3b5e2884a6
Model: zhenting/evolm-1B-160BT-cpt-MixedFW8FM42-sftep1-sampled500k_first100k_qwen7b Source: Original Platform
1135 lines
27 KiB
JSON
1135 lines
27 KiB
JSON
{
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1563,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.006397952655150352,
|
|
"grad_norm": 2.921719264032125,
|
|
"learning_rate": 6.369426751592357e-07,
|
|
"loss": 0.6483,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.012795905310300703,
|
|
"grad_norm": 2.5418030171136805,
|
|
"learning_rate": 1.2738853503184715e-06,
|
|
"loss": 0.6318,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.019193857965451054,
|
|
"grad_norm": 1.7815131166709655,
|
|
"learning_rate": 1.9108280254777074e-06,
|
|
"loss": 0.5449,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.025591810620601407,
|
|
"grad_norm": 1.3393628581068209,
|
|
"learning_rate": 2.547770700636943e-06,
|
|
"loss": 0.4472,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.03198976327575176,
|
|
"grad_norm": 1.2577561453510053,
|
|
"learning_rate": 3.1847133757961785e-06,
|
|
"loss": 0.3985,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.03838771593090211,
|
|
"grad_norm": 0.9148245180631123,
|
|
"learning_rate": 3.821656050955415e-06,
|
|
"loss": 0.3552,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.044785668586052464,
|
|
"grad_norm": 0.8930060191753311,
|
|
"learning_rate": 4.45859872611465e-06,
|
|
"loss": 0.334,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.05118362124120281,
|
|
"grad_norm": 1.008448100358152,
|
|
"learning_rate": 5.095541401273886e-06,
|
|
"loss": 0.324,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.05758157389635317,
|
|
"grad_norm": 0.8501502028134984,
|
|
"learning_rate": 5.732484076433121e-06,
|
|
"loss": 0.3072,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.06397952655150352,
|
|
"grad_norm": 0.8802430610781631,
|
|
"learning_rate": 6.369426751592357e-06,
|
|
"loss": 0.2956,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.07037747920665387,
|
|
"grad_norm": 0.915060821877453,
|
|
"learning_rate": 7.006369426751593e-06,
|
|
"loss": 0.2934,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.07677543186180422,
|
|
"grad_norm": 0.9069466967459935,
|
|
"learning_rate": 7.64331210191083e-06,
|
|
"loss": 0.2863,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.08317338451695458,
|
|
"grad_norm": 0.8940748388374813,
|
|
"learning_rate": 8.280254777070064e-06,
|
|
"loss": 0.2816,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.08957133717210493,
|
|
"grad_norm": 0.874644405636949,
|
|
"learning_rate": 8.9171974522293e-06,
|
|
"loss": 0.2675,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.09596928982725528,
|
|
"grad_norm": 0.8886191985454133,
|
|
"learning_rate": 9.554140127388536e-06,
|
|
"loss": 0.2704,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.10236724248240563,
|
|
"grad_norm": 0.9154797827979856,
|
|
"learning_rate": 9.999887666317538e-06,
|
|
"loss": 0.2556,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.10876519513755598,
|
|
"grad_norm": 0.8555186445038075,
|
|
"learning_rate": 9.997890763487869e-06,
|
|
"loss": 0.258,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.11516314779270634,
|
|
"grad_norm": 0.9311100334834356,
|
|
"learning_rate": 9.993398704133318e-06,
|
|
"loss": 0.2622,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.12156110044785669,
|
|
"grad_norm": 0.9085099596846602,
|
|
"learning_rate": 9.986413730878168e-06,
|
|
"loss": 0.2558,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.12795905310300704,
|
|
"grad_norm": 0.8771430204132477,
|
|
"learning_rate": 9.976939330913801e-06,
|
|
"loss": 0.2548,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.1343570057581574,
|
|
"grad_norm": 0.7736840774856885,
|
|
"learning_rate": 9.96498023425774e-06,
|
|
"loss": 0.2485,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.14075495841330773,
|
|
"grad_norm": 0.8001039033106894,
|
|
"learning_rate": 9.95054241139223e-06,
|
|
"loss": 0.2405,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.1471529110684581,
|
|
"grad_norm": 0.9442073202593417,
|
|
"learning_rate": 9.933633070283512e-06,
|
|
"loss": 0.245,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.15355086372360843,
|
|
"grad_norm": 0.8650872476606288,
|
|
"learning_rate": 9.914260652783323e-06,
|
|
"loss": 0.2351,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.1599488163787588,
|
|
"grad_norm": 0.8231465831071424,
|
|
"learning_rate": 9.892434830414354e-06,
|
|
"loss": 0.231,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.16634676903390916,
|
|
"grad_norm": 0.8421530599227623,
|
|
"learning_rate": 9.868166499541824e-06,
|
|
"loss": 0.2296,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.1727447216890595,
|
|
"grad_norm": 0.8649013439254571,
|
|
"learning_rate": 9.841467775933566e-06,
|
|
"loss": 0.2299,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.17914267434420986,
|
|
"grad_norm": 0.805808988849827,
|
|
"learning_rate": 9.812351988711312e-06,
|
|
"loss": 0.2311,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.1855406269993602,
|
|
"grad_norm": 0.9418928279447264,
|
|
"learning_rate": 9.780833673696255e-06,
|
|
"loss": 0.2323,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.19193857965451055,
|
|
"grad_norm": 0.8853188272484068,
|
|
"learning_rate": 9.746928566152148e-06,
|
|
"loss": 0.23,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.19833653230966092,
|
|
"grad_norm": 0.8213336112041623,
|
|
"learning_rate": 9.710653592929595e-06,
|
|
"loss": 0.222,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.20473448496481125,
|
|
"grad_norm": 0.8608256269231939,
|
|
"learning_rate": 9.672026864015476e-06,
|
|
"loss": 0.221,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.21113243761996162,
|
|
"grad_norm": 0.8418067711806069,
|
|
"learning_rate": 9.631067663491663e-06,
|
|
"loss": 0.2191,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.21753039027511195,
|
|
"grad_norm": 0.7845284987136818,
|
|
"learning_rate": 9.587796439907609e-06,
|
|
"loss": 0.217,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.22392834293026231,
|
|
"grad_norm": 0.8047684481827264,
|
|
"learning_rate": 9.542234796071577e-06,
|
|
"loss": 0.2232,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.23032629558541268,
|
|
"grad_norm": 0.8291836227784779,
|
|
"learning_rate": 9.4944054782656e-06,
|
|
"loss": 0.2145,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.236724248240563,
|
|
"grad_norm": 0.7992401593879032,
|
|
"learning_rate": 9.444332364889603e-06,
|
|
"loss": 0.2194,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.24312220089571338,
|
|
"grad_norm": 0.8450873820776212,
|
|
"learning_rate": 9.392040454540284e-06,
|
|
"loss": 0.2134,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.2495201535508637,
|
|
"grad_norm": 0.967399584058993,
|
|
"learning_rate": 9.337555853530785e-06,
|
|
"loss": 0.2108,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.2559181062060141,
|
|
"grad_norm": 0.8772050788302971,
|
|
"learning_rate": 9.280905762857315e-06,
|
|
"loss": 0.2115,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.26231605886116444,
|
|
"grad_norm": 0.8319365647969784,
|
|
"learning_rate": 9.222118464619278e-06,
|
|
"loss": 0.2122,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.2687140115163148,
|
|
"grad_norm": 0.8666958950512132,
|
|
"learning_rate": 9.161223307899659e-06,
|
|
"loss": 0.2045,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.2751119641714651,
|
|
"grad_norm": 0.8061480778705524,
|
|
"learning_rate": 9.09825069411274e-06,
|
|
"loss": 0.2057,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.28150991682661547,
|
|
"grad_norm": 0.8470645205499479,
|
|
"learning_rate": 9.033232061826428e-06,
|
|
"loss": 0.2033,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.28790786948176583,
|
|
"grad_norm": 0.8243507574943671,
|
|
"learning_rate": 8.96619987106682e-06,
|
|
"loss": 0.2104,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.2943058221369162,
|
|
"grad_norm": 0.8238623019564721,
|
|
"learning_rate": 8.897187587112783e-06,
|
|
"loss": 0.2007,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.30070377479206656,
|
|
"grad_norm": 0.8928681882880483,
|
|
"learning_rate": 8.826229663788688e-06,
|
|
"loss": 0.2011,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 0.30710172744721687,
|
|
"grad_norm": 0.8275747287175418,
|
|
"learning_rate": 8.753361526263622e-06,
|
|
"loss": 0.2085,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.31349968010236723,
|
|
"grad_norm": 0.8253140582094273,
|
|
"learning_rate": 8.67861955336566e-06,
|
|
"loss": 0.2016,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 0.3198976327575176,
|
|
"grad_norm": 0.9078921610234565,
|
|
"learning_rate": 8.602041059420017e-06,
|
|
"loss": 0.2028,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.32629558541266795,
|
|
"grad_norm": 0.8193590764099334,
|
|
"learning_rate": 8.523664275620185e-06,
|
|
"loss": 0.2038,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 0.3326935380678183,
|
|
"grad_norm": 0.751476028561934,
|
|
"learning_rate": 8.443528330941322e-06,
|
|
"loss": 0.1943,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.3390914907229686,
|
|
"grad_norm": 0.8116372043807868,
|
|
"learning_rate": 8.361673232605408e-06,
|
|
"loss": 0.1986,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 0.345489443378119,
|
|
"grad_norm": 0.7651945791801316,
|
|
"learning_rate": 8.27813984610799e-06,
|
|
"loss": 0.1969,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.35188739603326935,
|
|
"grad_norm": 0.7682455631964752,
|
|
"learning_rate": 8.19296987481639e-06,
|
|
"loss": 0.1983,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.3582853486884197,
|
|
"grad_norm": 0.7979712275711189,
|
|
"learning_rate": 8.106205839149653e-06,
|
|
"loss": 0.1892,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.3646833013435701,
|
|
"grad_norm": 0.780173946861971,
|
|
"learning_rate": 8.017891055350563e-06,
|
|
"loss": 0.1909,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 0.3710812539987204,
|
|
"grad_norm": 0.8094771534181999,
|
|
"learning_rate": 7.928069613860357e-06,
|
|
"loss": 0.1889,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.37747920665387075,
|
|
"grad_norm": 0.75546721405013,
|
|
"learning_rate": 7.836786357306943e-06,
|
|
"loss": 0.1883,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 0.3838771593090211,
|
|
"grad_norm": 0.8959844719353787,
|
|
"learning_rate": 7.744086858117565e-06,
|
|
"loss": 0.1927,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.3902751119641715,
|
|
"grad_norm": 0.8475662152796308,
|
|
"learning_rate": 7.650017395767149e-06,
|
|
"loss": 0.1852,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 0.39667306461932184,
|
|
"grad_norm": 0.7905827964197675,
|
|
"learning_rate": 7.554624933673638e-06,
|
|
"loss": 0.1896,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.40307101727447214,
|
|
"grad_norm": 0.8472922946532878,
|
|
"learning_rate": 7.457957095751896e-06,
|
|
"loss": 0.1872,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 0.4094689699296225,
|
|
"grad_norm": 0.8745381636530021,
|
|
"learning_rate": 7.3600621426378515e-06,
|
|
"loss": 0.1822,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.41586692258477287,
|
|
"grad_norm": 0.8501419353340188,
|
|
"learning_rate": 7.260988947594759e-06,
|
|
"loss": 0.1884,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.42226487523992323,
|
|
"grad_norm": 0.7996033956322788,
|
|
"learning_rate": 7.160786972113627e-06,
|
|
"loss": 0.1878,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.4286628278950736,
|
|
"grad_norm": 0.7888941818549653,
|
|
"learning_rate": 7.059506241219964e-06,
|
|
"loss": 0.1798,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 0.4350607805502239,
|
|
"grad_norm": 0.8007578444352214,
|
|
"learning_rate": 6.957197318499187e-06,
|
|
"loss": 0.184,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.44145873320537427,
|
|
"grad_norm": 0.816075763538245,
|
|
"learning_rate": 6.853911280853168e-06,
|
|
"loss": 0.1833,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 0.44785668586052463,
|
|
"grad_norm": 0.7709414985344181,
|
|
"learning_rate": 6.749699693000495e-06,
|
|
"loss": 0.1903,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.454254638515675,
|
|
"grad_norm": 0.7786538472546455,
|
|
"learning_rate": 6.6446145817332105e-06,
|
|
"loss": 0.1802,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 0.46065259117082535,
|
|
"grad_norm": 0.8210825749622989,
|
|
"learning_rate": 6.538708409942854e-06,
|
|
"loss": 0.179,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.46705054382597566,
|
|
"grad_norm": 0.8335982343820388,
|
|
"learning_rate": 6.4320340504287825e-06,
|
|
"loss": 0.1844,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 0.473448496481126,
|
|
"grad_norm": 0.9092201104993919,
|
|
"learning_rate": 6.324644759501869e-06,
|
|
"loss": 0.1824,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.4798464491362764,
|
|
"grad_norm": 0.8474724095639653,
|
|
"learning_rate": 6.2165941503966995e-06,
|
|
"loss": 0.1825,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.48624440179142675,
|
|
"grad_norm": 0.7737493984100262,
|
|
"learning_rate": 6.107936166505615e-06,
|
|
"loss": 0.1844,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.4926423544465771,
|
|
"grad_norm": 0.823135549887481,
|
|
"learning_rate": 5.998725054447904e-06,
|
|
"loss": 0.1741,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 0.4990403071017274,
|
|
"grad_norm": 0.7956378071989458,
|
|
"learning_rate": 5.889015336987614e-06,
|
|
"loss": 0.1776,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.5054382597568778,
|
|
"grad_norm": 0.8208243838737195,
|
|
"learning_rate": 5.778861785813508e-06,
|
|
"loss": 0.1709,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 0.5118362124120281,
|
|
"grad_norm": 0.9471998158646897,
|
|
"learning_rate": 5.6683193941947365e-06,
|
|
"loss": 0.1802,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.5182341650671785,
|
|
"grad_norm": 0.856851837620091,
|
|
"learning_rate": 5.5574433495259015e-06,
|
|
"loss": 0.1737,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 0.5246321177223289,
|
|
"grad_norm": 0.784645392482337,
|
|
"learning_rate": 5.446289005775185e-06,
|
|
"loss": 0.1769,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.5310300703774792,
|
|
"grad_norm": 0.7939588212021522,
|
|
"learning_rate": 5.334911855849334e-06,
|
|
"loss": 0.1699,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 0.5374280230326296,
|
|
"grad_norm": 0.7819404523459293,
|
|
"learning_rate": 5.2233675038892815e-06,
|
|
"loss": 0.1667,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.5438259756877799,
|
|
"grad_norm": 0.800765321243524,
|
|
"learning_rate": 5.111711637510216e-06,
|
|
"loss": 0.1778,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.5502239283429302,
|
|
"grad_norm": 0.7649782365235427,
|
|
"learning_rate": 5e-06,
|
|
"loss": 0.172,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.5566218809980806,
|
|
"grad_norm": 0.8935574128726307,
|
|
"learning_rate": 4.888288362489786e-06,
|
|
"loss": 0.1776,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 0.5630198336532309,
|
|
"grad_norm": 0.8482953894780804,
|
|
"learning_rate": 4.776632496110721e-06,
|
|
"loss": 0.1761,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.5694177863083814,
|
|
"grad_norm": 0.8430889361386534,
|
|
"learning_rate": 4.665088144150666e-06,
|
|
"loss": 0.1709,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 0.5758157389635317,
|
|
"grad_norm": 0.9053402619259454,
|
|
"learning_rate": 4.553710994224816e-06,
|
|
"loss": 0.1675,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.582213691618682,
|
|
"grad_norm": 0.836247331632949,
|
|
"learning_rate": 4.442556650474099e-06,
|
|
"loss": 0.1705,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 0.5886116442738324,
|
|
"grad_norm": 0.7837535565178672,
|
|
"learning_rate": 4.331680605805264e-06,
|
|
"loss": 0.17,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.5950095969289827,
|
|
"grad_norm": 0.7352841697690848,
|
|
"learning_rate": 4.221138214186493e-06,
|
|
"loss": 0.1717,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 0.6014075495841331,
|
|
"grad_norm": 0.7708645313805899,
|
|
"learning_rate": 4.110984663012388e-06,
|
|
"loss": 0.1709,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.6078055022392834,
|
|
"grad_norm": 0.7961654860598167,
|
|
"learning_rate": 4.001274945552098e-06,
|
|
"loss": 0.1722,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.6142034548944337,
|
|
"grad_norm": 0.7447107570479293,
|
|
"learning_rate": 3.892063833494387e-06,
|
|
"loss": 0.1684,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.6206014075495841,
|
|
"grad_norm": 0.7903348849847861,
|
|
"learning_rate": 3.783405849603302e-06,
|
|
"loss": 0.1657,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 0.6269993602047345,
|
|
"grad_norm": 0.8737100063534712,
|
|
"learning_rate": 3.675355240498133e-06,
|
|
"loss": 0.1641,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.6333973128598849,
|
|
"grad_norm": 0.7730316106501531,
|
|
"learning_rate": 3.567965949571219e-06,
|
|
"loss": 0.1676,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 0.6397952655150352,
|
|
"grad_norm": 0.7999574437385771,
|
|
"learning_rate": 3.4612915900571493e-06,
|
|
"loss": 0.1689,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.6461932181701855,
|
|
"grad_norm": 0.8280872795355375,
|
|
"learning_rate": 3.355385418266792e-06,
|
|
"loss": 0.1628,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 0.6525911708253359,
|
|
"grad_norm": 0.8096675245508481,
|
|
"learning_rate": 3.2503003069995057e-06,
|
|
"loss": 0.1693,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.6589891234804862,
|
|
"grad_norm": 0.7780142464577352,
|
|
"learning_rate": 3.1460887191468324e-06,
|
|
"loss": 0.1676,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 0.6653870761356366,
|
|
"grad_norm": 0.8046193448491429,
|
|
"learning_rate": 3.042802681500814e-06,
|
|
"loss": 0.1686,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.6717850287907869,
|
|
"grad_norm": 0.7848193624451667,
|
|
"learning_rate": 2.9404937587800374e-06,
|
|
"loss": 0.1598,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.6781829814459372,
|
|
"grad_norm": 0.8099947568674603,
|
|
"learning_rate": 2.839213027886373e-06,
|
|
"loss": 0.1596,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.6845809341010877,
|
|
"grad_norm": 0.837143415239909,
|
|
"learning_rate": 2.7390110524052415e-06,
|
|
"loss": 0.167,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 0.690978886756238,
|
|
"grad_norm": 0.7781762020588363,
|
|
"learning_rate": 2.6399378573621493e-06,
|
|
"loss": 0.1658,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.6973768394113884,
|
|
"grad_norm": 0.863979118680859,
|
|
"learning_rate": 2.5420429042481054e-06,
|
|
"loss": 0.1712,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 0.7037747920665387,
|
|
"grad_norm": 0.8701727964415887,
|
|
"learning_rate": 2.445375066326362e-06,
|
|
"loss": 0.1627,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.710172744721689,
|
|
"grad_norm": 0.7147714500791674,
|
|
"learning_rate": 2.349982604232851e-06,
|
|
"loss": 0.1659,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 0.7165706973768394,
|
|
"grad_norm": 0.8519730966708728,
|
|
"learning_rate": 2.255913141882436e-06,
|
|
"loss": 0.1585,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 0.7229686500319897,
|
|
"grad_norm": 0.818102769388926,
|
|
"learning_rate": 2.163213642693059e-06,
|
|
"loss": 0.1647,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 0.7293666026871402,
|
|
"grad_norm": 0.7982865785711878,
|
|
"learning_rate": 2.0719303861396435e-06,
|
|
"loss": 0.1612,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 0.7357645553422905,
|
|
"grad_norm": 0.7894748084247665,
|
|
"learning_rate": 1.982108944649441e-06,
|
|
"loss": 0.1647,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.7421625079974408,
|
|
"grad_norm": 0.7777189002121984,
|
|
"learning_rate": 1.8937941608503484e-06,
|
|
"loss": 0.1611,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 0.7485604606525912,
|
|
"grad_norm": 0.8573964650662292,
|
|
"learning_rate": 1.8070301251836108e-06,
|
|
"loss": 0.1572,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 0.7549584133077415,
|
|
"grad_norm": 0.828180686527211,
|
|
"learning_rate": 1.721860153892011e-06,
|
|
"loss": 0.1646,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 0.7613563659628919,
|
|
"grad_norm": 0.7673100213386458,
|
|
"learning_rate": 1.6383267673945925e-06,
|
|
"loss": 0.1638,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 0.7677543186180422,
|
|
"grad_norm": 0.7980588121937006,
|
|
"learning_rate": 1.55647166905868e-06,
|
|
"loss": 0.1615,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.7741522712731925,
|
|
"grad_norm": 0.8416722320217981,
|
|
"learning_rate": 1.4763357243798154e-06,
|
|
"loss": 0.1605,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 0.780550223928343,
|
|
"grad_norm": 0.8079816411214684,
|
|
"learning_rate": 1.3979589405799865e-06,
|
|
"loss": 0.1599,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 0.7869481765834933,
|
|
"grad_norm": 0.7902784854915059,
|
|
"learning_rate": 1.321380446634342e-06,
|
|
"loss": 0.1624,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 0.7933461292386437,
|
|
"grad_norm": 0.7247178667276454,
|
|
"learning_rate": 1.246638473736378e-06,
|
|
"loss": 0.1612,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 0.799744081893794,
|
|
"grad_norm": 0.8349282846680564,
|
|
"learning_rate": 1.1737703362113134e-06,
|
|
"loss": 0.1582,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.8061420345489443,
|
|
"grad_norm": 0.7758747971349048,
|
|
"learning_rate": 1.1028124128872191e-06,
|
|
"loss": 0.1596,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 0.8125399872040947,
|
|
"grad_norm": 0.7491710921891626,
|
|
"learning_rate": 1.033800128933179e-06,
|
|
"loss": 0.1577,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 0.818937939859245,
|
|
"grad_norm": 0.7671028274511632,
|
|
"learning_rate": 9.667679381735706e-07,
|
|
"loss": 0.1671,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 0.8253358925143954,
|
|
"grad_norm": 0.7600814570026183,
|
|
"learning_rate": 9.017493058872623e-07,
|
|
"loss": 0.167,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 0.8317338451695457,
|
|
"grad_norm": 0.8123107698189449,
|
|
"learning_rate": 8.387766921003427e-07,
|
|
"loss": 0.161,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.838131797824696,
|
|
"grad_norm": 0.8343346354987566,
|
|
"learning_rate": 7.77881535380724e-07,
|
|
"loss": 0.159,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 0.8445297504798465,
|
|
"grad_norm": 0.7970234220031028,
|
|
"learning_rate": 7.190942371426862e-07,
|
|
"loss": 0.1607,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 0.8509277031349968,
|
|
"grad_norm": 0.8033298859243498,
|
|
"learning_rate": 6.624441464692161e-07,
|
|
"loss": 0.1579,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 0.8573256557901472,
|
|
"grad_norm": 0.7556300050803251,
|
|
"learning_rate": 6.07959545459717e-07,
|
|
"loss": 0.1549,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 0.8637236084452975,
|
|
"grad_norm": 0.8954691634947354,
|
|
"learning_rate": 5.55667635110399e-07,
|
|
"loss": 0.1648,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.8701215611004478,
|
|
"grad_norm": 0.7565296989489972,
|
|
"learning_rate": 5.055945217344004e-07,
|
|
"loss": 0.1649,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 0.8765195137555982,
|
|
"grad_norm": 0.8665145517705075,
|
|
"learning_rate": 4.5776520392842473e-07,
|
|
"loss": 0.1556,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 0.8829174664107485,
|
|
"grad_norm": 0.7971559437186148,
|
|
"learning_rate": 4.122035600923913e-07,
|
|
"loss": 0.1598,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 0.889315419065899,
|
|
"grad_norm": 0.7888135120323285,
|
|
"learning_rate": 3.6893233650833916e-07,
|
|
"loss": 0.161,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 0.8957133717210493,
|
|
"grad_norm": 0.8101852282848508,
|
|
"learning_rate": 3.2797313598452506e-07,
|
|
"loss": 0.1515,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.9021113243761996,
|
|
"grad_norm": 0.7573183367175816,
|
|
"learning_rate": 2.893464070704055e-07,
|
|
"loss": 0.1516,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 0.90850927703135,
|
|
"grad_norm": 0.7739598773716162,
|
|
"learning_rate": 2.53071433847854e-07,
|
|
"loss": 0.1642,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 0.9149072296865003,
|
|
"grad_norm": 0.7571146407052398,
|
|
"learning_rate": 2.1916632630374579e-07,
|
|
"loss": 0.1487,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 0.9213051823416507,
|
|
"grad_norm": 0.7545363679623804,
|
|
"learning_rate": 1.876480112886886e-07,
|
|
"loss": 0.1528,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 0.927703134996801,
|
|
"grad_norm": 0.7763887847298273,
|
|
"learning_rate": 1.5853222406643555e-07,
|
|
"loss": 0.1625,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.9341010876519513,
|
|
"grad_norm": 0.8511608028334107,
|
|
"learning_rate": 1.318335004581761e-07,
|
|
"loss": 0.1588,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 0.9404990403071017,
|
|
"grad_norm": 0.7718837723403269,
|
|
"learning_rate": 1.0756516958564667e-07,
|
|
"loss": 0.1627,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 0.946896992962252,
|
|
"grad_norm": 0.8086316020268939,
|
|
"learning_rate": 8.573934721667731e-08,
|
|
"loss": 0.1606,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 0.9532949456174025,
|
|
"grad_norm": 0.7932864945707389,
|
|
"learning_rate": 6.636692971648873e-08,
|
|
"loss": 0.1569,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 0.9596928982725528,
|
|
"grad_norm": 0.827992022510174,
|
|
"learning_rate": 4.9457588607772497e-08,
|
|
"loss": 0.1552,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.9660908509277031,
|
|
"grad_norm": 0.8088713583459414,
|
|
"learning_rate": 3.501976574226018e-08,
|
|
"loss": 0.1607,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 0.9724888035828535,
|
|
"grad_norm": 0.8835505171126804,
|
|
"learning_rate": 2.3060669086199526e-08,
|
|
"loss": 0.1576,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 0.9788867562380038,
|
|
"grad_norm": 0.8088348599596843,
|
|
"learning_rate": 1.3586269121833028e-08,
|
|
"loss": 0.1584,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 0.9852847088931542,
|
|
"grad_norm": 0.8101451300553387,
|
|
"learning_rate": 6.6012958666827886e-09,
|
|
"loss": 0.1682,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 0.9916826615483045,
|
|
"grad_norm": 0.8161568968787135,
|
|
"learning_rate": 2.1092365121305745e-09,
|
|
"loss": 0.1562,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.9980806142034548,
|
|
"grad_norm": 0.8089380625311189,
|
|
"learning_rate": 1.1233368246321708e-10,
|
|
"loss": 0.1588,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"step": 1563,
|
|
"total_flos": 51865560023040.0,
|
|
"train_loss": 0.20229712361261315,
|
|
"train_runtime": 1308.1188,
|
|
"train_samples_per_second": 76.446,
|
|
"train_steps_per_second": 1.195
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 1563,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 1000,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 51865560023040.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|