test-mlm-wwm2 / trainer_state.json
hxl520
Update from test-mlm-wwm2
8fa1142
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 90939,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 4.972788227098572e-05,
"loss": 2.3513,
"step": 500
},
{
"epoch": 0.03,
"learning_rate": 4.945576454197144e-05,
"loss": 2.3368,
"step": 1000
},
{
"epoch": 0.05,
"learning_rate": 4.9183646812957165e-05,
"loss": 2.3554,
"step": 1500
},
{
"epoch": 0.07,
"learning_rate": 4.891152908394288e-05,
"loss": 2.4827,
"step": 2000
},
{
"epoch": 0.08,
"learning_rate": 4.86394113549286e-05,
"loss": 2.5032,
"step": 2500
},
{
"epoch": 0.1,
"learning_rate": 4.836729362591432e-05,
"loss": 2.4807,
"step": 3000
},
{
"epoch": 0.11,
"learning_rate": 4.809517589690004e-05,
"loss": 2.4771,
"step": 3500
},
{
"epoch": 0.13,
"learning_rate": 4.782305816788576e-05,
"loss": 2.432,
"step": 4000
},
{
"epoch": 0.15,
"learning_rate": 4.7550940438871475e-05,
"loss": 2.4253,
"step": 4500
},
{
"epoch": 0.16,
"learning_rate": 4.7278822709857194e-05,
"loss": 2.396,
"step": 5000
},
{
"epoch": 0.18,
"learning_rate": 4.700670498084291e-05,
"loss": 2.3774,
"step": 5500
},
{
"epoch": 0.2,
"learning_rate": 4.673458725182863e-05,
"loss": 2.3601,
"step": 6000
},
{
"epoch": 0.21,
"learning_rate": 4.646246952281435e-05,
"loss": 2.3444,
"step": 6500
},
{
"epoch": 0.23,
"learning_rate": 4.6190351793800074e-05,
"loss": 2.3522,
"step": 7000
},
{
"epoch": 0.24,
"learning_rate": 4.591823406478579e-05,
"loss": 2.4588,
"step": 7500
},
{
"epoch": 0.26,
"learning_rate": 4.564611633577151e-05,
"loss": 2.3588,
"step": 8000
},
{
"epoch": 0.28,
"learning_rate": 4.537399860675723e-05,
"loss": 2.3733,
"step": 8500
},
{
"epoch": 0.29,
"learning_rate": 4.510188087774295e-05,
"loss": 2.3612,
"step": 9000
},
{
"epoch": 0.31,
"learning_rate": 4.482976314872867e-05,
"loss": 2.413,
"step": 9500
},
{
"epoch": 0.33,
"learning_rate": 4.455764541971439e-05,
"loss": 2.3861,
"step": 10000
},
{
"epoch": 0.34,
"learning_rate": 4.428552769070011e-05,
"loss": 2.3726,
"step": 10500
},
{
"epoch": 0.36,
"learning_rate": 4.401340996168583e-05,
"loss": 2.328,
"step": 11000
},
{
"epoch": 0.38,
"learning_rate": 4.374129223267154e-05,
"loss": 2.3552,
"step": 11500
},
{
"epoch": 0.39,
"learning_rate": 4.346917450365726e-05,
"loss": 2.4816,
"step": 12000
},
{
"epoch": 0.41,
"learning_rate": 4.3197056774642984e-05,
"loss": 2.3795,
"step": 12500
},
{
"epoch": 0.42,
"learning_rate": 4.29249390456287e-05,
"loss": 2.3067,
"step": 13000
},
{
"epoch": 0.44,
"learning_rate": 4.265282131661442e-05,
"loss": 2.3856,
"step": 13500
},
{
"epoch": 0.46,
"learning_rate": 4.238070358760014e-05,
"loss": 2.2952,
"step": 14000
},
{
"epoch": 0.47,
"learning_rate": 4.210858585858586e-05,
"loss": 2.2996,
"step": 14500
},
{
"epoch": 0.49,
"learning_rate": 4.183646812957158e-05,
"loss": 2.3925,
"step": 15000
},
{
"epoch": 0.51,
"learning_rate": 4.15643504005573e-05,
"loss": 2.3015,
"step": 15500
},
{
"epoch": 0.52,
"learning_rate": 4.129223267154302e-05,
"loss": 2.4006,
"step": 16000
},
{
"epoch": 0.54,
"learning_rate": 4.102011494252874e-05,
"loss": 2.3463,
"step": 16500
},
{
"epoch": 0.56,
"learning_rate": 4.0747997213514457e-05,
"loss": 2.3236,
"step": 17000
},
{
"epoch": 0.57,
"learning_rate": 4.047587948450018e-05,
"loss": 2.3262,
"step": 17500
},
{
"epoch": 0.59,
"learning_rate": 4.02037617554859e-05,
"loss": 2.3238,
"step": 18000
},
{
"epoch": 0.6,
"learning_rate": 3.993164402647161e-05,
"loss": 2.185,
"step": 18500
},
{
"epoch": 0.62,
"learning_rate": 3.965952629745733e-05,
"loss": 2.2306,
"step": 19000
},
{
"epoch": 0.64,
"learning_rate": 3.938740856844305e-05,
"loss": 2.2768,
"step": 19500
},
{
"epoch": 0.65,
"learning_rate": 3.911529083942877e-05,
"loss": 2.2603,
"step": 20000
},
{
"epoch": 0.67,
"learning_rate": 3.884317311041449e-05,
"loss": 2.3123,
"step": 20500
},
{
"epoch": 0.69,
"learning_rate": 3.857105538140021e-05,
"loss": 2.2436,
"step": 21000
},
{
"epoch": 0.7,
"learning_rate": 3.829893765238593e-05,
"loss": 2.3416,
"step": 21500
},
{
"epoch": 0.72,
"learning_rate": 3.802681992337165e-05,
"loss": 2.2964,
"step": 22000
},
{
"epoch": 0.73,
"learning_rate": 3.7754702194357366e-05,
"loss": 2.4278,
"step": 22500
},
{
"epoch": 0.75,
"learning_rate": 3.748258446534309e-05,
"loss": 2.3682,
"step": 23000
},
{
"epoch": 0.77,
"learning_rate": 3.721046673632881e-05,
"loss": 2.3585,
"step": 23500
},
{
"epoch": 0.78,
"learning_rate": 3.693834900731453e-05,
"loss": 2.2861,
"step": 24000
},
{
"epoch": 0.8,
"learning_rate": 3.666623127830025e-05,
"loss": 2.2447,
"step": 24500
},
{
"epoch": 0.82,
"learning_rate": 3.6394113549285965e-05,
"loss": 2.3573,
"step": 25000
},
{
"epoch": 0.83,
"learning_rate": 3.6121995820271684e-05,
"loss": 2.2819,
"step": 25500
},
{
"epoch": 0.85,
"learning_rate": 3.58498780912574e-05,
"loss": 2.3119,
"step": 26000
},
{
"epoch": 0.87,
"learning_rate": 3.542979359790629e-05,
"loss": 2.2534,
"step": 26500
},
{
"epoch": 0.89,
"learning_rate": 3.5154884043149804e-05,
"loss": 2.2352,
"step": 27000
},
{
"epoch": 0.91,
"learning_rate": 3.4879974488393316e-05,
"loss": 2.2077,
"step": 27500
},
{
"epoch": 0.92,
"learning_rate": 3.4605064933636835e-05,
"loss": 2.2063,
"step": 28000
},
{
"epoch": 0.94,
"learning_rate": 3.4330155378880354e-05,
"loss": 2.2603,
"step": 28500
},
{
"epoch": 0.96,
"learning_rate": 3.4055245824123866e-05,
"loss": 2.2869,
"step": 29000
},
{
"epoch": 0.97,
"learning_rate": 3.378033626936738e-05,
"loss": 2.2388,
"step": 29500
},
{
"epoch": 0.99,
"learning_rate": 3.35054267146109e-05,
"loss": 2.2192,
"step": 30000
},
{
"epoch": 1.01,
"learning_rate": 3.323051715985441e-05,
"loss": 2.2518,
"step": 30500
},
{
"epoch": 1.02,
"learning_rate": 3.295560760509793e-05,
"loss": 2.2317,
"step": 31000
},
{
"epoch": 1.04,
"learning_rate": 3.268069805034144e-05,
"loss": 2.1664,
"step": 31500
},
{
"epoch": 1.06,
"learning_rate": 3.240578849558495e-05,
"loss": 2.2451,
"step": 32000
},
{
"epoch": 1.07,
"learning_rate": 3.213087894082847e-05,
"loss": 2.1932,
"step": 32500
},
{
"epoch": 1.09,
"learning_rate": 3.185596938607198e-05,
"loss": 2.1175,
"step": 33000
},
{
"epoch": 1.11,
"learning_rate": 3.1581059831315495e-05,
"loss": 2.1072,
"step": 33500
},
{
"epoch": 1.12,
"learning_rate": 3.1306150276559014e-05,
"loss": 2.2148,
"step": 34000
},
{
"epoch": 1.14,
"learning_rate": 3.103124072180253e-05,
"loss": 2.1475,
"step": 34500
},
{
"epoch": 1.15,
"learning_rate": 3.0756331167046045e-05,
"loss": 2.133,
"step": 35000
},
{
"epoch": 1.17,
"learning_rate": 3.0481421612289557e-05,
"loss": 2.0703,
"step": 35500
},
{
"epoch": 1.19,
"learning_rate": 3.0206512057533072e-05,
"loss": 2.1113,
"step": 36000
},
{
"epoch": 1.2,
"learning_rate": 2.993160250277659e-05,
"loss": 2.0876,
"step": 36500
},
{
"epoch": 1.22,
"learning_rate": 2.9656692948020103e-05,
"loss": 2.1706,
"step": 37000
},
{
"epoch": 1.24,
"learning_rate": 2.938178339326362e-05,
"loss": 2.1105,
"step": 37500
},
{
"epoch": 1.25,
"learning_rate": 2.910687383850713e-05,
"loss": 2.105,
"step": 38000
},
{
"epoch": 1.27,
"learning_rate": 2.8831964283750646e-05,
"loss": 2.144,
"step": 38500
},
{
"epoch": 1.29,
"learning_rate": 2.8557054728994165e-05,
"loss": 2.1656,
"step": 39000
},
{
"epoch": 1.3,
"learning_rate": 2.8282145174237677e-05,
"loss": 2.073,
"step": 39500
},
{
"epoch": 1.32,
"learning_rate": 2.8007235619481192e-05,
"loss": 2.1128,
"step": 40000
},
{
"epoch": 1.34,
"learning_rate": 2.7732326064724708e-05,
"loss": 2.028,
"step": 40500
},
{
"epoch": 1.35,
"learning_rate": 2.745741650996822e-05,
"loss": 2.1078,
"step": 41000
},
{
"epoch": 1.37,
"learning_rate": 2.7182506955211735e-05,
"loss": 2.0751,
"step": 41500
},
{
"epoch": 1.39,
"learning_rate": 2.6907597400455254e-05,
"loss": 2.0514,
"step": 42000
},
{
"epoch": 1.4,
"learning_rate": 2.6632687845698766e-05,
"loss": 2.1142,
"step": 42500
},
{
"epoch": 1.42,
"learning_rate": 2.6357778290942282e-05,
"loss": 2.0525,
"step": 43000
},
{
"epoch": 1.44,
"learning_rate": 2.6082868736185794e-05,
"loss": 2.1227,
"step": 43500
},
{
"epoch": 1.45,
"learning_rate": 2.580795918142931e-05,
"loss": 2.0927,
"step": 44000
},
{
"epoch": 1.47,
"learning_rate": 2.553304962667283e-05,
"loss": 2.0817,
"step": 44500
},
{
"epoch": 1.48,
"learning_rate": 2.525814007191634e-05,
"loss": 2.0319,
"step": 45000
},
{
"epoch": 1.5,
"learning_rate": 2.4983230517159856e-05,
"loss": 2.0231,
"step": 45500
},
{
"epoch": 1.52,
"learning_rate": 2.4708320962403368e-05,
"loss": 2.0348,
"step": 46000
},
{
"epoch": 1.53,
"learning_rate": 2.4433411407646887e-05,
"loss": 2.012,
"step": 46500
},
{
"epoch": 1.55,
"learning_rate": 2.41585018528904e-05,
"loss": 2.0662,
"step": 47000
},
{
"epoch": 1.57,
"learning_rate": 2.3883592298133914e-05,
"loss": 2.0736,
"step": 47500
},
{
"epoch": 1.58,
"learning_rate": 2.360868274337743e-05,
"loss": 2.0574,
"step": 48000
},
{
"epoch": 1.6,
"learning_rate": 2.3333773188620945e-05,
"loss": 2.1095,
"step": 48500
},
{
"epoch": 1.62,
"learning_rate": 2.305886363386446e-05,
"loss": 2.0707,
"step": 49000
},
{
"epoch": 1.63,
"learning_rate": 2.2783954079107976e-05,
"loss": 2.0368,
"step": 49500
},
{
"epoch": 1.65,
"learning_rate": 2.250904452435149e-05,
"loss": 2.1298,
"step": 50000
},
{
"epoch": 1.67,
"learning_rate": 2.2234134969595004e-05,
"loss": 2.0072,
"step": 50500
},
{
"epoch": 1.68,
"learning_rate": 2.195922541483852e-05,
"loss": 1.9771,
"step": 51000
},
{
"epoch": 1.7,
"learning_rate": 2.168431586008203e-05,
"loss": 1.9216,
"step": 51500
},
{
"epoch": 1.72,
"learning_rate": 2.140940630532555e-05,
"loss": 1.9915,
"step": 52000
},
{
"epoch": 1.73,
"learning_rate": 2.1134496750569062e-05,
"loss": 2.0122,
"step": 52500
},
{
"epoch": 1.75,
"learning_rate": 2.085958719581258e-05,
"loss": 2.0255,
"step": 53000
},
{
"epoch": 1.76,
"learning_rate": 2.0584677641056093e-05,
"loss": 1.9464,
"step": 53500
},
{
"epoch": 1.78,
"learning_rate": 2.030976808629961e-05,
"loss": 1.9671,
"step": 54000
},
{
"epoch": 1.8,
"learning_rate": 2.0034858531543124e-05,
"loss": 1.968,
"step": 54500
},
{
"epoch": 1.81,
"learning_rate": 1.975994897678664e-05,
"loss": 2.0287,
"step": 55000
},
{
"epoch": 1.83,
"learning_rate": 1.948503942203015e-05,
"loss": 1.9584,
"step": 55500
},
{
"epoch": 1.85,
"learning_rate": 1.9210129867273667e-05,
"loss": 1.9247,
"step": 56000
},
{
"epoch": 1.86,
"learning_rate": 1.8935220312517183e-05,
"loss": 1.9425,
"step": 56500
},
{
"epoch": 1.88,
"learning_rate": 1.8660310757760698e-05,
"loss": 1.9221,
"step": 57000
},
{
"epoch": 1.9,
"learning_rate": 1.8385401203004214e-05,
"loss": 1.9912,
"step": 57500
},
{
"epoch": 1.91,
"learning_rate": 1.8110491648247726e-05,
"loss": 1.9346,
"step": 58000
},
{
"epoch": 1.93,
"learning_rate": 1.7835582093491245e-05,
"loss": 2.0059,
"step": 58500
},
{
"epoch": 1.95,
"learning_rate": 1.7560672538734757e-05,
"loss": 1.9797,
"step": 59000
},
{
"epoch": 1.96,
"learning_rate": 1.7285762983978272e-05,
"loss": 2.023,
"step": 59500
},
{
"epoch": 1.98,
"learning_rate": 1.7010853429221788e-05,
"loss": 1.9738,
"step": 60000
},
{
"epoch": 2.0,
"learning_rate": 1.6735943874465303e-05,
"loss": 1.951,
"step": 60500
},
{
"epoch": 2.01,
"learning_rate": 1.646103431970882e-05,
"loss": 1.8491,
"step": 61000
},
{
"epoch": 2.03,
"learning_rate": 1.618612476495233e-05,
"loss": 1.8754,
"step": 61500
},
{
"epoch": 2.05,
"learning_rate": 1.5911215210195846e-05,
"loss": 1.856,
"step": 62000
},
{
"epoch": 2.06,
"learning_rate": 1.563630565543936e-05,
"loss": 1.8521,
"step": 62500
},
{
"epoch": 2.08,
"learning_rate": 1.5361396100682877e-05,
"loss": 1.7808,
"step": 63000
},
{
"epoch": 2.09,
"learning_rate": 1.508648654592639e-05,
"loss": 1.8407,
"step": 63500
},
{
"epoch": 2.11,
"learning_rate": 1.4811576991169906e-05,
"loss": 1.85,
"step": 64000
},
{
"epoch": 2.13,
"learning_rate": 1.453666743641342e-05,
"loss": 1.8037,
"step": 64500
},
{
"epoch": 2.14,
"learning_rate": 1.4261757881656937e-05,
"loss": 1.7651,
"step": 65000
},
{
"epoch": 2.16,
"learning_rate": 1.3986848326900451e-05,
"loss": 1.8384,
"step": 65500
},
{
"epoch": 2.18,
"learning_rate": 1.3711938772143965e-05,
"loss": 1.8137,
"step": 66000
},
{
"epoch": 2.19,
"learning_rate": 1.3437029217387482e-05,
"loss": 1.8169,
"step": 66500
},
{
"epoch": 2.21,
"learning_rate": 1.3162119662630996e-05,
"loss": 1.8162,
"step": 67000
},
{
"epoch": 2.23,
"learning_rate": 1.288721010787451e-05,
"loss": 1.8186,
"step": 67500
},
{
"epoch": 2.24,
"learning_rate": 1.2612300553118025e-05,
"loss": 1.707,
"step": 68000
},
{
"epoch": 2.26,
"learning_rate": 1.2337390998361539e-05,
"loss": 1.7471,
"step": 68500
},
{
"epoch": 2.28,
"learning_rate": 1.2062481443605054e-05,
"loss": 1.775,
"step": 69000
},
{
"epoch": 2.29,
"learning_rate": 1.178757188884857e-05,
"loss": 1.8094,
"step": 69500
},
{
"epoch": 2.31,
"learning_rate": 1.1512662334092083e-05,
"loss": 1.7498,
"step": 70000
},
{
"epoch": 2.33,
"learning_rate": 1.1237752779335599e-05,
"loss": 1.8305,
"step": 70500
},
{
"epoch": 2.34,
"learning_rate": 1.0962843224579114e-05,
"loss": 1.7538,
"step": 71000
},
{
"epoch": 2.36,
"learning_rate": 1.068793366982263e-05,
"loss": 1.7574,
"step": 71500
},
{
"epoch": 2.38,
"learning_rate": 1.0413024115066144e-05,
"loss": 1.7935,
"step": 72000
},
{
"epoch": 2.39,
"learning_rate": 1.0138114560309659e-05,
"loss": 1.774,
"step": 72500
},
{
"epoch": 2.41,
"learning_rate": 9.863205005553173e-06,
"loss": 1.7352,
"step": 73000
},
{
"epoch": 2.42,
"learning_rate": 9.588295450796688e-06,
"loss": 1.7788,
"step": 73500
},
{
"epoch": 2.44,
"learning_rate": 9.313385896040202e-06,
"loss": 1.7613,
"step": 74000
},
{
"epoch": 2.46,
"learning_rate": 9.038476341283717e-06,
"loss": 1.8059,
"step": 74500
},
{
"epoch": 2.47,
"learning_rate": 8.763566786527233e-06,
"loss": 1.7423,
"step": 75000
},
{
"epoch": 2.49,
"learning_rate": 8.488657231770748e-06,
"loss": 1.7693,
"step": 75500
},
{
"epoch": 2.51,
"learning_rate": 8.213747677014262e-06,
"loss": 1.776,
"step": 76000
},
{
"epoch": 2.52,
"learning_rate": 7.938838122257778e-06,
"loss": 1.7052,
"step": 76500
},
{
"epoch": 2.54,
"learning_rate": 7.663928567501293e-06,
"loss": 1.8079,
"step": 77000
},
{
"epoch": 2.56,
"learning_rate": 7.389019012744808e-06,
"loss": 1.7739,
"step": 77500
},
{
"epoch": 2.57,
"learning_rate": 7.1141094579883215e-06,
"loss": 1.6852,
"step": 78000
},
{
"epoch": 2.59,
"learning_rate": 6.839199903231837e-06,
"loss": 1.7038,
"step": 78500
},
{
"epoch": 2.61,
"learning_rate": 6.5642903484753525e-06,
"loss": 1.7107,
"step": 79000
},
{
"epoch": 2.62,
"learning_rate": 6.289380793718867e-06,
"loss": 1.7392,
"step": 79500
},
{
"epoch": 2.64,
"learning_rate": 6.014471238962382e-06,
"loss": 1.6904,
"step": 80000
},
{
"epoch": 2.66,
"learning_rate": 5.739561684205896e-06,
"loss": 1.8101,
"step": 80500
},
{
"epoch": 2.67,
"learning_rate": 5.464652129449412e-06,
"loss": 1.6975,
"step": 81000
},
{
"epoch": 2.69,
"learning_rate": 5.1897425746929264e-06,
"loss": 1.7233,
"step": 81500
},
{
"epoch": 2.71,
"learning_rate": 4.914833019936441e-06,
"loss": 1.755,
"step": 82000
},
{
"epoch": 2.72,
"learning_rate": 4.639923465179956e-06,
"loss": 1.6656,
"step": 82500
},
{
"epoch": 2.74,
"learning_rate": 4.365013910423471e-06,
"loss": 1.7117,
"step": 83000
},
{
"epoch": 2.75,
"learning_rate": 4.090104355666986e-06,
"loss": 1.7476,
"step": 83500
},
{
"epoch": 2.77,
"learning_rate": 3.815194800910501e-06,
"loss": 1.7272,
"step": 84000
},
{
"epoch": 2.79,
"learning_rate": 3.5402852461540154e-06,
"loss": 1.6628,
"step": 84500
},
{
"epoch": 2.8,
"learning_rate": 3.2653756913975305e-06,
"loss": 1.7535,
"step": 85000
},
{
"epoch": 2.82,
"learning_rate": 2.990466136641045e-06,
"loss": 1.68,
"step": 85500
},
{
"epoch": 2.84,
"learning_rate": 2.71555658188456e-06,
"loss": 1.6641,
"step": 86000
},
{
"epoch": 2.85,
"learning_rate": 2.4406470271280748e-06,
"loss": 1.6485,
"step": 86500
},
{
"epoch": 2.87,
"learning_rate": 2.16573747237159e-06,
"loss": 1.6617,
"step": 87000
},
{
"epoch": 2.89,
"learning_rate": 1.8908279176151046e-06,
"loss": 1.6926,
"step": 87500
},
{
"epoch": 2.9,
"learning_rate": 1.6159183628586195e-06,
"loss": 1.6692,
"step": 88000
},
{
"epoch": 2.92,
"learning_rate": 1.3410088081021345e-06,
"loss": 1.734,
"step": 88500
},
{
"epoch": 2.94,
"learning_rate": 1.0660992533456494e-06,
"loss": 1.6983,
"step": 89000
},
{
"epoch": 2.95,
"learning_rate": 7.911896985891643e-07,
"loss": 1.6775,
"step": 89500
},
{
"epoch": 2.97,
"learning_rate": 5.16280143832679e-07,
"loss": 1.6666,
"step": 90000
},
{
"epoch": 2.99,
"learning_rate": 2.4137058907619396e-07,
"loss": 1.7055,
"step": 90500
},
{
"epoch": 3.0,
"step": 90939,
"total_flos": 1.6358060886932256e+16,
"train_loss": 1.3765672623247966,
"train_runtime": 7341.0199,
"train_samples_per_second": 99.099,
"train_steps_per_second": 12.388
}
],
"max_steps": 90939,
"num_train_epochs": 3,
"total_flos": 1.6358060886932256e+16,
"trial_name": null,
"trial_params": null
}