Borealis-ASR-4B-10k / trainer_state.json
AlexWortega's picture
Upload checkpoint-10000
a767f08 verified
Raw
History Blame Contribute Delete
35.8 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.331707964308223,
"eval_steps": 15000000,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0016585398215411152,
"grad_norm": 5.2646002769470215,
"learning_rate": 1.9503781345362876e-06,
"loss": 3.606,
"step": 50
},
{
"epoch": 0.0033170796430822304,
"grad_norm": 11.763270378112793,
"learning_rate": 3.940559904471274e-06,
"loss": 2.6238,
"step": 100
},
{
"epoch": 0.004975619464623346,
"grad_norm": 6.007935047149658,
"learning_rate": 5.930741674406262e-06,
"loss": 2.5644,
"step": 150
},
{
"epoch": 0.006634159286164461,
"grad_norm": 13.862035751342773,
"learning_rate": 7.920923444341248e-06,
"loss": 2.4601,
"step": 200
},
{
"epoch": 0.008292699107705576,
"grad_norm": 6.1305341720581055,
"learning_rate": 9.911105214276236e-06,
"loss": 2.4099,
"step": 250
},
{
"epoch": 0.009951238929246691,
"grad_norm": 9.804327964782715,
"learning_rate": 1.1901286984211225e-05,
"loss": 2.3409,
"step": 300
},
{
"epoch": 0.011609778750787806,
"grad_norm": 6.416324138641357,
"learning_rate": 1.3891468754146211e-05,
"loss": 2.2951,
"step": 350
},
{
"epoch": 0.013268318572328922,
"grad_norm": 9.413233757019043,
"learning_rate": 1.5881650524081196e-05,
"loss": 2.0864,
"step": 400
},
{
"epoch": 0.014926858393870037,
"grad_norm": 4.721926212310791,
"learning_rate": 1.7871832294016186e-05,
"loss": 0.7346,
"step": 450
},
{
"epoch": 0.016585398215411152,
"grad_norm": 3.284379720687866,
"learning_rate": 1.9862014063951173e-05,
"loss": 0.4516,
"step": 500
},
{
"epoch": 0.01824393803695227,
"grad_norm": 2.671109199523926,
"learning_rate": 2.185219583388616e-05,
"loss": 0.3903,
"step": 550
},
{
"epoch": 0.019902477858493382,
"grad_norm": 2.644299268722534,
"learning_rate": 2.384237760382115e-05,
"loss": 0.3512,
"step": 600
},
{
"epoch": 0.0215610176800345,
"grad_norm": 2.1769394874572754,
"learning_rate": 2.5832559373756132e-05,
"loss": 0.3298,
"step": 650
},
{
"epoch": 0.023219557501575613,
"grad_norm": 2.0835165977478027,
"learning_rate": 2.7822741143691122e-05,
"loss": 0.3162,
"step": 700
},
{
"epoch": 0.02487809732311673,
"grad_norm": 2.069329023361206,
"learning_rate": 2.981292291362611e-05,
"loss": 0.3185,
"step": 750
},
{
"epoch": 0.026536637144657843,
"grad_norm": 1.8482165336608887,
"learning_rate": 3.1803104683561096e-05,
"loss": 0.3057,
"step": 800
},
{
"epoch": 0.02819517696619896,
"grad_norm": 1.8083229064941406,
"learning_rate": 3.379328645349608e-05,
"loss": 0.2857,
"step": 850
},
{
"epoch": 0.029853716787740073,
"grad_norm": 1.7274185419082642,
"learning_rate": 3.578346822343107e-05,
"loss": 0.3166,
"step": 900
},
{
"epoch": 0.03151225660928119,
"grad_norm": 1.6723179817199707,
"learning_rate": 3.777364999336606e-05,
"loss": 0.2981,
"step": 950
},
{
"epoch": 0.033170796430822304,
"grad_norm": 1.9198503494262695,
"learning_rate": 3.976383176330104e-05,
"loss": 0.2818,
"step": 1000
},
{
"epoch": 0.03482933625236342,
"grad_norm": 1.680993914604187,
"learning_rate": 4.1754013533236035e-05,
"loss": 0.2909,
"step": 1050
},
{
"epoch": 0.03648787607390454,
"grad_norm": 1.6936819553375244,
"learning_rate": 4.374419530317102e-05,
"loss": 0.2871,
"step": 1100
},
{
"epoch": 0.03814641589544565,
"grad_norm": 1.460769534111023,
"learning_rate": 4.5734377073106e-05,
"loss": 0.2734,
"step": 1150
},
{
"epoch": 0.039804955716986765,
"grad_norm": 1.5563299655914307,
"learning_rate": 4.7724558843040995e-05,
"loss": 0.2883,
"step": 1200
},
{
"epoch": 0.04146349553852788,
"grad_norm": 1.4941402673721313,
"learning_rate": 4.971474061297598e-05,
"loss": 0.2858,
"step": 1250
},
{
"epoch": 0.043122035360069,
"grad_norm": 1.2444454431533813,
"learning_rate": 5.170492238291097e-05,
"loss": 0.2626,
"step": 1300
},
{
"epoch": 0.04478057518161011,
"grad_norm": 1.2899850606918335,
"learning_rate": 5.3695104152845955e-05,
"loss": 0.29,
"step": 1350
},
{
"epoch": 0.046439115003151225,
"grad_norm": 1.6327451467514038,
"learning_rate": 5.568528592278094e-05,
"loss": 0.2897,
"step": 1400
},
{
"epoch": 0.04809765482469234,
"grad_norm": 1.2772693634033203,
"learning_rate": 5.767546769271593e-05,
"loss": 0.2926,
"step": 1450
},
{
"epoch": 0.04975619464623346,
"grad_norm": 1.2572505474090576,
"learning_rate": 5.966564946265092e-05,
"loss": 0.2816,
"step": 1500
},
{
"epoch": 0.05141473446777457,
"grad_norm": 1.1974629163742065,
"learning_rate": 6.16558312325859e-05,
"loss": 0.2772,
"step": 1550
},
{
"epoch": 0.053073274289315686,
"grad_norm": 1.1929668188095093,
"learning_rate": 6.364601300252089e-05,
"loss": 0.289,
"step": 1600
},
{
"epoch": 0.0547318141108568,
"grad_norm": 1.2904795408248901,
"learning_rate": 6.563619477245587e-05,
"loss": 0.2792,
"step": 1650
},
{
"epoch": 0.05639035393239792,
"grad_norm": 1.169552206993103,
"learning_rate": 6.762637654239086e-05,
"loss": 0.2814,
"step": 1700
},
{
"epoch": 0.05804889375393903,
"grad_norm": 1.1359589099884033,
"learning_rate": 6.961655831232586e-05,
"loss": 0.2936,
"step": 1750
},
{
"epoch": 0.05970743357548015,
"grad_norm": 1.2144523859024048,
"learning_rate": 7.160674008226085e-05,
"loss": 0.272,
"step": 1800
},
{
"epoch": 0.06136597339702126,
"grad_norm": 1.1847233772277832,
"learning_rate": 7.359692185219582e-05,
"loss": 0.2635,
"step": 1850
},
{
"epoch": 0.06302451321856238,
"grad_norm": 1.231399655342102,
"learning_rate": 7.558710362213082e-05,
"loss": 0.2772,
"step": 1900
},
{
"epoch": 0.0646830530401035,
"grad_norm": 1.0221542119979858,
"learning_rate": 7.757728539206581e-05,
"loss": 0.2694,
"step": 1950
},
{
"epoch": 0.06634159286164461,
"grad_norm": 1.091098427772522,
"learning_rate": 7.956746716200078e-05,
"loss": 0.289,
"step": 2000
},
{
"epoch": 0.06800013268318572,
"grad_norm": 1.1065690517425537,
"learning_rate": 8.155764893193577e-05,
"loss": 0.2974,
"step": 2050
},
{
"epoch": 0.06965867250472683,
"grad_norm": 0.9319403767585754,
"learning_rate": 8.354783070187077e-05,
"loss": 0.2819,
"step": 2100
},
{
"epoch": 0.07131721232626795,
"grad_norm": 1.1582205295562744,
"learning_rate": 8.553801247180575e-05,
"loss": 0.276,
"step": 2150
},
{
"epoch": 0.07297575214780908,
"grad_norm": 0.9972165822982788,
"learning_rate": 8.752819424174074e-05,
"loss": 0.2814,
"step": 2200
},
{
"epoch": 0.07463429196935019,
"grad_norm": 1.2905242443084717,
"learning_rate": 8.951837601167573e-05,
"loss": 0.2688,
"step": 2250
},
{
"epoch": 0.0762928317908913,
"grad_norm": 0.944217324256897,
"learning_rate": 9.150855778161071e-05,
"loss": 0.2786,
"step": 2300
},
{
"epoch": 0.07795137161243242,
"grad_norm": 0.9599469900131226,
"learning_rate": 9.349873955154569e-05,
"loss": 0.2677,
"step": 2350
},
{
"epoch": 0.07960991143397353,
"grad_norm": 1.0232480764389038,
"learning_rate": 9.54889213214807e-05,
"loss": 0.2751,
"step": 2400
},
{
"epoch": 0.08126845125551464,
"grad_norm": 0.8944919109344482,
"learning_rate": 9.747910309141567e-05,
"loss": 0.2883,
"step": 2450
},
{
"epoch": 0.08292699107705576,
"grad_norm": 0.8229078650474548,
"learning_rate": 9.946928486135066e-05,
"loss": 0.2783,
"step": 2500
},
{
"epoch": 0.08458553089859687,
"grad_norm": 1.0346490144729614,
"learning_rate": 0.00010145946663128565,
"loss": 0.2774,
"step": 2550
},
{
"epoch": 0.086244070720138,
"grad_norm": 1.1092647314071655,
"learning_rate": 0.00010344964840122063,
"loss": 0.254,
"step": 2600
},
{
"epoch": 0.08790261054167911,
"grad_norm": 1.0459065437316895,
"learning_rate": 0.00010543983017115562,
"loss": 0.2907,
"step": 2650
},
{
"epoch": 0.08956115036322022,
"grad_norm": 0.8515977263450623,
"learning_rate": 0.00010743001194109062,
"loss": 0.2767,
"step": 2700
},
{
"epoch": 0.09121969018476134,
"grad_norm": 0.8556401133537292,
"learning_rate": 0.0001094201937110256,
"loss": 0.2657,
"step": 2750
},
{
"epoch": 0.09287823000630245,
"grad_norm": 0.9892629384994507,
"learning_rate": 0.00011141037548096058,
"loss": 0.2635,
"step": 2800
},
{
"epoch": 0.09453676982784356,
"grad_norm": 0.8110876679420471,
"learning_rate": 0.00011340055725089557,
"loss": 0.2783,
"step": 2850
},
{
"epoch": 0.09619530964938468,
"grad_norm": 0.9971825480461121,
"learning_rate": 0.00011539073902083055,
"loss": 0.2723,
"step": 2900
},
{
"epoch": 0.09785384947092579,
"grad_norm": 0.8159605860710144,
"learning_rate": 0.00011738092079076555,
"loss": 0.2904,
"step": 2950
},
{
"epoch": 0.09951238929246692,
"grad_norm": 0.8371638655662537,
"learning_rate": 0.00011937110256070054,
"loss": 0.2725,
"step": 3000
},
{
"epoch": 0.10117092911400803,
"grad_norm": 0.9144733548164368,
"learning_rate": 0.00012136128433063553,
"loss": 0.2733,
"step": 3050
},
{
"epoch": 0.10282946893554915,
"grad_norm": 0.9184954762458801,
"learning_rate": 0.0001233514661005705,
"loss": 0.2778,
"step": 3100
},
{
"epoch": 0.10448800875709026,
"grad_norm": 0.760306179523468,
"learning_rate": 0.00012534164787050549,
"loss": 0.2714,
"step": 3150
},
{
"epoch": 0.10614654857863137,
"grad_norm": 0.7679548263549805,
"learning_rate": 0.00012733182964044047,
"loss": 0.2607,
"step": 3200
},
{
"epoch": 0.10780508840017249,
"grad_norm": 0.899803876876831,
"learning_rate": 0.00012932201141037549,
"loss": 0.2572,
"step": 3250
},
{
"epoch": 0.1094636282217136,
"grad_norm": 0.7616294026374817,
"learning_rate": 0.00013131219318031047,
"loss": 0.2733,
"step": 3300
},
{
"epoch": 0.11112216804325471,
"grad_norm": 0.8336801528930664,
"learning_rate": 0.00013330237495024543,
"loss": 0.274,
"step": 3350
},
{
"epoch": 0.11278070786479584,
"grad_norm": 0.7640217542648315,
"learning_rate": 0.00013529255672018042,
"loss": 0.2784,
"step": 3400
},
{
"epoch": 0.11443924768633695,
"grad_norm": 0.7056282758712769,
"learning_rate": 0.0001372827384901154,
"loss": 0.2639,
"step": 3450
},
{
"epoch": 0.11609778750787807,
"grad_norm": 0.8658528327941895,
"learning_rate": 0.00013927292026005042,
"loss": 0.2681,
"step": 3500
},
{
"epoch": 0.11775632732941918,
"grad_norm": 0.802803099155426,
"learning_rate": 0.0001412631020299854,
"loss": 0.2963,
"step": 3550
},
{
"epoch": 0.1194148671509603,
"grad_norm": 0.8917680978775024,
"learning_rate": 0.0001432532837999204,
"loss": 0.2826,
"step": 3600
},
{
"epoch": 0.12107340697250141,
"grad_norm": 0.7213064432144165,
"learning_rate": 0.00014524346556985538,
"loss": 0.2758,
"step": 3650
},
{
"epoch": 0.12273194679404252,
"grad_norm": 0.6630449891090393,
"learning_rate": 0.00014723364733979034,
"loss": 0.2669,
"step": 3700
},
{
"epoch": 0.12439048661558363,
"grad_norm": 0.9187581539154053,
"learning_rate": 0.00014922382910972532,
"loss": 0.2664,
"step": 3750
},
{
"epoch": 0.12604902643712476,
"grad_norm": 0.7035880088806152,
"learning_rate": 0.0001512140108796603,
"loss": 0.2813,
"step": 3800
},
{
"epoch": 0.12770756625866586,
"grad_norm": 0.9671415686607361,
"learning_rate": 0.0001532041926495953,
"loss": 0.2892,
"step": 3850
},
{
"epoch": 0.129366106080207,
"grad_norm": 0.809038519859314,
"learning_rate": 0.00015519437441953028,
"loss": 0.2662,
"step": 3900
},
{
"epoch": 0.1310246459017481,
"grad_norm": 0.8309035301208496,
"learning_rate": 0.0001571845561894653,
"loss": 0.2567,
"step": 3950
},
{
"epoch": 0.13268318572328922,
"grad_norm": 0.7343366146087646,
"learning_rate": 0.00015917473795940028,
"loss": 0.2789,
"step": 4000
},
{
"epoch": 0.13434172554483034,
"grad_norm": 0.6523250341415405,
"learning_rate": 0.00016116491972933527,
"loss": 0.2568,
"step": 4050
},
{
"epoch": 0.13600026536637144,
"grad_norm": 0.6583340167999268,
"learning_rate": 0.00016315510149927026,
"loss": 0.2694,
"step": 4100
},
{
"epoch": 0.13765880518791257,
"grad_norm": 0.674410879611969,
"learning_rate": 0.00016514528326920524,
"loss": 0.2839,
"step": 4150
},
{
"epoch": 0.13931734500945367,
"grad_norm": 0.9803613424301147,
"learning_rate": 0.00016713546503914023,
"loss": 0.2789,
"step": 4200
},
{
"epoch": 0.1409758848309948,
"grad_norm": 0.6113103032112122,
"learning_rate": 0.00016912564680907522,
"loss": 0.263,
"step": 4250
},
{
"epoch": 0.1426344246525359,
"grad_norm": 0.6843391060829163,
"learning_rate": 0.0001711158285790102,
"loss": 0.2673,
"step": 4300
},
{
"epoch": 0.14429296447407702,
"grad_norm": 0.7611909508705139,
"learning_rate": 0.0001731060103489452,
"loss": 0.2628,
"step": 4350
},
{
"epoch": 0.14595150429561815,
"grad_norm": 0.7338739037513733,
"learning_rate": 0.00017509619211888015,
"loss": 0.2728,
"step": 4400
},
{
"epoch": 0.14761004411715925,
"grad_norm": 0.6578771471977234,
"learning_rate": 0.00017708637388881514,
"loss": 0.2503,
"step": 4450
},
{
"epoch": 0.14926858393870038,
"grad_norm": 0.6933615803718567,
"learning_rate": 0.00017907655565875018,
"loss": 0.2655,
"step": 4500
},
{
"epoch": 0.15092712376024148,
"grad_norm": 0.574677586555481,
"learning_rate": 0.00018106673742868514,
"loss": 0.2545,
"step": 4550
},
{
"epoch": 0.1525856635817826,
"grad_norm": 0.5872885584831238,
"learning_rate": 0.00018305691919862012,
"loss": 0.2728,
"step": 4600
},
{
"epoch": 0.1542442034033237,
"grad_norm": 0.7030934691429138,
"learning_rate": 0.0001850471009685551,
"loss": 0.2691,
"step": 4650
},
{
"epoch": 0.15590274322486483,
"grad_norm": 0.6045774817466736,
"learning_rate": 0.0001870372827384901,
"loss": 0.2676,
"step": 4700
},
{
"epoch": 0.15756128304640593,
"grad_norm": 0.6607248783111572,
"learning_rate": 0.00018902746450842508,
"loss": 0.2706,
"step": 4750
},
{
"epoch": 0.15921982286794706,
"grad_norm": 0.6813399195671082,
"learning_rate": 0.00019101764627836007,
"loss": 0.2602,
"step": 4800
},
{
"epoch": 0.16087836268948819,
"grad_norm": 0.6294276714324951,
"learning_rate": 0.00019300782804829506,
"loss": 0.2861,
"step": 4850
},
{
"epoch": 0.16253690251102929,
"grad_norm": 0.7689523696899414,
"learning_rate": 0.00019499800981823004,
"loss": 0.2919,
"step": 4900
},
{
"epoch": 0.1641954423325704,
"grad_norm": 0.6293140053749084,
"learning_rate": 0.00019698819158816503,
"loss": 0.2625,
"step": 4950
},
{
"epoch": 0.1658539821541115,
"grad_norm": 0.6510013937950134,
"learning_rate": 0.00019897837335810002,
"loss": 0.2645,
"step": 5000
},
{
"epoch": 0.16751252197565264,
"grad_norm": 0.655305027961731,
"learning_rate": 0.00020096855512803503,
"loss": 0.2899,
"step": 5050
},
{
"epoch": 0.16917106179719374,
"grad_norm": 0.6357027888298035,
"learning_rate": 0.00020295873689797002,
"loss": 0.2841,
"step": 5100
},
{
"epoch": 0.17082960161873487,
"grad_norm": 0.6427227258682251,
"learning_rate": 0.000204948918667905,
"loss": 0.2904,
"step": 5150
},
{
"epoch": 0.172488141440276,
"grad_norm": 0.5456770658493042,
"learning_rate": 0.00020693910043784,
"loss": 0.2606,
"step": 5200
},
{
"epoch": 0.1741466812618171,
"grad_norm": 0.6049277186393738,
"learning_rate": 0.00020892928220777495,
"loss": 0.2789,
"step": 5250
},
{
"epoch": 0.17580522108335822,
"grad_norm": 0.604241669178009,
"learning_rate": 0.00021091946397770994,
"loss": 0.2855,
"step": 5300
},
{
"epoch": 0.17746376090489932,
"grad_norm": 0.5764302015304565,
"learning_rate": 0.00021290964574764492,
"loss": 0.2724,
"step": 5350
},
{
"epoch": 0.17912230072644045,
"grad_norm": 0.5926195979118347,
"learning_rate": 0.0002148998275175799,
"loss": 0.2693,
"step": 5400
},
{
"epoch": 0.18078084054798155,
"grad_norm": 0.550534188747406,
"learning_rate": 0.0002168900092875149,
"loss": 0.2739,
"step": 5450
},
{
"epoch": 0.18243938036952267,
"grad_norm": 0.9231746792793274,
"learning_rate": 0.00021888019105744988,
"loss": 0.2713,
"step": 5500
},
{
"epoch": 0.18409792019106377,
"grad_norm": 0.6074857115745544,
"learning_rate": 0.00022087037282738487,
"loss": 0.2628,
"step": 5550
},
{
"epoch": 0.1857564600126049,
"grad_norm": 0.5123354196548462,
"learning_rate": 0.00022286055459731988,
"loss": 0.2752,
"step": 5600
},
{
"epoch": 0.18741499983414603,
"grad_norm": 0.5391439199447632,
"learning_rate": 0.00022485073636725487,
"loss": 0.2772,
"step": 5650
},
{
"epoch": 0.18907353965568713,
"grad_norm": 0.5493964552879333,
"learning_rate": 0.00022684091813718986,
"loss": 0.2656,
"step": 5700
},
{
"epoch": 0.19073207947722826,
"grad_norm": 0.531689465045929,
"learning_rate": 0.00022883109990712484,
"loss": 0.2652,
"step": 5750
},
{
"epoch": 0.19239061929876936,
"grad_norm": 0.6406611204147339,
"learning_rate": 0.00023082128167705983,
"loss": 0.2807,
"step": 5800
},
{
"epoch": 0.19404915912031048,
"grad_norm": 0.5147985219955444,
"learning_rate": 0.00023281146344699481,
"loss": 0.2859,
"step": 5850
},
{
"epoch": 0.19570769894185158,
"grad_norm": 0.5153710246086121,
"learning_rate": 0.0002348016452169298,
"loss": 0.2529,
"step": 5900
},
{
"epoch": 0.1973662387633927,
"grad_norm": 0.5277467966079712,
"learning_rate": 0.00023679182698686476,
"loss": 0.2731,
"step": 5950
},
{
"epoch": 0.19902477858493384,
"grad_norm": 0.480101615190506,
"learning_rate": 0.00023878200875679975,
"loss": 0.2784,
"step": 6000
},
{
"epoch": 0.20068331840647494,
"grad_norm": 0.5489948987960815,
"learning_rate": 0.00024077219052673473,
"loss": 0.2968,
"step": 6050
},
{
"epoch": 0.20234185822801606,
"grad_norm": 0.6148312091827393,
"learning_rate": 0.00024276237229666972,
"loss": 0.2731,
"step": 6100
},
{
"epoch": 0.20400039804955716,
"grad_norm": 0.5747624039649963,
"learning_rate": 0.00024475255406660476,
"loss": 0.2701,
"step": 6150
},
{
"epoch": 0.2056589378710983,
"grad_norm": 0.6135842800140381,
"learning_rate": 0.00024674273583653975,
"loss": 0.2691,
"step": 6200
},
{
"epoch": 0.2073174776926394,
"grad_norm": 0.5454925298690796,
"learning_rate": 0.00024873291760647473,
"loss": 0.2737,
"step": 6250
},
{
"epoch": 0.20897601751418052,
"grad_norm": 0.6004371047019958,
"learning_rate": 0.0002507230993764097,
"loss": 0.288,
"step": 6300
},
{
"epoch": 0.21063455733572162,
"grad_norm": 0.5263451933860779,
"learning_rate": 0.0002527132811463447,
"loss": 0.2752,
"step": 6350
},
{
"epoch": 0.21229309715726274,
"grad_norm": 0.5507785677909851,
"learning_rate": 0.00025470346291627964,
"loss": 0.2749,
"step": 6400
},
{
"epoch": 0.21395163697880387,
"grad_norm": 0.5042358636856079,
"learning_rate": 0.0002566936446862146,
"loss": 0.2773,
"step": 6450
},
{
"epoch": 0.21561017680034497,
"grad_norm": 0.4615878164768219,
"learning_rate": 0.0002586838264561496,
"loss": 0.2782,
"step": 6500
},
{
"epoch": 0.2172687166218861,
"grad_norm": 0.5063010454177856,
"learning_rate": 0.0002606740082260846,
"loss": 0.2805,
"step": 6550
},
{
"epoch": 0.2189272564434272,
"grad_norm": 0.5367996096611023,
"learning_rate": 0.0002626641899960196,
"loss": 0.2689,
"step": 6600
},
{
"epoch": 0.22058579626496833,
"grad_norm": 0.4879344701766968,
"learning_rate": 0.0002646543717659546,
"loss": 0.2745,
"step": 6650
},
{
"epoch": 0.22224433608650943,
"grad_norm": 0.504068911075592,
"learning_rate": 0.00026664455353588956,
"loss": 0.2685,
"step": 6700
},
{
"epoch": 0.22390287590805055,
"grad_norm": 0.4813278913497925,
"learning_rate": 0.0002686347353058246,
"loss": 0.261,
"step": 6750
},
{
"epoch": 0.22556141572959168,
"grad_norm": 0.46839526295661926,
"learning_rate": 0.0002706249170757596,
"loss": 0.2702,
"step": 6800
},
{
"epoch": 0.22721995555113278,
"grad_norm": 0.5233504176139832,
"learning_rate": 0.0002726150988456946,
"loss": 0.2736,
"step": 6850
},
{
"epoch": 0.2288784953726739,
"grad_norm": 0.5277813673019409,
"learning_rate": 0.00027460528061562956,
"loss": 0.2772,
"step": 6900
},
{
"epoch": 0.230537035194215,
"grad_norm": 0.6029180884361267,
"learning_rate": 0.00027659546238556455,
"loss": 0.2768,
"step": 6950
},
{
"epoch": 0.23219557501575613,
"grad_norm": 0.575391948223114,
"learning_rate": 0.00027858564415549953,
"loss": 0.265,
"step": 7000
},
{
"epoch": 0.23385411483729723,
"grad_norm": 0.4373869299888611,
"learning_rate": 0.0002805758259254345,
"loss": 0.2657,
"step": 7050
},
{
"epoch": 0.23551265465883836,
"grad_norm": 0.8708984851837158,
"learning_rate": 0.00028256600769536945,
"loss": 0.2754,
"step": 7100
},
{
"epoch": 0.23717119448037946,
"grad_norm": 0.5389964580535889,
"learning_rate": 0.00028455618946530444,
"loss": 0.2654,
"step": 7150
},
{
"epoch": 0.2388297343019206,
"grad_norm": 0.5265533328056335,
"learning_rate": 0.0002865463712352394,
"loss": 0.2943,
"step": 7200
},
{
"epoch": 0.24048827412346171,
"grad_norm": 0.43175622820854187,
"learning_rate": 0.0002885365530051744,
"loss": 0.2597,
"step": 7250
},
{
"epoch": 0.24214681394500281,
"grad_norm": 0.42523327469825745,
"learning_rate": 0.00029052673477510945,
"loss": 0.281,
"step": 7300
},
{
"epoch": 0.24380535376654394,
"grad_norm": 0.4567892551422119,
"learning_rate": 0.00029251691654504444,
"loss": 0.2776,
"step": 7350
},
{
"epoch": 0.24546389358808504,
"grad_norm": 0.4635256826877594,
"learning_rate": 0.0002945070983149794,
"loss": 0.2608,
"step": 7400
},
{
"epoch": 0.24712243340962617,
"grad_norm": 0.43223607540130615,
"learning_rate": 0.0002964972800849144,
"loss": 0.2577,
"step": 7450
},
{
"epoch": 0.24878097323116727,
"grad_norm": 0.4766380190849304,
"learning_rate": 0.0002984874618548494,
"loss": 0.2717,
"step": 7500
},
{
"epoch": 0.2504395130527084,
"grad_norm": 0.47827085852622986,
"learning_rate": 0.0002999999948018406,
"loss": 0.2748,
"step": 7550
},
{
"epoch": 0.2520980528742495,
"grad_norm": 0.5068143606185913,
"learning_rate": 0.000299999861238043,
"loss": 0.2786,
"step": 7600
},
{
"epoch": 0.25375659269579065,
"grad_norm": 0.48091641068458557,
"learning_rate": 0.0002999995471827833,
"loss": 0.2623,
"step": 7650
},
{
"epoch": 0.2554151325173317,
"grad_norm": 0.5682414770126343,
"learning_rate": 0.00029999905263643944,
"loss": 0.2784,
"step": 7700
},
{
"epoch": 0.25707367233887285,
"grad_norm": 0.42487502098083496,
"learning_rate": 0.00029999837759960636,
"loss": 0.2632,
"step": 7750
},
{
"epoch": 0.258732212160414,
"grad_norm": 0.3851436674594879,
"learning_rate": 0.00029999752207309643,
"loss": 0.262,
"step": 7800
},
{
"epoch": 0.2603907519819551,
"grad_norm": 0.4501873552799225,
"learning_rate": 0.000299996486057939,
"loss": 0.2579,
"step": 7850
},
{
"epoch": 0.2620492918034962,
"grad_norm": 0.42192208766937256,
"learning_rate": 0.00029999526955538073,
"loss": 0.2651,
"step": 7900
},
{
"epoch": 0.2637078316250373,
"grad_norm": 0.3992311656475067,
"learning_rate": 0.0002999938725668854,
"loss": 0.2687,
"step": 7950
},
{
"epoch": 0.26536637144657843,
"grad_norm": 0.4475759267807007,
"learning_rate": 0.000299992295094134,
"loss": 0.262,
"step": 8000
},
{
"epoch": 0.26702491126811956,
"grad_norm": 1.0701886415481567,
"learning_rate": 0.00029999053713902464,
"loss": 0.2702,
"step": 8050
},
{
"epoch": 0.2686834510896607,
"grad_norm": 0.7080543637275696,
"learning_rate": 0.00029998859870367263,
"loss": 0.2752,
"step": 8100
},
{
"epoch": 0.27034199091120176,
"grad_norm": 0.47246304154396057,
"learning_rate": 0.0002999864797904105,
"loss": 0.2708,
"step": 8150
},
{
"epoch": 0.2720005307327429,
"grad_norm": 0.3701266348361969,
"learning_rate": 0.0002999841804017878,
"loss": 0.2607,
"step": 8200
},
{
"epoch": 0.273659070554284,
"grad_norm": 0.41229259967803955,
"learning_rate": 0.00029998170054057135,
"loss": 0.2618,
"step": 8250
},
{
"epoch": 0.27531761037582514,
"grad_norm": 0.41077643632888794,
"learning_rate": 0.00029997904020974517,
"loss": 0.2554,
"step": 8300
},
{
"epoch": 0.27697615019736627,
"grad_norm": 0.36561328172683716,
"learning_rate": 0.0002999761994125103,
"loss": 0.2764,
"step": 8350
},
{
"epoch": 0.27863469001890734,
"grad_norm": 0.4316375255584717,
"learning_rate": 0.00029997317815228506,
"loss": 0.2678,
"step": 8400
},
{
"epoch": 0.28029322984044847,
"grad_norm": 0.4210038483142853,
"learning_rate": 0.0002999699764327049,
"loss": 0.2462,
"step": 8450
},
{
"epoch": 0.2819517696619896,
"grad_norm": 0.4256073236465454,
"learning_rate": 0.0002999665942576223,
"loss": 0.271,
"step": 8500
},
{
"epoch": 0.2836103094835307,
"grad_norm": 0.40567606687545776,
"learning_rate": 0.000299963031631107,
"loss": 0.2574,
"step": 8550
},
{
"epoch": 0.2852688493050718,
"grad_norm": 0.3660525381565094,
"learning_rate": 0.00029995928855744577,
"loss": 0.2697,
"step": 8600
},
{
"epoch": 0.2869273891266129,
"grad_norm": 0.44151097536087036,
"learning_rate": 0.0002999553650411427,
"loss": 0.2508,
"step": 8650
},
{
"epoch": 0.28858592894815405,
"grad_norm": 0.38013362884521484,
"learning_rate": 0.00029995126108691865,
"loss": 0.237,
"step": 8700
},
{
"epoch": 0.2902444687696952,
"grad_norm": 0.44534966349601746,
"learning_rate": 0.00029994697669971204,
"loss": 0.2575,
"step": 8750
},
{
"epoch": 0.2919030085912363,
"grad_norm": 0.40814974904060364,
"learning_rate": 0.00029994251188467813,
"loss": 0.2494,
"step": 8800
},
{
"epoch": 0.2935615484127774,
"grad_norm": 0.3318258821964264,
"learning_rate": 0.0002999378666471892,
"loss": 0.2381,
"step": 8850
},
{
"epoch": 0.2952200882343185,
"grad_norm": 0.43739181756973267,
"learning_rate": 0.00029993304099283483,
"loss": 0.2388,
"step": 8900
},
{
"epoch": 0.29687862805585963,
"grad_norm": 0.3652569055557251,
"learning_rate": 0.00029992803492742166,
"loss": 0.2678,
"step": 8950
},
{
"epoch": 0.29853716787740076,
"grad_norm": 0.37581753730773926,
"learning_rate": 0.0002999228484569734,
"loss": 0.258,
"step": 9000
},
{
"epoch": 0.3001957076989418,
"grad_norm": 0.36842867732048035,
"learning_rate": 0.0002999174815877307,
"loss": 0.241,
"step": 9050
},
{
"epoch": 0.30185424752048295,
"grad_norm": 0.36468687653541565,
"learning_rate": 0.0002999119343261515,
"loss": 0.2601,
"step": 9100
},
{
"epoch": 0.3035127873420241,
"grad_norm": 0.44215506315231323,
"learning_rate": 0.0002999062066789106,
"loss": 0.2577,
"step": 9150
},
{
"epoch": 0.3051713271635652,
"grad_norm": 0.379226952791214,
"learning_rate": 0.00029990029865290007,
"loss": 0.2509,
"step": 9200
},
{
"epoch": 0.30682986698510634,
"grad_norm": 0.3723479211330414,
"learning_rate": 0.0002998942102552288,
"loss": 0.242,
"step": 9250
},
{
"epoch": 0.3084884068066474,
"grad_norm": 0.34578976035118103,
"learning_rate": 0.00029988794149322283,
"loss": 0.2791,
"step": 9300
},
{
"epoch": 0.31014694662818854,
"grad_norm": 0.33455216884613037,
"learning_rate": 0.0002998814923744253,
"loss": 0.2618,
"step": 9350
},
{
"epoch": 0.31180548644972966,
"grad_norm": 0.3505556583404541,
"learning_rate": 0.0002998748629065962,
"loss": 0.2416,
"step": 9400
},
{
"epoch": 0.3134640262712708,
"grad_norm": 0.3335639536380768,
"learning_rate": 0.00029986805309771276,
"loss": 0.2321,
"step": 9450
},
{
"epoch": 0.31512256609281186,
"grad_norm": 0.31904950737953186,
"learning_rate": 0.00029986106295596884,
"loss": 0.2405,
"step": 9500
},
{
"epoch": 0.316781105914353,
"grad_norm": 0.41776758432388306,
"learning_rate": 0.0002998538924897757,
"loss": 0.2359,
"step": 9550
},
{
"epoch": 0.3184396457358941,
"grad_norm": 0.3445201814174652,
"learning_rate": 0.0002998465417077613,
"loss": 0.2451,
"step": 9600
},
{
"epoch": 0.32009818555743524,
"grad_norm": 0.36852407455444336,
"learning_rate": 0.0002998390106187708,
"loss": 0.238,
"step": 9650
},
{
"epoch": 0.32175672537897637,
"grad_norm": 0.3637458384037018,
"learning_rate": 0.00029983129923186614,
"loss": 0.2286,
"step": 9700
},
{
"epoch": 0.32341526520051744,
"grad_norm": 0.35523203015327454,
"learning_rate": 0.00029982340755632615,
"loss": 0.2649,
"step": 9750
},
{
"epoch": 0.32507380502205857,
"grad_norm": 0.4552542567253113,
"learning_rate": 0.00029981533560164683,
"loss": 0.2462,
"step": 9800
},
{
"epoch": 0.3267323448435997,
"grad_norm": 0.3182304799556732,
"learning_rate": 0.0002998070833775409,
"loss": 0.259,
"step": 9850
},
{
"epoch": 0.3283908846651408,
"grad_norm": 0.3835619390010834,
"learning_rate": 0.0002997986508939382,
"loss": 0.2628,
"step": 9900
},
{
"epoch": 0.33004942448668195,
"grad_norm": 0.3364975154399872,
"learning_rate": 0.00029979003816098514,
"loss": 0.2512,
"step": 9950
},
{
"epoch": 0.331707964308223,
"grad_norm": 0.3546120822429657,
"learning_rate": 0.0002997812451890454,
"loss": 0.2312,
"step": 10000
}
],
"logging_steps": 50,
"max_steps": 150735,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 10000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}