252a46083a
Model: pre-to-post-olmo/math-1b-sft-numinamath-bs512-from-step80000 Source: Original Platform
625 lines
15 KiB
JSON
625 lines
15 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1679,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.011911852293031567,
|
|
"grad_norm": 2.4980649948120117,
|
|
"learning_rate": 3.7254901960784316e-06,
|
|
"loss": 1.9374,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.023823704586063133,
|
|
"grad_norm": 0.7357017397880554,
|
|
"learning_rate": 7.647058823529411e-06,
|
|
"loss": 1.5063,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.0357355568790947,
|
|
"grad_norm": 0.28605589270591736,
|
|
"learning_rate": 9.999404197526844e-06,
|
|
"loss": 0.9789,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.047647409172126266,
|
|
"grad_norm": 0.16456548869609833,
|
|
"learning_rate": 9.99270305028903e-06,
|
|
"loss": 0.7814,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.05955926146515783,
|
|
"grad_norm": 0.12905146181583405,
|
|
"learning_rate": 9.978566016361393e-06,
|
|
"loss": 0.7043,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.0714711137581894,
|
|
"grad_norm": 0.11757974326610565,
|
|
"learning_rate": 9.95701415074788e-06,
|
|
"loss": 0.667,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.08338296605122096,
|
|
"grad_norm": 0.10049410909414291,
|
|
"learning_rate": 9.928079551738542e-06,
|
|
"loss": 0.6471,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.09529481834425253,
|
|
"grad_norm": 0.10536422580480576,
|
|
"learning_rate": 9.891805313103928e-06,
|
|
"loss": 0.6362,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.1072066706372841,
|
|
"grad_norm": 0.09555111825466156,
|
|
"learning_rate": 9.848245459913323e-06,
|
|
"loss": 0.6173,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.11911852293031566,
|
|
"grad_norm": 0.09140878915786743,
|
|
"learning_rate": 9.797464868072489e-06,
|
|
"loss": 0.6172,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.13103037522334723,
|
|
"grad_norm": 0.08511078357696533,
|
|
"learning_rate": 9.739539167700649e-06,
|
|
"loss": 0.6081,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.1429422275163788,
|
|
"grad_norm": 0.08799832314252853,
|
|
"learning_rate": 9.674554630490727e-06,
|
|
"loss": 0.5972,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.15485407980941035,
|
|
"grad_norm": 0.08326159417629242,
|
|
"learning_rate": 9.60260804122051e-06,
|
|
"loss": 0.5892,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.16676593210244192,
|
|
"grad_norm": 0.0890437588095665,
|
|
"learning_rate": 9.523806553606157e-06,
|
|
"loss": 0.5889,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.1786777843954735,
|
|
"grad_norm": 0.08911486715078354,
|
|
"learning_rate": 9.438267530712727e-06,
|
|
"loss": 0.5879,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.19058963668850507,
|
|
"grad_norm": 0.08892590552568436,
|
|
"learning_rate": 9.34611837015938e-06,
|
|
"loss": 0.5824,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.20250148898153664,
|
|
"grad_norm": 0.09178025275468826,
|
|
"learning_rate": 9.247496314379634e-06,
|
|
"loss": 0.5799,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.2144133412745682,
|
|
"grad_norm": 0.082877516746521,
|
|
"learning_rate": 9.142548246219212e-06,
|
|
"loss": 0.5694,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.22632519356759975,
|
|
"grad_norm": 0.081816166639328,
|
|
"learning_rate": 9.031430470175958e-06,
|
|
"loss": 0.5702,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.23823704586063132,
|
|
"grad_norm": 0.08267427980899811,
|
|
"learning_rate": 8.914308479607592e-06,
|
|
"loss": 0.5622,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.25014889815366287,
|
|
"grad_norm": 0.08150000870227814,
|
|
"learning_rate": 8.791356710254026e-06,
|
|
"loss": 0.561,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.26206075044669447,
|
|
"grad_norm": 0.08462639898061752,
|
|
"learning_rate": 8.662758280441353e-06,
|
|
"loss": 0.5626,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.273972602739726,
|
|
"grad_norm": 0.08133713901042938,
|
|
"learning_rate": 8.528704718354392e-06,
|
|
"loss": 0.5624,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.2858844550327576,
|
|
"grad_norm": 0.08033239841461182,
|
|
"learning_rate": 8.389395676784025e-06,
|
|
"loss": 0.5592,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.29779630732578916,
|
|
"grad_norm": 0.07994557917118073,
|
|
"learning_rate": 8.245038635774144e-06,
|
|
"loss": 0.5564,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.3097081596188207,
|
|
"grad_norm": 0.08364811539649963,
|
|
"learning_rate": 8.095848593611045e-06,
|
|
"loss": 0.5525,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.3216200119118523,
|
|
"grad_norm": 0.08064120262861252,
|
|
"learning_rate": 7.942047746615561e-06,
|
|
"loss": 0.553,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.33353186420488384,
|
|
"grad_norm": 0.08826671540737152,
|
|
"learning_rate": 7.783865158214768e-06,
|
|
"loss": 0.5507,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.34544371649791544,
|
|
"grad_norm": 0.0798407793045044,
|
|
"learning_rate": 7.621536417786159e-06,
|
|
"loss": 0.5459,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.357355568790947,
|
|
"grad_norm": 0.08206599205732346,
|
|
"learning_rate": 7.455303289782414e-06,
|
|
"loss": 0.5535,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.36926742108397853,
|
|
"grad_norm": 0.08513249456882477,
|
|
"learning_rate": 7.285413353659285e-06,
|
|
"loss": 0.5502,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.38117927337701013,
|
|
"grad_norm": 0.08413926512002945,
|
|
"learning_rate": 7.1121196351429235e-06,
|
|
"loss": 0.5502,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.3930911256700417,
|
|
"grad_norm": 0.0805065929889679,
|
|
"learning_rate": 6.935680229385788e-06,
|
|
"loss": 0.5442,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.4050029779630733,
|
|
"grad_norm": 0.0822906345129013,
|
|
"learning_rate": 6.7563579165724e-06,
|
|
"loss": 0.549,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.4169148302561048,
|
|
"grad_norm": 0.08298345655202866,
|
|
"learning_rate": 6.574419770547455e-06,
|
|
"loss": 0.5413,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.4288266825491364,
|
|
"grad_norm": 0.08241568505764008,
|
|
"learning_rate": 6.390136761049138e-06,
|
|
"loss": 0.5398,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.44073853484216796,
|
|
"grad_norm": 0.0785905048251152,
|
|
"learning_rate": 6.2037833501401236e-06,
|
|
"loss": 0.544,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.4526503871351995,
|
|
"grad_norm": 0.08352043479681015,
|
|
"learning_rate": 6.015637083437249e-06,
|
|
"loss": 0.5412,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.4645622394282311,
|
|
"grad_norm": 0.07776437699794769,
|
|
"learning_rate": 5.825978176748709e-06,
|
|
"loss": 0.5424,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.47647409172126265,
|
|
"grad_norm": 0.07624149322509766,
|
|
"learning_rate": 5.635089098734394e-06,
|
|
"loss": 0.5405,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.48838594401429425,
|
|
"grad_norm": 0.07820378243923187,
|
|
"learning_rate": 5.443254150210946e-06,
|
|
"loss": 0.5453,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.5002977963073257,
|
|
"grad_norm": 0.0802140012383461,
|
|
"learning_rate": 5.250759040728082e-06,
|
|
"loss": 0.5381,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.5122096486003573,
|
|
"grad_norm": 0.07368727028369904,
|
|
"learning_rate": 5.057890463046825e-06,
|
|
"loss": 0.5404,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.5241215008933889,
|
|
"grad_norm": 0.07918524742126465,
|
|
"learning_rate": 4.864935666153388e-06,
|
|
"loss": 0.5403,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.5360333531864205,
|
|
"grad_norm": 0.07988528162240982,
|
|
"learning_rate": 4.672182027444649e-06,
|
|
"loss": 0.5332,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.547945205479452,
|
|
"grad_norm": 0.07852893322706223,
|
|
"learning_rate": 4.4799166247223576e-06,
|
|
"loss": 0.5344,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.5598570577724836,
|
|
"grad_norm": 0.08311598002910614,
|
|
"learning_rate": 4.2884258086335755e-06,
|
|
"loss": 0.534,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.5717689100655152,
|
|
"grad_norm": 0.08012568205595016,
|
|
"learning_rate": 4.09799477619407e-06,
|
|
"loss": 0.5358,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.5836807623585467,
|
|
"grad_norm": 0.08017480373382568,
|
|
"learning_rate": 3.908907146029905e-06,
|
|
"loss": 0.53,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.5955926146515783,
|
|
"grad_norm": 0.07982592284679413,
|
|
"learning_rate": 3.721444535969774e-06,
|
|
"loss": 0.5377,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.6075044669446099,
|
|
"grad_norm": 0.07628551125526428,
|
|
"learning_rate": 3.5358861436172487e-06,
|
|
"loss": 0.5331,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.6194163192376414,
|
|
"grad_norm": 0.0765288844704628,
|
|
"learning_rate": 3.352508330527581e-06,
|
|
"loss": 0.5279,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.631328171530673,
|
|
"grad_norm": 0.07875853776931763,
|
|
"learning_rate": 3.1715842106083704e-06,
|
|
"loss": 0.5332,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.6432400238237046,
|
|
"grad_norm": 0.076643206179142,
|
|
"learning_rate": 2.993383243357134e-06,
|
|
"loss": 0.5322,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.6551518761167362,
|
|
"grad_norm": 0.07596970349550247,
|
|
"learning_rate": 2.8181708325415603e-06,
|
|
"loss": 0.531,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.6670637284097677,
|
|
"grad_norm": 0.07834818959236145,
|
|
"learning_rate": 2.646207930920175e-06,
|
|
"loss": 0.533,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 0.6789755807027993,
|
|
"grad_norm": 0.07674740254878998,
|
|
"learning_rate": 2.4777506515921222e-06,
|
|
"loss": 0.5346,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 0.6908874329958309,
|
|
"grad_norm": 0.07472624629735947,
|
|
"learning_rate": 2.31304988655489e-06,
|
|
"loss": 0.5288,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 0.7027992852888624,
|
|
"grad_norm": 0.0729040876030922,
|
|
"learning_rate": 2.152350933038097e-06,
|
|
"loss": 0.5246,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 0.714711137581894,
|
|
"grad_norm": 0.07641004025936127,
|
|
"learning_rate": 1.995893128169846e-06,
|
|
"loss": 0.5284,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.7266229898749256,
|
|
"grad_norm": 0.08119594305753708,
|
|
"learning_rate": 1.8439094925197593e-06,
|
|
"loss": 0.5258,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 0.7385348421679571,
|
|
"grad_norm": 0.07687598466873169,
|
|
"learning_rate": 1.6966263830495939e-06,
|
|
"loss": 0.5317,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 0.7504466944609887,
|
|
"grad_norm": 0.07267104089260101,
|
|
"learning_rate": 1.554263155988287e-06,
|
|
"loss": 0.5334,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 0.7623585467540203,
|
|
"grad_norm": 0.07765422761440277,
|
|
"learning_rate": 1.4170318401335858e-06,
|
|
"loss": 0.5252,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 0.7742703990470519,
|
|
"grad_norm": 0.07622691988945007,
|
|
"learning_rate": 1.2851368210667426e-06,
|
|
"loss": 0.5286,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.7861822513400833,
|
|
"grad_norm": 0.07585112005472183,
|
|
"learning_rate": 1.158774536750698e-06,
|
|
"loss": 0.5278,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 0.798094103633115,
|
|
"grad_norm": 0.07797589153051376,
|
|
"learning_rate": 1.038133184965001e-06,
|
|
"loss": 0.5256,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 0.8100059559261465,
|
|
"grad_norm": 0.07552938163280487,
|
|
"learning_rate": 9.233924430133024e-07,
|
|
"loss": 0.5352,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 0.821917808219178,
|
|
"grad_norm": 0.07989319413900375,
|
|
"learning_rate": 8.147232001207894e-07,
|
|
"loss": 0.5335,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 0.8338296605122096,
|
|
"grad_norm": 0.07479098439216614,
|
|
"learning_rate": 7.122873029201732e-07,
|
|
"loss": 0.527,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.8457415128052412,
|
|
"grad_norm": 0.07334813475608826,
|
|
"learning_rate": 6.162373144052636e-07,
|
|
"loss": 0.5329,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 0.8576533650982728,
|
|
"grad_norm": 0.07817459106445312,
|
|
"learning_rate": 5.267162867111387e-07,
|
|
"loss": 0.5311,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 0.8695652173913043,
|
|
"grad_norm": 0.07533743977546692,
|
|
"learning_rate": 4.43857548059321e-07,
|
|
"loss": 0.5249,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 0.8814770696843359,
|
|
"grad_norm": 0.07640492916107178,
|
|
"learning_rate": 3.6778450418527735e-07,
|
|
"loss": 0.5349,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 0.8933889219773675,
|
|
"grad_norm": 0.07609718292951584,
|
|
"learning_rate": 2.9861045454396943e-07,
|
|
"loss": 0.531,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.905300774270399,
|
|
"grad_norm": 0.07761748880147934,
|
|
"learning_rate": 2.3643842356722124e-07,
|
|
"loss": 0.5377,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 0.9172126265634306,
|
|
"grad_norm": 0.07907618582248688,
|
|
"learning_rate": 1.8136100722418003e-07,
|
|
"loss": 0.523,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 0.9291244788564622,
|
|
"grad_norm": 0.07826044410467148,
|
|
"learning_rate": 1.3346023511343842e-07,
|
|
"loss": 0.5285,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 0.9410363311494937,
|
|
"grad_norm": 0.07943511754274368,
|
|
"learning_rate": 9.280744829216982e-08,
|
|
"loss": 0.5363,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 0.9529481834425253,
|
|
"grad_norm": 0.07584235072135925,
|
|
"learning_rate": 5.9463193024269505e-08,
|
|
"loss": 0.5273,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.9648600357355569,
|
|
"grad_norm": 0.07421740889549255,
|
|
"learning_rate": 3.347713060571855e-08,
|
|
"loss": 0.5307,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 0.9767718880285885,
|
|
"grad_norm": 0.07467110455036163,
|
|
"learning_rate": 1.4887963401492011e-08,
|
|
"loss": 0.529,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 0.98868374032162,
|
|
"grad_norm": 0.0778217762708664,
|
|
"learning_rate": 3.7233772041533888e-09,
|
|
"loss": 0.5298,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"step": 1679,
|
|
"total_flos": 9.42165996526266e+18,
|
|
"train_loss": 0.5869280341979363,
|
|
"train_runtime": 4960.6559,
|
|
"train_samples_per_second": 173.262,
|
|
"train_steps_per_second": 0.338
|
|
}
|
|
],
|
|
"logging_steps": 20,
|
|
"max_steps": 1679,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 9.42165996526266e+18,
|
|
"train_batch_size": 32,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|