hp_ablations_gemma_adambeta2_0.98 / trainer_state.json
sedrickkeh's picture
End of training
ac7bc36 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9991537376586743,
"eval_steps": 500,
"global_step": 1329,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.022566995768688293,
"grad_norm": 1.106562668405512,
"learning_rate": 5e-06,
"loss": 0.7437,
"step": 10
},
{
"epoch": 0.045133991537376586,
"grad_norm": 0.9579301203585069,
"learning_rate": 5e-06,
"loss": 0.6804,
"step": 20
},
{
"epoch": 0.06770098730606489,
"grad_norm": 0.6720710771592706,
"learning_rate": 5e-06,
"loss": 0.663,
"step": 30
},
{
"epoch": 0.09026798307475317,
"grad_norm": 0.6664805039251297,
"learning_rate": 5e-06,
"loss": 0.6406,
"step": 40
},
{
"epoch": 0.11283497884344147,
"grad_norm": 0.6367313460995002,
"learning_rate": 5e-06,
"loss": 0.634,
"step": 50
},
{
"epoch": 0.13540197461212977,
"grad_norm": 0.6120015931358381,
"learning_rate": 5e-06,
"loss": 0.6333,
"step": 60
},
{
"epoch": 0.15796897038081806,
"grad_norm": 0.6468942788509799,
"learning_rate": 5e-06,
"loss": 0.629,
"step": 70
},
{
"epoch": 0.18053596614950634,
"grad_norm": 0.679108726345772,
"learning_rate": 5e-06,
"loss": 0.6307,
"step": 80
},
{
"epoch": 0.20310296191819463,
"grad_norm": 0.680799148527619,
"learning_rate": 5e-06,
"loss": 0.6284,
"step": 90
},
{
"epoch": 0.22566995768688294,
"grad_norm": 0.7208249773354808,
"learning_rate": 5e-06,
"loss": 0.6275,
"step": 100
},
{
"epoch": 0.24823695345557123,
"grad_norm": 0.5883661778221817,
"learning_rate": 5e-06,
"loss": 0.6181,
"step": 110
},
{
"epoch": 0.27080394922425954,
"grad_norm": 0.6251345541177512,
"learning_rate": 5e-06,
"loss": 0.6175,
"step": 120
},
{
"epoch": 0.2933709449929478,
"grad_norm": 0.7385638462100499,
"learning_rate": 5e-06,
"loss": 0.6188,
"step": 130
},
{
"epoch": 0.3159379407616361,
"grad_norm": 0.7960379744739738,
"learning_rate": 5e-06,
"loss": 0.6137,
"step": 140
},
{
"epoch": 0.3385049365303244,
"grad_norm": 0.8310164398220077,
"learning_rate": 5e-06,
"loss": 0.6141,
"step": 150
},
{
"epoch": 0.3610719322990127,
"grad_norm": 0.58030004573728,
"learning_rate": 5e-06,
"loss": 0.6162,
"step": 160
},
{
"epoch": 0.383638928067701,
"grad_norm": 0.6518191185210568,
"learning_rate": 5e-06,
"loss": 0.6122,
"step": 170
},
{
"epoch": 0.40620592383638926,
"grad_norm": 0.6327708449088391,
"learning_rate": 5e-06,
"loss": 0.6089,
"step": 180
},
{
"epoch": 0.4287729196050776,
"grad_norm": 0.5822395493242641,
"learning_rate": 5e-06,
"loss": 0.6034,
"step": 190
},
{
"epoch": 0.4513399153737659,
"grad_norm": 0.6467856785277318,
"learning_rate": 5e-06,
"loss": 0.6099,
"step": 200
},
{
"epoch": 0.47390691114245415,
"grad_norm": 0.5818508149943973,
"learning_rate": 5e-06,
"loss": 0.6065,
"step": 210
},
{
"epoch": 0.49647390691114246,
"grad_norm": 0.5836244522967193,
"learning_rate": 5e-06,
"loss": 0.5982,
"step": 220
},
{
"epoch": 0.5190409026798307,
"grad_norm": 0.6168234928958146,
"learning_rate": 5e-06,
"loss": 0.6071,
"step": 230
},
{
"epoch": 0.5416078984485191,
"grad_norm": 0.6516114810453338,
"learning_rate": 5e-06,
"loss": 0.606,
"step": 240
},
{
"epoch": 0.5641748942172073,
"grad_norm": 0.5950322087384775,
"learning_rate": 5e-06,
"loss": 0.6,
"step": 250
},
{
"epoch": 0.5867418899858956,
"grad_norm": 0.7451472316728437,
"learning_rate": 5e-06,
"loss": 0.6026,
"step": 260
},
{
"epoch": 0.609308885754584,
"grad_norm": 0.6121557213958919,
"learning_rate": 5e-06,
"loss": 0.5981,
"step": 270
},
{
"epoch": 0.6318758815232722,
"grad_norm": 0.5994773210552294,
"learning_rate": 5e-06,
"loss": 0.6047,
"step": 280
},
{
"epoch": 0.6544428772919605,
"grad_norm": 0.5926672661358988,
"learning_rate": 5e-06,
"loss": 0.5981,
"step": 290
},
{
"epoch": 0.6770098730606487,
"grad_norm": 0.6538562481309866,
"learning_rate": 5e-06,
"loss": 0.6041,
"step": 300
},
{
"epoch": 0.6995768688293371,
"grad_norm": 0.5616466106542526,
"learning_rate": 5e-06,
"loss": 0.5974,
"step": 310
},
{
"epoch": 0.7221438645980254,
"grad_norm": 0.618462003583941,
"learning_rate": 5e-06,
"loss": 0.6007,
"step": 320
},
{
"epoch": 0.7447108603667136,
"grad_norm": 0.5687356941697601,
"learning_rate": 5e-06,
"loss": 0.5963,
"step": 330
},
{
"epoch": 0.767277856135402,
"grad_norm": 0.6426107325596253,
"learning_rate": 5e-06,
"loss": 0.5957,
"step": 340
},
{
"epoch": 0.7898448519040903,
"grad_norm": 0.8487173900730957,
"learning_rate": 5e-06,
"loss": 0.5972,
"step": 350
},
{
"epoch": 0.8124118476727785,
"grad_norm": 0.7180246951320329,
"learning_rate": 5e-06,
"loss": 0.5962,
"step": 360
},
{
"epoch": 0.8349788434414669,
"grad_norm": 0.655933551399716,
"learning_rate": 5e-06,
"loss": 0.6032,
"step": 370
},
{
"epoch": 0.8575458392101551,
"grad_norm": 0.6371289837710932,
"learning_rate": 5e-06,
"loss": 0.5999,
"step": 380
},
{
"epoch": 0.8801128349788434,
"grad_norm": 0.5729445889093481,
"learning_rate": 5e-06,
"loss": 0.5939,
"step": 390
},
{
"epoch": 0.9026798307475318,
"grad_norm": 0.554750173557113,
"learning_rate": 5e-06,
"loss": 0.5962,
"step": 400
},
{
"epoch": 0.92524682651622,
"grad_norm": 0.5844610515507608,
"learning_rate": 5e-06,
"loss": 0.5942,
"step": 410
},
{
"epoch": 0.9478138222849083,
"grad_norm": 0.6232694956021368,
"learning_rate": 5e-06,
"loss": 0.5948,
"step": 420
},
{
"epoch": 0.9703808180535967,
"grad_norm": 0.5990980113102288,
"learning_rate": 5e-06,
"loss": 0.5889,
"step": 430
},
{
"epoch": 0.9929478138222849,
"grad_norm": 0.6690879102852272,
"learning_rate": 5e-06,
"loss": 0.5936,
"step": 440
},
{
"epoch": 0.9997179125528914,
"eval_loss": 0.5914917588233948,
"eval_runtime": 689.7056,
"eval_samples_per_second": 17.313,
"eval_steps_per_second": 0.542,
"step": 443
},
{
"epoch": 1.0155148095909732,
"grad_norm": 0.7179426488760348,
"learning_rate": 5e-06,
"loss": 0.601,
"step": 450
},
{
"epoch": 1.0380818053596614,
"grad_norm": 0.7564212979451005,
"learning_rate": 5e-06,
"loss": 0.5286,
"step": 460
},
{
"epoch": 1.0606488011283497,
"grad_norm": 0.6230515219807028,
"learning_rate": 5e-06,
"loss": 0.5432,
"step": 470
},
{
"epoch": 1.0832157968970382,
"grad_norm": 0.5875806608173868,
"learning_rate": 5e-06,
"loss": 0.5306,
"step": 480
},
{
"epoch": 1.1057827926657264,
"grad_norm": 0.5997676348493596,
"learning_rate": 5e-06,
"loss": 0.5406,
"step": 490
},
{
"epoch": 1.1283497884344147,
"grad_norm": 0.5814937507596145,
"learning_rate": 5e-06,
"loss": 0.5393,
"step": 500
},
{
"epoch": 1.150916784203103,
"grad_norm": 0.5963612022825507,
"learning_rate": 5e-06,
"loss": 0.5404,
"step": 510
},
{
"epoch": 1.1734837799717912,
"grad_norm": 0.603838193506647,
"learning_rate": 5e-06,
"loss": 0.5329,
"step": 520
},
{
"epoch": 1.1960507757404795,
"grad_norm": 0.6703022537084004,
"learning_rate": 5e-06,
"loss": 0.5391,
"step": 530
},
{
"epoch": 1.2186177715091677,
"grad_norm": 0.60367766192998,
"learning_rate": 5e-06,
"loss": 0.5365,
"step": 540
},
{
"epoch": 1.2411847672778562,
"grad_norm": 0.6241677396345536,
"learning_rate": 5e-06,
"loss": 0.5414,
"step": 550
},
{
"epoch": 1.2637517630465445,
"grad_norm": 0.5904073954783137,
"learning_rate": 5e-06,
"loss": 0.5432,
"step": 560
},
{
"epoch": 1.2863187588152327,
"grad_norm": 0.6340431122600575,
"learning_rate": 5e-06,
"loss": 0.5406,
"step": 570
},
{
"epoch": 1.308885754583921,
"grad_norm": 0.5704148848793037,
"learning_rate": 5e-06,
"loss": 0.5404,
"step": 580
},
{
"epoch": 1.3314527503526092,
"grad_norm": 0.5496784474652162,
"learning_rate": 5e-06,
"loss": 0.5405,
"step": 590
},
{
"epoch": 1.3540197461212977,
"grad_norm": 0.6422298085041148,
"learning_rate": 5e-06,
"loss": 0.5458,
"step": 600
},
{
"epoch": 1.376586741889986,
"grad_norm": 0.5455850816911257,
"learning_rate": 5e-06,
"loss": 0.5385,
"step": 610
},
{
"epoch": 1.3991537376586742,
"grad_norm": 0.5671880293017345,
"learning_rate": 5e-06,
"loss": 0.5427,
"step": 620
},
{
"epoch": 1.4217207334273625,
"grad_norm": 0.5382281075653461,
"learning_rate": 5e-06,
"loss": 0.5408,
"step": 630
},
{
"epoch": 1.4442877291960508,
"grad_norm": 0.6529715193484273,
"learning_rate": 5e-06,
"loss": 0.5404,
"step": 640
},
{
"epoch": 1.466854724964739,
"grad_norm": 0.5537305420882022,
"learning_rate": 5e-06,
"loss": 0.5413,
"step": 650
},
{
"epoch": 1.4894217207334273,
"grad_norm": 0.6210033660917779,
"learning_rate": 5e-06,
"loss": 0.5451,
"step": 660
},
{
"epoch": 1.5119887165021155,
"grad_norm": 0.5565754286025256,
"learning_rate": 5e-06,
"loss": 0.5398,
"step": 670
},
{
"epoch": 1.5345557122708038,
"grad_norm": 0.5381225505106777,
"learning_rate": 5e-06,
"loss": 0.5341,
"step": 680
},
{
"epoch": 1.5571227080394923,
"grad_norm": 0.5773905544863132,
"learning_rate": 5e-06,
"loss": 0.5363,
"step": 690
},
{
"epoch": 1.5796897038081805,
"grad_norm": 0.5769719221347809,
"learning_rate": 5e-06,
"loss": 0.5404,
"step": 700
},
{
"epoch": 1.6022566995768688,
"grad_norm": 0.5533757752872863,
"learning_rate": 5e-06,
"loss": 0.5431,
"step": 710
},
{
"epoch": 1.6248236953455573,
"grad_norm": 0.6442176398164877,
"learning_rate": 5e-06,
"loss": 0.5512,
"step": 720
},
{
"epoch": 1.6473906911142455,
"grad_norm": 0.536095505533598,
"learning_rate": 5e-06,
"loss": 0.5345,
"step": 730
},
{
"epoch": 1.6699576868829338,
"grad_norm": 0.6522995385264682,
"learning_rate": 5e-06,
"loss": 0.5502,
"step": 740
},
{
"epoch": 1.692524682651622,
"grad_norm": 0.6541763499758874,
"learning_rate": 5e-06,
"loss": 0.5418,
"step": 750
},
{
"epoch": 1.7150916784203103,
"grad_norm": 0.5755443728719877,
"learning_rate": 5e-06,
"loss": 0.5407,
"step": 760
},
{
"epoch": 1.7376586741889986,
"grad_norm": 0.6681662106026035,
"learning_rate": 5e-06,
"loss": 0.5414,
"step": 770
},
{
"epoch": 1.7602256699576868,
"grad_norm": 0.5602122988289128,
"learning_rate": 5e-06,
"loss": 0.5371,
"step": 780
},
{
"epoch": 1.782792665726375,
"grad_norm": 0.6443163196122946,
"learning_rate": 5e-06,
"loss": 0.5407,
"step": 790
},
{
"epoch": 1.8053596614950633,
"grad_norm": 0.6579511730704369,
"learning_rate": 5e-06,
"loss": 0.5358,
"step": 800
},
{
"epoch": 1.8279266572637518,
"grad_norm": 0.5314700702921354,
"learning_rate": 5e-06,
"loss": 0.5464,
"step": 810
},
{
"epoch": 1.85049365303244,
"grad_norm": 0.6693001540329377,
"learning_rate": 5e-06,
"loss": 0.5337,
"step": 820
},
{
"epoch": 1.8730606488011283,
"grad_norm": 0.5923366245983457,
"learning_rate": 5e-06,
"loss": 0.5459,
"step": 830
},
{
"epoch": 1.8956276445698168,
"grad_norm": 0.5746590039094727,
"learning_rate": 5e-06,
"loss": 0.5458,
"step": 840
},
{
"epoch": 1.918194640338505,
"grad_norm": 0.5449016184415011,
"learning_rate": 5e-06,
"loss": 0.5338,
"step": 850
},
{
"epoch": 1.9407616361071933,
"grad_norm": 0.6838679558088623,
"learning_rate": 5e-06,
"loss": 0.5389,
"step": 860
},
{
"epoch": 1.9633286318758816,
"grad_norm": 0.6874608927396291,
"learning_rate": 5e-06,
"loss": 0.5454,
"step": 870
},
{
"epoch": 1.9858956276445698,
"grad_norm": 0.6068228126793352,
"learning_rate": 5e-06,
"loss": 0.5436,
"step": 880
},
{
"epoch": 1.9994358251057829,
"eval_loss": 0.5872226357460022,
"eval_runtime": 687.8631,
"eval_samples_per_second": 17.36,
"eval_steps_per_second": 0.544,
"step": 886
},
{
"epoch": 2.008462623413258,
"grad_norm": 0.9392594820523761,
"learning_rate": 5e-06,
"loss": 0.5673,
"step": 890
},
{
"epoch": 2.0310296191819464,
"grad_norm": 0.7231338117823009,
"learning_rate": 5e-06,
"loss": 0.479,
"step": 900
},
{
"epoch": 2.0535966149506346,
"grad_norm": 0.7538417405966139,
"learning_rate": 5e-06,
"loss": 0.4807,
"step": 910
},
{
"epoch": 2.076163610719323,
"grad_norm": 0.615051591066965,
"learning_rate": 5e-06,
"loss": 0.473,
"step": 920
},
{
"epoch": 2.098730606488011,
"grad_norm": 0.6303042321293323,
"learning_rate": 5e-06,
"loss": 0.4804,
"step": 930
},
{
"epoch": 2.1212976022566994,
"grad_norm": 0.6126432163883374,
"learning_rate": 5e-06,
"loss": 0.4843,
"step": 940
},
{
"epoch": 2.143864598025388,
"grad_norm": 0.6144634421530719,
"learning_rate": 5e-06,
"loss": 0.486,
"step": 950
},
{
"epoch": 2.1664315937940763,
"grad_norm": 0.6127415554957332,
"learning_rate": 5e-06,
"loss": 0.482,
"step": 960
},
{
"epoch": 2.1889985895627646,
"grad_norm": 0.6899522979446271,
"learning_rate": 5e-06,
"loss": 0.48,
"step": 970
},
{
"epoch": 2.211565585331453,
"grad_norm": 0.6352864289512354,
"learning_rate": 5e-06,
"loss": 0.475,
"step": 980
},
{
"epoch": 2.234132581100141,
"grad_norm": 0.5919868650847235,
"learning_rate": 5e-06,
"loss": 0.4749,
"step": 990
},
{
"epoch": 2.2566995768688294,
"grad_norm": 0.6663802191849582,
"learning_rate": 5e-06,
"loss": 0.4837,
"step": 1000
},
{
"epoch": 2.2792665726375176,
"grad_norm": 0.6084128153521763,
"learning_rate": 5e-06,
"loss": 0.4899,
"step": 1010
},
{
"epoch": 2.301833568406206,
"grad_norm": 0.6928904950138666,
"learning_rate": 5e-06,
"loss": 0.4842,
"step": 1020
},
{
"epoch": 2.324400564174894,
"grad_norm": 0.6123725626600022,
"learning_rate": 5e-06,
"loss": 0.4885,
"step": 1030
},
{
"epoch": 2.3469675599435824,
"grad_norm": 0.6152856773064957,
"learning_rate": 5e-06,
"loss": 0.4865,
"step": 1040
},
{
"epoch": 2.3695345557122707,
"grad_norm": 0.5928077081638156,
"learning_rate": 5e-06,
"loss": 0.4841,
"step": 1050
},
{
"epoch": 2.392101551480959,
"grad_norm": 0.5825006611511295,
"learning_rate": 5e-06,
"loss": 0.4834,
"step": 1060
},
{
"epoch": 2.414668547249647,
"grad_norm": 0.6995506108743696,
"learning_rate": 5e-06,
"loss": 0.4844,
"step": 1070
},
{
"epoch": 2.4372355430183354,
"grad_norm": 0.6859804193253511,
"learning_rate": 5e-06,
"loss": 0.4877,
"step": 1080
},
{
"epoch": 2.459802538787024,
"grad_norm": 0.5875532797041796,
"learning_rate": 5e-06,
"loss": 0.4858,
"step": 1090
},
{
"epoch": 2.4823695345557124,
"grad_norm": 0.5995493923282224,
"learning_rate": 5e-06,
"loss": 0.489,
"step": 1100
},
{
"epoch": 2.5049365303244007,
"grad_norm": 0.5864121474051133,
"learning_rate": 5e-06,
"loss": 0.4922,
"step": 1110
},
{
"epoch": 2.527503526093089,
"grad_norm": 0.6646236430324739,
"learning_rate": 5e-06,
"loss": 0.483,
"step": 1120
},
{
"epoch": 2.550070521861777,
"grad_norm": 0.5863190712954558,
"learning_rate": 5e-06,
"loss": 0.488,
"step": 1130
},
{
"epoch": 2.5726375176304654,
"grad_norm": 0.5716627525683331,
"learning_rate": 5e-06,
"loss": 0.4863,
"step": 1140
},
{
"epoch": 2.5952045133991537,
"grad_norm": 0.7390382086747569,
"learning_rate": 5e-06,
"loss": 0.4872,
"step": 1150
},
{
"epoch": 2.617771509167842,
"grad_norm": 0.6159110202696495,
"learning_rate": 5e-06,
"loss": 0.4909,
"step": 1160
},
{
"epoch": 2.64033850493653,
"grad_norm": 0.5839945192664074,
"learning_rate": 5e-06,
"loss": 0.48,
"step": 1170
},
{
"epoch": 2.6629055007052185,
"grad_norm": 0.6059242185177918,
"learning_rate": 5e-06,
"loss": 0.4907,
"step": 1180
},
{
"epoch": 2.685472496473907,
"grad_norm": 0.5603477452119345,
"learning_rate": 5e-06,
"loss": 0.4915,
"step": 1190
},
{
"epoch": 2.7080394922425954,
"grad_norm": 0.6696129009218381,
"learning_rate": 5e-06,
"loss": 0.4901,
"step": 1200
},
{
"epoch": 2.7306064880112837,
"grad_norm": 0.5830483872814973,
"learning_rate": 5e-06,
"loss": 0.4912,
"step": 1210
},
{
"epoch": 2.753173483779972,
"grad_norm": 0.5933628600686228,
"learning_rate": 5e-06,
"loss": 0.4879,
"step": 1220
},
{
"epoch": 2.77574047954866,
"grad_norm": 0.5761818796204454,
"learning_rate": 5e-06,
"loss": 0.4945,
"step": 1230
},
{
"epoch": 2.7983074753173485,
"grad_norm": 0.6500783963121674,
"learning_rate": 5e-06,
"loss": 0.4929,
"step": 1240
},
{
"epoch": 2.8208744710860367,
"grad_norm": 0.5673086128991506,
"learning_rate": 5e-06,
"loss": 0.4929,
"step": 1250
},
{
"epoch": 2.843441466854725,
"grad_norm": 0.5798462048796974,
"learning_rate": 5e-06,
"loss": 0.4948,
"step": 1260
},
{
"epoch": 2.8660084626234132,
"grad_norm": 0.5969025403904411,
"learning_rate": 5e-06,
"loss": 0.4919,
"step": 1270
},
{
"epoch": 2.8885754583921015,
"grad_norm": 0.550948638132691,
"learning_rate": 5e-06,
"loss": 0.49,
"step": 1280
},
{
"epoch": 2.9111424541607898,
"grad_norm": 0.5828653470350499,
"learning_rate": 5e-06,
"loss": 0.4936,
"step": 1290
},
{
"epoch": 2.933709449929478,
"grad_norm": 0.5839820299754724,
"learning_rate": 5e-06,
"loss": 0.4892,
"step": 1300
},
{
"epoch": 2.9562764456981663,
"grad_norm": 0.638479735690827,
"learning_rate": 5e-06,
"loss": 0.4932,
"step": 1310
},
{
"epoch": 2.9788434414668545,
"grad_norm": 0.5926061391624452,
"learning_rate": 5e-06,
"loss": 0.49,
"step": 1320
},
{
"epoch": 2.9991537376586743,
"eval_loss": 0.6000288128852844,
"eval_runtime": 688.8385,
"eval_samples_per_second": 17.335,
"eval_steps_per_second": 0.543,
"step": 1329
},
{
"epoch": 2.9991537376586743,
"step": 1329,
"total_flos": 5064195066298368.0,
"train_loss": 0.5476739688238943,
"train_runtime": 121175.2413,
"train_samples_per_second": 5.617,
"train_steps_per_second": 0.011
}
],
"logging_steps": 10,
"max_steps": 1329,
"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": 5064195066298368.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}