5030007362365723,
|
|
"learning_rate": 3.176253748099335e-06,
|
|
"loss": 0.9258022308349609,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2600,
|
|
"token_acc": 0.7362502555714577,
|
|
"train_speed(iter/s)": 0.464837
|
|
},
|
|
{
|
|
"epoch": 1.8945454545454545,
|
|
"grad_norm": 3.8884432315826416,
|
|
"learning_rate": 3.1579902002965057e-06,
|
|
"loss": 1.0826837539672851,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2605,
|
|
"token_acc": 0.7276667923105918,
|
|
"train_speed(iter/s)": 0.464868
|
|
},
|
|
{
|
|
"epoch": 1.8981818181818182,
|
|
"grad_norm": 4.530186176300049,
|
|
"learning_rate": 3.13975504436577e-06,
|
|
"loss": 1.0418546676635743,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2610,
|
|
"token_acc": 0.760082304526749,
|
|
"train_speed(iter/s)": 0.464853
|
|
},
|
|
{
|
|
"epoch": 1.9018181818181819,
|
|
"grad_norm": 3.1030256748199463,
|
|
"learning_rate": 3.1215485613752394e-06,
|
|
"loss": 0.9568429946899414,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2615,
|
|
"token_acc": 0.7639067376537544,
|
|
"train_speed(iter/s)": 0.464815
|
|
},
|
|
{
|
|
"epoch": 1.9054545454545453,
|
|
"grad_norm": 3.6175343990325928,
|
|
"learning_rate": 3.103371031951072e-06,
|
|
"loss": 0.8990876197814941,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2620,
|
|
"token_acc": 0.7014793552660632,
|
|
"train_speed(iter/s)": 0.464751
|
|
},
|
|
{
|
|
"epoch": 1.9090909090909092,
|
|
"grad_norm": 3.7534470558166504,
|
|
"learning_rate": 3.0852227362731525e-06,
|
|
"loss": 1.0102110862731934,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2625,
|
|
"token_acc": 0.7285464098073555,
|
|
"train_speed(iter/s)": 0.464805
|
|
},
|
|
{
|
|
"epoch": 1.9127272727272726,
|
|
"grad_norm": 4.0336785316467285,
|
|
"learning_rate": 3.0671039540707704e-06,
|
|
"loss": 0.9847513198852539,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2630,
|
|
"token_acc": 0.7320584443489472,
|
|
"train_speed(iter/s)": 0.464847
|
|
},
|
|
{
|
|
"epoch": 1.9163636363636365,
|
|
"grad_norm": 3.6290838718414307,
|
|
"learning_rate": 3.049014964618302e-06,
|
|
"loss": 1.0085186958312988,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2635,
|
|
"token_acc": 0.7455083491862186,
|
|
"train_speed(iter/s)": 0.464907
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"grad_norm": 3.8827333450317383,
|
|
"learning_rate": 3.0309560467309203e-06,
|
|
"loss": 0.971209716796875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2640,
|
|
"token_acc": 0.7537816657554219,
|
|
"train_speed(iter/s)": 0.464862
|
|
},
|
|
{
|
|
"epoch": 1.9236363636363636,
|
|
"grad_norm": 3.2638890743255615,
|
|
"learning_rate": 3.012927478760285e-06,
|
|
"loss": 0.9603887557983398,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2645,
|
|
"token_acc": 0.7753898019384745,
|
|
"train_speed(iter/s)": 0.464866
|
|
},
|
|
{
|
|
"epoch": 1.9272727272727272,
|
|
"grad_norm": 3.629682779312134,
|
|
"learning_rate": 2.99492953859026e-06,
|
|
"loss": 0.9840577125549317,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2650,
|
|
"token_acc": 0.6956521739130435,
|
|
"train_speed(iter/s)": 0.464902
|
|
},
|
|
{
|
|
"epoch": 1.930909090909091,
|
|
"grad_norm": 4.383025646209717,
|
|
"learning_rate": 2.9769625036326198e-06,
|
|
"loss": 1.034652328491211,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2655,
|
|
"token_acc": 0.7468574743763295,
|
|
"train_speed(iter/s)": 0.4649
|
|
},
|
|
{
|
|
"epoch": 1.9345454545454546,
|
|
"grad_norm": 3.8552260398864746,
|
|
"learning_rate": 2.9590266508227878e-06,
|
|
"loss": 1.0464018821716308,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2660,
|
|
"token_acc": 0.7078343313373253,
|
|
"train_speed(iter/s)": 0.46497
|
|
},
|
|
{
|
|
"epoch": 1.9381818181818182,
|
|
"grad_norm": 3.4079573154449463,
|
|
"learning_rate": 2.9411222566155583e-06,
|
|
"loss": 0.9897180557250976,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2665,
|
|
"token_acc": 0.7533512064343163,
|
|
"train_speed(iter/s)": 0.464956
|
|
},
|
|
{
|
|
"epoch": 1.9418181818181819,
|
|
"grad_norm": 3.910155773162842,
|
|
"learning_rate": 2.9232495969808334e-06,
|
|
"loss": 0.8761829376220703,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2670,
|
|
"token_acc": 0.7727180732645214,
|
|
"train_speed(iter/s)": 0.464948
|
|
},
|
|
{
|
|
"epoch": 1.9454545454545453,
|
|
"grad_norm": 3.8864212036132812,
|
|
"learning_rate": 2.9054089473993787e-06,
|
|
"loss": 0.9600641250610351,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2675,
|
|
"token_acc": 0.7343532684283728,
|
|
"train_speed(iter/s)": 0.464956
|
|
},
|
|
{
|
|
"epoch": 1.9490909090909092,
|
|
"grad_norm": 4.038954257965088,
|
|
"learning_rate": 2.887600582858572e-06,
|
|
"loss": 0.9259279251098633,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2680,
|
|
"token_acc": 0.7577170775038863,
|
|
"train_speed(iter/s)": 0.464916
|
|
},
|
|
{
|
|
"epoch": 1.9527272727272726,
|
|
"grad_norm": 4.611103534698486,
|
|
"learning_rate": 2.869824777848159e-06,
|
|
"loss": 0.9913465499877929,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2685,
|
|
"token_acc": 0.7248833210513388,
|
|
"train_speed(iter/s)": 0.464943
|
|
},
|
|
{
|
|
"epoch": 1.9563636363636365,
|
|
"grad_norm": 3.434164047241211,
|
|
"learning_rate": 2.852081806356033e-06,
|
|
"loss": 0.9562204360961915,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2690,
|
|
"token_acc": 0.7549668874172185,
|
|
"train_speed(iter/s)": 0.464941
|
|
},
|
|
{
|
|
"epoch": 1.96,
|
|
"grad_norm": 4.066905498504639,
|
|
"learning_rate": 2.834371941864006e-06,
|
|
"loss": 0.9627280235290527,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2695,
|
|
"token_acc": 0.7466131460110387,
|
|
"train_speed(iter/s)": 0.465
|
|
},
|
|
{
|
|
"epoch": 1.9636363636363636,
|
|
"grad_norm": 4.238523006439209,
|
|
"learning_rate": 2.8166954573435964e-06,
|
|
"loss": 0.9915515899658203,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2700,
|
|
"token_acc": 0.727447216890595,
|
|
"train_speed(iter/s)": 0.465016
|
|
},
|
|
{
|
|
"epoch": 1.9672727272727273,
|
|
"grad_norm": 3.670944929122925,
|
|
"learning_rate": 2.7990526252518123e-06,
|
|
"loss": 0.9599745750427247,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2705,
|
|
"token_acc": 0.7029633113828786,
|
|
"train_speed(iter/s)": 0.465006
|
|
},
|
|
{
|
|
"epoch": 1.970909090909091,
|
|
"grad_norm": 3.1716055870056152,
|
|
"learning_rate": 2.781443717526966e-06,
|
|
"loss": 0.9637611389160157,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2710,
|
|
"token_acc": 0.8069548261293468,
|
|
"train_speed(iter/s)": 0.464986
|
|
},
|
|
{
|
|
"epoch": 1.9745454545454546,
|
|
"grad_norm": 3.7300829887390137,
|
|
"learning_rate": 2.7638690055844765e-06,
|
|
"loss": 0.925160026550293,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2715,
|
|
"token_acc": 0.7350575961747446,
|
|
"train_speed(iter/s)": 0.46499
|
|
},
|
|
{
|
|
"epoch": 1.978181818181818,
|
|
"grad_norm": 3.949653148651123,
|
|
"learning_rate": 2.7463287603126763e-06,
|
|
"loss": 0.9642942428588868,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2720,
|
|
"token_acc": 0.789059706828745,
|
|
"train_speed(iter/s)": 0.464998
|
|
},
|
|
{
|
|
"epoch": 1.981818181818182,
|
|
"grad_norm": 3.9616222381591797,
|
|
"learning_rate": 2.728823252068653e-06,
|
|
"loss": 0.8890161514282227,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2725,
|
|
"token_acc": 0.7694375119984642,
|
|
"train_speed(iter/s)": 0.465007
|
|
},
|
|
{
|
|
"epoch": 1.9854545454545454,
|
|
"grad_norm": 3.559417724609375,
|
|
"learning_rate": 2.711352750674072e-06,
|
|
"loss": 0.9697627067565918,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2730,
|
|
"token_acc": 0.7073286052009456,
|
|
"train_speed(iter/s)": 0.465034
|
|
},
|
|
{
|
|
"epoch": 1.9890909090909092,
|
|
"grad_norm": 3.9709830284118652,
|
|
"learning_rate": 2.6939175254110194e-06,
|
|
"loss": 0.990437126159668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2735,
|
|
"token_acc": 0.6944444444444444,
|
|
"train_speed(iter/s)": 0.46505
|
|
},
|
|
{
|
|
"epoch": 1.9927272727272727,
|
|
"grad_norm": 4.004317760467529,
|
|
"learning_rate": 2.6765178450178503e-06,
|
|
"loss": 0.9779638290405274,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2740,
|
|
"token_acc": 0.7541723049165538,
|
|
"train_speed(iter/s)": 0.46506
|
|
},
|
|
{
|
|
"epoch": 1.9963636363636363,
|
|
"grad_norm": 4.03814172744751,
|
|
"learning_rate": 2.6591539776850494e-06,
|
|
"loss": 0.9912105560302734,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2745,
|
|
"token_acc": 0.7429111531190926,
|
|
"train_speed(iter/s)": 0.465063
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"grad_norm": 3.923563241958618,
|
|
"learning_rate": 2.6418261910510978e-06,
|
|
"loss": 1.0082598686218263,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2750,
|
|
"token_acc": 0.7344060389352404,
|
|
"train_speed(iter/s)": 0.465082
|
|
},
|
|
{
|
|
"epoch": 2.0036363636363634,
|
|
"grad_norm": 3.0618035793304443,
|
|
"learning_rate": 2.6245347521983394e-06,
|
|
"loss": 0.6273193359375,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2755,
|
|
"token_acc": 0.8602760446209113,
|
|
"train_speed(iter/s)": 0.45798
|
|
},
|
|
{
|
|
"epoch": 2.0072727272727273,
|
|
"grad_norm": 3.0331130027770996,
|
|
"learning_rate": 2.607279927648875e-06,
|
|
"loss": 0.5837445259094238,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2760,
|
|
"token_acc": 0.8547550432276657,
|
|
"train_speed(iter/s)": 0.458002
|
|
},
|
|
{
|
|
"epoch": 2.0109090909090908,
|
|
"grad_norm": 4.053404808044434,
|
|
"learning_rate": 2.590061983360452e-06,
|
|
"loss": 0.6628801822662354,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2765,
|
|
"token_acc": 0.860643821391485,
|
|
"train_speed(iter/s)": 0.458016
|
|
},
|
|
{
|
|
"epoch": 2.0145454545454546,
|
|
"grad_norm": 3.389713764190674,
|
|
"learning_rate": 2.5728811847223555e-06,
|
|
"loss": 0.5978437423706054,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2770,
|
|
"token_acc": 0.8469597754911132,
|
|
"train_speed(iter/s)": 0.458034
|
|
},
|
|
{
|
|
"epoch": 2.018181818181818,
|
|
"grad_norm": 4.968929290771484,
|
|
"learning_rate": 2.555737796551332e-06,
|
|
"loss": 0.5556485176086425,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2775,
|
|
"token_acc": 0.8376184032476319,
|
|
"train_speed(iter/s)": 0.458085
|
|
},
|
|
{
|
|
"epoch": 2.021818181818182,
|
|
"grad_norm": 4.230146408081055,
|
|
"learning_rate": 2.5386320830874995e-06,
|
|
"loss": 0.5672976493835449,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2780,
|
|
"token_acc": 0.8402439024390244,
|
|
"train_speed(iter/s)": 0.458105
|
|
},
|
|
{
|
|
"epoch": 2.0254545454545454,
|
|
"grad_norm": 3.9281728267669678,
|
|
"learning_rate": 2.5215643079902704e-06,
|
|
"loss": 0.6257860660552979,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2785,
|
|
"token_acc": 0.8389728821694264,
|
|
"train_speed(iter/s)": 0.458146
|
|
},
|
|
{
|
|
"epoch": 2.0290909090909093,
|
|
"grad_norm": 4.398049831390381,
|
|
"learning_rate": 2.5045347343342984e-06,
|
|
"loss": 0.530487060546875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2790,
|
|
"token_acc": 0.865845192641444,
|
|
"train_speed(iter/s)": 0.458119
|
|
},
|
|
{
|
|
"epoch": 2.0327272727272727,
|
|
"grad_norm": 4.044005870819092,
|
|
"learning_rate": 2.4875436246054187e-06,
|
|
"loss": 0.5579940795898437,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2795,
|
|
"token_acc": 0.8288086156607584,
|
|
"train_speed(iter/s)": 0.458164
|
|
},
|
|
{
|
|
"epoch": 2.036363636363636,
|
|
"grad_norm": 3.480076789855957,
|
|
"learning_rate": 2.470591240696599e-06,
|
|
"loss": 0.5503246307373046,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2800,
|
|
"token_acc": 0.8433308494783904,
|
|
"train_speed(iter/s)": 0.458125
|
|
},
|
|
{
|
|
"epoch": 2.04,
|
|
"grad_norm": 3.8780994415283203,
|
|
"learning_rate": 2.4536778439039045e-06,
|
|
"loss": 0.569395637512207,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2805,
|
|
"token_acc": 0.8455534531693472,
|
|
"train_speed(iter/s)": 0.458151
|
|
},
|
|
{
|
|
"epoch": 2.0436363636363635,
|
|
"grad_norm": 3.4846231937408447,
|
|
"learning_rate": 2.4368036949224733e-06,
|
|
"loss": 0.5532159328460693,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2810,
|
|
"token_acc": 0.8399535423925668,
|
|
"train_speed(iter/s)": 0.458204
|
|
},
|
|
{
|
|
"epoch": 2.0472727272727274,
|
|
"grad_norm": 4.17202615737915,
|
|
"learning_rate": 2.419969053842498e-06,
|
|
"loss": 0.5738080978393555,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2815,
|
|
"token_acc": 0.8741489581184237,
|
|
"train_speed(iter/s)": 0.458185
|
|
},
|
|
{
|
|
"epoch": 2.050909090909091,
|
|
"grad_norm": 3.1433913707733154,
|
|
"learning_rate": 2.4031741801452095e-06,
|
|
"loss": 0.5935293197631836,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2820,
|
|
"token_acc": 0.8103585657370518,
|
|
"train_speed(iter/s)": 0.458209
|
|
},
|
|
{
|
|
"epoch": 2.0545454545454547,
|
|
"grad_norm": 4.294399738311768,
|
|
"learning_rate": 2.386419332698888e-06,
|
|
"loss": 0.5968852043151855,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2825,
|
|
"token_acc": 0.8341388400702988,
|
|
"train_speed(iter/s)": 0.458263
|
|
},
|
|
{
|
|
"epoch": 2.058181818181818,
|
|
"grad_norm": 3.853757619857788,
|
|
"learning_rate": 2.369704769754868e-06,
|
|
"loss": 0.5755794048309326,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2830,
|
|
"token_acc": 0.8497652582159625,
|
|
"train_speed(iter/s)": 0.458303
|
|
},
|
|
{
|
|
"epoch": 2.061818181818182,
|
|
"grad_norm": 3.72324275970459,
|
|
"learning_rate": 2.3530307489435523e-06,
|
|
"loss": 0.5538017272949218,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2835,
|
|
"token_acc": 0.834593572778828,
|
|
"train_speed(iter/s)": 0.458371
|
|
},
|
|
{
|
|
"epoch": 2.0654545454545454,
|
|
"grad_norm": 3.630908250808716,
|
|
"learning_rate": 2.3363975272704515e-06,
|
|
"loss": 0.5382684707641602,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2840,
|
|
"token_acc": 0.8325254483761513,
|
|
"train_speed(iter/s)": 0.458392
|
|
},
|
|
{
|
|
"epoch": 2.0690909090909093,
|
|
"grad_norm": 4.250785827636719,
|
|
"learning_rate": 2.319805361112217e-06,
|
|
"loss": 0.5941808700561524,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2845,
|
|
"token_acc": 0.8257705479452054,
|
|
"train_speed(iter/s)": 0.458415
|
|
},
|
|
{
|
|
"epoch": 2.0727272727272728,
|
|
"grad_norm": 3.9917092323303223,
|
|
"learning_rate": 2.303254506212687e-06,
|
|
"loss": 0.5847339153289794,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2850,
|
|
"token_acc": 0.8336499321573948,
|
|
"train_speed(iter/s)": 0.458488
|
|
},
|
|
{
|
|
"epoch": 2.076363636363636,
|
|
"grad_norm": 4.361266613006592,
|
|
"learning_rate": 2.2867452176789495e-06,
|
|
"loss": 0.5764480590820312,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2855,
|
|
"token_acc": 0.8504418762746431,
|
|
"train_speed(iter/s)": 0.458449
|
|
},
|
|
{
|
|
"epoch": 2.08,
|
|
"grad_norm": 3.822355270385742,
|
|
"learning_rate": 2.270277749977408e-06,
|
|
"loss": 0.5814784049987793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2860,
|
|
"token_acc": 0.8517949702438088,
|
|
"train_speed(iter/s)": 0.458441
|
|
},
|
|
{
|
|
"epoch": 2.0836363636363635,
|
|
"grad_norm": 4.25092077255249,
|
|
"learning_rate": 2.253852356929861e-06,
|
|
"loss": 0.5927882671356202,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2865,
|
|
"token_acc": 0.8643165946125707,
|
|
"train_speed(iter/s)": 0.458426
|
|
},
|
|
{
|
|
"epoch": 2.0872727272727274,
|
|
"grad_norm": 3.814818859100342,
|
|
"learning_rate": 2.237469291709582e-06,
|
|
"loss": 0.5530673027038574,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2870,
|
|
"token_acc": 0.8843641114982579,
|
|
"train_speed(iter/s)": 0.458492
|
|
},
|
|
{
|
|
"epoch": 2.090909090909091,
|
|
"grad_norm": 3.3128604888916016,
|
|
"learning_rate": 2.2211288068374304e-06,
|
|
"loss": 0.5221426010131835,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2875,
|
|
"token_acc": 0.8021608643457383,
|
|
"train_speed(iter/s)": 0.458514
|
|
},
|
|
{
|
|
"epoch": 2.0945454545454547,
|
|
"grad_norm": 3.841860055923462,
|
|
"learning_rate": 2.2048311541779483e-06,
|
|
"loss": 0.5207252025604248,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2880,
|
|
"token_acc": 0.8762656946132037,
|
|
"train_speed(iter/s)": 0.458498
|
|
},
|
|
{
|
|
"epoch": 2.098181818181818,
|
|
"grad_norm": 4.4180145263671875,
|
|
"learning_rate": 2.188576584935483e-06,
|
|
"loss": 0.5573165893554688,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2885,
|
|
"token_acc": 0.8565947242206235,
|
|
"train_speed(iter/s)": 0.458576
|
|
},
|
|
{
|
|
"epoch": 2.101818181818182,
|
|
"grad_norm": 3.850353956222534,
|
|
"learning_rate": 2.1723653496503156e-06,
|
|
"loss": 0.5557028770446777,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2890,
|
|
"token_acc": 0.8416609628108601,
|
|
"train_speed(iter/s)": 0.4586
|
|
},
|
|
{
|
|
"epoch": 2.1054545454545455,
|
|
"grad_norm": 4.130335807800293,
|
|
"learning_rate": 2.1561976981947974e-06,
|
|
"loss": 0.5635205268859863,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2895,
|
|
"token_acc": 0.8624701195219123,
|
|
"train_speed(iter/s)": 0.458599
|
|
},
|
|
{
|
|
"epoch": 2.109090909090909,
|
|
"grad_norm": 3.935039520263672,
|
|
"learning_rate": 2.1400738797694954e-06,
|
|
"loss": 0.5272860527038574,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2900,
|
|
"token_acc": 0.8585553419390277,
|
|
"train_speed(iter/s)": 0.458627
|
|
},
|
|
{
|
|
"epoch": 2.112727272727273,
|
|
"grad_norm": 3.9505932331085205,
|
|
"learning_rate": 2.1239941428993593e-06,
|
|
"loss": 0.5397134780883789,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2905,
|
|
"token_acc": 0.8265644955300128,
|
|
"train_speed(iter/s)": 0.458668
|
|
},
|
|
{
|
|
"epoch": 2.1163636363636362,
|
|
"grad_norm": 3.789020299911499,
|
|
"learning_rate": 2.1079587354298865e-06,
|
|
"loss": 0.5284680366516114,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2910,
|
|
"token_acc": 0.8370165745856354,
|
|
"train_speed(iter/s)": 0.458724
|
|
},
|
|
{
|
|
"epoch": 2.12,
|
|
"grad_norm": 4.089132785797119,
|
|
"learning_rate": 2.0919679045233004e-06,
|
|
"loss": 0.5436430454254151,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2915,
|
|
"token_acc": 0.8373091603053435,
|
|
"train_speed(iter/s)": 0.458761
|
|
},
|
|
{
|
|
"epoch": 2.1236363636363635,
|
|
"grad_norm": 4.195389270782471,
|
|
"learning_rate": 2.0760218966547403e-06,
|
|
"loss": 0.49860348701477053,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2920,
|
|
"token_acc": 0.8677836092046831,
|
|
"train_speed(iter/s)": 0.458759
|
|
},
|
|
{
|
|
"epoch": 2.1272727272727274,
|
|
"grad_norm": 4.3651862144470215,
|
|
"learning_rate": 2.0601209576084675e-06,
|
|
"loss": 0.5386879444122314,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2925,
|
|
"token_acc": 0.8342798141455859,
|
|
"train_speed(iter/s)": 0.458777
|
|
},
|
|
{
|
|
"epoch": 2.130909090909091,
|
|
"grad_norm": 4.288598537445068,
|
|
"learning_rate": 2.0442653324740747e-06,
|
|
"loss": 0.5425639629364014,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2930,
|
|
"token_acc": 0.8665203511572227,
|
|
"train_speed(iter/s)": 0.4588
|
|
},
|
|
{
|
|
"epoch": 2.1345454545454547,
|
|
"grad_norm": 3.187136650085449,
|
|
"learning_rate": 2.028455265642701e-06,
|
|
"loss": 0.5388356208801269,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2935,
|
|
"token_acc": 0.8361738389433319,
|
|
"train_speed(iter/s)": 0.458812
|
|
},
|
|
{
|
|
"epoch": 2.138181818181818,
|
|
"grad_norm": 3.3392493724823,
|
|
"learning_rate": 2.0126910008032775e-06,
|
|
"loss": 0.49450435638427737,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2940,
|
|
"token_acc": 0.8647859922178989,
|
|
"train_speed(iter/s)": 0.458795
|
|
},
|
|
{
|
|
"epoch": 2.1418181818181816,
|
|
"grad_norm": 3.969228506088257,
|
|
"learning_rate": 1.9969727809387645e-06,
|
|
"loss": 0.5926756858825684,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2945,
|
|
"token_acc": 0.8393939393939394,
|
|
"train_speed(iter/s)": 0.45886
|
|
},
|
|
{
|
|
"epoch": 2.1454545454545455,
|
|
"grad_norm": 3.8145792484283447,
|
|
"learning_rate": 1.981300848322401e-06,
|
|
"loss": 0.5568334102630615,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2950,
|
|
"token_acc": 0.8597499155119973,
|
|
"train_speed(iter/s)": 0.458896
|
|
},
|
|
{
|
|
"epoch": 2.149090909090909,
|
|
"grad_norm": 3.3614907264709473,
|
|
"learning_rate": 1.9656754445139838e-06,
|
|
"loss": 0.5743967056274414,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2955,
|
|
"token_acc": 0.8292061179898034,
|
|
"train_speed(iter/s)": 0.458934
|
|
},
|
|
{
|
|
"epoch": 2.152727272727273,
|
|
"grad_norm": 3.790090322494507,
|
|
"learning_rate": 1.9500968103561306e-06,
|
|
"loss": 0.5191942691802979,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2960,
|
|
"token_acc": 0.8744884038199181,
|
|
"train_speed(iter/s)": 0.458944
|
|
},
|
|
{
|
|
"epoch": 2.1563636363636363,
|
|
"grad_norm": 3.9427294731140137,
|
|
"learning_rate": 1.9345651859705794e-06,
|
|
"loss": 0.5102787971496582,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2965,
|
|
"token_acc": 0.8653482011361245,
|
|
"train_speed(iter/s)": 0.458942
|
|
},
|
|
{
|
|
"epoch": 2.16,
|
|
"grad_norm": 3.465834140777588,
|
|
"learning_rate": 1.9190808107544743e-06,
|
|
"loss": 0.5655551433563233,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2970,
|
|
"token_acc": 0.8827075439200827,
|
|
"train_speed(iter/s)": 0.458901
|
|
},
|
|
{
|
|
"epoch": 2.1636363636363636,
|
|
"grad_norm": 4.011518478393555,
|
|
"learning_rate": 1.903643923376689e-06,
|
|
"loss": 0.5846277236938476,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2975,
|
|
"token_acc": 0.8455261676983681,
|
|
"train_speed(iter/s)": 0.458965
|
|
},
|
|
{
|
|
"epoch": 2.1672727272727275,
|
|
"grad_norm": 4.44923734664917,
|
|
"learning_rate": 1.8882547617741414e-06,
|
|
"loss": 0.5695096969604492,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2980,
|
|
"token_acc": 0.8611372180451128,
|
|
"train_speed(iter/s)": 0.459009
|
|
},
|
|
{
|
|
"epoch": 2.170909090909091,
|
|
"grad_norm": 4.384074687957764,
|
|
"learning_rate": 1.8729135631481222e-06,
|
|
"loss": 0.5614648818969726,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2985,
|
|
"token_acc": 0.8193957968476357,
|
|
"train_speed(iter/s)": 0.459036
|
|
},
|
|
{
|
|
"epoch": 2.174545454545455,
|
|
"grad_norm": 3.7020986080169678,
|
|
"learning_rate": 1.8576205639606482e-06,
|
|
"loss": 0.5832395553588867,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2990,
|
|
"token_acc": 0.8895298024820836,
|
|
"train_speed(iter/s)": 0.459045
|
|
},
|
|
{
|
|
"epoch": 2.178181818181818,
|
|
"grad_norm": 3.8569514751434326,
|
|
"learning_rate": 1.8423759999308132e-06,
|
|
"loss": 0.5824544906616211,
|
|
"memory(GiB)": 64.44,
|
|
"step": 2995,
|
|
"token_acc": 0.8710670314637483,
|
|
"train_speed(iter/s)": 0.459023
|
|
},
|
|
{
|
|
"epoch": 2.1818181818181817,
|
|
"grad_norm": 3.4723124504089355,
|
|
"learning_rate": 1.8271801060311485e-06,
|
|
"loss": 0.5585835456848145,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3000,
|
|
"token_acc": 0.8206177132337902,
|
|
"train_speed(iter/s)": 0.459065
|
|
},
|
|
{
|
|
"epoch": 2.1854545454545455,
|
|
"grad_norm": 3.9203343391418457,
|
|
"learning_rate": 1.8120331164840116e-06,
|
|
"loss": 0.5572877883911133,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3005,
|
|
"token_acc": 0.8126218323586745,
|
|
"train_speed(iter/s)": 0.459105
|
|
},
|
|
{
|
|
"epoch": 2.189090909090909,
|
|
"grad_norm": 3.1291725635528564,
|
|
"learning_rate": 1.7969352647579707e-06,
|
|
"loss": 0.4823342800140381,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3010,
|
|
"token_acc": 0.8701427438540841,
|
|
"train_speed(iter/s)": 0.459041
|
|
},
|
|
{
|
|
"epoch": 2.192727272727273,
|
|
"grad_norm": 4.418652057647705,
|
|
"learning_rate": 1.781886783564209e-06,
|
|
"loss": 0.5473741054534912,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3015,
|
|
"token_acc": 0.8813121272365805,
|
|
"train_speed(iter/s)": 0.459057
|
|
},
|
|
{
|
|
"epoch": 2.1963636363636363,
|
|
"grad_norm": 4.301779747009277,
|
|
"learning_rate": 1.766887904852928e-06,
|
|
"loss": 0.5795573234558106,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3020,
|
|
"token_acc": 0.8556416881998278,
|
|
"train_speed(iter/s)": 0.459081
|
|
},
|
|
{
|
|
"epoch": 2.2,
|
|
"grad_norm": 3.995138168334961,
|
|
"learning_rate": 1.7519388598097876e-06,
|
|
"loss": 0.5912516593933106,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3025,
|
|
"token_acc": 0.8219990871748061,
|
|
"train_speed(iter/s)": 0.459103
|
|
},
|
|
{
|
|
"epoch": 2.2036363636363636,
|
|
"grad_norm": 4.0938262939453125,
|
|
"learning_rate": 1.7370398788523319e-06,
|
|
"loss": 0.52225341796875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3030,
|
|
"token_acc": 0.8836140888208269,
|
|
"train_speed(iter/s)": 0.45914
|
|
},
|
|
{
|
|
"epoch": 2.207272727272727,
|
|
"grad_norm": 4.07735538482666,
|
|
"learning_rate": 1.7221911916264405e-06,
|
|
"loss": 0.5614495754241944,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3035,
|
|
"token_acc": 0.8265093684941013,
|
|
"train_speed(iter/s)": 0.459167
|
|
},
|
|
{
|
|
"epoch": 2.210909090909091,
|
|
"grad_norm": 4.394158840179443,
|
|
"learning_rate": 1.7073930270027905e-06,
|
|
"loss": 0.5650532722473145,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3040,
|
|
"token_acc": 0.8541278708876474,
|
|
"train_speed(iter/s)": 0.459208
|
|
},
|
|
{
|
|
"epoch": 2.2145454545454544,
|
|
"grad_norm": 3.5056724548339844,
|
|
"learning_rate": 1.6926456130733277e-06,
|
|
"loss": 0.5250924110412598,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3045,
|
|
"token_acc": 0.8541409147095179,
|
|
"train_speed(iter/s)": 0.459263
|
|
},
|
|
{
|
|
"epoch": 2.2181818181818183,
|
|
"grad_norm": 3.697410821914673,
|
|
"learning_rate": 1.6779491771477457e-06,
|
|
"loss": 0.5638864517211915,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3050,
|
|
"token_acc": 0.8667886550777676,
|
|
"train_speed(iter/s)": 0.459288
|
|
},
|
|
{
|
|
"epoch": 2.2218181818181817,
|
|
"grad_norm": 4.366413116455078,
|
|
"learning_rate": 1.6633039457499933e-06,
|
|
"loss": 0.527287483215332,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3055,
|
|
"token_acc": 0.8699014111734004,
|
|
"train_speed(iter/s)": 0.459302
|
|
},
|
|
{
|
|
"epoch": 2.2254545454545456,
|
|
"grad_norm": 4.17656946182251,
|
|
"learning_rate": 1.6487101446147758e-06,
|
|
"loss": 0.5217004776000976,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3060,
|
|
"token_acc": 0.8625788643533123,
|
|
"train_speed(iter/s)": 0.459271
|
|
},
|
|
{
|
|
"epoch": 2.229090909090909,
|
|
"grad_norm": 4.382239818572998,
|
|
"learning_rate": 1.634167998684072e-06,
|
|
"loss": 0.5332613945007324,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3065,
|
|
"token_acc": 0.8863858961802155,
|
|
"train_speed(iter/s)": 0.459241
|
|
},
|
|
{
|
|
"epoch": 2.232727272727273,
|
|
"grad_norm": 3.959284782409668,
|
|
"learning_rate": 1.6196777321036756e-06,
|
|
"loss": 0.6081136226654053,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3070,
|
|
"token_acc": 0.8399458972046889,
|
|
"train_speed(iter/s)": 0.459281
|
|
},
|
|
{
|
|
"epoch": 2.2363636363636363,
|
|
"grad_norm": 3.482314348220825,
|
|
"learning_rate": 1.6052395682197363e-06,
|
|
"loss": 0.5248013019561768,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3075,
|
|
"token_acc": 0.8643042350907519,
|
|
"train_speed(iter/s)": 0.459336
|
|
},
|
|
{
|
|
"epoch": 2.24,
|
|
"grad_norm": 4.430671691894531,
|
|
"learning_rate": 1.590853729575319e-06,
|
|
"loss": 0.6185288429260254,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3080,
|
|
"token_acc": 0.8582693085321472,
|
|
"train_speed(iter/s)": 0.459367
|
|
},
|
|
{
|
|
"epoch": 2.2436363636363637,
|
|
"grad_norm": 4.537671089172363,
|
|
"learning_rate": 1.5765204379069654e-06,
|
|
"loss": 0.6154137134552002,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3085,
|
|
"token_acc": 0.8374798639699302,
|
|
"train_speed(iter/s)": 0.459393
|
|
},
|
|
{
|
|
"epoch": 2.247272727272727,
|
|
"grad_norm": 3.5833494663238525,
|
|
"learning_rate": 1.5622399141412898e-06,
|
|
"loss": 0.5241693019866943,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3090,
|
|
"token_acc": 0.8457924836601307,
|
|
"train_speed(iter/s)": 0.459391
|
|
},
|
|
{
|
|
"epoch": 2.250909090909091,
|
|
"grad_norm": 4.317352294921875,
|
|
"learning_rate": 1.5480123783915645e-06,
|
|
"loss": 0.5752131462097168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3095,
|
|
"token_acc": 0.8522615535889873,
|
|
"train_speed(iter/s)": 0.459459
|
|
},
|
|
{
|
|
"epoch": 2.2545454545454544,
|
|
"grad_norm": 3.6059021949768066,
|
|
"learning_rate": 1.5338380499543276e-06,
|
|
"loss": 0.5613752365112304,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3100,
|
|
"token_acc": 0.8562077241981235,
|
|
"train_speed(iter/s)": 0.459508
|
|
},
|
|
{
|
|
"epoch": 2.2581818181818183,
|
|
"grad_norm": 5.299020290374756,
|
|
"learning_rate": 1.5197171473060075e-06,
|
|
"loss": 0.5874448299407959,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3105,
|
|
"token_acc": 0.8576022048690859,
|
|
"train_speed(iter/s)": 0.459539
|
|
},
|
|
{
|
|
"epoch": 2.2618181818181817,
|
|
"grad_norm": 4.414374351501465,
|
|
"learning_rate": 1.5056498880995524e-06,
|
|
"loss": 0.571533203125,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3110,
|
|
"token_acc": 0.839841010053776,
|
|
"train_speed(iter/s)": 0.459539
|
|
},
|
|
{
|
|
"epoch": 2.2654545454545456,
|
|
"grad_norm": 3.932117223739624,
|
|
"learning_rate": 1.491636489161072e-06,
|
|
"loss": 0.516119384765625,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3115,
|
|
"token_acc": 0.8690719732192672,
|
|
"train_speed(iter/s)": 0.45957
|
|
},
|
|
{
|
|
"epoch": 2.269090909090909,
|
|
"grad_norm": 3.662917137145996,
|
|
"learning_rate": 1.4776771664865036e-06,
|
|
"loss": 0.5404162883758545,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3120,
|
|
"token_acc": 0.8421052631578947,
|
|
"train_speed(iter/s)": 0.459592
|
|
},
|
|
{
|
|
"epoch": 2.2727272727272725,
|
|
"grad_norm": 4.036909103393555,
|
|
"learning_rate": 1.4637721352382772e-06,
|
|
"loss": 0.5555058479309082,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3125,
|
|
"token_acc": 0.8584229390681004,
|
|
"train_speed(iter/s)": 0.459626
|
|
},
|
|
{
|
|
"epoch": 2.2763636363636364,
|
|
"grad_norm": 3.80786395072937,
|
|
"learning_rate": 1.449921609742001e-06,
|
|
"loss": 0.5728797912597656,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3130,
|
|
"token_acc": 0.8645363654185454,
|
|
"train_speed(iter/s)": 0.459659
|
|
},
|
|
{
|
|
"epoch": 2.2800000000000002,
|
|
"grad_norm": 3.886758327484131,
|
|
"learning_rate": 1.4361258034831533e-06,
|
|
"loss": 0.5380870819091796,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3135,
|
|
"token_acc": 0.8600067957866123,
|
|
"train_speed(iter/s)": 0.459649
|
|
},
|
|
{
|
|
"epoch": 2.2836363636363637,
|
|
"grad_norm": 3.8579964637756348,
|
|
"learning_rate": 1.4223849291038e-06,
|
|
"loss": 0.5628611087799072,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3140,
|
|
"token_acc": 0.8469529085872576,
|
|
"train_speed(iter/s)": 0.459697
|
|
},
|
|
{
|
|
"epoch": 2.287272727272727,
|
|
"grad_norm": 4.657957553863525,
|
|
"learning_rate": 1.4086991983993132e-06,
|
|
"loss": 0.5323144912719726,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3145,
|
|
"token_acc": 0.865530303030303,
|
|
"train_speed(iter/s)": 0.459743
|
|
},
|
|
{
|
|
"epoch": 2.290909090909091,
|
|
"grad_norm": 4.06402063369751,
|
|
"learning_rate": 1.3950688223151009e-06,
|
|
"loss": 0.5702249526977539,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3150,
|
|
"token_acc": 0.8512775715221664,
|
|
"train_speed(iter/s)": 0.459758
|
|
},
|
|
{
|
|
"epoch": 2.2945454545454544,
|
|
"grad_norm": 3.851614475250244,
|
|
"learning_rate": 1.3814940109433678e-06,
|
|
"loss": 0.5616402626037598,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3155,
|
|
"token_acc": 0.8460471567267683,
|
|
"train_speed(iter/s)": 0.459824
|
|
},
|
|
{
|
|
"epoch": 2.2981818181818183,
|
|
"grad_norm": 4.075229644775391,
|
|
"learning_rate": 1.367974973519866e-06,
|
|
"loss": 0.6098086357116699,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3160,
|
|
"token_acc": 0.8224387924887093,
|
|
"train_speed(iter/s)": 0.459885
|
|
},
|
|
{
|
|
"epoch": 2.3018181818181818,
|
|
"grad_norm": 4.408036708831787,
|
|
"learning_rate": 1.3545119184206778e-06,
|
|
"loss": 0.584037971496582,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3165,
|
|
"token_acc": 0.8570526315789474,
|
|
"train_speed(iter/s)": 0.459931
|
|
},
|
|
{
|
|
"epoch": 2.3054545454545456,
|
|
"grad_norm": 3.7412238121032715,
|
|
"learning_rate": 1.3411050531589942e-06,
|
|
"loss": 0.5043732643127441,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3170,
|
|
"token_acc": 0.8607445008460237,
|
|
"train_speed(iter/s)": 0.459923
|
|
},
|
|
{
|
|
"epoch": 2.309090909090909,
|
|
"grad_norm": 4.605288028717041,
|
|
"learning_rate": 1.3277545843819278e-06,
|
|
"loss": 0.5658152103424072,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3175,
|
|
"token_acc": 0.8707877705869226,
|
|
"train_speed(iter/s)": 0.459985
|
|
},
|
|
{
|
|
"epoch": 2.3127272727272725,
|
|
"grad_norm": 3.898077964782715,
|
|
"learning_rate": 1.314460717867322e-06,
|
|
"loss": 0.5670787334442139,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3180,
|
|
"token_acc": 0.8559006211180125,
|
|
"train_speed(iter/s)": 0.459978
|
|
},
|
|
{
|
|
"epoch": 2.3163636363636364,
|
|
"grad_norm": 3.6667439937591553,
|
|
"learning_rate": 1.3012236585205773e-06,
|
|
"loss": 0.6092076301574707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3185,
|
|
"token_acc": 0.81068524970964,
|
|
"train_speed(iter/s)": 0.460006
|
|
},
|
|
{
|
|
"epoch": 2.32,
|
|
"grad_norm": 3.862675905227661,
|
|
"learning_rate": 1.2880436103714962e-06,
|
|
"loss": 0.503480339050293,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3190,
|
|
"token_acc": 0.8837376891664664,
|
|
"train_speed(iter/s)": 0.460016
|
|
},
|
|
{
|
|
"epoch": 2.3236363636363637,
|
|
"grad_norm": 3.480212926864624,
|
|
"learning_rate": 1.2749207765711398e-06,
|
|
"loss": 0.4982267379760742,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3195,
|
|
"token_acc": 0.8575321136629038,
|
|
"train_speed(iter/s)": 0.46002
|
|
},
|
|
{
|
|
"epoch": 2.327272727272727,
|
|
"grad_norm": 4.333157539367676,
|
|
"learning_rate": 1.2618553593886885e-06,
|
|
"loss": 0.52719144821167,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3200,
|
|
"token_acc": 0.8505338078291815,
|
|
"train_speed(iter/s)": 0.46006
|
|
},
|
|
{
|
|
"epoch": 2.330909090909091,
|
|
"grad_norm": 3.873192071914673,
|
|
"learning_rate": 1.2488475602083345e-06,
|
|
"loss": 0.5189596176147461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3205,
|
|
"token_acc": 0.8571148569188554,
|
|
"train_speed(iter/s)": 0.460068
|
|
},
|
|
{
|
|
"epoch": 2.3345454545454545,
|
|
"grad_norm": 3.650998115539551,
|
|
"learning_rate": 1.2358975795261746e-06,
|
|
"loss": 0.5396578788757325,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3210,
|
|
"token_acc": 0.8491717523975588,
|
|
"train_speed(iter/s)": 0.460102
|
|
},
|
|
{
|
|
"epoch": 2.3381818181818184,
|
|
"grad_norm": 4.723164081573486,
|
|
"learning_rate": 1.2230056169471122e-06,
|
|
"loss": 0.5817994117736817,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3215,
|
|
"token_acc": 0.844311377245509,
|
|
"train_speed(iter/s)": 0.460144
|
|
},
|
|
{
|
|
"epoch": 2.341818181818182,
|
|
"grad_norm": 4.452855110168457,
|
|
"learning_rate": 1.2101718711817928e-06,
|
|
"loss": 0.5918424606323243,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3220,
|
|
"token_acc": 0.8243565932608119,
|
|
"train_speed(iter/s)": 0.460201
|
|
},
|
|
{
|
|
"epoch": 2.3454545454545457,
|
|
"grad_norm": 4.132510185241699,
|
|
"learning_rate": 1.1973965400435344e-06,
|
|
"loss": 0.5174155712127686,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3225,
|
|
"token_acc": 0.8721298495645289,
|
|
"train_speed(iter/s)": 0.460229
|
|
},
|
|
{
|
|
"epoch": 2.349090909090909,
|
|
"grad_norm": 4.476179599761963,
|
|
"learning_rate": 1.1846798204452754e-06,
|
|
"loss": 0.5070356845855712,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3230,
|
|
"token_acc": 0.8643922509857707,
|
|
"train_speed(iter/s)": 0.460216
|
|
},
|
|
{
|
|
"epoch": 2.3527272727272726,
|
|
"grad_norm": 4.049841403961182,
|
|
"learning_rate": 1.1720219083965478e-06,
|
|
"loss": 0.5351714134216309,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3235,
|
|
"token_acc": 0.8564006650041562,
|
|
"train_speed(iter/s)": 0.460236
|
|
},
|
|
{
|
|
"epoch": 2.3563636363636364,
|
|
"grad_norm": 4.259440898895264,
|
|
"learning_rate": 1.1594229990004508e-06,
|
|
"loss": 0.5463751792907715,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3240,
|
|
"token_acc": 0.8500355366027008,
|
|
"train_speed(iter/s)": 0.460272
|
|
},
|
|
{
|
|
"epoch": 2.36,
|
|
"grad_norm": 3.7025094032287598,
|
|
"learning_rate": 1.1468832864506458e-06,
|
|
"loss": 0.5029178619384765,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3245,
|
|
"token_acc": 0.875223271242664,
|
|
"train_speed(iter/s)": 0.460289
|
|
},
|
|
{
|
|
"epoch": 2.3636363636363638,
|
|
"grad_norm": 4.097626686096191,
|
|
"learning_rate": 1.134402964028356e-06,
|
|
"loss": 0.5285816192626953,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3250,
|
|
"token_acc": 0.8633448445979279,
|
|
"train_speed(iter/s)": 0.460296
|
|
},
|
|
{
|
|
"epoch": 2.367272727272727,
|
|
"grad_norm": 4.1264801025390625,
|
|
"learning_rate": 1.1219822240993994e-06,
|
|
"loss": 0.5510709762573243,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3255,
|
|
"token_acc": 0.8759833410458121,
|
|
"train_speed(iter/s)": 0.4603
|
|
},
|
|
{
|
|
"epoch": 2.370909090909091,
|
|
"grad_norm": 4.0943379402160645,
|
|
"learning_rate": 1.109621258111217e-06,
|
|
"loss": 0.5874659538269043,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3260,
|
|
"token_acc": 0.8329943043124491,
|
|
"train_speed(iter/s)": 0.460293
|
|
},
|
|
{
|
|
"epoch": 2.3745454545454545,
|
|
"grad_norm": 3.5104446411132812,
|
|
"learning_rate": 1.0973202565899172e-06,
|
|
"loss": 0.5906445503234863,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3265,
|
|
"token_acc": 0.8552338530066815,
|
|
"train_speed(iter/s)": 0.460302
|
|
},
|
|
{
|
|
"epoch": 2.378181818181818,
|
|
"grad_norm": 4.113192081451416,
|
|
"learning_rate": 1.0850794091373502e-06,
|
|
"loss": 0.5461389541625976,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3270,
|
|
"token_acc": 0.8535174472849412,
|
|
"train_speed(iter/s)": 0.460292
|
|
},
|
|
{
|
|
"epoch": 2.381818181818182,
|
|
"grad_norm": 3.782274007797241,
|
|
"learning_rate": 1.0728989044281779e-06,
|
|
"loss": 0.5787214279174805,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3275,
|
|
"token_acc": 0.8531020177258156,
|
|
"train_speed(iter/s)": 0.460331
|
|
},
|
|
{
|
|
"epoch": 2.3854545454545453,
|
|
"grad_norm": 3.9185631275177,
|
|
"learning_rate": 1.060778930206965e-06,
|
|
"loss": 0.5424469470977783,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3280,
|
|
"token_acc": 0.863598844407759,
|
|
"train_speed(iter/s)": 0.460373
|
|
},
|
|
{
|
|
"epoch": 2.389090909090909,
|
|
"grad_norm": 3.632422924041748,
|
|
"learning_rate": 1.04871967328529e-06,
|
|
"loss": 0.5419508457183838,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3285,
|
|
"token_acc": 0.8583182821593418,
|
|
"train_speed(iter/s)": 0.460396
|
|
},
|
|
{
|
|
"epoch": 2.3927272727272726,
|
|
"grad_norm": 3.406076192855835,
|
|
"learning_rate": 1.0367213195388649e-06,
|
|
"loss": 0.5859408378601074,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3290,
|
|
"token_acc": 0.8361921097770154,
|
|
"train_speed(iter/s)": 0.460399
|
|
},
|
|
{
|
|
"epoch": 2.3963636363636365,
|
|
"grad_norm": 4.544106960296631,
|
|
"learning_rate": 1.0247840539046672e-06,
|
|
"loss": 0.5752851486206054,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3295,
|
|
"token_acc": 0.8585009949148795,
|
|
"train_speed(iter/s)": 0.460426
|
|
},
|
|
{
|
|
"epoch": 2.4,
|
|
"grad_norm": 4.9750142097473145,
|
|
"learning_rate": 1.012908060378089e-06,
|
|
"loss": 0.532856035232544,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3300,
|
|
"token_acc": 0.8393331722638319,
|
|
"train_speed(iter/s)": 0.460466
|
|
},
|
|
{
|
|
"epoch": 2.403636363636364,
|
|
"grad_norm": 3.526374101638794,
|
|
"learning_rate": 1.001093522010106e-06,
|
|
"loss": 0.5794241905212403,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3305,
|
|
"token_acc": 0.8321299638989169,
|
|
"train_speed(iter/s)": 0.460506
|
|
},
|
|
{
|
|
"epoch": 2.4072727272727272,
|
|
"grad_norm": 4.099818706512451,
|
|
"learning_rate": 9.89340620904452e-07,
|
|
"loss": 0.5658499240875244,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3310,
|
|
"token_acc": 0.85,
|
|
"train_speed(iter/s)": 0.46052
|
|
},
|
|
{
|
|
"epoch": 2.410909090909091,
|
|
"grad_norm": 3.6997690200805664,
|
|
"learning_rate": 9.776495382148132e-07,
|
|
"loss": 0.5323297500610351,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3315,
|
|
"token_acc": 0.8670710133939853,
|
|
"train_speed(iter/s)": 0.460555
|
|
},
|
|
{
|
|
"epoch": 2.4145454545454546,
|
|
"grad_norm": 4.329352855682373,
|
|
"learning_rate": 9.66020454142037e-07,
|
|
"loss": 0.6215848922729492,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3320,
|
|
"token_acc": 0.8559788144224894,
|
|
"train_speed(iter/s)": 0.460585
|
|
},
|
|
{
|
|
"epoch": 2.418181818181818,
|
|
"grad_norm": 4.3700361251831055,
|
|
"learning_rate": 9.544535479313494e-07,
|
|
"loss": 0.5541114330291748,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3325,
|
|
"token_acc": 0.8365769128232471,
|
|
"train_speed(iter/s)": 0.460599
|
|
},
|
|
{
|
|
"epoch": 2.421818181818182,
|
|
"grad_norm": 3.6950056552886963,
|
|
"learning_rate": 9.429489978695994e-07,
|
|
"loss": 0.5436219215393067,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3330,
|
|
"token_acc": 0.8593533487297922,
|
|
"train_speed(iter/s)": 0.460639
|
|
},
|
|
{
|
|
"epoch": 2.4254545454545453,
|
|
"grad_norm": 3.716181755065918,
|
|
"learning_rate": 9.315069812825073e-07,
|
|
"loss": 0.6107286930084228,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3335,
|
|
"token_acc": 0.855609756097561,
|
|
"train_speed(iter/s)": 0.460656
|
|
},
|
|
{
|
|
"epoch": 2.429090909090909,
|
|
"grad_norm": 4.002308368682861,
|
|
"learning_rate": 9.20127674531932e-07,
|
|
"loss": 0.5892920970916748,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3340,
|
|
"token_acc": 0.8453924110299317,
|
|
"train_speed(iter/s)": 0.4607
|
|
},
|
|
{
|
|
"epoch": 2.4327272727272726,
|
|
"grad_norm": 3.5489273071289062,
|
|
"learning_rate": 9.088112530131527e-07,
|
|
"loss": 0.5960803508758545,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3345,
|
|
"token_acc": 0.8484493856056173,
|
|
"train_speed(iter/s)": 0.46075
|
|
},
|
|
{
|
|
"epoch": 2.4363636363636365,
|
|
"grad_norm": 3.402630090713501,
|
|
"learning_rate": 8.975578911521616e-07,
|
|
"loss": 0.5366581439971924,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3350,
|
|
"token_acc": 0.8674698795180723,
|
|
"train_speed(iter/s)": 0.460789
|
|
},
|
|
{
|
|
"epoch": 2.44,
|
|
"grad_norm": 4.412635803222656,
|
|
"learning_rate": 8.863677624029837e-07,
|
|
"loss": 0.6205951690673828,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3355,
|
|
"token_acc": 0.8476459890887048,
|
|
"train_speed(iter/s)": 0.460848
|
|
},
|
|
{
|
|
"epoch": 2.443636363636364,
|
|
"grad_norm": 3.683108329772949,
|
|
"learning_rate": 8.752410392449984e-07,
|
|
"loss": 0.5838453769683838,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3360,
|
|
"token_acc": 0.853091684434968,
|
|
"train_speed(iter/s)": 0.460866
|
|
},
|
|
{
|
|
"epoch": 2.4472727272727273,
|
|
"grad_norm": 5.493093967437744,
|
|
"learning_rate": 8.641778931802763e-07,
|
|
"loss": 0.5790290832519531,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3365,
|
|
"token_acc": 0.8514484861685907,
|
|
"train_speed(iter/s)": 0.46087
|
|
},
|
|
{
|
|
"epoch": 2.450909090909091,
|
|
"grad_norm": 4.477910995483398,
|
|
"learning_rate": 8.531784947309458e-07,
|
|
"loss": 0.5288788795471191,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3370,
|
|
"token_acc": 0.8701589889930698,
|
|
"train_speed(iter/s)": 0.460883
|
|
},
|
|
{
|
|
"epoch": 2.4545454545454546,
|
|
"grad_norm": 3.794621706008911,
|
|
"learning_rate": 8.422430134365584e-07,
|
|
"loss": 0.5916152000427246,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3375,
|
|
"token_acc": 0.8327744243237201,
|
|
"train_speed(iter/s)": 0.460914
|
|
},
|
|
{
|
|
"epoch": 2.458181818181818,
|
|
"grad_norm": 3.1803672313690186,
|
|
"learning_rate": 8.31371617851473e-07,
|
|
"loss": 0.5673966407775879,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3380,
|
|
"token_acc": 0.8521260841971652,
|
|
"train_speed(iter/s)": 0.460936
|
|
},
|
|
{
|
|
"epoch": 2.461818181818182,
|
|
"grad_norm": 3.857543706893921,
|
|
"learning_rate": 8.20564475542267e-07,
|
|
"loss": 0.5137066841125488,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3385,
|
|
"token_acc": 0.8740687334775294,
|
|
"train_speed(iter/s)": 0.460965
|
|
},
|
|
{
|
|
"epoch": 2.4654545454545453,
|
|
"grad_norm": 4.353023052215576,
|
|
"learning_rate": 8.098217530851454e-07,
|
|
"loss": 0.513277816772461,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3390,
|
|
"token_acc": 0.8848245180425112,
|
|
"train_speed(iter/s)": 0.46101
|
|
},
|
|
{
|
|
"epoch": 2.4690909090909092,
|
|
"grad_norm": 3.8349452018737793,
|
|
"learning_rate": 7.991436160633742e-07,
|
|
"loss": 0.534281873703003,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3395,
|
|
"token_acc": 0.8546059933407325,
|
|
"train_speed(iter/s)": 0.461028
|
|
},
|
|
{
|
|
"epoch": 2.4727272727272727,
|
|
"grad_norm": 3.646422863006592,
|
|
"learning_rate": 7.885302290647328e-07,
|
|
"loss": 0.47332277297973635,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3400,
|
|
"token_acc": 0.8866328257191202,
|
|
"train_speed(iter/s)": 0.461025
|
|
},
|
|
{
|
|
"epoch": 2.4763636363636365,
|
|
"grad_norm": 4.512187480926514,
|
|
"learning_rate": 7.779817556789731e-07,
|
|
"loss": 0.5604523658752442,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3405,
|
|
"token_acc": 0.8550799381124291,
|
|
"train_speed(iter/s)": 0.461041
|
|
},
|
|
{
|
|
"epoch": 2.48,
|
|
"grad_norm": 3.4519810676574707,
|
|
"learning_rate": 7.674983584953e-07,
|
|
"loss": 0.548017930984497,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3410,
|
|
"token_acc": 0.8271052631578948,
|
|
"train_speed(iter/s)": 0.461029
|
|
},
|
|
{
|
|
"epoch": 2.4836363636363634,
|
|
"grad_norm": 4.231698989868164,
|
|
"learning_rate": 7.570801990998617e-07,
|
|
"loss": 0.5751005649566651,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3415,
|
|
"token_acc": 0.8549730524252818,
|
|
"train_speed(iter/s)": 0.461082
|
|
},
|
|
{
|
|
"epoch": 2.4872727272727273,
|
|
"grad_norm": 3.640695810317993,
|
|
"learning_rate": 7.467274380732647e-07,
|
|
"loss": 0.4948701858520508,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3420,
|
|
"token_acc": 0.8915375446960667,
|
|
"train_speed(iter/s)": 0.461057
|
|
},
|
|
{
|
|
"epoch": 2.4909090909090907,
|
|
"grad_norm": 4.212837219238281,
|
|
"learning_rate": 7.364402349880961e-07,
|
|
"loss": 0.6127988815307617,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3425,
|
|
"token_acc": 0.8273244781783681,
|
|
"train_speed(iter/s)": 0.461077
|
|
},
|
|
{
|
|
"epoch": 2.4945454545454546,
|
|
"grad_norm": 4.170472621917725,
|
|
"learning_rate": 7.262187484064598e-07,
|
|
"loss": 0.5763627052307129,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3430,
|
|
"token_acc": 0.8165927471954083,
|
|
"train_speed(iter/s)": 0.461074
|
|
},
|
|
{
|
|
"epoch": 2.498181818181818,
|
|
"grad_norm": 4.051548004150391,
|
|
"learning_rate": 7.160631358775411e-07,
|
|
"loss": 0.5574194431304932,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3435,
|
|
"token_acc": 0.8442812982998454,
|
|
"train_speed(iter/s)": 0.461084
|
|
},
|
|
{
|
|
"epoch": 2.501818181818182,
|
|
"grad_norm": 3.917257070541382,
|
|
"learning_rate": 7.059735539351714e-07,
|
|
"loss": 0.5000098705291748,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3440,
|
|
"token_acc": 0.8565266742338252,
|
|
"train_speed(iter/s)": 0.461092
|
|
},
|
|
{
|
|
"epoch": 2.5054545454545454,
|
|
"grad_norm": 3.361421823501587,
|
|
"learning_rate": 6.95950158095417e-07,
|
|
"loss": 0.564305305480957,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3445,
|
|
"token_acc": 0.8359232175502742,
|
|
"train_speed(iter/s)": 0.461116
|
|
},
|
|
{
|
|
"epoch": 2.509090909090909,
|
|
"grad_norm": 3.9124817848205566,
|
|
"learning_rate": 6.859931028541849e-07,
|
|
"loss": 0.5879053592681884,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3450,
|
|
"token_acc": 0.8065608465608466,
|
|
"train_speed(iter/s)": 0.46113
|
|
},
|
|
{
|
|
"epoch": 2.5127272727272727,
|
|
"grad_norm": 4.365234851837158,
|
|
"learning_rate": 6.761025416848377e-07,
|
|
"loss": 0.5042540550231933,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3455,
|
|
"token_acc": 0.8511663286004056,
|
|
"train_speed(iter/s)": 0.461186
|
|
},
|
|
{
|
|
"epoch": 2.5163636363636366,
|
|
"grad_norm": 3.8373539447784424,
|
|
"learning_rate": 6.662786270358296e-07,
|
|
"loss": 0.5439183235168457,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3460,
|
|
"token_acc": 0.85625,
|
|
"train_speed(iter/s)": 0.461196
|
|
},
|
|
{
|
|
"epoch": 2.52,
|
|
"grad_norm": 4.121663570404053,
|
|
"learning_rate": 6.565215103283584e-07,
|
|
"loss": 0.5968369960784912,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3465,
|
|
"token_acc": 0.867917155138726,
|
|
"train_speed(iter/s)": 0.461253
|
|
},
|
|
{
|
|
"epoch": 2.5236363636363635,
|
|
"grad_norm": 4.577036380767822,
|
|
"learning_rate": 6.468313419540295e-07,
|
|
"loss": 0.6250477790832519,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3470,
|
|
"token_acc": 0.7919896640826873,
|
|
"train_speed(iter/s)": 0.461277
|
|
},
|
|
{
|
|
"epoch": 2.5272727272727273,
|
|
"grad_norm": 3.807965040206909,
|
|
"learning_rate": 6.372082712725352e-07,
|
|
"loss": 0.518894100189209,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3475,
|
|
"token_acc": 0.8521634615384616,
|
|
"train_speed(iter/s)": 0.461252
|
|
},
|
|
{
|
|
"epoch": 2.5309090909090908,
|
|
"grad_norm": 3.836399555206299,
|
|
"learning_rate": 6.276524466093592e-07,
|
|
"loss": 0.5360546112060547,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3480,
|
|
"token_acc": 0.8698093689966011,
|
|
"train_speed(iter/s)": 0.461221
|
|
},
|
|
{
|
|
"epoch": 2.5345454545454547,
|
|
"grad_norm": 3.81571102142334,
|
|
"learning_rate": 6.181640152534845e-07,
|
|
"loss": 0.5474881649017334,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3485,
|
|
"token_acc": 0.8732005521593375,
|
|
"train_speed(iter/s)": 0.461203
|
|
},
|
|
{
|
|
"epoch": 2.538181818181818,
|
|
"grad_norm": 3.869523286819458,
|
|
"learning_rate": 6.087431234551278e-07,
|
|
"loss": 0.5763465881347656,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3490,
|
|
"token_acc": 0.8603756505996832,
|
|
"train_speed(iter/s)": 0.461205
|
|
},
|
|
{
|
|
"epoch": 2.541818181818182,
|
|
"grad_norm": 4.300631999969482,
|
|
"learning_rate": 5.993899164234795e-07,
|
|
"loss": 0.5737009048461914,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3495,
|
|
"token_acc": 0.853713260040386,
|
|
"train_speed(iter/s)": 0.461229
|
|
},
|
|
{
|
|
"epoch": 2.5454545454545454,
|
|
"grad_norm": 4.0556416511535645,
|
|
"learning_rate": 5.901045383244708e-07,
|
|
"loss": 0.558495044708252,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3500,
|
|
"token_acc": 0.8296205630354957,
|
|
"train_speed(iter/s)": 0.461267
|
|
},
|
|
{
|
|
"epoch": 2.549090909090909,
|
|
"grad_norm": 4.591411590576172,
|
|
"learning_rate": 5.808871322785503e-07,
|
|
"loss": 0.537141990661621,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3505,
|
|
"token_acc": 0.8630566122174983,
|
|
"train_speed(iter/s)": 0.461298
|
|
},
|
|
{
|
|
"epoch": 2.5527272727272727,
|
|
"grad_norm": 3.8997533321380615,
|
|
"learning_rate": 5.717378403584766e-07,
|
|
"loss": 0.5735316753387452,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3510,
|
|
"token_acc": 0.8538715769593956,
|
|
"train_speed(iter/s)": 0.461339
|
|
},
|
|
{
|
|
"epoch": 2.5563636363636366,
|
|
"grad_norm": 4.364303112030029,
|
|
"learning_rate": 5.626568035871282e-07,
|
|
"loss": 0.610892391204834,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3515,
|
|
"token_acc": 0.8688324377137274,
|
|
"train_speed(iter/s)": 0.461371
|
|
},
|
|
{
|
|
"epoch": 2.56,
|
|
"grad_norm": 4.005393028259277,
|
|
"learning_rate": 5.536441619353322e-07,
|
|
"loss": 0.5376082420349121,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3520,
|
|
"token_acc": 0.8538071065989847,
|
|
"train_speed(iter/s)": 0.461379
|
|
},
|
|
{
|
|
"epoch": 2.5636363636363635,
|
|
"grad_norm": 3.683354616165161,
|
|
"learning_rate": 5.447000543197067e-07,
|
|
"loss": 0.47856903076171875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3525,
|
|
"token_acc": 0.8590619307832422,
|
|
"train_speed(iter/s)": 0.461392
|
|
},
|
|
{
|
|
"epoch": 2.5672727272727274,
|
|
"grad_norm": 3.6826767921447754,
|
|
"learning_rate": 5.358246186005139e-07,
|
|
"loss": 0.5937067031860351,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3530,
|
|
"token_acc": 0.8394269831088305,
|
|
"train_speed(iter/s)": 0.461397
|
|
},
|
|
{
|
|
"epoch": 2.570909090909091,
|
|
"grad_norm": 3.76254940032959,
|
|
"learning_rate": 5.270179915795442e-07,
|
|
"loss": 0.5464666366577149,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3535,
|
|
"token_acc": 0.8444878048780488,
|
|
"train_speed(iter/s)": 0.461425
|
|
},
|
|
{
|
|
"epoch": 2.5745454545454547,
|
|
"grad_norm": 4.266871452331543,
|
|
"learning_rate": 5.182803089980032e-07,
|
|
"loss": 0.5382218837738038,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3540,
|
|
"token_acc": 0.8665244191559981,
|
|
"train_speed(iter/s)": 0.461445
|
|
},
|
|
{
|
|
"epoch": 2.578181818181818,
|
|
"grad_norm": 4.570847988128662,
|
|
"learning_rate": 5.096117055344146e-07,
|
|
"loss": 0.5921483039855957,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3545,
|
|
"token_acc": 0.849323753169907,
|
|
"train_speed(iter/s)": 0.461488
|
|
},
|
|
{
|
|
"epoch": 2.581818181818182,
|
|
"grad_norm": 3.771512985229492,
|
|
"learning_rate": 5.010123148025536e-07,
|
|
"loss": 0.5578415870666504,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3550,
|
|
"token_acc": 0.8299770583736936,
|
|
"train_speed(iter/s)": 0.461505
|
|
},
|
|
{
|
|
"epoch": 2.5854545454545454,
|
|
"grad_norm": 3.8370325565338135,
|
|
"learning_rate": 4.924822693493819e-07,
|
|
"loss": 0.5817060470581055,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3555,
|
|
"token_acc": 0.8504863153132776,
|
|
"train_speed(iter/s)": 0.461528
|
|
},
|
|
{
|
|
"epoch": 2.589090909090909,
|
|
"grad_norm": 4.462881565093994,
|
|
"learning_rate": 4.840217006530063e-07,
|
|
"loss": 0.5203249454498291,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3560,
|
|
"token_acc": 0.8685489470030086,
|
|
"train_speed(iter/s)": 0.461554
|
|
},
|
|
{
|
|
"epoch": 2.5927272727272728,
|
|
"grad_norm": 3.998481035232544,
|
|
"learning_rate": 4.7563073912064826e-07,
|
|
"loss": 0.5419306278228759,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3565,
|
|
"token_acc": 0.8826725699335882,
|
|
"train_speed(iter/s)": 0.461575
|
|
},
|
|
{
|
|
"epoch": 2.596363636363636,
|
|
"grad_norm": 3.884657859802246,
|
|
"learning_rate": 4.6730951408664126e-07,
|
|
"loss": 0.6358286380767822,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3570,
|
|
"token_acc": 0.8244274809160306,
|
|
"train_speed(iter/s)": 0.461616
|
|
},
|
|
{
|
|
"epoch": 2.6,
|
|
"grad_norm": 4.17387056350708,
|
|
"learning_rate": 4.590581538104316e-07,
|
|
"loss": 0.5894208431243897,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3575,
|
|
"token_acc": 0.8605391636595203,
|
|
"train_speed(iter/s)": 0.461646
|
|
},
|
|
{
|
|
"epoch": 2.6036363636363635,
|
|
"grad_norm": 3.776859998703003,
|
|
"learning_rate": 4.5087678547460157e-07,
|
|
"loss": 0.5507136344909668,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3580,
|
|
"token_acc": 0.8549187725631769,
|
|
"train_speed(iter/s)": 0.461652
|
|
},
|
|
{
|
|
"epoch": 2.6072727272727274,
|
|
"grad_norm": 3.787224769592285,
|
|
"learning_rate": 4.42765535182913e-07,
|
|
"loss": 0.5116243362426758,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3585,
|
|
"token_acc": 0.8763873775843308,
|
|
"train_speed(iter/s)": 0.461668
|
|
},
|
|
{
|
|
"epoch": 2.610909090909091,
|
|
"grad_norm": 3.734219551086426,
|
|
"learning_rate": 4.3472452795836116e-07,
|
|
"loss": 0.5455838203430176,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3590,
|
|
"token_acc": 0.8290505815526102,
|
|
"train_speed(iter/s)": 0.461668
|
|
},
|
|
{
|
|
"epoch": 2.6145454545454543,
|
|
"grad_norm": 3.9400975704193115,
|
|
"learning_rate": 4.2675388774124525e-07,
|
|
"loss": 0.507750129699707,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3595,
|
|
"token_acc": 0.8634268956849602,
|
|
"train_speed(iter/s)": 0.461716
|
|
},
|
|
{
|
|
"epoch": 2.618181818181818,
|
|
"grad_norm": 4.625056743621826,
|
|
"learning_rate": 4.1885373738726254e-07,
|
|
"loss": 0.5847376346588135,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3600,
|
|
"token_acc": 0.8611609814482346,
|
|
"train_speed(iter/s)": 0.461741
|
|
},
|
|
{
|
|
"epoch": 2.621818181818182,
|
|
"grad_norm": 3.855128765106201,
|
|
"learning_rate": 4.1102419866561217e-07,
|
|
"loss": 0.4857977867126465,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3605,
|
|
"token_acc": 0.8603732950466619,
|
|
"train_speed(iter/s)": 0.461714
|
|
},
|
|
{
|
|
"epoch": 2.6254545454545455,
|
|
"grad_norm": 3.6857378482818604,
|
|
"learning_rate": 4.0326539225712e-07,
|
|
"loss": 0.5717270374298096,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3610,
|
|
"token_acc": 0.868129841386942,
|
|
"train_speed(iter/s)": 0.461706
|
|
},
|
|
{
|
|
"epoch": 2.629090909090909,
|
|
"grad_norm": 3.903055191040039,
|
|
"learning_rate": 3.9557743775237523e-07,
|
|
"loss": 0.5993773937225342,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3615,
|
|
"token_acc": 0.8062308236960113,
|
|
"train_speed(iter/s)": 0.461738
|
|
},
|
|
{
|
|
"epoch": 2.632727272727273,
|
|
"grad_norm": 3.270725727081299,
|
|
"learning_rate": 3.87960453649891e-07,
|
|
"loss": 0.5760551929473877,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3620,
|
|
"token_acc": 0.8619246861924686,
|
|
"train_speed(iter/s)": 0.461774
|
|
},
|
|
{
|
|
"epoch": 2.6363636363636362,
|
|
"grad_norm": 4.472327709197998,
|
|
"learning_rate": 3.8041455735427446e-07,
|
|
"loss": 0.5789799690246582,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3625,
|
|
"token_acc": 0.8526070492280804,
|
|
"train_speed(iter/s)": 0.461738
|
|
},
|
|
{
|
|
"epoch": 2.64,
|
|
"grad_norm": 4.271474838256836,
|
|
"learning_rate": 3.7293986517441927e-07,
|
|
"loss": 0.5459742069244384,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3630,
|
|
"token_acc": 0.862533692722372,
|
|
"train_speed(iter/s)": 0.461757
|
|
},
|
|
{
|
|
"epoch": 2.6436363636363636,
|
|
"grad_norm": 4.120786190032959,
|
|
"learning_rate": 3.655364923217136e-07,
|
|
"loss": 0.5433560371398926,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3635,
|
|
"token_acc": 0.8354055654233274,
|
|
"train_speed(iter/s)": 0.461783
|
|
},
|
|
{
|
|
"epoch": 2.6472727272727274,
|
|
"grad_norm": 4.382391929626465,
|
|
"learning_rate": 3.5820455290826285e-07,
|
|
"loss": 0.582338285446167,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3640,
|
|
"token_acc": 0.8313772455089821,
|
|
"train_speed(iter/s)": 0.461828
|
|
},
|
|
{
|
|
"epoch": 2.650909090909091,
|
|
"grad_norm": 3.799560785293579,
|
|
"learning_rate": 3.509441599451285e-07,
|
|
"loss": 0.5547786712646484,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3645,
|
|
"token_acc": 0.8674388674388674,
|
|
"train_speed(iter/s)": 0.461851
|
|
},
|
|
{
|
|
"epoch": 2.6545454545454543,
|
|
"grad_norm": 3.852847099304199,
|
|
"learning_rate": 3.437554253405906e-07,
|
|
"loss": 0.5601497650146484,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3650,
|
|
"token_acc": 0.8693735498839907,
|
|
"train_speed(iter/s)": 0.461886
|
|
},
|
|
{
|
|
"epoch": 2.658181818181818,
|
|
"grad_norm": 4.284626483917236,
|
|
"learning_rate": 3.366384598984229e-07,
|
|
"loss": 0.5693297386169434,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3655,
|
|
"token_acc": 0.8661684782608695,
|
|
"train_speed(iter/s)": 0.461922
|
|
},
|
|
{
|
|
"epoch": 2.661818181818182,
|
|
"grad_norm": 4.123050212860107,
|
|
"learning_rate": 3.295933733161788e-07,
|
|
"loss": 0.5483355045318603,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3660,
|
|
"token_acc": 0.8760154738878143,
|
|
"train_speed(iter/s)": 0.461945
|
|
},
|
|
{
|
|
"epoch": 2.6654545454545455,
|
|
"grad_norm": 3.7868292331695557,
|
|
"learning_rate": 3.2262027418350773e-07,
|
|
"loss": 0.5330245971679688,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3665,
|
|
"token_acc": 0.859638673813778,
|
|
"train_speed(iter/s)": 0.461908
|
|
},
|
|
{
|
|
"epoch": 2.669090909090909,
|
|
"grad_norm": 3.9128313064575195,
|
|
"learning_rate": 3.157192699804784e-07,
|
|
"loss": 0.5247835636138916,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3670,
|
|
"token_acc": 0.8648218196297034,
|
|
"train_speed(iter/s)": 0.461888
|
|
},
|
|
{
|
|
"epoch": 2.672727272727273,
|
|
"grad_norm": 3.6926794052124023,
|
|
"learning_rate": 3.0889046707592185e-07,
|
|
"loss": 0.5355826854705811,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3675,
|
|
"token_acc": 0.8548972188633616,
|
|
"train_speed(iter/s)": 0.461913
|
|
},
|
|
{
|
|
"epoch": 2.6763636363636363,
|
|
"grad_norm": 3.795276641845703,
|
|
"learning_rate": 3.021339707257903e-07,
|
|
"loss": 0.5026448249816895,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3680,
|
|
"token_acc": 0.8815384615384615,
|
|
"train_speed(iter/s)": 0.461933
|
|
},
|
|
{
|
|
"epoch": 2.68,
|
|
"grad_norm": 3.9286348819732666,
|
|
"learning_rate": 2.9544988507153984e-07,
|
|
"loss": 0.5500066757202149,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3685,
|
|
"token_acc": 0.8544303797468354,
|
|
"train_speed(iter/s)": 0.46193
|
|
},
|
|
{
|
|
"epoch": 2.6836363636363636,
|
|
"grad_norm": 3.818805694580078,
|
|
"learning_rate": 2.8883831313852086e-07,
|
|
"loss": 0.5495880126953125,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3690,
|
|
"token_acc": 0.8208686712927268,
|
|
"train_speed(iter/s)": 0.461966
|
|
},
|
|
{
|
|
"epoch": 2.6872727272727275,
|
|
"grad_norm": 3.5088865756988525,
|
|
"learning_rate": 2.822993568343907e-07,
|
|
"loss": 0.5506268978118897,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3695,
|
|
"token_acc": 0.8467027559055118,
|
|
"train_speed(iter/s)": 0.461986
|
|
},
|
|
{
|
|
"epoch": 2.690909090909091,
|
|
"grad_norm": 4.072029113769531,
|
|
"learning_rate": 2.7583311694754424e-07,
|
|
"loss": 0.5637264251708984,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3700,
|
|
"token_acc": 0.826214405360134,
|
|
"train_speed(iter/s)": 0.462012
|
|
},
|
|
{
|
|
"epoch": 2.6945454545454544,
|
|
"grad_norm": 3.523855447769165,
|
|
"learning_rate": 2.694396931455606e-07,
|
|
"loss": 0.5251345634460449,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3705,
|
|
"token_acc": 0.8783808313668542,
|
|
"train_speed(iter/s)": 0.462002
|
|
},
|
|
{
|
|
"epoch": 2.6981818181818182,
|
|
"grad_norm": 4.599588871002197,
|
|
"learning_rate": 2.631191839736635e-07,
|
|
"loss": 0.5488700866699219,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3710,
|
|
"token_acc": 0.8507853403141361,
|
|
"train_speed(iter/s)": 0.462039
|
|
},
|
|
{
|
|
"epoch": 2.7018181818181817,
|
|
"grad_norm": 3.0808658599853516,
|
|
"learning_rate": 2.5687168685320784e-07,
|
|
"loss": 0.5053971767425537,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3715,
|
|
"token_acc": 0.8607594936708861,
|
|
"train_speed(iter/s)": 0.462017
|
|
},
|
|
{
|
|
"epoch": 2.7054545454545456,
|
|
"grad_norm": 3.8653998374938965,
|
|
"learning_rate": 2.506972980801731e-07,
|
|
"loss": 0.5484971523284912,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3720,
|
|
"token_acc": 0.8357475426797724,
|
|
"train_speed(iter/s)": 0.462026
|
|
},
|
|
{
|
|
"epoch": 2.709090909090909,
|
|
"grad_norm": 3.6488606929779053,
|
|
"learning_rate": 2.4459611282368346e-07,
|
|
"loss": 0.513886833190918,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3725,
|
|
"token_acc": 0.8707535121328225,
|
|
"train_speed(iter/s)": 0.462036
|
|
},
|
|
{
|
|
"epoch": 2.712727272727273,
|
|
"grad_norm": 3.738938808441162,
|
|
"learning_rate": 2.3856822512453414e-07,
|
|
"loss": 0.5665274620056152,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3730,
|
|
"token_acc": 0.8454545454545455,
|
|
"train_speed(iter/s)": 0.462015
|
|
},
|
|
{
|
|
"epoch": 2.7163636363636363,
|
|
"grad_norm": 3.5817694664001465,
|
|
"learning_rate": 2.3261372789375124e-07,
|
|
"loss": 0.5293080806732178,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3735,
|
|
"token_acc": 0.83125,
|
|
"train_speed(iter/s)": 0.462051
|
|
},
|
|
{
|
|
"epoch": 2.7199999999999998,
|
|
"grad_norm": 4.456275463104248,
|
|
"learning_rate": 2.2673271291115184e-07,
|
|
"loss": 0.5739312171936035,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3740,
|
|
"token_acc": 0.8260763804427147,
|
|
"train_speed(iter/s)": 0.462047
|
|
},
|
|
{
|
|
"epoch": 2.7236363636363636,
|
|
"grad_norm": 3.6416001319885254,
|
|
"learning_rate": 2.2092527082393235e-07,
|
|
"loss": 0.5760965347290039,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3745,
|
|
"token_acc": 0.846979530703944,
|
|
"train_speed(iter/s)": 0.462087
|
|
},
|
|
{
|
|
"epoch": 2.7272727272727275,
|
|
"grad_norm": 4.468888282775879,
|
|
"learning_rate": 2.151914911452724e-07,
|
|
"loss": 0.5503689765930175,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3750,
|
|
"token_acc": 0.8618996798292423,
|
|
"train_speed(iter/s)": 0.462101
|
|
},
|
|
{
|
|
"epoch": 2.730909090909091,
|
|
"grad_norm": 4.282381534576416,
|
|
"learning_rate": 2.0953146225295263e-07,
|
|
"loss": 0.5355560779571533,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3755,
|
|
"token_acc": 0.8769058295964126,
|
|
"train_speed(iter/s)": 0.462125
|
|
},
|
|
{
|
|
"epoch": 2.7345454545454544,
|
|
"grad_norm": 3.994455099105835,
|
|
"learning_rate": 2.039452713879958e-07,
|
|
"loss": 0.5964118003845215,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3760,
|
|
"token_acc": 0.8107826810990841,
|
|
"train_speed(iter/s)": 0.462116
|
|
},
|
|
{
|
|
"epoch": 2.7381818181818183,
|
|
"grad_norm": 3.3923728466033936,
|
|
"learning_rate": 1.9843300465331728e-07,
|
|
"loss": 0.4777844429016113,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3765,
|
|
"token_acc": 0.8853016142735769,
|
|
"train_speed(iter/s)": 0.462119
|
|
},
|
|
{
|
|
"epoch": 2.7418181818181817,
|
|
"grad_norm": 4.135165691375732,
|
|
"learning_rate": 1.9299474701240384e-07,
|
|
"loss": 0.5326554298400878,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3770,
|
|
"token_acc": 0.8466922555881689,
|
|
"train_speed(iter/s)": 0.462148
|
|
},
|
|
{
|
|
"epoch": 2.7454545454545456,
|
|
"grad_norm": 4.381650924682617,
|
|
"learning_rate": 1.87630582287997e-07,
|
|
"loss": 0.5507936477661133,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3775,
|
|
"token_acc": 0.8197317572175494,
|
|
"train_speed(iter/s)": 0.462168
|
|
},
|
|
{
|
|
"epoch": 2.749090909090909,
|
|
"grad_norm": 3.7157084941864014,
|
|
"learning_rate": 1.8234059316080855e-07,
|
|
"loss": 0.5473185539245605,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3780,
|
|
"token_acc": 0.8370760769935839,
|
|
"train_speed(iter/s)": 0.462175
|
|
},
|
|
{
|
|
"epoch": 2.752727272727273,
|
|
"grad_norm": 3.659472942352295,
|
|
"learning_rate": 1.771248611682408e-07,
|
|
"loss": 0.596890640258789,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3785,
|
|
"token_acc": 0.8329901624342255,
|
|
"train_speed(iter/s)": 0.462191
|
|
},
|
|
{
|
|
"epoch": 2.7563636363636363,
|
|
"grad_norm": 3.8994619846343994,
|
|
"learning_rate": 1.7198346670313228e-07,
|
|
"loss": 0.5438637256622314,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3790,
|
|
"token_acc": 0.8527472527472527,
|
|
"train_speed(iter/s)": 0.462186
|
|
},
|
|
{
|
|
"epoch": 2.76,
|
|
"grad_norm": 3.3449583053588867,
|
|
"learning_rate": 1.6691648901251633e-07,
|
|
"loss": 0.5252495288848877,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3795,
|
|
"token_acc": 0.8635451505016722,
|
|
"train_speed(iter/s)": 0.462206
|
|
},
|
|
{
|
|
"epoch": 2.7636363636363637,
|
|
"grad_norm": 3.654770612716675,
|
|
"learning_rate": 1.619240061964028e-07,
|
|
"loss": 0.5614541053771973,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3800,
|
|
"token_acc": 0.8361153262518968,
|
|
"train_speed(iter/s)": 0.462186
|
|
},
|
|
{
|
|
"epoch": 2.767272727272727,
|
|
"grad_norm": 4.122347831726074,
|
|
"learning_rate": 1.5700609520657328e-07,
|
|
"loss": 0.5514491081237793,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3805,
|
|
"token_acc": 0.8334858188472095,
|
|
"train_speed(iter/s)": 0.462199
|
|
},
|
|
{
|
|
"epoch": 2.770909090909091,
|
|
"grad_norm": 3.933711290359497,
|
|
"learning_rate": 1.5216283184539103e-07,
|
|
"loss": 0.5302469253540039,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3810,
|
|
"token_acc": 0.8600960534558363,
|
|
"train_speed(iter/s)": 0.462167
|
|
},
|
|
{
|
|
"epoch": 2.7745454545454544,
|
|
"grad_norm": 3.4784810543060303,
|
|
"learning_rate": 1.4739429076463863e-07,
|
|
"loss": 0.5577369213104248,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3815,
|
|
"token_acc": 0.8665246028671058,
|
|
"train_speed(iter/s)": 0.462204
|
|
},
|
|
{
|
|
"epoch": 2.7781818181818183,
|
|
"grad_norm": 3.944777488708496,
|
|
"learning_rate": 1.427005454643654e-07,
|
|
"loss": 0.5139852523803711,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3820,
|
|
"token_acc": 0.8726895119418484,
|
|
"train_speed(iter/s)": 0.462202
|
|
},
|
|
{
|
|
"epoch": 2.7818181818181817,
|
|
"grad_norm": 3.965413808822632,
|
|
"learning_rate": 1.380816682917502e-07,
|
|
"loss": 0.5491145133972168,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3825,
|
|
"token_acc": 0.839851024208566,
|
|
"train_speed(iter/s)": 0.462205
|
|
},
|
|
{
|
|
"epoch": 2.785454545454545,
|
|
"grad_norm": 4.084675312042236,
|
|
"learning_rate": 1.3353773043999263e-07,
|
|
"loss": 0.5206490516662597,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3830,
|
|
"token_acc": 0.8489866534849234,
|
|
"train_speed(iter/s)": 0.462216
|
|
},
|
|
{
|
|
"epoch": 2.789090909090909,
|
|
"grad_norm": 4.1376142501831055,
|
|
"learning_rate": 1.290688019472125e-07,
|
|
"loss": 0.621760082244873,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3835,
|
|
"token_acc": 0.8129857107044371,
|
|
"train_speed(iter/s)": 0.462244
|
|
},
|
|
{
|
|
"epoch": 2.792727272727273,
|
|
"grad_norm": 4.061305522918701,
|
|
"learning_rate": 1.2467495169536993e-07,
|
|
"loss": 0.548912763595581,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3840,
|
|
"token_acc": 0.8436461794019934,
|
|
"train_speed(iter/s)": 0.462284
|
|
},
|
|
{
|
|
"epoch": 2.7963636363636364,
|
|
"grad_norm": 4.353764057159424,
|
|
"learning_rate": 1.203562474092046e-07,
|
|
"loss": 0.5282083988189697,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3845,
|
|
"token_acc": 0.8687521251275077,
|
|
"train_speed(iter/s)": 0.462265
|
|
},
|
|
{
|
|
"epoch": 2.8,
|
|
"grad_norm": 4.023250579833984,
|
|
"learning_rate": 1.161127556551922e-07,
|
|
"loss": 0.5132468223571778,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3850,
|
|
"token_acc": 0.8647706422018349,
|
|
"train_speed(iter/s)": 0.462302
|
|
},
|
|
{
|
|
"epoch": 2.8036363636363637,
|
|
"grad_norm": 3.5195140838623047,
|
|
"learning_rate": 1.1194454184051685e-07,
|
|
"loss": 0.5616849899291992,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3855,
|
|
"token_acc": 0.8549200087661626,
|
|
"train_speed(iter/s)": 0.462312
|
|
},
|
|
{
|
|
"epoch": 2.807272727272727,
|
|
"grad_norm": 3.75972580909729,
|
|
"learning_rate": 1.0785167021206466e-07,
|
|
"loss": 0.5320147037506103,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3860,
|
|
"token_acc": 0.861925477586533,
|
|
"train_speed(iter/s)": 0.462331
|
|
},
|
|
{
|
|
"epoch": 2.810909090909091,
|
|
"grad_norm": 4.204853534698486,
|
|
"learning_rate": 1.0383420385543241e-07,
|
|
"loss": 0.44639225006103517,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3865,
|
|
"token_acc": 0.870872908186341,
|
|
"train_speed(iter/s)": 0.462335
|
|
},
|
|
{
|
|
"epoch": 2.8145454545454545,
|
|
"grad_norm": 3.4746897220611572,
|
|
"learning_rate": 9.989220469395599e-08,
|
|
"loss": 0.5511418342590332,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3870,
|
|
"token_acc": 0.8400713436385255,
|
|
"train_speed(iter/s)": 0.46236
|
|
},
|
|
{
|
|
"epoch": 2.8181818181818183,
|
|
"grad_norm": 3.238222122192383,
|
|
"learning_rate": 9.602573348775512e-08,
|
|
"loss": 0.4952839851379395,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3875,
|
|
"token_acc": 0.8549369290996085,
|
|
"train_speed(iter/s)": 0.462351
|
|
},
|
|
{
|
|
"epoch": 2.821818181818182,
|
|
"grad_norm": 4.106288433074951,
|
|
"learning_rate": 9.223484983279685e-08,
|
|
"loss": 0.5499021530151367,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3880,
|
|
"token_acc": 0.853290676416819,
|
|
"train_speed(iter/s)": 0.462371
|
|
},
|
|
{
|
|
"epoch": 2.825454545454545,
|
|
"grad_norm": 3.5807745456695557,
|
|
"learning_rate": 8.85196121599774e-08,
|
|
"loss": 0.5587635517120362,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3885,
|
|
"token_acc": 0.8456197672044109,
|
|
"train_speed(iter/s)": 0.462358
|
|
},
|
|
{
|
|
"epoch": 2.829090909090909,
|
|
"grad_norm": 4.102653980255127,
|
|
"learning_rate": 8.488007773422291e-08,
|
|
"loss": 0.5749258041381836,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3890,
|
|
"token_acc": 0.8238822246455835,
|
|
"train_speed(iter/s)": 0.462379
|
|
},
|
|
{
|
|
"epoch": 2.832727272727273,
|
|
"grad_norm": 3.7965023517608643,
|
|
"learning_rate": 8.131630265360235e-08,
|
|
"loss": 0.5211573123931885,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3895,
|
|
"token_acc": 0.8566985114419018,
|
|
"train_speed(iter/s)": 0.462397
|
|
},
|
|
{
|
|
"epoch": 2.8363636363636364,
|
|
"grad_norm": 4.140139102935791,
|
|
"learning_rate": 7.782834184846877e-08,
|
|
"loss": 0.5460617065429687,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3900,
|
|
"token_acc": 0.8294644909216272,
|
|
"train_speed(iter/s)": 0.462449
|
|
},
|
|
{
|
|
"epoch": 2.84,
|
|
"grad_norm": 4.100113868713379,
|
|
"learning_rate": 7.441624908060774e-08,
|
|
"loss": 0.65106201171875,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3905,
|
|
"token_acc": 0.8127829560585885,
|
|
"train_speed(iter/s)": 0.462479
|
|
},
|
|
{
|
|
"epoch": 2.8436363636363637,
|
|
"grad_norm": 4.006384372711182,
|
|
"learning_rate": 7.108007694241137e-08,
|
|
"loss": 0.6106512069702148,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3910,
|
|
"token_acc": 0.8437872644884331,
|
|
"train_speed(iter/s)": 0.462509
|
|
},
|
|
{
|
|
"epoch": 2.847272727272727,
|
|
"grad_norm": 4.023671627044678,
|
|
"learning_rate": 6.781987685606672e-08,
|
|
"loss": 0.5270663738250733,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3915,
|
|
"token_acc": 0.8609300304215558,
|
|
"train_speed(iter/s)": 0.46252
|
|
},
|
|
{
|
|
"epoch": 2.850909090909091,
|
|
"grad_norm": 3.586935043334961,
|
|
"learning_rate": 6.463569907276424e-08,
|
|
"loss": 0.5477522850036621,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3920,
|
|
"token_acc": 0.870264247055078,
|
|
"train_speed(iter/s)": 0.462498
|
|
},
|
|
{
|
|
"epoch": 2.8545454545454545,
|
|
"grad_norm": 3.2990100383758545,
|
|
"learning_rate": 6.15275926719211e-08,
|
|
"loss": 0.4375779151916504,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3925,
|
|
"token_acc": 0.8727871552079045,
|
|
"train_speed(iter/s)": 0.462446
|
|
},
|
|
{
|
|
"epoch": 2.8581818181818184,
|
|
"grad_norm": 4.358846664428711,
|
|
"learning_rate": 5.849560556042577e-08,
|
|
"loss": 0.5151421070098877,
|
|
"memory(GiB)": 64.44,
|
|
"step": 3930,
|
|
"token_acc": 0.8831139355424421,
|
|
"train_speed(iter/s)": 0.462481
|
|
},
|
|
{
|
|
"epoch": 2.861818181818182,
|
|
"grad_norm": 4.679753303527832,
|
|
"learning_rate": 5.5539784471901335e-08,
|
|
"loss": 0.5724036693572998,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3935,
|
|
"token_acc": 0.8513093652907234,
|
|
"train_speed(iter/s)": 0.462487
|
|
},
|
|
{
|
|
"epoch": 2.8654545454545453,
|
|
"grad_norm": 3.894798755645752,
|
|
"learning_rate": 5.2660174965983305e-08,
|
|
"loss": 0.5823453903198242,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3940,
|
|
"token_acc": 0.8236692015209125,
|
|
"train_speed(iter/s)": 0.462532
|
|
},
|
|
{
|
|
"epoch": 2.869090909090909,
|
|
"grad_norm": 3.816575288772583,
|
|
"learning_rate": 4.985682142761572e-08,
|
|
"loss": 0.5283288955688477,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3945,
|
|
"token_acc": 0.8284527972027972,
|
|
"train_speed(iter/s)": 0.46255
|
|
},
|
|
{
|
|
"epoch": 2.8727272727272726,
|
|
"grad_norm": 4.2328901290893555,
|
|
"learning_rate": 4.712976706637229e-08,
|
|
"loss": 0.5628486156463623,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3950,
|
|
"token_acc": 0.8680026138096275,
|
|
"train_speed(iter/s)": 0.462573
|
|
},
|
|
{
|
|
"epoch": 2.8763636363636365,
|
|
"grad_norm": 3.2980473041534424,
|
|
"learning_rate": 4.4479053915784664e-08,
|
|
"loss": 0.5245370864868164,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3955,
|
|
"token_acc": 0.8287955699123212,
|
|
"train_speed(iter/s)": 0.462531
|
|
},
|
|
{
|
|
"epoch": 2.88,
|
|
"grad_norm": 4.490112781524658,
|
|
"learning_rate": 4.190472283269742e-08,
|
|
"loss": 0.6101532936096191,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3960,
|
|
"token_acc": 0.8576118178087813,
|
|
"train_speed(iter/s)": 0.462573
|
|
},
|
|
{
|
|
"epoch": 2.8836363636363638,
|
|
"grad_norm": 3.7872161865234375,
|
|
"learning_rate": 3.940681349663911e-08,
|
|
"loss": 0.532585334777832,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3965,
|
|
"token_acc": 0.8381294964028777,
|
|
"train_speed(iter/s)": 0.46259
|
|
},
|
|
{
|
|
"epoch": 2.887272727272727,
|
|
"grad_norm": 3.914175033569336,
|
|
"learning_rate": 3.69853644092083e-08,
|
|
"loss": 0.5524998664855957,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3970,
|
|
"token_acc": 0.8546465448768864,
|
|
"train_speed(iter/s)": 0.462616
|
|
},
|
|
{
|
|
"epoch": 2.8909090909090907,
|
|
"grad_norm": 3.7957468032836914,
|
|
"learning_rate": 3.4640412893480744e-08,
|
|
"loss": 0.5505013465881348,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3975,
|
|
"token_acc": 0.8538713195201745,
|
|
"train_speed(iter/s)": 0.462637
|
|
},
|
|
{
|
|
"epoch": 2.8945454545454545,
|
|
"grad_norm": 4.858002662658691,
|
|
"learning_rate": 3.2371995093435915e-08,
|
|
"loss": 0.5586944580078125,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3980,
|
|
"token_acc": 0.8747870528109029,
|
|
"train_speed(iter/s)": 0.462665
|
|
},
|
|
{
|
|
"epoch": 2.8981818181818184,
|
|
"grad_norm": 3.603170871734619,
|
|
"learning_rate": 3.018014597339913e-08,
|
|
"loss": 0.538426685333252,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3985,
|
|
"token_acc": 0.8521528998242531,
|
|
"train_speed(iter/s)": 0.462638
|
|
},
|
|
{
|
|
"epoch": 2.901818181818182,
|
|
"grad_norm": 4.4803853034973145,
|
|
"learning_rate": 2.8064899317499227e-08,
|
|
"loss": 0.5476878643035888,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3990,
|
|
"token_acc": 0.8301707779886148,
|
|
"train_speed(iter/s)": 0.462675
|
|
},
|
|
{
|
|
"epoch": 2.9054545454545453,
|
|
"grad_norm": 3.489401340484619,
|
|
"learning_rate": 2.6026287729153943e-08,
|
|
"loss": 0.5317276954650879,
|
|
"memory(GiB)": 64.47,
|
|
"step": 3995,
|
|
"token_acc": 0.8414748268762867,
|
|
"train_speed(iter/s)": 0.462658
|
|
},
|
|
{
|
|
"epoch": 2.909090909090909,
|
|
"grad_norm": 4.480035781860352,
|
|
"learning_rate": 2.4064342630562566e-08,
|
|
"loss": 0.5574604988098144,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4000,
|
|
"token_acc": 0.8581411600279525,
|
|
"train_speed(iter/s)": 0.462665
|
|
},
|
|
{
|
|
"epoch": 2.9127272727272726,
|
|
"grad_norm": 3.872577428817749,
|
|
"learning_rate": 2.2179094262223533e-08,
|
|
"loss": 0.6005492210388184,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4005,
|
|
"token_acc": 0.8393406593406594,
|
|
"train_speed(iter/s)": 0.462673
|
|
},
|
|
{
|
|
"epoch": 2.9163636363636365,
|
|
"grad_norm": 3.6794826984405518,
|
|
"learning_rate": 2.0370571682467588e-08,
|
|
"loss": 0.565433120727539,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4010,
|
|
"token_acc": 0.8388540734109221,
|
|
"train_speed(iter/s)": 0.462676
|
|
},
|
|
{
|
|
"epoch": 2.92,
|
|
"grad_norm": 3.7139086723327637,
|
|
"learning_rate": 1.863880276701091e-08,
|
|
"loss": 0.471858549118042,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4015,
|
|
"token_acc": 0.8674345958609918,
|
|
"train_speed(iter/s)": 0.462673
|
|
},
|
|
{
|
|
"epoch": 2.923636363636364,
|
|
"grad_norm": 3.3604671955108643,
|
|
"learning_rate": 1.6983814208523797e-08,
|
|
"loss": 0.54185791015625,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4020,
|
|
"token_acc": 0.8667071197411004,
|
|
"train_speed(iter/s)": 0.462647
|
|
},
|
|
{
|
|
"epoch": 2.9272727272727272,
|
|
"grad_norm": 3.7538976669311523,
|
|
"learning_rate": 1.540563151622154e-08,
|
|
"loss": 0.51304030418396,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4025,
|
|
"token_acc": 0.8918128654970761,
|
|
"train_speed(iter/s)": 0.462672
|
|
},
|
|
{
|
|
"epoch": 2.9309090909090907,
|
|
"grad_norm": 4.008846282958984,
|
|
"learning_rate": 1.3904279015469202e-08,
|
|
"loss": 0.5321861743927002,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4030,
|
|
"token_acc": 0.87595850248083,
|
|
"train_speed(iter/s)": 0.462678
|
|
},
|
|
{
|
|
"epoch": 2.9345454545454546,
|
|
"grad_norm": 4.978183269500732,
|
|
"learning_rate": 1.2479779847408001e-08,
|
|
"loss": 0.5494882583618164,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4035,
|
|
"token_acc": 0.8528012684989429,
|
|
"train_speed(iter/s)": 0.462721
|
|
},
|
|
{
|
|
"epoch": 2.9381818181818184,
|
|
"grad_norm": 4.310401439666748,
|
|
"learning_rate": 1.1132155968597291e-08,
|
|
"loss": 0.550199031829834,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4040,
|
|
"token_acc": 0.8362573099415205,
|
|
"train_speed(iter/s)": 0.462766
|
|
},
|
|
{
|
|
"epoch": 2.941818181818182,
|
|
"grad_norm": 3.632244348526001,
|
|
"learning_rate": 9.861428150677032e-09,
|
|
"loss": 0.5241313934326172,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4045,
|
|
"token_acc": 0.8391878172588833,
|
|
"train_speed(iter/s)": 0.462773
|
|
},
|
|
{
|
|
"epoch": 2.9454545454545453,
|
|
"grad_norm": 4.239510536193848,
|
|
"learning_rate": 8.667615980048061e-09,
|
|
"loss": 0.5985838890075683,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4050,
|
|
"token_acc": 0.8517462580185318,
|
|
"train_speed(iter/s)": 0.462769
|
|
},
|
|
{
|
|
"epoch": 2.949090909090909,
|
|
"grad_norm": 3.496842622756958,
|
|
"learning_rate": 7.55073785756899e-09,
|
|
"loss": 0.5512666702270508,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4055,
|
|
"token_acc": 0.8383963628848936,
|
|
"train_speed(iter/s)": 0.462776
|
|
},
|
|
{
|
|
"epoch": 2.9527272727272726,
|
|
"grad_norm": 3.777428150177002,
|
|
"learning_rate": 6.51081099827422e-09,
|
|
"loss": 0.5256683349609375,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4060,
|
|
"token_acc": 0.837874659400545,
|
|
"train_speed(iter/s)": 0.462788
|
|
},
|
|
{
|
|
"epoch": 2.9563636363636365,
|
|
"grad_norm": 4.25238037109375,
|
|
"learning_rate": 5.547851431105811e-09,
|
|
"loss": 0.5643882751464844,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4065,
|
|
"token_acc": 0.8642220019821606,
|
|
"train_speed(iter/s)": 0.462802
|
|
},
|
|
{
|
|
"epoch": 2.96,
|
|
"grad_norm": 4.666615962982178,
|
|
"learning_rate": 4.661873998670907e-09,
|
|
"loss": 0.5404569625854492,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4070,
|
|
"token_acc": 0.908675799086758,
|
|
"train_speed(iter/s)": 0.462793
|
|
},
|
|
{
|
|
"epoch": 2.963636363636364,
|
|
"grad_norm": 3.5408148765563965,
|
|
"learning_rate": 3.852892357006921e-09,
|
|
"loss": 0.5230738639831543,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4075,
|
|
"token_acc": 0.8577827547592385,
|
|
"train_speed(iter/s)": 0.46281
|
|
},
|
|
{
|
|
"epoch": 2.9672727272727273,
|
|
"grad_norm": 4.09399938583374,
|
|
"learning_rate": 3.120918975377252e-09,
|
|
"loss": 0.5126709938049316,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4080,
|
|
"token_acc": 0.8629802805448262,
|
|
"train_speed(iter/s)": 0.462828
|
|
},
|
|
{
|
|
"epoch": 2.9709090909090907,
|
|
"grad_norm": 4.4036149978637695,
|
|
"learning_rate": 2.465965136074222e-09,
|
|
"loss": 0.5580806732177734,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4085,
|
|
"token_acc": 0.8440065681444991,
|
|
"train_speed(iter/s)": 0.462857
|
|
},
|
|
{
|
|
"epoch": 2.9745454545454546,
|
|
"grad_norm": 4.547696113586426,
|
|
"learning_rate": 1.8880409342492134e-09,
|
|
"loss": 0.5888032913208008,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4090,
|
|
"token_acc": 0.8312101910828026,
|
|
"train_speed(iter/s)": 0.462889
|
|
},
|
|
{
|
|
"epoch": 2.978181818181818,
|
|
"grad_norm": 3.8731865882873535,
|
|
"learning_rate": 1.3871552777527941e-09,
|
|
"loss": 0.5509771823883056,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4095,
|
|
"token_acc": 0.8560301507537689,
|
|
"train_speed(iter/s)": 0.462928
|
|
},
|
|
{
|
|
"epoch": 2.981818181818182,
|
|
"grad_norm": 4.096609115600586,
|
|
"learning_rate": 9.633158870014924e-10,
|
|
"loss": 0.5540934085845948,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4100,
|
|
"token_acc": 0.8637532133676092,
|
|
"train_speed(iter/s)": 0.462962
|
|
},
|
|
{
|
|
"epoch": 2.9854545454545454,
|
|
"grad_norm": 3.3930113315582275,
|
|
"learning_rate": 6.165292948556723e-10,
|
|
"loss": 0.5454161643981934,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4105,
|
|
"token_acc": 0.8733822260569456,
|
|
"train_speed(iter/s)": 0.462982
|
|
},
|
|
{
|
|
"epoch": 2.9890909090909092,
|
|
"grad_norm": 4.123258113861084,
|
|
"learning_rate": 3.468008465207229e-10,
|
|
"loss": 0.5773702621459961,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4110,
|
|
"token_acc": 0.8334710743801653,
|
|
"train_speed(iter/s)": 0.462997
|
|
},
|
|
{
|
|
"epoch": 2.9927272727272727,
|
|
"grad_norm": 4.4657416343688965,
|
|
"learning_rate": 1.5413469946434778e-10,
|
|
"loss": 0.5174553871154786,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4115,
|
|
"token_acc": 0.8584444444444445,
|
|
"train_speed(iter/s)": 0.462982
|
|
},
|
|
{
|
|
"epoch": 2.996363636363636,
|
|
"grad_norm": 4.009359836578369,
|
|
"learning_rate": 3.8533823351616376e-11,
|
|
"loss": 0.5238511085510253,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4120,
|
|
"token_acc": 0.8342233009708738,
|
|
"train_speed(iter/s)": 0.463014
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"grad_norm": 3.5021286010742188,
|
|
"learning_rate": 0.0,
|
|
"loss": 0.5140440940856934,
|
|
"memory(GiB)": 64.47,
|
|
"step": 4125,
|
|
"token_acc": 0.8927929726492314,
|
|
"train_speed(iter/s)": 0.46301
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4125,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 3,
|
|
"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": 1.4148570216262533e+18,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|