| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 9871, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010130685847431872, | |
| "grad_norm": 7.154277324676514, | |
| "learning_rate": 2.532928064842959e-07, | |
| "loss": 6.7612, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.020261371694863743, | |
| "grad_norm": 7.9137983322143555, | |
| "learning_rate": 5.065856129685918e-07, | |
| "loss": 6.7189, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.030392057542295615, | |
| "grad_norm": 7.134788990020752, | |
| "learning_rate": 7.598784194528875e-07, | |
| "loss": 6.6275, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.040522743389727486, | |
| "grad_norm": 5.295099258422852, | |
| "learning_rate": 1.0131712259371835e-06, | |
| "loss": 6.2129, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.050653429237159354, | |
| "grad_norm": 4.518360614776611, | |
| "learning_rate": 1.2664640324214794e-06, | |
| "loss": 5.6904, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06078411508459123, | |
| "grad_norm": 2.872621774673462, | |
| "learning_rate": 1.519756838905775e-06, | |
| "loss": 5.1702, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0709148009320231, | |
| "grad_norm": 4.534188270568848, | |
| "learning_rate": 1.7730496453900712e-06, | |
| "loss": 4.8866, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.08104548677945497, | |
| "grad_norm": 1.8225231170654297, | |
| "learning_rate": 2.026342451874367e-06, | |
| "loss": 4.7456, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.09117617262688683, | |
| "grad_norm": 1.6859124898910522, | |
| "learning_rate": 2.279635258358663e-06, | |
| "loss": 4.5977, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.10130685847431871, | |
| "grad_norm": 1.448459506034851, | |
| "learning_rate": 2.532928064842959e-06, | |
| "loss": 4.5081, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11143754432175058, | |
| "grad_norm": 7.997034549713135, | |
| "learning_rate": 2.7862208713272543e-06, | |
| "loss": 4.4333, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.12156823016918246, | |
| "grad_norm": 10.502978324890137, | |
| "learning_rate": 3.03951367781155e-06, | |
| "loss": 4.3472, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.13169891601661432, | |
| "grad_norm": 1.39786958694458, | |
| "learning_rate": 3.2928064842958464e-06, | |
| "loss": 4.2736, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.1418296018640462, | |
| "grad_norm": 1.8371859788894653, | |
| "learning_rate": 3.5460992907801423e-06, | |
| "loss": 4.1658, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.15196028771147807, | |
| "grad_norm": 1.7692155838012695, | |
| "learning_rate": 3.799392097264438e-06, | |
| "loss": 4.121, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16209097355890995, | |
| "grad_norm": 2.3330609798431396, | |
| "learning_rate": 4.052684903748734e-06, | |
| "loss": 3.9781, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.17222165940634182, | |
| "grad_norm": 9.345723152160645, | |
| "learning_rate": 4.3059777102330295e-06, | |
| "loss": 3.8582, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.18235234525377367, | |
| "grad_norm": 4.135860919952393, | |
| "learning_rate": 4.559270516717326e-06, | |
| "loss": 3.6802, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.19248303110120554, | |
| "grad_norm": 3.1482667922973633, | |
| "learning_rate": 4.812563323201621e-06, | |
| "loss": 3.4976, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.20261371694863742, | |
| "grad_norm": 3.647608995437622, | |
| "learning_rate": 5.065856129685918e-06, | |
| "loss": 3.3249, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2127444027960693, | |
| "grad_norm": 4.777210235595703, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 3.1602, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.22287508864350117, | |
| "grad_norm": 3.177557945251465, | |
| "learning_rate": 5.5724417426545085e-06, | |
| "loss": 2.9832, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.23300577449093304, | |
| "grad_norm": 2.4862775802612305, | |
| "learning_rate": 5.825734549138805e-06, | |
| "loss": 2.86, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.24313646033836492, | |
| "grad_norm": 2.4592087268829346, | |
| "learning_rate": 6.0790273556231e-06, | |
| "loss": 2.76, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.25326714618579677, | |
| "grad_norm": 4.776676654815674, | |
| "learning_rate": 6.3323201621073974e-06, | |
| "loss": 2.7536, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.26339783203322864, | |
| "grad_norm": 3.8928306102752686, | |
| "learning_rate": 6.585612968591693e-06, | |
| "loss": 2.6681, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.2735285178806605, | |
| "grad_norm": 3.821058511734009, | |
| "learning_rate": 6.838905775075988e-06, | |
| "loss": 2.6202, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.2836592037280924, | |
| "grad_norm": 2.162001609802246, | |
| "learning_rate": 7.092198581560285e-06, | |
| "loss": 2.5764, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.29378988957552427, | |
| "grad_norm": 1.753183364868164, | |
| "learning_rate": 7.34549138804458e-06, | |
| "loss": 2.5782, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.30392057542295614, | |
| "grad_norm": 1.6950722932815552, | |
| "learning_rate": 7.598784194528876e-06, | |
| "loss": 2.5569, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.314051261270388, | |
| "grad_norm": 1.6093318462371826, | |
| "learning_rate": 7.852077001013173e-06, | |
| "loss": 2.498, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3241819471178199, | |
| "grad_norm": 1.3843328952789307, | |
| "learning_rate": 8.105369807497468e-06, | |
| "loss": 2.4841, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.33431263296525177, | |
| "grad_norm": 2.4331817626953125, | |
| "learning_rate": 8.358662613981764e-06, | |
| "loss": 2.4389, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.34444331881268364, | |
| "grad_norm": 1.6857964992523193, | |
| "learning_rate": 8.611955420466059e-06, | |
| "loss": 2.4589, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3545740046601155, | |
| "grad_norm": 1.323832631111145, | |
| "learning_rate": 8.865248226950355e-06, | |
| "loss": 2.3911, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.36470469050754734, | |
| "grad_norm": 1.2381794452667236, | |
| "learning_rate": 9.118541033434652e-06, | |
| "loss": 2.4117, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.3748353763549792, | |
| "grad_norm": 1.2798967361450195, | |
| "learning_rate": 9.371833839918947e-06, | |
| "loss": 2.3574, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.3849660622024111, | |
| "grad_norm": 1.5534555912017822, | |
| "learning_rate": 9.625126646403243e-06, | |
| "loss": 2.3676, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.39509674804984296, | |
| "grad_norm": 2.2549705505371094, | |
| "learning_rate": 9.878419452887538e-06, | |
| "loss": 2.3417, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.40522743389727484, | |
| "grad_norm": 1.787307620048523, | |
| "learning_rate": 1.0131712259371835e-05, | |
| "loss": 2.3361, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.4153581197447067, | |
| "grad_norm": 1.382817029953003, | |
| "learning_rate": 1.038500506585613e-05, | |
| "loss": 2.321, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4254888055921386, | |
| "grad_norm": 2.0577666759490967, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "loss": 2.316, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.43561949143957046, | |
| "grad_norm": 1.5490138530731201, | |
| "learning_rate": 1.0891590678824722e-05, | |
| "loss": 2.2916, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.44575017728700234, | |
| "grad_norm": 1.516607642173767, | |
| "learning_rate": 1.1144883485309017e-05, | |
| "loss": 2.2755, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.4558808631344342, | |
| "grad_norm": 1.2727792263031006, | |
| "learning_rate": 1.1398176291793314e-05, | |
| "loss": 2.2622, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4660115489818661, | |
| "grad_norm": 2.546395778656006, | |
| "learning_rate": 1.165146909827761e-05, | |
| "loss": 2.2575, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.47614223482929796, | |
| "grad_norm": 20.72391128540039, | |
| "learning_rate": 1.1904761904761905e-05, | |
| "loss": 2.2558, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.48627292067672984, | |
| "grad_norm": 2.8161871433258057, | |
| "learning_rate": 1.21580547112462e-05, | |
| "loss": 2.2573, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4964036065241617, | |
| "grad_norm": 2.429856300354004, | |
| "learning_rate": 1.2411347517730496e-05, | |
| "loss": 2.2347, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5065342923715935, | |
| "grad_norm": 1.5232312679290771, | |
| "learning_rate": 1.2664640324214795e-05, | |
| "loss": 2.2388, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5166649782190255, | |
| "grad_norm": 1.2436437606811523, | |
| "learning_rate": 1.291793313069909e-05, | |
| "loss": 2.2174, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5267956640664573, | |
| "grad_norm": 1.0977882146835327, | |
| "learning_rate": 1.3171225937183386e-05, | |
| "loss": 2.2086, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5369263499138892, | |
| "grad_norm": 1.1018918752670288, | |
| "learning_rate": 1.3424518743667681e-05, | |
| "loss": 2.2273, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.547057035761321, | |
| "grad_norm": 1.9143567085266113, | |
| "learning_rate": 1.3677811550151977e-05, | |
| "loss": 2.217, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.557187721608753, | |
| "grad_norm": 1.243256688117981, | |
| "learning_rate": 1.3931104356636274e-05, | |
| "loss": 2.1944, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.5673184074561848, | |
| "grad_norm": 1.2262539863586426, | |
| "learning_rate": 1.418439716312057e-05, | |
| "loss": 2.199, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.5774490933036166, | |
| "grad_norm": 1.6061089038848877, | |
| "learning_rate": 1.4437689969604865e-05, | |
| "loss": 2.1886, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5875797791510485, | |
| "grad_norm": 1.6153846979141235, | |
| "learning_rate": 1.469098277608916e-05, | |
| "loss": 2.2086, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.5977104649984804, | |
| "grad_norm": 1.023937702178955, | |
| "learning_rate": 1.4944275582573456e-05, | |
| "loss": 2.1795, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.6078411508459123, | |
| "grad_norm": 1.7739474773406982, | |
| "learning_rate": 1.5197568389057753e-05, | |
| "loss": 2.198, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6179718366933441, | |
| "grad_norm": 1.8380703926086426, | |
| "learning_rate": 1.545086119554205e-05, | |
| "loss": 2.1816, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.628102522540776, | |
| "grad_norm": 1.173989176750183, | |
| "learning_rate": 1.5704154002026345e-05, | |
| "loss": 2.1804, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6382332083882079, | |
| "grad_norm": 1.0197932720184326, | |
| "learning_rate": 1.595744680851064e-05, | |
| "loss": 2.1791, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.6483638942356398, | |
| "grad_norm": 1.3487520217895508, | |
| "learning_rate": 1.6210739614994936e-05, | |
| "loss": 2.1597, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.6584945800830716, | |
| "grad_norm": 1.2801405191421509, | |
| "learning_rate": 1.6464032421479232e-05, | |
| "loss": 2.1587, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.6686252659305035, | |
| "grad_norm": 1.4226802587509155, | |
| "learning_rate": 1.6717325227963527e-05, | |
| "loss": 2.1462, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.6787559517779354, | |
| "grad_norm": 6.232682228088379, | |
| "learning_rate": 1.6970618034447823e-05, | |
| "loss": 2.1429, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.6888866376253673, | |
| "grad_norm": 1.2026145458221436, | |
| "learning_rate": 1.7223910840932118e-05, | |
| "loss": 2.1549, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6990173234727991, | |
| "grad_norm": 1.0961929559707642, | |
| "learning_rate": 1.7477203647416414e-05, | |
| "loss": 2.138, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.709148009320231, | |
| "grad_norm": 1.0841082334518433, | |
| "learning_rate": 1.773049645390071e-05, | |
| "loss": 2.1235, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7192786951676629, | |
| "grad_norm": 6.760639667510986, | |
| "learning_rate": 1.7983789260385008e-05, | |
| "loss": 2.1259, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7294093810150947, | |
| "grad_norm": 1.358530044555664, | |
| "learning_rate": 1.8237082066869303e-05, | |
| "loss": 2.1259, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.7395400668625266, | |
| "grad_norm": 0.9958035349845886, | |
| "learning_rate": 1.84903748733536e-05, | |
| "loss": 2.1364, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.7496707527099584, | |
| "grad_norm": 1.4952771663665771, | |
| "learning_rate": 1.8743667679837894e-05, | |
| "loss": 2.1037, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.7598014385573904, | |
| "grad_norm": 1.4222711324691772, | |
| "learning_rate": 1.899696048632219e-05, | |
| "loss": 2.1181, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7699321244048222, | |
| "grad_norm": 0.9418506622314453, | |
| "learning_rate": 1.9250253292806485e-05, | |
| "loss": 2.1146, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.7800628102522541, | |
| "grad_norm": 1.6261628866195679, | |
| "learning_rate": 1.950354609929078e-05, | |
| "loss": 2.1089, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.7901934960996859, | |
| "grad_norm": 2.184122085571289, | |
| "learning_rate": 1.9756838905775076e-05, | |
| "loss": 2.1036, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8003241819471179, | |
| "grad_norm": 3.901762008666992, | |
| "learning_rate": 1.999979757969855e-05, | |
| "loss": 2.1042, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.8104548677945497, | |
| "grad_norm": 4.365869522094727, | |
| "learning_rate": 1.9863475200884386e-05, | |
| "loss": 2.1132, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8205855536419816, | |
| "grad_norm": 2.3704495429992676, | |
| "learning_rate": 1.9478106818608973e-05, | |
| "loss": 2.0985, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.8307162394894134, | |
| "grad_norm": 4.145130634307861, | |
| "learning_rate": 1.8853422720981873e-05, | |
| "loss": 2.1131, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.8408469253368454, | |
| "grad_norm": 1.318753719329834, | |
| "learning_rate": 1.800519575426379e-05, | |
| "loss": 2.1137, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.8509776111842772, | |
| "grad_norm": 1.3615083694458008, | |
| "learning_rate": 1.6954843069285113e-05, | |
| "loss": 2.1177, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.861108297031709, | |
| "grad_norm": 1.8678381443023682, | |
| "learning_rate": 1.5728885353034063e-05, | |
| "loss": 2.0999, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.8712389828791409, | |
| "grad_norm": 1.5799565315246582, | |
| "learning_rate": 1.4358277199447007e-05, | |
| "loss": 2.0918, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.8813696687265727, | |
| "grad_norm": 0.9612336754798889, | |
| "learning_rate": 1.2877625527080568e-05, | |
| "loss": 2.1174, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.8915003545740047, | |
| "grad_norm": 1.934597134590149, | |
| "learning_rate": 1.1324315778084788e-05, | |
| "loss": 2.1028, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.9016310404214365, | |
| "grad_norm": 2.4133620262145996, | |
| "learning_rate": 9.737567961355728e-06, | |
| "loss": 2.0958, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.9117617262688684, | |
| "grad_norm": 2.853616237640381, | |
| "learning_rate": 8.157446374132335e-06, | |
| "loss": 2.0875, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9218924121163002, | |
| "grad_norm": 1.1536306142807007, | |
| "learning_rate": 6.623848005890046e-06, | |
| "loss": 2.1093, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.9320230979637322, | |
| "grad_norm": 3.1503701210021973, | |
| "learning_rate": 5.175495166640958e-06, | |
| "loss": 2.0838, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.942153783811164, | |
| "grad_norm": 1.7781630754470825, | |
| "learning_rate": 3.848957775087207e-06, | |
| "loss": 2.0971, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.9522844696585959, | |
| "grad_norm": 3.86759614944458, | |
| "learning_rate": 2.6777299931828482e-06, | |
| "loss": 2.0888, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.9624151555060277, | |
| "grad_norm": 2.273083209991455, | |
| "learning_rate": 1.6913845214503065e-06, | |
| "loss": 2.0841, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.9725458413534597, | |
| "grad_norm": 5.05561637878418, | |
| "learning_rate": 9.148259085161204e-07, | |
| "loss": 2.0874, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.9826765272008915, | |
| "grad_norm": 5.530516147613525, | |
| "learning_rate": 3.676617282896666e-07, | |
| "loss": 2.0993, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.9928072130483234, | |
| "grad_norm": 1.7245455980300903, | |
| "learning_rate": 6.37075021310396e-08, | |
| "loss": 2.0655, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 9871, | |
| "total_flos": 1.1780246832876093e+18, | |
| "train_loss": 2.774943617143574, | |
| "train_runtime": 2530.0612, | |
| "train_samples_per_second": 31.21, | |
| "train_steps_per_second": 3.901 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 9871, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1780246832876093e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |