task-13-google-gemma-2-2b / trainer_state.json
ManyingZ's picture
Upload folder using huggingface_hub
1cfc045 verified
raw
history blame
45.8 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 24.405125076266017,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.09762050030506407,
"grad_norm": 1.6181267499923706,
"learning_rate": 4e-05,
"loss": 2.5822,
"step": 20
},
{
"epoch": 0.19524100061012814,
"grad_norm": 1.536891222000122,
"learning_rate": 8e-05,
"loss": 2.2602,
"step": 40
},
{
"epoch": 0.2928615009151922,
"grad_norm": 1.1685199737548828,
"learning_rate": 0.00012,
"loss": 1.7209,
"step": 60
},
{
"epoch": 0.3904820012202563,
"grad_norm": 1.5937044620513916,
"learning_rate": 0.00016,
"loss": 1.6235,
"step": 80
},
{
"epoch": 0.4881025015253203,
"grad_norm": 2.1078925132751465,
"learning_rate": 0.0002,
"loss": 1.5259,
"step": 100
},
{
"epoch": 0.5857230018303844,
"grad_norm": 2.369110345840454,
"learning_rate": 0.0001998688524590164,
"loss": 1.5145,
"step": 120
},
{
"epoch": 0.6833435021354485,
"grad_norm": 2.531956434249878,
"learning_rate": 0.0001997377049180328,
"loss": 1.4408,
"step": 140
},
{
"epoch": 0.7809640024405126,
"grad_norm": 2.254030466079712,
"learning_rate": 0.00019960655737704918,
"loss": 1.3881,
"step": 160
},
{
"epoch": 0.8785845027455765,
"grad_norm": 1.8673783540725708,
"learning_rate": 0.0001994754098360656,
"loss": 1.3656,
"step": 180
},
{
"epoch": 0.9762050030506406,
"grad_norm": 2.6715195178985596,
"learning_rate": 0.00019934426229508198,
"loss": 1.2812,
"step": 200
},
{
"epoch": 1.0738255033557047,
"grad_norm": 1.7909507751464844,
"learning_rate": 0.00019921311475409837,
"loss": 1.3006,
"step": 220
},
{
"epoch": 1.1714460036607688,
"grad_norm": 2.6789026260375977,
"learning_rate": 0.00019908196721311476,
"loss": 1.2214,
"step": 240
},
{
"epoch": 1.2690665039658329,
"grad_norm": 2.88253116607666,
"learning_rate": 0.00019895081967213115,
"loss": 1.2806,
"step": 260
},
{
"epoch": 1.366687004270897,
"grad_norm": 1.9062846899032593,
"learning_rate": 0.00019881967213114757,
"loss": 1.2342,
"step": 280
},
{
"epoch": 1.4643075045759608,
"grad_norm": 2.622042179107666,
"learning_rate": 0.00019868852459016393,
"loss": 1.1657,
"step": 300
},
{
"epoch": 1.561928004881025,
"grad_norm": 2.6899402141571045,
"learning_rate": 0.00019855737704918035,
"loss": 1.1497,
"step": 320
},
{
"epoch": 1.659548505186089,
"grad_norm": 3.025324821472168,
"learning_rate": 0.00019842622950819674,
"loss": 1.0834,
"step": 340
},
{
"epoch": 1.757169005491153,
"grad_norm": 2.729680061340332,
"learning_rate": 0.00019829508196721313,
"loss": 1.1031,
"step": 360
},
{
"epoch": 1.8547895057962172,
"grad_norm": 3.1579582691192627,
"learning_rate": 0.00019816393442622951,
"loss": 1.1726,
"step": 380
},
{
"epoch": 1.9524100061012812,
"grad_norm": 2.698084592819214,
"learning_rate": 0.0001980327868852459,
"loss": 1.1241,
"step": 400
},
{
"epoch": 2.0500305064063453,
"grad_norm": 2.4257960319519043,
"learning_rate": 0.00019790163934426232,
"loss": 1.0619,
"step": 420
},
{
"epoch": 2.1476510067114094,
"grad_norm": 3.122441291809082,
"learning_rate": 0.00019777049180327868,
"loss": 1.0413,
"step": 440
},
{
"epoch": 2.2452715070164735,
"grad_norm": 3.2882649898529053,
"learning_rate": 0.0001976393442622951,
"loss": 0.9949,
"step": 460
},
{
"epoch": 2.3428920073215376,
"grad_norm": 2.3513290882110596,
"learning_rate": 0.0001975081967213115,
"loss": 1.0627,
"step": 480
},
{
"epoch": 2.4405125076266017,
"grad_norm": 2.9709153175354004,
"learning_rate": 0.00019737704918032788,
"loss": 1.0612,
"step": 500
},
{
"epoch": 2.5381330079316657,
"grad_norm": 4.2656569480896,
"learning_rate": 0.00019724590163934427,
"loss": 0.9615,
"step": 520
},
{
"epoch": 2.63575350823673,
"grad_norm": 2.7959163188934326,
"learning_rate": 0.00019711475409836066,
"loss": 0.9915,
"step": 540
},
{
"epoch": 2.733374008541794,
"grad_norm": 2.7856175899505615,
"learning_rate": 0.00019698360655737707,
"loss": 1.0299,
"step": 560
},
{
"epoch": 2.830994508846858,
"grad_norm": 3.0926871299743652,
"learning_rate": 0.00019685245901639344,
"loss": 0.9642,
"step": 580
},
{
"epoch": 2.9286150091519216,
"grad_norm": 4.457515239715576,
"learning_rate": 0.00019672131147540985,
"loss": 0.9922,
"step": 600
},
{
"epoch": 3.026235509456986,
"grad_norm": 3.0343308448791504,
"learning_rate": 0.00019659016393442624,
"loss": 0.9777,
"step": 620
},
{
"epoch": 3.1238560097620502,
"grad_norm": 3.12493896484375,
"learning_rate": 0.00019645901639344263,
"loss": 0.8765,
"step": 640
},
{
"epoch": 3.221476510067114,
"grad_norm": 3.1015610694885254,
"learning_rate": 0.00019632786885245902,
"loss": 0.8882,
"step": 660
},
{
"epoch": 3.319097010372178,
"grad_norm": 3.334591865539551,
"learning_rate": 0.0001961967213114754,
"loss": 0.8958,
"step": 680
},
{
"epoch": 3.416717510677242,
"grad_norm": 3.3690192699432373,
"learning_rate": 0.00019606557377049183,
"loss": 0.954,
"step": 700
},
{
"epoch": 3.514338010982306,
"grad_norm": 3.070910930633545,
"learning_rate": 0.00019593442622950822,
"loss": 0.8964,
"step": 720
},
{
"epoch": 3.61195851128737,
"grad_norm": 3.369615077972412,
"learning_rate": 0.0001958032786885246,
"loss": 0.8758,
"step": 740
},
{
"epoch": 3.7095790115924343,
"grad_norm": 2.3834476470947266,
"learning_rate": 0.000195672131147541,
"loss": 0.9307,
"step": 760
},
{
"epoch": 3.8071995118974984,
"grad_norm": 2.924922466278076,
"learning_rate": 0.00019554098360655738,
"loss": 0.8911,
"step": 780
},
{
"epoch": 3.9048200122025625,
"grad_norm": 3.321655035018921,
"learning_rate": 0.00019540983606557377,
"loss": 0.86,
"step": 800
},
{
"epoch": 4.002440512507627,
"grad_norm": 4.9272074699401855,
"learning_rate": 0.00019527868852459016,
"loss": 0.9365,
"step": 820
},
{
"epoch": 4.100061012812691,
"grad_norm": 3.4161181449890137,
"learning_rate": 0.00019514754098360658,
"loss": 0.804,
"step": 840
},
{
"epoch": 4.197681513117755,
"grad_norm": 2.86344051361084,
"learning_rate": 0.00019501639344262297,
"loss": 0.8029,
"step": 860
},
{
"epoch": 4.295302013422819,
"grad_norm": 3.9372735023498535,
"learning_rate": 0.00019488524590163936,
"loss": 0.7969,
"step": 880
},
{
"epoch": 4.392922513727883,
"grad_norm": 3.2105510234832764,
"learning_rate": 0.00019475409836065575,
"loss": 0.7838,
"step": 900
},
{
"epoch": 4.490543014032947,
"grad_norm": 3.1704697608947754,
"learning_rate": 0.00019462295081967214,
"loss": 0.8246,
"step": 920
},
{
"epoch": 4.588163514338011,
"grad_norm": 2.9047462940216064,
"learning_rate": 0.00019449180327868855,
"loss": 0.79,
"step": 940
},
{
"epoch": 4.685784014643075,
"grad_norm": 3.0593173503875732,
"learning_rate": 0.00019436065573770491,
"loss": 0.8105,
"step": 960
},
{
"epoch": 4.783404514948139,
"grad_norm": 3.4904775619506836,
"learning_rate": 0.00019422950819672133,
"loss": 0.864,
"step": 980
},
{
"epoch": 4.881025015253203,
"grad_norm": 3.080754280090332,
"learning_rate": 0.00019409836065573772,
"loss": 0.8163,
"step": 1000
},
{
"epoch": 4.978645515558267,
"grad_norm": 3.663107395172119,
"learning_rate": 0.0001939672131147541,
"loss": 0.8192,
"step": 1020
},
{
"epoch": 5.0762660158633315,
"grad_norm": 3.986875534057617,
"learning_rate": 0.0001938360655737705,
"loss": 0.7943,
"step": 1040
},
{
"epoch": 5.173886516168396,
"grad_norm": 3.140963554382324,
"learning_rate": 0.0001937049180327869,
"loss": 0.7274,
"step": 1060
},
{
"epoch": 5.27150701647346,
"grad_norm": 4.659877300262451,
"learning_rate": 0.0001935737704918033,
"loss": 0.7494,
"step": 1080
},
{
"epoch": 5.369127516778524,
"grad_norm": 4.54330587387085,
"learning_rate": 0.00019344262295081967,
"loss": 0.6986,
"step": 1100
},
{
"epoch": 5.466748017083588,
"grad_norm": 3.893850088119507,
"learning_rate": 0.00019331147540983608,
"loss": 0.7555,
"step": 1120
},
{
"epoch": 5.564368517388652,
"grad_norm": 3.24857759475708,
"learning_rate": 0.00019318032786885247,
"loss": 0.721,
"step": 1140
},
{
"epoch": 5.661989017693716,
"grad_norm": 4.82593297958374,
"learning_rate": 0.00019304918032786886,
"loss": 0.7523,
"step": 1160
},
{
"epoch": 5.75960951799878,
"grad_norm": 3.055907726287842,
"learning_rate": 0.00019291803278688525,
"loss": 0.7719,
"step": 1180
},
{
"epoch": 5.857230018303844,
"grad_norm": 3.192716121673584,
"learning_rate": 0.00019278688524590164,
"loss": 0.7732,
"step": 1200
},
{
"epoch": 5.954850518608908,
"grad_norm": 3.196037769317627,
"learning_rate": 0.00019265573770491806,
"loss": 0.745,
"step": 1220
},
{
"epoch": 6.052471018913972,
"grad_norm": 3.7806878089904785,
"learning_rate": 0.00019252459016393442,
"loss": 0.713,
"step": 1240
},
{
"epoch": 6.150091519219036,
"grad_norm": 3.6102254390716553,
"learning_rate": 0.00019239344262295084,
"loss": 0.6063,
"step": 1260
},
{
"epoch": 6.2477120195241005,
"grad_norm": 4.120365142822266,
"learning_rate": 0.00019226229508196723,
"loss": 0.6771,
"step": 1280
},
{
"epoch": 6.345332519829164,
"grad_norm": 3.118666887283325,
"learning_rate": 0.00019213114754098362,
"loss": 0.7173,
"step": 1300
},
{
"epoch": 6.442953020134228,
"grad_norm": 3.8055977821350098,
"learning_rate": 0.000192,
"loss": 0.6609,
"step": 1320
},
{
"epoch": 6.540573520439292,
"grad_norm": 3.4207704067230225,
"learning_rate": 0.0001918688524590164,
"loss": 0.7084,
"step": 1340
},
{
"epoch": 6.638194020744356,
"grad_norm": 3.812415838241577,
"learning_rate": 0.0001917377049180328,
"loss": 0.744,
"step": 1360
},
{
"epoch": 6.73581452104942,
"grad_norm": 4.302225112915039,
"learning_rate": 0.00019160655737704917,
"loss": 0.667,
"step": 1380
},
{
"epoch": 6.833435021354484,
"grad_norm": 3.3958733081817627,
"learning_rate": 0.0001914754098360656,
"loss": 0.6769,
"step": 1400
},
{
"epoch": 6.931055521659548,
"grad_norm": 3.6200642585754395,
"learning_rate": 0.00019134426229508198,
"loss": 0.7494,
"step": 1420
},
{
"epoch": 7.028676021964612,
"grad_norm": 3.2580292224884033,
"learning_rate": 0.00019121311475409837,
"loss": 0.6252,
"step": 1440
},
{
"epoch": 7.126296522269676,
"grad_norm": 3.581437826156616,
"learning_rate": 0.00019108196721311476,
"loss": 0.6072,
"step": 1460
},
{
"epoch": 7.22391702257474,
"grad_norm": 3.7347512245178223,
"learning_rate": 0.00019095081967213115,
"loss": 0.5715,
"step": 1480
},
{
"epoch": 7.3215375228798045,
"grad_norm": 3.564328193664551,
"learning_rate": 0.00019081967213114756,
"loss": 0.5937,
"step": 1500
},
{
"epoch": 7.419158023184869,
"grad_norm": 3.8675971031188965,
"learning_rate": 0.00019068852459016395,
"loss": 0.6441,
"step": 1520
},
{
"epoch": 7.516778523489933,
"grad_norm": 3.7394795417785645,
"learning_rate": 0.00019055737704918034,
"loss": 0.6345,
"step": 1540
},
{
"epoch": 7.614399023794997,
"grad_norm": 3.7391791343688965,
"learning_rate": 0.00019042622950819673,
"loss": 0.6334,
"step": 1560
},
{
"epoch": 7.712019524100061,
"grad_norm": 3.2816712856292725,
"learning_rate": 0.00019029508196721312,
"loss": 0.6448,
"step": 1580
},
{
"epoch": 7.809640024405125,
"grad_norm": 3.741111993789673,
"learning_rate": 0.0001901639344262295,
"loss": 0.6605,
"step": 1600
},
{
"epoch": 7.907260524710189,
"grad_norm": 3.5533151626586914,
"learning_rate": 0.0001900327868852459,
"loss": 0.6434,
"step": 1620
},
{
"epoch": 8.004881025015253,
"grad_norm": 3.243546724319458,
"learning_rate": 0.00018990163934426232,
"loss": 0.6923,
"step": 1640
},
{
"epoch": 8.102501525320317,
"grad_norm": 3.860666513442993,
"learning_rate": 0.0001897704918032787,
"loss": 0.5619,
"step": 1660
},
{
"epoch": 8.200122025625381,
"grad_norm": 3.4826905727386475,
"learning_rate": 0.0001896393442622951,
"loss": 0.5446,
"step": 1680
},
{
"epoch": 8.297742525930445,
"grad_norm": 5.119688034057617,
"learning_rate": 0.00018950819672131148,
"loss": 0.5525,
"step": 1700
},
{
"epoch": 8.39536302623551,
"grad_norm": 3.46353816986084,
"learning_rate": 0.00018937704918032787,
"loss": 0.5346,
"step": 1720
},
{
"epoch": 8.492983526540574,
"grad_norm": 4.458425045013428,
"learning_rate": 0.0001892459016393443,
"loss": 0.6127,
"step": 1740
},
{
"epoch": 8.590604026845638,
"grad_norm": 3.592191457748413,
"learning_rate": 0.00018911475409836065,
"loss": 0.5721,
"step": 1760
},
{
"epoch": 8.688224527150702,
"grad_norm": 4.131028652191162,
"learning_rate": 0.00018898360655737707,
"loss": 0.5799,
"step": 1780
},
{
"epoch": 8.785845027455766,
"grad_norm": 3.6534175872802734,
"learning_rate": 0.00018885245901639346,
"loss": 0.6198,
"step": 1800
},
{
"epoch": 8.88346552776083,
"grad_norm": 3.0888559818267822,
"learning_rate": 0.00018872131147540985,
"loss": 0.6309,
"step": 1820
},
{
"epoch": 8.981086028065894,
"grad_norm": 2.7711477279663086,
"learning_rate": 0.00018859016393442624,
"loss": 0.601,
"step": 1840
},
{
"epoch": 9.078706528370958,
"grad_norm": 2.7783164978027344,
"learning_rate": 0.00018845901639344263,
"loss": 0.5696,
"step": 1860
},
{
"epoch": 9.176327028676022,
"grad_norm": 3.6661510467529297,
"learning_rate": 0.00018832786885245904,
"loss": 0.5214,
"step": 1880
},
{
"epoch": 9.273947528981086,
"grad_norm": 3.4614779949188232,
"learning_rate": 0.0001881967213114754,
"loss": 0.5181,
"step": 1900
},
{
"epoch": 9.37156802928615,
"grad_norm": 3.470071315765381,
"learning_rate": 0.00018806557377049182,
"loss": 0.5196,
"step": 1920
},
{
"epoch": 9.469188529591214,
"grad_norm": 3.686056613922119,
"learning_rate": 0.0001879344262295082,
"loss": 0.5432,
"step": 1940
},
{
"epoch": 9.566809029896278,
"grad_norm": 3.4668374061584473,
"learning_rate": 0.0001878032786885246,
"loss": 0.5502,
"step": 1960
},
{
"epoch": 9.664429530201343,
"grad_norm": 4.547699928283691,
"learning_rate": 0.000187672131147541,
"loss": 0.5482,
"step": 1980
},
{
"epoch": 9.762050030506407,
"grad_norm": 3.784132957458496,
"learning_rate": 0.00018754098360655738,
"loss": 0.516,
"step": 2000
},
{
"epoch": 9.85967053081147,
"grad_norm": 4.160193920135498,
"learning_rate": 0.0001874098360655738,
"loss": 0.5658,
"step": 2020
},
{
"epoch": 9.957291031116535,
"grad_norm": 4.115555286407471,
"learning_rate": 0.00018727868852459016,
"loss": 0.5775,
"step": 2040
},
{
"epoch": 10.054911531421599,
"grad_norm": 3.361625909805298,
"learning_rate": 0.00018714754098360657,
"loss": 0.5105,
"step": 2060
},
{
"epoch": 10.152532031726663,
"grad_norm": 3.663980484008789,
"learning_rate": 0.00018701639344262296,
"loss": 0.4556,
"step": 2080
},
{
"epoch": 10.250152532031727,
"grad_norm": 3.6515650749206543,
"learning_rate": 0.00018688524590163935,
"loss": 0.4937,
"step": 2100
},
{
"epoch": 10.347773032336791,
"grad_norm": 3.2449493408203125,
"learning_rate": 0.00018675409836065574,
"loss": 0.481,
"step": 2120
},
{
"epoch": 10.445393532641855,
"grad_norm": 4.262176513671875,
"learning_rate": 0.00018662295081967213,
"loss": 0.4559,
"step": 2140
},
{
"epoch": 10.54301403294692,
"grad_norm": 3.885936737060547,
"learning_rate": 0.00018649180327868855,
"loss": 0.5263,
"step": 2160
},
{
"epoch": 10.640634533251983,
"grad_norm": 3.676922559738159,
"learning_rate": 0.00018636065573770494,
"loss": 0.4942,
"step": 2180
},
{
"epoch": 10.738255033557047,
"grad_norm": 4.31233024597168,
"learning_rate": 0.00018622950819672133,
"loss": 0.5339,
"step": 2200
},
{
"epoch": 10.835875533862112,
"grad_norm": 3.410269260406494,
"learning_rate": 0.00018609836065573772,
"loss": 0.5365,
"step": 2220
},
{
"epoch": 10.933496034167176,
"grad_norm": 3.114283800125122,
"learning_rate": 0.0001859672131147541,
"loss": 0.5252,
"step": 2240
},
{
"epoch": 11.03111653447224,
"grad_norm": 3.4641740322113037,
"learning_rate": 0.0001858360655737705,
"loss": 0.4993,
"step": 2260
},
{
"epoch": 11.128737034777304,
"grad_norm": 5.0364203453063965,
"learning_rate": 0.00018570491803278688,
"loss": 0.4097,
"step": 2280
},
{
"epoch": 11.226357535082368,
"grad_norm": 3.811703681945801,
"learning_rate": 0.0001855737704918033,
"loss": 0.4265,
"step": 2300
},
{
"epoch": 11.323978035387432,
"grad_norm": 3.528463125228882,
"learning_rate": 0.0001854426229508197,
"loss": 0.4542,
"step": 2320
},
{
"epoch": 11.421598535692496,
"grad_norm": 3.013249158859253,
"learning_rate": 0.00018531147540983608,
"loss": 0.4824,
"step": 2340
},
{
"epoch": 11.51921903599756,
"grad_norm": 3.7065834999084473,
"learning_rate": 0.00018518032786885247,
"loss": 0.4724,
"step": 2360
},
{
"epoch": 11.616839536302624,
"grad_norm": 3.271639347076416,
"learning_rate": 0.00018504918032786886,
"loss": 0.4731,
"step": 2380
},
{
"epoch": 11.714460036607688,
"grad_norm": 3.738567590713501,
"learning_rate": 0.00018491803278688527,
"loss": 0.5024,
"step": 2400
},
{
"epoch": 11.812080536912752,
"grad_norm": 3.5277130603790283,
"learning_rate": 0.00018478688524590164,
"loss": 0.4552,
"step": 2420
},
{
"epoch": 11.909701037217816,
"grad_norm": 3.909186840057373,
"learning_rate": 0.00018465573770491805,
"loss": 0.4454,
"step": 2440
},
{
"epoch": 12.00732153752288,
"grad_norm": 3.629971981048584,
"learning_rate": 0.00018452459016393444,
"loss": 0.4928,
"step": 2460
},
{
"epoch": 12.104942037827945,
"grad_norm": 4.8608245849609375,
"learning_rate": 0.0001843934426229508,
"loss": 0.395,
"step": 2480
},
{
"epoch": 12.202562538133009,
"grad_norm": 3.377361536026001,
"learning_rate": 0.00018426229508196722,
"loss": 0.4124,
"step": 2500
},
{
"epoch": 12.300183038438073,
"grad_norm": 3.1160948276519775,
"learning_rate": 0.0001841311475409836,
"loss": 0.4317,
"step": 2520
},
{
"epoch": 12.397803538743137,
"grad_norm": 4.747882843017578,
"learning_rate": 0.00018400000000000003,
"loss": 0.4054,
"step": 2540
},
{
"epoch": 12.495424039048201,
"grad_norm": 3.0192835330963135,
"learning_rate": 0.0001838688524590164,
"loss": 0.4384,
"step": 2560
},
{
"epoch": 12.593044539353265,
"grad_norm": 3.397606372833252,
"learning_rate": 0.0001837377049180328,
"loss": 0.4095,
"step": 2580
},
{
"epoch": 12.690665039658327,
"grad_norm": 3.571641445159912,
"learning_rate": 0.0001836065573770492,
"loss": 0.4526,
"step": 2600
},
{
"epoch": 12.788285539963393,
"grad_norm": 3.9279720783233643,
"learning_rate": 0.00018347540983606558,
"loss": 0.4447,
"step": 2620
},
{
"epoch": 12.885906040268456,
"grad_norm": 5.057145118713379,
"learning_rate": 0.00018334426229508197,
"loss": 0.4285,
"step": 2640
},
{
"epoch": 12.98352654057352,
"grad_norm": 4.515413761138916,
"learning_rate": 0.00018321311475409836,
"loss": 0.4614,
"step": 2660
},
{
"epoch": 13.081147040878584,
"grad_norm": 5.738150119781494,
"learning_rate": 0.00018308196721311478,
"loss": 0.3594,
"step": 2680
},
{
"epoch": 13.178767541183648,
"grad_norm": 4.123675346374512,
"learning_rate": 0.00018295081967213114,
"loss": 0.3604,
"step": 2700
},
{
"epoch": 13.276388041488712,
"grad_norm": 4.020540237426758,
"learning_rate": 0.00018281967213114756,
"loss": 0.3455,
"step": 2720
},
{
"epoch": 13.374008541793776,
"grad_norm": 3.946470260620117,
"learning_rate": 0.00018268852459016395,
"loss": 0.3936,
"step": 2740
},
{
"epoch": 13.47162904209884,
"grad_norm": 3.4082658290863037,
"learning_rate": 0.00018255737704918034,
"loss": 0.3853,
"step": 2760
},
{
"epoch": 13.569249542403904,
"grad_norm": 3.8602256774902344,
"learning_rate": 0.00018242622950819673,
"loss": 0.4138,
"step": 2780
},
{
"epoch": 13.666870042708968,
"grad_norm": 3.4819118976593018,
"learning_rate": 0.00018229508196721312,
"loss": 0.4235,
"step": 2800
},
{
"epoch": 13.764490543014032,
"grad_norm": 4.620802402496338,
"learning_rate": 0.00018216393442622953,
"loss": 0.3917,
"step": 2820
},
{
"epoch": 13.862111043319096,
"grad_norm": 4.1257500648498535,
"learning_rate": 0.00018203278688524592,
"loss": 0.4289,
"step": 2840
},
{
"epoch": 13.95973154362416,
"grad_norm": 4.032405376434326,
"learning_rate": 0.0001819016393442623,
"loss": 0.4501,
"step": 2860
},
{
"epoch": 14.057352043929225,
"grad_norm": 3.9458096027374268,
"learning_rate": 0.0001817704918032787,
"loss": 0.3863,
"step": 2880
},
{
"epoch": 14.154972544234289,
"grad_norm": 5.01477575302124,
"learning_rate": 0.0001816393442622951,
"loss": 0.3421,
"step": 2900
},
{
"epoch": 14.252593044539353,
"grad_norm": 3.396898031234741,
"learning_rate": 0.00018150819672131148,
"loss": 0.3438,
"step": 2920
},
{
"epoch": 14.350213544844417,
"grad_norm": 4.596593856811523,
"learning_rate": 0.00018137704918032787,
"loss": 0.3435,
"step": 2940
},
{
"epoch": 14.44783404514948,
"grad_norm": 3.6386988162994385,
"learning_rate": 0.00018124590163934429,
"loss": 0.3703,
"step": 2960
},
{
"epoch": 14.545454545454545,
"grad_norm": 3.3389110565185547,
"learning_rate": 0.00018111475409836067,
"loss": 0.3727,
"step": 2980
},
{
"epoch": 14.643075045759609,
"grad_norm": 4.675887107849121,
"learning_rate": 0.00018098360655737704,
"loss": 0.3972,
"step": 3000
},
{
"epoch": 14.740695546064673,
"grad_norm": 3.3281068801879883,
"learning_rate": 0.00018085245901639345,
"loss": 0.38,
"step": 3020
},
{
"epoch": 14.838316046369737,
"grad_norm": 3.9550278186798096,
"learning_rate": 0.00018072131147540984,
"loss": 0.3995,
"step": 3040
},
{
"epoch": 14.935936546674801,
"grad_norm": 2.93839168548584,
"learning_rate": 0.00018059016393442626,
"loss": 0.3885,
"step": 3060
},
{
"epoch": 15.033557046979865,
"grad_norm": 3.583588123321533,
"learning_rate": 0.00018045901639344262,
"loss": 0.3407,
"step": 3080
},
{
"epoch": 15.13117754728493,
"grad_norm": 3.6306777000427246,
"learning_rate": 0.00018032786885245904,
"loss": 0.3016,
"step": 3100
},
{
"epoch": 15.228798047589994,
"grad_norm": 3.278693437576294,
"learning_rate": 0.00018019672131147543,
"loss": 0.292,
"step": 3120
},
{
"epoch": 15.326418547895058,
"grad_norm": 2.9507830142974854,
"learning_rate": 0.0001800655737704918,
"loss": 0.3204,
"step": 3140
},
{
"epoch": 15.424039048200122,
"grad_norm": 2.957294464111328,
"learning_rate": 0.0001799344262295082,
"loss": 0.3446,
"step": 3160
},
{
"epoch": 15.521659548505186,
"grad_norm": 3.4953787326812744,
"learning_rate": 0.0001798032786885246,
"loss": 0.3307,
"step": 3180
},
{
"epoch": 15.61928004881025,
"grad_norm": 3.349458694458008,
"learning_rate": 0.000179672131147541,
"loss": 0.3448,
"step": 3200
},
{
"epoch": 15.716900549115314,
"grad_norm": 4.927302837371826,
"learning_rate": 0.00017954098360655737,
"loss": 0.3661,
"step": 3220
},
{
"epoch": 15.814521049420378,
"grad_norm": 4.214022636413574,
"learning_rate": 0.0001794098360655738,
"loss": 0.3595,
"step": 3240
},
{
"epoch": 15.912141549725442,
"grad_norm": 3.7332723140716553,
"learning_rate": 0.00017927868852459018,
"loss": 0.3916,
"step": 3260
},
{
"epoch": 16.009762050030506,
"grad_norm": 3.1130378246307373,
"learning_rate": 0.00017914754098360657,
"loss": 0.365,
"step": 3280
},
{
"epoch": 16.107382550335572,
"grad_norm": 3.9815053939819336,
"learning_rate": 0.00017901639344262296,
"loss": 0.2466,
"step": 3300
},
{
"epoch": 16.205003050640634,
"grad_norm": 4.437533855438232,
"learning_rate": 0.00017888524590163935,
"loss": 0.2917,
"step": 3320
},
{
"epoch": 16.3026235509457,
"grad_norm": 3.869553565979004,
"learning_rate": 0.00017875409836065576,
"loss": 0.2982,
"step": 3340
},
{
"epoch": 16.400244051250763,
"grad_norm": 3.1284611225128174,
"learning_rate": 0.00017862295081967213,
"loss": 0.3146,
"step": 3360
},
{
"epoch": 16.49786455155583,
"grad_norm": 3.837712049484253,
"learning_rate": 0.00017849180327868852,
"loss": 0.3129,
"step": 3380
},
{
"epoch": 16.59548505186089,
"grad_norm": 3.348344326019287,
"learning_rate": 0.00017836065573770493,
"loss": 0.3379,
"step": 3400
},
{
"epoch": 16.693105552165953,
"grad_norm": 3.809512138366699,
"learning_rate": 0.00017822950819672132,
"loss": 0.3193,
"step": 3420
},
{
"epoch": 16.79072605247102,
"grad_norm": 3.3745222091674805,
"learning_rate": 0.0001780983606557377,
"loss": 0.3308,
"step": 3440
},
{
"epoch": 16.888346552776085,
"grad_norm": 4.550785541534424,
"learning_rate": 0.0001779672131147541,
"loss": 0.3609,
"step": 3460
},
{
"epoch": 16.985967053081147,
"grad_norm": 4.031665802001953,
"learning_rate": 0.00017783606557377052,
"loss": 0.3648,
"step": 3480
},
{
"epoch": 17.08358755338621,
"grad_norm": 3.0923168659210205,
"learning_rate": 0.0001777049180327869,
"loss": 0.2731,
"step": 3500
},
{
"epoch": 17.181208053691275,
"grad_norm": 3.291416883468628,
"learning_rate": 0.00017757377049180327,
"loss": 0.2829,
"step": 3520
},
{
"epoch": 17.278828553996338,
"grad_norm": 3.059995651245117,
"learning_rate": 0.00017744262295081969,
"loss": 0.3047,
"step": 3540
},
{
"epoch": 17.376449054301403,
"grad_norm": 3.1089327335357666,
"learning_rate": 0.00017731147540983607,
"loss": 0.3092,
"step": 3560
},
{
"epoch": 17.474069554606466,
"grad_norm": 3.9796876907348633,
"learning_rate": 0.00017718032786885246,
"loss": 0.2647,
"step": 3580
},
{
"epoch": 17.57169005491153,
"grad_norm": 3.587038040161133,
"learning_rate": 0.00017704918032786885,
"loss": 0.308,
"step": 3600
},
{
"epoch": 17.669310555216594,
"grad_norm": 3.7032790184020996,
"learning_rate": 0.00017691803278688527,
"loss": 0.3101,
"step": 3620
},
{
"epoch": 17.76693105552166,
"grad_norm": 3.7440781593322754,
"learning_rate": 0.00017678688524590166,
"loss": 0.2922,
"step": 3640
},
{
"epoch": 17.864551555826722,
"grad_norm": 3.4123542308807373,
"learning_rate": 0.00017665573770491802,
"loss": 0.3097,
"step": 3660
},
{
"epoch": 17.962172056131788,
"grad_norm": 3.958204507827759,
"learning_rate": 0.00017652459016393444,
"loss": 0.2955,
"step": 3680
},
{
"epoch": 18.05979255643685,
"grad_norm": 4.549073696136475,
"learning_rate": 0.00017639344262295083,
"loss": 0.2846,
"step": 3700
},
{
"epoch": 18.157413056741916,
"grad_norm": 3.5509791374206543,
"learning_rate": 0.00017626229508196724,
"loss": 0.2651,
"step": 3720
},
{
"epoch": 18.25503355704698,
"grad_norm": 4.044325828552246,
"learning_rate": 0.0001761311475409836,
"loss": 0.242,
"step": 3740
},
{
"epoch": 18.352654057352044,
"grad_norm": 3.255535125732422,
"learning_rate": 0.00017600000000000002,
"loss": 0.2857,
"step": 3760
},
{
"epoch": 18.450274557657107,
"grad_norm": 3.6761083602905273,
"learning_rate": 0.0001758688524590164,
"loss": 0.285,
"step": 3780
},
{
"epoch": 18.547895057962172,
"grad_norm": 4.1334381103515625,
"learning_rate": 0.00017573770491803277,
"loss": 0.2648,
"step": 3800
},
{
"epoch": 18.645515558267235,
"grad_norm": 3.5164451599121094,
"learning_rate": 0.0001756065573770492,
"loss": 0.2654,
"step": 3820
},
{
"epoch": 18.7431360585723,
"grad_norm": 3.96533465385437,
"learning_rate": 0.00017547540983606558,
"loss": 0.2862,
"step": 3840
},
{
"epoch": 18.840756558877363,
"grad_norm": 3.932554006576538,
"learning_rate": 0.000175344262295082,
"loss": 0.2803,
"step": 3860
},
{
"epoch": 18.93837705918243,
"grad_norm": 4.888127326965332,
"learning_rate": 0.00017521311475409836,
"loss": 0.2902,
"step": 3880
},
{
"epoch": 19.03599755948749,
"grad_norm": 3.475789785385132,
"learning_rate": 0.00017508196721311475,
"loss": 0.2815,
"step": 3900
},
{
"epoch": 19.133618059792557,
"grad_norm": 2.57926607131958,
"learning_rate": 0.00017495081967213116,
"loss": 0.21,
"step": 3920
},
{
"epoch": 19.23123856009762,
"grad_norm": 3.3455469608306885,
"learning_rate": 0.00017481967213114753,
"loss": 0.232,
"step": 3940
},
{
"epoch": 19.328859060402685,
"grad_norm": 3.574331045150757,
"learning_rate": 0.00017468852459016394,
"loss": 0.2506,
"step": 3960
},
{
"epoch": 19.426479560707747,
"grad_norm": 4.422223091125488,
"learning_rate": 0.00017455737704918033,
"loss": 0.2494,
"step": 3980
},
{
"epoch": 19.524100061012813,
"grad_norm": 3.8312060832977295,
"learning_rate": 0.00017442622950819675,
"loss": 0.2758,
"step": 4000
},
{
"epoch": 19.621720561317876,
"grad_norm": 3.464089870452881,
"learning_rate": 0.0001742950819672131,
"loss": 0.2628,
"step": 4020
},
{
"epoch": 19.71934106162294,
"grad_norm": 4.096322536468506,
"learning_rate": 0.0001741639344262295,
"loss": 0.2651,
"step": 4040
},
{
"epoch": 19.816961561928004,
"grad_norm": 5.72092866897583,
"learning_rate": 0.00017403278688524592,
"loss": 0.29,
"step": 4060
},
{
"epoch": 19.91458206223307,
"grad_norm": 3.7189135551452637,
"learning_rate": 0.0001739016393442623,
"loss": 0.2715,
"step": 4080
},
{
"epoch": 20.012202562538132,
"grad_norm": 3.4471986293792725,
"learning_rate": 0.0001737704918032787,
"loss": 0.2762,
"step": 4100
},
{
"epoch": 20.109823062843198,
"grad_norm": 3.046600580215454,
"learning_rate": 0.00017363934426229509,
"loss": 0.2219,
"step": 4120
},
{
"epoch": 20.20744356314826,
"grad_norm": 3.6281490325927734,
"learning_rate": 0.0001735081967213115,
"loss": 0.231,
"step": 4140
},
{
"epoch": 20.305064063453326,
"grad_norm": 2.8515024185180664,
"learning_rate": 0.00017337704918032786,
"loss": 0.2212,
"step": 4160
},
{
"epoch": 20.40268456375839,
"grad_norm": 2.9778804779052734,
"learning_rate": 0.00017324590163934425,
"loss": 0.2327,
"step": 4180
},
{
"epoch": 20.500305064063454,
"grad_norm": 2.92470121383667,
"learning_rate": 0.00017311475409836067,
"loss": 0.2378,
"step": 4200
},
{
"epoch": 20.597925564368516,
"grad_norm": 3.8426191806793213,
"learning_rate": 0.00017298360655737706,
"loss": 0.2652,
"step": 4220
},
{
"epoch": 20.695546064673582,
"grad_norm": 3.8123021125793457,
"learning_rate": 0.00017285245901639345,
"loss": 0.2446,
"step": 4240
},
{
"epoch": 20.793166564978645,
"grad_norm": 3.650644540786743,
"learning_rate": 0.00017272131147540984,
"loss": 0.2331,
"step": 4260
},
{
"epoch": 20.89078706528371,
"grad_norm": 4.259769916534424,
"learning_rate": 0.00017259016393442625,
"loss": 0.2595,
"step": 4280
},
{
"epoch": 20.988407565588773,
"grad_norm": 4.065052509307861,
"learning_rate": 0.00017245901639344264,
"loss": 0.2672,
"step": 4300
},
{
"epoch": 21.08602806589384,
"grad_norm": 3.565068244934082,
"learning_rate": 0.000172327868852459,
"loss": 0.1997,
"step": 4320
},
{
"epoch": 21.1836485661989,
"grad_norm": 3.036924362182617,
"learning_rate": 0.00017219672131147542,
"loss": 0.1936,
"step": 4340
},
{
"epoch": 21.281269066503967,
"grad_norm": 3.2253048419952393,
"learning_rate": 0.0001720655737704918,
"loss": 0.2099,
"step": 4360
},
{
"epoch": 21.37888956680903,
"grad_norm": 3.6241676807403564,
"learning_rate": 0.0001719344262295082,
"loss": 0.2232,
"step": 4380
},
{
"epoch": 21.476510067114095,
"grad_norm": 3.3272531032562256,
"learning_rate": 0.0001718032786885246,
"loss": 0.2103,
"step": 4400
},
{
"epoch": 21.574130567419157,
"grad_norm": 3.6076695919036865,
"learning_rate": 0.00017167213114754098,
"loss": 0.2323,
"step": 4420
},
{
"epoch": 21.671751067724223,
"grad_norm": 3.642751455307007,
"learning_rate": 0.0001715409836065574,
"loss": 0.2385,
"step": 4440
},
{
"epoch": 21.769371568029285,
"grad_norm": 3.1882801055908203,
"learning_rate": 0.00017140983606557376,
"loss": 0.2516,
"step": 4460
},
{
"epoch": 21.86699206833435,
"grad_norm": 3.4758126735687256,
"learning_rate": 0.00017127868852459018,
"loss": 0.2365,
"step": 4480
},
{
"epoch": 21.964612568639414,
"grad_norm": 3.498697519302368,
"learning_rate": 0.00017114754098360656,
"loss": 0.2436,
"step": 4500
},
{
"epoch": 22.06223306894448,
"grad_norm": 3.926394462585449,
"learning_rate": 0.00017101639344262298,
"loss": 0.2101,
"step": 4520
},
{
"epoch": 22.15985356924954,
"grad_norm": 2.982520580291748,
"learning_rate": 0.00017088524590163934,
"loss": 0.1655,
"step": 4540
},
{
"epoch": 22.257474069554608,
"grad_norm": 3.843905448913574,
"learning_rate": 0.00017075409836065573,
"loss": 0.1896,
"step": 4560
},
{
"epoch": 22.35509456985967,
"grad_norm": 2.8714027404785156,
"learning_rate": 0.00017062295081967215,
"loss": 0.2113,
"step": 4580
},
{
"epoch": 22.452715070164736,
"grad_norm": 3.552220344543457,
"learning_rate": 0.0001704918032786885,
"loss": 0.2311,
"step": 4600
},
{
"epoch": 22.550335570469798,
"grad_norm": 4.077561378479004,
"learning_rate": 0.00017036065573770493,
"loss": 0.2205,
"step": 4620
},
{
"epoch": 22.647956070774864,
"grad_norm": 3.475970506668091,
"learning_rate": 0.00017022950819672132,
"loss": 0.2047,
"step": 4640
},
{
"epoch": 22.745576571079926,
"grad_norm": 2.9817659854888916,
"learning_rate": 0.00017009836065573773,
"loss": 0.2312,
"step": 4660
},
{
"epoch": 22.843197071384992,
"grad_norm": 2.8980283737182617,
"learning_rate": 0.0001699672131147541,
"loss": 0.2113,
"step": 4680
},
{
"epoch": 22.940817571690054,
"grad_norm": 2.9917445182800293,
"learning_rate": 0.00016983606557377049,
"loss": 0.2334,
"step": 4700
},
{
"epoch": 23.03843807199512,
"grad_norm": 2.9001080989837646,
"learning_rate": 0.0001697049180327869,
"loss": 0.2096,
"step": 4720
},
{
"epoch": 23.136058572300183,
"grad_norm": 2.7678308486938477,
"learning_rate": 0.0001695737704918033,
"loss": 0.18,
"step": 4740
},
{
"epoch": 23.23367907260525,
"grad_norm": 4.077008247375488,
"learning_rate": 0.00016944262295081968,
"loss": 0.175,
"step": 4760
},
{
"epoch": 23.33129957291031,
"grad_norm": 3.806175708770752,
"learning_rate": 0.00016931147540983607,
"loss": 0.1962,
"step": 4780
},
{
"epoch": 23.428920073215377,
"grad_norm": 3.203763246536255,
"learning_rate": 0.00016918032786885249,
"loss": 0.2072,
"step": 4800
},
{
"epoch": 23.52654057352044,
"grad_norm": 3.717864513397217,
"learning_rate": 0.00016904918032786885,
"loss": 0.203,
"step": 4820
},
{
"epoch": 23.624161073825505,
"grad_norm": 3.1602675914764404,
"learning_rate": 0.00016891803278688524,
"loss": 0.1925,
"step": 4840
},
{
"epoch": 23.721781574130567,
"grad_norm": 4.2820000648498535,
"learning_rate": 0.00016878688524590165,
"loss": 0.2037,
"step": 4860
},
{
"epoch": 23.819402074435633,
"grad_norm": 3.9633703231811523,
"learning_rate": 0.00016865573770491804,
"loss": 0.2143,
"step": 4880
},
{
"epoch": 23.917022574740695,
"grad_norm": 2.801804542541504,
"learning_rate": 0.00016852459016393443,
"loss": 0.1938,
"step": 4900
},
{
"epoch": 24.01464307504576,
"grad_norm": 2.719308614730835,
"learning_rate": 0.00016839344262295082,
"loss": 0.215,
"step": 4920
},
{
"epoch": 24.112263575350823,
"grad_norm": 3.0437123775482178,
"learning_rate": 0.0001682622950819672,
"loss": 0.1806,
"step": 4940
},
{
"epoch": 24.20988407565589,
"grad_norm": 3.314267158508301,
"learning_rate": 0.00016813114754098363,
"loss": 0.169,
"step": 4960
},
{
"epoch": 24.30750457596095,
"grad_norm": 2.26737117767334,
"learning_rate": 0.000168,
"loss": 0.1704,
"step": 4980
},
{
"epoch": 24.405125076266017,
"grad_norm": 4.0000715255737305,
"learning_rate": 0.0001678688524590164,
"loss": 0.176,
"step": 5000
}
],
"logging_steps": 20,
"max_steps": 30600,
"num_input_tokens_seen": 0,
"num_train_epochs": 150,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 6.574589658966067e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}