Files
math-1b-sft-numinamath-bs51…/trainer_state.json

625 lines
15 KiB
JSON
Raw Normal View History

{
"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": 2.319519281387329,
"learning_rate": 3.7254901960784316e-06,
"loss": 1.6734,
"step": 20
},
{
"epoch": 0.023823704586063133,
"grad_norm": 0.5233764052391052,
"learning_rate": 7.647058823529411e-06,
"loss": 1.1925,
"step": 40
},
{
"epoch": 0.0357355568790947,
"grad_norm": 0.19094516336917877,
"learning_rate": 9.999404197526844e-06,
"loss": 0.798,
"step": 60
},
{
"epoch": 0.047647409172126266,
"grad_norm": 0.12445113062858582,
"learning_rate": 9.99270305028903e-06,
"loss": 0.6976,
"step": 80
},
{
"epoch": 0.05955926146515783,
"grad_norm": 0.11021372675895691,
"learning_rate": 9.978566016361393e-06,
"loss": 0.663,
"step": 100
},
{
"epoch": 0.0714711137581894,
"grad_norm": 0.10500118881464005,
"learning_rate": 9.95701415074788e-06,
"loss": 0.6451,
"step": 120
},
{
"epoch": 0.08338296605122096,
"grad_norm": 0.09714240580797195,
"learning_rate": 9.928079551738542e-06,
"loss": 0.6341,
"step": 140
},
{
"epoch": 0.09529481834425253,
"grad_norm": 0.10310710966587067,
"learning_rate": 9.891805313103928e-06,
"loss": 0.6293,
"step": 160
},
{
"epoch": 0.1072066706372841,
"grad_norm": 0.09910018742084503,
"learning_rate": 9.848245459913323e-06,
"loss": 0.6142,
"step": 180
},
{
"epoch": 0.11911852293031566,
"grad_norm": 0.09675974398851395,
"learning_rate": 9.797464868072489e-06,
"loss": 0.6166,
"step": 200
},
{
"epoch": 0.13103037522334723,
"grad_norm": 0.09194370359182358,
"learning_rate": 9.739539167700649e-06,
"loss": 0.6094,
"step": 220
},
{
"epoch": 0.1429422275163788,
"grad_norm": 0.09696715325117111,
"learning_rate": 9.674554630490727e-06,
"loss": 0.5995,
"step": 240
},
{
"epoch": 0.15485407980941035,
"grad_norm": 0.08973082154989243,
"learning_rate": 9.60260804122051e-06,
"loss": 0.5949,
"step": 260
},
{
"epoch": 0.16676593210244192,
"grad_norm": 0.09478715062141418,
"learning_rate": 9.523806553606157e-06,
"loss": 0.5945,
"step": 280
},
{
"epoch": 0.1786777843954735,
"grad_norm": 0.09915424883365631,
"learning_rate": 9.438267530712727e-06,
"loss": 0.5944,
"step": 300
},
{
"epoch": 0.19058963668850507,
"grad_norm": 0.09569481760263443,
"learning_rate": 9.34611837015938e-06,
"loss": 0.59,
"step": 320
},
{
"epoch": 0.20250148898153664,
"grad_norm": 0.09609861671924591,
"learning_rate": 9.247496314379634e-06,
"loss": 0.5876,
"step": 340
},
{
"epoch": 0.2144133412745682,
"grad_norm": 0.09400948882102966,
"learning_rate": 9.142548246219212e-06,
"loss": 0.5788,
"step": 360
},
{
"epoch": 0.22632519356759975,
"grad_norm": 0.08953362703323364,
"learning_rate": 9.031430470175958e-06,
"loss": 0.5799,
"step": 380
},
{
"epoch": 0.23823704586063132,
"grad_norm": 0.09138830751180649,
"learning_rate": 8.914308479607592e-06,
"loss": 0.5722,
"step": 400
},
{
"epoch": 0.25014889815366287,
"grad_norm": 0.09195993840694427,
"learning_rate": 8.791356710254026e-06,
"loss": 0.5715,
"step": 420
},
{
"epoch": 0.26206075044669447,
"grad_norm": 0.0933992937207222,
"learning_rate": 8.662758280441353e-06,
"loss": 0.5735,
"step": 440
},
{
"epoch": 0.273972602739726,
"grad_norm": 0.09093426913022995,
"learning_rate": 8.528704718354392e-06,
"loss": 0.5728,
"step": 460
},
{
"epoch": 0.2858844550327576,
"grad_norm": 0.09080294519662857,
"learning_rate": 8.389395676784025e-06,
"loss": 0.5702,
"step": 480
},
{
"epoch": 0.29779630732578916,
"grad_norm": 0.09120159596204758,
"learning_rate": 8.245038635774144e-06,
"loss": 0.5677,
"step": 500
},
{
"epoch": 0.3097081596188207,
"grad_norm": 0.09126840531826019,
"learning_rate": 8.095848593611045e-06,
"loss": 0.5643,
"step": 520
},
{
"epoch": 0.3216200119118523,
"grad_norm": 0.09190158545970917,
"learning_rate": 7.942047746615561e-06,
"loss": 0.5648,
"step": 540
},
{
"epoch": 0.33353186420488384,
"grad_norm": 0.09290704131126404,
"learning_rate": 7.783865158214768e-06,
"loss": 0.5623,
"step": 560
},
{
"epoch": 0.34544371649791544,
"grad_norm": 0.09177636355161667,
"learning_rate": 7.621536417786159e-06,
"loss": 0.5576,
"step": 580
},
{
"epoch": 0.357355568790947,
"grad_norm": 0.09153669327497482,
"learning_rate": 7.455303289782414e-06,
"loss": 0.5664,
"step": 600
},
{
"epoch": 0.36926742108397853,
"grad_norm": 0.08856385946273804,
"learning_rate": 7.285413353659285e-06,
"loss": 0.562,
"step": 620
},
{
"epoch": 0.38117927337701013,
"grad_norm": 0.08958728611469269,
"learning_rate": 7.1121196351429235e-06,
"loss": 0.5631,
"step": 640
},
{
"epoch": 0.3930911256700417,
"grad_norm": 0.08628527820110321,
"learning_rate": 6.935680229385788e-06,
"loss": 0.5567,
"step": 660
},
{
"epoch": 0.4050029779630733,
"grad_norm": 0.09060859680175781,
"learning_rate": 6.7563579165724e-06,
"loss": 0.5618,
"step": 680
},
{
"epoch": 0.4169148302561048,
"grad_norm": 0.0899747982621193,
"learning_rate": 6.574419770547455e-06,
"loss": 0.5539,
"step": 700
},
{
"epoch": 0.4288266825491364,
"grad_norm": 0.08697455376386642,
"learning_rate": 6.390136761049138e-06,
"loss": 0.5528,
"step": 720
},
{
"epoch": 0.44073853484216796,
"grad_norm": 0.08819800615310669,
"learning_rate": 6.2037833501401236e-06,
"loss": 0.5567,
"step": 740
},
{
"epoch": 0.4526503871351995,
"grad_norm": 0.09015385061502457,
"learning_rate": 6.015637083437249e-06,
"loss": 0.5541,
"step": 760
},
{
"epoch": 0.4645622394282311,
"grad_norm": 0.08507155627012253,
"learning_rate": 5.825978176748709e-06,
"loss": 0.5551,
"step": 780
},
{
"epoch": 0.47647409172126265,
"grad_norm": 0.08560214191675186,
"learning_rate": 5.635089098734394e-06,
"loss": 0.5533,
"step": 800
},
{
"epoch": 0.48838594401429425,
"grad_norm": 0.08728593587875366,
"learning_rate": 5.443254150210946e-06,
"loss": 0.5579,
"step": 820
},
{
"epoch": 0.5002977963073257,
"grad_norm": 0.09012866020202637,
"learning_rate": 5.250759040728082e-06,
"loss": 0.5511,
"step": 840
},
{
"epoch": 0.5122096486003573,
"grad_norm": 0.08543788641691208,
"learning_rate": 5.057890463046825e-06,
"loss": 0.5536,
"step": 860
},
{
"epoch": 0.5241215008933889,
"grad_norm": 0.08579913526773453,
"learning_rate": 4.864935666153388e-06,
"loss": 0.5532,
"step": 880
},
{
"epoch": 0.5360333531864205,
"grad_norm": 0.09294642508029938,
"learning_rate": 4.672182027444649e-06,
"loss": 0.5465,
"step": 900
},
{
"epoch": 0.547945205479452,
"grad_norm": 0.0893787369132042,
"learning_rate": 4.4799166247223576e-06,
"loss": 0.5477,
"step": 920
},
{
"epoch": 0.5598570577724836,
"grad_norm": 0.09061983972787857,
"learning_rate": 4.2884258086335755e-06,
"loss": 0.5471,
"step": 940
},
{
"epoch": 0.5717689100655152,
"grad_norm": 0.08709097653627396,
"learning_rate": 4.09799477619407e-06,
"loss": 0.5487,
"step": 960
},
{
"epoch": 0.5836807623585467,
"grad_norm": 0.09192101657390594,
"learning_rate": 3.908907146029905e-06,
"loss": 0.5433,
"step": 980
},
{
"epoch": 0.5955926146515783,
"grad_norm": 0.08835067600011826,
"learning_rate": 3.721444535969774e-06,
"loss": 0.5508,
"step": 1000
},
{
"epoch": 0.6075044669446099,
"grad_norm": 0.08473947644233704,
"learning_rate": 3.5358861436172487e-06,
"loss": 0.5464,
"step": 1020
},
{
"epoch": 0.6194163192376414,
"grad_norm": 0.08449548482894897,
"learning_rate": 3.352508330527581e-06,
"loss": 0.5416,
"step": 1040
},
{
"epoch": 0.631328171530673,
"grad_norm": 0.08983448892831802,
"learning_rate": 3.1715842106083704e-06,
"loss": 0.5468,
"step": 1060
},
{
"epoch": 0.6432400238237046,
"grad_norm": 0.08712620288133621,
"learning_rate": 2.993383243357134e-06,
"loss": 0.5456,
"step": 1080
},
{
"epoch": 0.6551518761167362,
"grad_norm": 0.08562883734703064,
"learning_rate": 2.8181708325415603e-06,
"loss": 0.5446,
"step": 1100
},
{
"epoch": 0.6670637284097677,
"grad_norm": 0.08872649073600769,
"learning_rate": 2.646207930920175e-06,
"loss": 0.5463,
"step": 1120
},
{
"epoch": 0.6789755807027993,
"grad_norm": 0.08730657398700714,
"learning_rate": 2.4777506515921222e-06,
"loss": 0.548,
"step": 1140
},
{
"epoch": 0.6908874329958309,
"grad_norm": 0.08714567869901657,
"learning_rate": 2.31304988655489e-06,
"loss": 0.5421,
"step": 1160
},
{
"epoch": 0.7027992852888624,
"grad_norm": 0.0823967382311821,
"learning_rate": 2.152350933038097e-06,
"loss": 0.538,
"step": 1180
},
{
"epoch": 0.714711137581894,
"grad_norm": 0.08490780740976334,
"learning_rate": 1.995893128169846e-06,
"loss": 0.5417,
"step": 1200
},
{
"epoch": 0.7266229898749256,
"grad_norm": 0.08957539498806,
"learning_rate": 1.8439094925197593e-06,
"loss": 0.5396,
"step": 1220
},
{
"epoch": 0.7385348421679571,
"grad_norm": 0.08784908056259155,
"learning_rate": 1.6966263830495939e-06,
"loss": 0.5455,
"step": 1240
},
{
"epoch": 0.7504466944609887,
"grad_norm": 0.08189205080270767,
"learning_rate": 1.554263155988287e-06,
"loss": 0.5472,
"step": 1260
},
{
"epoch": 0.7623585467540203,
"grad_norm": 0.08474428951740265,
"learning_rate": 1.4170318401335858e-06,
"loss": 0.5386,
"step": 1280
},
{
"epoch": 0.7742703990470519,
"grad_norm": 0.08598953485488892,
"learning_rate": 1.2851368210667426e-06,
"loss": 0.5424,
"step": 1300
},
{
"epoch": 0.7861822513400833,
"grad_norm": 0.08535321056842804,
"learning_rate": 1.158774536750698e-06,
"loss": 0.5412,
"step": 1320
},
{
"epoch": 0.798094103633115,
"grad_norm": 0.08556421101093292,
"learning_rate": 1.038133184965001e-06,
"loss": 0.5388,
"step": 1340
},
{
"epoch": 0.8100059559261465,
"grad_norm": 0.08341170847415924,
"learning_rate": 9.233924430133024e-07,
"loss": 0.5485,
"step": 1360
},
{
"epoch": 0.821917808219178,
"grad_norm": 0.08838168531656265,
"learning_rate": 8.147232001207894e-07,
"loss": 0.5469,
"step": 1380
},
{
"epoch": 0.8338296605122096,
"grad_norm": 0.08617264032363892,
"learning_rate": 7.122873029201732e-07,
"loss": 0.5403,
"step": 1400
},
{
"epoch": 0.8457415128052412,
"grad_norm": 0.0815279632806778,
"learning_rate": 6.162373144052636e-07,
"loss": 0.5465,
"step": 1420
},
{
"epoch": 0.8576533650982728,
"grad_norm": 0.08794637769460678,
"learning_rate": 5.267162867111387e-07,
"loss": 0.5446,
"step": 1440
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.08446835726499557,
"learning_rate": 4.43857548059321e-07,
"loss": 0.538,
"step": 1460
},
{
"epoch": 0.8814770696843359,
"grad_norm": 0.0861380472779274,
"learning_rate": 3.6778450418527735e-07,
"loss": 0.5484,
"step": 1480
},
{
"epoch": 0.8933889219773675,
"grad_norm": 0.08714042603969574,
"learning_rate": 2.9861045454396943e-07,
"loss": 0.5447,
"step": 1500
},
{
"epoch": 0.905300774270399,
"grad_norm": 0.08558403700590134,
"learning_rate": 2.3643842356722124e-07,
"loss": 0.5516,
"step": 1520
},
{
"epoch": 0.9172126265634306,
"grad_norm": 0.08699293434619904,
"learning_rate": 1.8136100722418003e-07,
"loss": 0.5365,
"step": 1540
},
{
"epoch": 0.9291244788564622,
"grad_norm": 0.0858757421374321,
"learning_rate": 1.3346023511343842e-07,
"loss": 0.5422,
"step": 1560
},
{
"epoch": 0.9410363311494937,
"grad_norm": 0.08714671432971954,
"learning_rate": 9.280744829216982e-08,
"loss": 0.5502,
"step": 1580
},
{
"epoch": 0.9529481834425253,
"grad_norm": 0.0844811275601387,
"learning_rate": 5.9463193024269505e-08,
"loss": 0.5405,
"step": 1600
},
{
"epoch": 0.9648600357355569,
"grad_norm": 0.08579962700605392,
"learning_rate": 3.347713060571855e-08,
"loss": 0.5444,
"step": 1620
},
{
"epoch": 0.9767718880285885,
"grad_norm": 0.08677230775356293,
"learning_rate": 1.4887963401492011e-08,
"loss": 0.5426,
"step": 1640
},
{
"epoch": 0.98868374032162,
"grad_norm": 0.08511412888765335,
"learning_rate": 3.7233772041533888e-09,
"loss": 0.5434,
"step": 1660
},
{
"epoch": 1.0,
"step": 1679,
"total_flos": 9.42165996526266e+18,
"train_loss": 0.5865233610470143,
"train_runtime": 4862.2389,
"train_samples_per_second": 176.769,
"train_steps_per_second": 0.345
}
],
"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
}