AllSparkv2-3B-V-P / trainer_state.json
ShaoRun's picture
Upload 17 files
99e78b6 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.999775734469612,
"eval_steps": 500,
"global_step": 2229,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008970621215519175,
"grad_norm": 1.2712721824645996,
"learning_rate": 5.970149253731343e-06,
"loss": 3.8614,
"step": 20
},
{
"epoch": 0.01794124243103835,
"grad_norm": 1.3492835760116577,
"learning_rate": 1.1940298507462686e-05,
"loss": 2.8143,
"step": 40
},
{
"epoch": 0.026911863646557524,
"grad_norm": 0.6147993206977844,
"learning_rate": 1.791044776119403e-05,
"loss": 2.7456,
"step": 60
},
{
"epoch": 0.0358824848620767,
"grad_norm": 0.7868013978004456,
"learning_rate": 1.999821584672887e-05,
"loss": 2.6969,
"step": 80
},
{
"epoch": 0.04485310607759587,
"grad_norm": 0.8422935009002686,
"learning_rate": 1.998850515736159e-05,
"loss": 2.7144,
"step": 100
},
{
"epoch": 0.05382372729311505,
"grad_norm": 0.5046640038490295,
"learning_rate": 1.9970358823117534e-05,
"loss": 2.6939,
"step": 120
},
{
"epoch": 0.06279434850863422,
"grad_norm": 0.6028425097465515,
"learning_rate": 1.994379216921594e-05,
"loss": 2.5943,
"step": 140
},
{
"epoch": 0.0717649697241534,
"grad_norm": 0.6477038264274597,
"learning_rate": 1.990882763213298e-05,
"loss": 2.6143,
"step": 160
},
{
"epoch": 0.08073559093967257,
"grad_norm": 0.3538740873336792,
"learning_rate": 1.986549474065333e-05,
"loss": 2.6754,
"step": 180
},
{
"epoch": 0.08970621215519174,
"grad_norm": 0.3921215534210205,
"learning_rate": 1.98138300909321e-05,
"loss": 2.7099,
"step": 200
},
{
"epoch": 0.09867683337071093,
"grad_norm": 0.39746612310409546,
"learning_rate": 1.9753877315588072e-05,
"loss": 2.4453,
"step": 220
},
{
"epoch": 0.1076474545862301,
"grad_norm": 0.5270581245422363,
"learning_rate": 1.9685687046854415e-05,
"loss": 2.5436,
"step": 240
},
{
"epoch": 0.11661807580174927,
"grad_norm": 0.4105972349643707,
"learning_rate": 1.9609316873817992e-05,
"loss": 2.6322,
"step": 260
},
{
"epoch": 0.12558869701726844,
"grad_norm": 0.33983883261680603,
"learning_rate": 1.952483129378333e-05,
"loss": 2.5725,
"step": 280
},
{
"epoch": 0.13455931823278763,
"grad_norm": 0.4139712154865265,
"learning_rate": 1.9432301657802378e-05,
"loss": 2.6455,
"step": 300
},
{
"epoch": 0.1435299394483068,
"grad_norm": 0.3298761546611786,
"learning_rate": 1.9331806110416027e-05,
"loss": 2.61,
"step": 320
},
{
"epoch": 0.15250056066382597,
"grad_norm": 0.3583829998970032,
"learning_rate": 1.922342952365829e-05,
"loss": 2.5551,
"step": 340
},
{
"epoch": 0.16147118187934514,
"grad_norm": 0.3369527757167816,
"learning_rate": 1.9107263425378873e-05,
"loss": 2.6248,
"step": 360
},
{
"epoch": 0.1704418030948643,
"grad_norm": 0.4316694736480713,
"learning_rate": 1.8983405921944686e-05,
"loss": 2.5037,
"step": 380
},
{
"epoch": 0.17941242431038348,
"grad_norm": 0.3557095527648926,
"learning_rate": 1.8851961615385542e-05,
"loss": 2.6168,
"step": 400
},
{
"epoch": 0.18838304552590268,
"grad_norm": 0.3330799341201782,
"learning_rate": 1.8713041515054065e-05,
"loss": 2.5813,
"step": 420
},
{
"epoch": 0.19735366674142185,
"grad_norm": 0.5107260346412659,
"learning_rate": 1.8566762943874376e-05,
"loss": 2.5269,
"step": 440
},
{
"epoch": 0.20632428795694102,
"grad_norm": 0.30552321672439575,
"learning_rate": 1.8413249439258743e-05,
"loss": 2.5806,
"step": 460
},
{
"epoch": 0.2152949091724602,
"grad_norm": 0.33228904008865356,
"learning_rate": 1.8252630648775874e-05,
"loss": 2.5864,
"step": 480
},
{
"epoch": 0.22426553038797936,
"grad_norm": 0.3951259255409241,
"learning_rate": 1.8085042220658993e-05,
"loss": 2.5511,
"step": 500
},
{
"epoch": 0.23323615160349853,
"grad_norm": 0.3492446839809418,
"learning_rate": 1.791062568924609e-05,
"loss": 2.5582,
"step": 520
},
{
"epoch": 0.24220677281901773,
"grad_norm": 0.3901676833629608,
"learning_rate": 1.7729528355449214e-05,
"loss": 2.5093,
"step": 540
},
{
"epoch": 0.25117739403453687,
"grad_norm": 0.41344988346099854,
"learning_rate": 1.7541903162353638e-05,
"loss": 2.5634,
"step": 560
},
{
"epoch": 0.26014801525005604,
"grad_norm": 0.3539174497127533,
"learning_rate": 1.734790856605204e-05,
"loss": 2.5236,
"step": 580
},
{
"epoch": 0.26911863646557527,
"grad_norm": 0.30354949831962585,
"learning_rate": 1.714770840182273e-05,
"loss": 2.4823,
"step": 600
},
{
"epoch": 0.27808925768109444,
"grad_norm": 0.3280078172683716,
"learning_rate": 1.6941471745764996e-05,
"loss": 2.5034,
"step": 620
},
{
"epoch": 0.2870598788966136,
"grad_norm": 0.33530184626579285,
"learning_rate": 1.672937277200837e-05,
"loss": 2.5895,
"step": 640
},
{
"epoch": 0.2960305001121328,
"grad_norm": 0.6118967533111572,
"learning_rate": 1.6511590605616423e-05,
"loss": 2.5025,
"step": 660
},
{
"epoch": 0.30500112132765195,
"grad_norm": 0.45741620659828186,
"learning_rate": 1.628830917130935e-05,
"loss": 2.5659,
"step": 680
},
{
"epoch": 0.3139717425431711,
"grad_norm": 0.3485761284828186,
"learning_rate": 1.6059717038133038e-05,
"loss": 2.6027,
"step": 700
},
{
"epoch": 0.3229423637586903,
"grad_norm": 0.42178016901016235,
"learning_rate": 1.5826007260205868e-05,
"loss": 2.4769,
"step": 720
},
{
"epoch": 0.33191298497420946,
"grad_norm": 0.4537988305091858,
"learning_rate": 1.5587377213677705e-05,
"loss": 2.4729,
"step": 740
},
{
"epoch": 0.3408836061897286,
"grad_norm": 0.355874627828598,
"learning_rate": 1.5344028430038764e-05,
"loss": 2.4891,
"step": 760
},
{
"epoch": 0.3498542274052478,
"grad_norm": 0.32584452629089355,
"learning_rate": 1.5096166425919176e-05,
"loss": 2.4743,
"step": 780
},
{
"epoch": 0.35882484862076697,
"grad_norm": 0.32193851470947266,
"learning_rate": 1.4844000529522942e-05,
"loss": 2.4749,
"step": 800
},
{
"epoch": 0.36779546983628614,
"grad_norm": 0.31153932213783264,
"learning_rate": 1.458774370384287e-05,
"loss": 2.5422,
"step": 820
},
{
"epoch": 0.37676609105180536,
"grad_norm": 0.3169076144695282,
"learning_rate": 1.4327612366805832e-05,
"loss": 2.4914,
"step": 840
},
{
"epoch": 0.38573671226732453,
"grad_norm": 0.2868211269378662,
"learning_rate": 1.4063826208500182e-05,
"loss": 2.5825,
"step": 860
},
{
"epoch": 0.3947073334828437,
"grad_norm": 0.3413689136505127,
"learning_rate": 1.3796608005639738e-05,
"loss": 2.5594,
"step": 880
},
{
"epoch": 0.4036779546983629,
"grad_norm": 0.3216472566127777,
"learning_rate": 1.352618343342098e-05,
"loss": 2.4943,
"step": 900
},
{
"epoch": 0.41264857591388204,
"grad_norm": 0.3140867054462433,
"learning_rate": 1.3252780874932395e-05,
"loss": 2.4842,
"step": 920
},
{
"epoch": 0.4216191971294012,
"grad_norm": 0.3276834487915039,
"learning_rate": 1.2976631228276894e-05,
"loss": 2.5065,
"step": 940
},
{
"epoch": 0.4305898183449204,
"grad_norm": 0.3846268951892853,
"learning_rate": 1.2697967711570243e-05,
"loss": 2.4219,
"step": 960
},
{
"epoch": 0.43956043956043955,
"grad_norm": 0.3238185942173004,
"learning_rate": 1.2417025665980114e-05,
"loss": 2.4617,
"step": 980
},
{
"epoch": 0.4485310607759587,
"grad_norm": 0.34462079405784607,
"learning_rate": 1.2134042356972175e-05,
"loss": 2.5461,
"step": 1000
},
{
"epoch": 0.4575016819914779,
"grad_norm": 0.31439366936683655,
"learning_rate": 1.1849256773931058e-05,
"loss": 2.3627,
"step": 1020
},
{
"epoch": 0.46647230320699706,
"grad_norm": 0.2591687738895416,
"learning_rate": 1.156290942832536e-05,
"loss": 2.4446,
"step": 1040
},
{
"epoch": 0.47544292442251623,
"grad_norm": 0.5144124031066895,
"learning_rate": 1.1275242150587254e-05,
"loss": 2.4891,
"step": 1060
},
{
"epoch": 0.48441354563803546,
"grad_norm": 0.39585819840431213,
"learning_rate": 1.0986497885878145e-05,
"loss": 2.4465,
"step": 1080
},
{
"epoch": 0.49338416685355463,
"grad_norm": 0.4279848039150238,
"learning_rate": 1.0696920488912923e-05,
"loss": 2.4889,
"step": 1100
},
{
"epoch": 0.5023547880690737,
"grad_norm": 0.35390934348106384,
"learning_rate": 1.0406754518016047e-05,
"loss": 2.5671,
"step": 1120
},
{
"epoch": 0.5113254092845929,
"grad_norm": 0.32814425230026245,
"learning_rate": 1.0116245028583418e-05,
"loss": 2.5396,
"step": 1140
},
{
"epoch": 0.5202960305001121,
"grad_norm": 0.2794135808944702,
"learning_rate": 9.825637366124458e-06,
"loss": 2.4186,
"step": 1160
},
{
"epoch": 0.5292666517156313,
"grad_norm": 0.3125162124633789,
"learning_rate": 9.535176959059171e-06,
"loss": 2.4878,
"step": 1180
},
{
"epoch": 0.5382372729311505,
"grad_norm": 0.3027559816837311,
"learning_rate": 9.245109111445189e-06,
"loss": 2.4423,
"step": 1200
},
{
"epoch": 0.5472078941466697,
"grad_norm": 0.3461575508117676,
"learning_rate": 8.95567879580984e-06,
"loss": 2.4254,
"step": 1220
},
{
"epoch": 0.5561785153621889,
"grad_norm": 0.4075671136379242,
"learning_rate": 8.667130446262214e-06,
"loss": 2.4678,
"step": 1240
},
{
"epoch": 0.565149136577708,
"grad_norm": 0.36932793259620667,
"learning_rate": 8.379707752059932e-06,
"loss": 2.443,
"step": 1260
},
{
"epoch": 0.5741197577932272,
"grad_norm": 0.3690980076789856,
"learning_rate": 8.093653451804987e-06,
"loss": 2.5446,
"step": 1280
},
{
"epoch": 0.5830903790087464,
"grad_norm": 0.33403080701828003,
"learning_rate": 7.809209128442408e-06,
"loss": 2.4948,
"step": 1300
},
{
"epoch": 0.5920610002242656,
"grad_norm": 0.3500635325908661,
"learning_rate": 7.52661500523497e-06,
"loss": 2.4717,
"step": 1320
},
{
"epoch": 0.6010316214397847,
"grad_norm": 0.3530026376247406,
"learning_rate": 7.246109742886156e-06,
"loss": 2.5245,
"step": 1340
},
{
"epoch": 0.6100022426553039,
"grad_norm": 0.35534512996673584,
"learning_rate": 6.967930237982793e-06,
"loss": 2.407,
"step": 1360
},
{
"epoch": 0.6189728638708231,
"grad_norm": 0.3208062946796417,
"learning_rate": 6.692311422927515e-06,
"loss": 2.4504,
"step": 1380
},
{
"epoch": 0.6279434850863422,
"grad_norm": 0.32474184036254883,
"learning_rate": 6.4194860675300695e-06,
"loss": 2.5117,
"step": 1400
},
{
"epoch": 0.6369141063018614,
"grad_norm": 0.3804394006729126,
"learning_rate": 6.149684582425013e-06,
"loss": 2.4671,
"step": 1420
},
{
"epoch": 0.6458847275173806,
"grad_norm": 0.35071736574172974,
"learning_rate": 5.883134824481786e-06,
"loss": 2.4593,
"step": 1440
},
{
"epoch": 0.6548553487328997,
"grad_norm": 0.348230242729187,
"learning_rate": 5.620061904371565e-06,
"loss": 2.436,
"step": 1460
},
{
"epoch": 0.6638259699484189,
"grad_norm": 0.42061659693717957,
"learning_rate": 5.360687996453348e-06,
"loss": 2.4705,
"step": 1480
},
{
"epoch": 0.6727965911639381,
"grad_norm": 0.3868522644042969,
"learning_rate": 5.105232151139895e-06,
"loss": 2.4973,
"step": 1500
},
{
"epoch": 0.6817672123794573,
"grad_norm": 0.37379515171051025,
"learning_rate": 4.853910109901901e-06,
"loss": 2.4178,
"step": 1520
},
{
"epoch": 0.6907378335949764,
"grad_norm": 0.38078969717025757,
"learning_rate": 4.606934123066739e-06,
"loss": 2.3759,
"step": 1540
},
{
"epoch": 0.6997084548104956,
"grad_norm": 0.44406720995903015,
"learning_rate": 4.3645127705655654e-06,
"loss": 2.507,
"step": 1560
},
{
"epoch": 0.7086790760260148,
"grad_norm": 0.3889126777648926,
"learning_rate": 4.126850785780199e-06,
"loss": 2.5821,
"step": 1580
},
{
"epoch": 0.7176496972415339,
"grad_norm": 0.362594336271286,
"learning_rate": 3.8941488826385855e-06,
"loss": 2.4599,
"step": 1600
},
{
"epoch": 0.7266203184570531,
"grad_norm": 0.3674083948135376,
"learning_rate": 3.6666035861047744e-06,
"loss": 2.4943,
"step": 1620
},
{
"epoch": 0.7355909396725723,
"grad_norm": 0.40499699115753174,
"learning_rate": 3.444407066206692e-06,
"loss": 2.4119,
"step": 1640
},
{
"epoch": 0.7445615608880914,
"grad_norm": 0.3226064443588257,
"learning_rate": 3.2277469757417403e-06,
"loss": 2.4375,
"step": 1660
},
{
"epoch": 0.7535321821036107,
"grad_norm": 0.36566492915153503,
"learning_rate": 3.0168062917974173e-06,
"loss": 2.4084,
"step": 1680
},
{
"epoch": 0.7625028033191299,
"grad_norm": 0.33492156863212585,
"learning_rate": 2.8117631612207084e-06,
"loss": 2.4378,
"step": 1700
},
{
"epoch": 0.7714734245346491,
"grad_norm": 0.3422945439815521,
"learning_rate": 2.6127907501667726e-06,
"loss": 2.5039,
"step": 1720
},
{
"epoch": 0.7804440457501682,
"grad_norm": 0.39878708124160767,
"learning_rate": 2.420057097854046e-06,
"loss": 2.4394,
"step": 1740
},
{
"epoch": 0.7894146669656874,
"grad_norm": 0.3516499698162079,
"learning_rate": 2.2337249746491695e-06,
"loss": 2.4093,
"step": 1760
},
{
"epoch": 0.7983852881812066,
"grad_norm": 0.3679056167602539,
"learning_rate": 2.0539517446016975e-06,
"loss": 2.3983,
"step": 1780
},
{
"epoch": 0.8073559093967257,
"grad_norm": 0.3900187313556671,
"learning_rate": 1.880889232544585e-06,
"loss": 2.4602,
"step": 1800
},
{
"epoch": 0.8163265306122449,
"grad_norm": 0.34092533588409424,
"learning_rate": 1.714683595872777e-06,
"loss": 2.4214,
"step": 1820
},
{
"epoch": 0.8252971518277641,
"grad_norm": 0.3562910854816437,
"learning_rate": 1.5554752011081332e-06,
"loss": 2.4335,
"step": 1840
},
{
"epoch": 0.8342677730432833,
"grad_norm": 0.39239928126335144,
"learning_rate": 1.4033985053549425e-06,
"loss": 2.3766,
"step": 1860
},
{
"epoch": 0.8432383942588024,
"grad_norm": 0.3768567144870758,
"learning_rate": 1.2585819427461564e-06,
"loss": 2.4227,
"step": 1880
},
{
"epoch": 0.8522090154743216,
"grad_norm": 0.378839910030365,
"learning_rate": 1.121147815976248e-06,
"loss": 2.3695,
"step": 1900
},
{
"epoch": 0.8611796366898408,
"grad_norm": 0.32945871353149414,
"learning_rate": 9.912121930122542e-07,
"loss": 2.3736,
"step": 1920
},
{
"epoch": 0.8701502579053599,
"grad_norm": 0.43066656589508057,
"learning_rate": 8.688848090702928e-07,
"loss": 2.3896,
"step": 1940
},
{
"epoch": 0.8791208791208791,
"grad_norm": 0.3874249756336212,
"learning_rate": 7.542689739403097e-07,
"loss": 2.4423,
"step": 1960
},
{
"epoch": 0.8880915003363983,
"grad_norm": 0.3411208391189575,
"learning_rate": 6.474614847373051e-07,
"loss": 2.4222,
"step": 1980
},
{
"epoch": 0.8970621215519174,
"grad_norm": 0.3745633661746979,
"learning_rate": 5.485525441527651e-07,
"loss": 2.3641,
"step": 2000
},
{
"epoch": 0.9060327427674366,
"grad_norm": 0.3834417462348938,
"learning_rate": 4.5762568427529795e-07,
"loss": 2.4481,
"step": 2020
},
{
"epoch": 0.9150033639829558,
"grad_norm": 0.3854876160621643,
"learning_rate": 3.747576960448551e-07,
"loss": 2.4281,
"step": 2040
},
{
"epoch": 0.923973985198475,
"grad_norm": 1.0627397298812866,
"learning_rate": 3.0001856440005307e-07,
"loss": 2.3635,
"step": 2060
},
{
"epoch": 0.9329446064139941,
"grad_norm": 0.38673698902130127,
"learning_rate": 2.3347140917344579e-07,
"loss": 2.5118,
"step": 2080
},
{
"epoch": 0.9419152276295133,
"grad_norm": 0.36718830466270447,
"learning_rate": 1.7517243178458486e-07,
"loss": 2.453,
"step": 2100
},
{
"epoch": 0.9508858488450325,
"grad_norm": 0.5423038601875305,
"learning_rate": 1.2517086777594112e-07,
"loss": 2.4404,
"step": 2120
},
{
"epoch": 0.9598564700605517,
"grad_norm": 0.33170536160469055,
"learning_rate": 8.35089452317639e-08,
"loss": 2.3524,
"step": 2140
},
{
"epoch": 0.9688270912760709,
"grad_norm": 0.44688114523887634,
"learning_rate": 5.022184911495864e-08,
"loss": 2.3764,
"step": 2160
},
{
"epoch": 0.9777977124915901,
"grad_norm": 0.43967434763908386,
"learning_rate": 2.5337691552156372e-08,
"loss": 2.427,
"step": 2180
},
{
"epoch": 0.9867683337071093,
"grad_norm": 0.44690561294555664,
"learning_rate": 8.877488092022823e-09,
"loss": 2.423,
"step": 2200
},
{
"epoch": 0.9957389549226284,
"grad_norm": 0.34528931975364685,
"learning_rate": 8.551399568945684e-10,
"loss": 2.4153,
"step": 2220
},
{
"epoch": 0.999775734469612,
"step": 2229,
"total_flos": 1.4099141814984376e+18,
"train_loss": 2.5103028606758784,
"train_runtime": 5031.0289,
"train_samples_per_second": 28.356,
"train_steps_per_second": 0.443
}
],
"logging_steps": 20,
"max_steps": 2229,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.4099141814984376e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}