| { | |
| "best_global_step": 12630, | |
| "best_metric": 0.8520642201834863, | |
| "best_model_checkpoint": "./mbert-sst2-finetuned\\checkpoint-12630", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 12630, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.023752969121140142, | |
| "grad_norm": 9.634153366088867, | |
| "learning_rate": 1.9849564528899446e-05, | |
| "loss": 0.666, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.047505938242280284, | |
| "grad_norm": 4.3779520988464355, | |
| "learning_rate": 1.969121140142518e-05, | |
| "loss": 0.6312, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07125890736342043, | |
| "grad_norm": 4.55907678604126, | |
| "learning_rate": 1.9532858273950912e-05, | |
| "loss": 0.5593, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.09501187648456057, | |
| "grad_norm": 7.7376017570495605, | |
| "learning_rate": 1.9374505146476644e-05, | |
| "loss": 0.5619, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.1187648456057007, | |
| "grad_norm": 8.003072738647461, | |
| "learning_rate": 1.9216152019002377e-05, | |
| "loss": 0.5221, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14251781472684086, | |
| "grad_norm": 14.349533081054688, | |
| "learning_rate": 1.905779889152811e-05, | |
| "loss": 0.5207, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.166270783847981, | |
| "grad_norm": 7.375957489013672, | |
| "learning_rate": 1.889944576405384e-05, | |
| "loss": 0.478, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.19002375296912113, | |
| "grad_norm": 7.584378242492676, | |
| "learning_rate": 1.8741092636579576e-05, | |
| "loss": 0.4883, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.21377672209026127, | |
| "grad_norm": 15.048194885253906, | |
| "learning_rate": 1.8582739509105305e-05, | |
| "loss": 0.496, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2375296912114014, | |
| "grad_norm": 9.2789945602417, | |
| "learning_rate": 1.8424386381631038e-05, | |
| "loss": 0.4595, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.26128266033254155, | |
| "grad_norm": 5.139932155609131, | |
| "learning_rate": 1.826603325415677e-05, | |
| "loss": 0.4463, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2850356294536817, | |
| "grad_norm": 9.257096290588379, | |
| "learning_rate": 1.8107680126682503e-05, | |
| "loss": 0.4465, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.3087885985748218, | |
| "grad_norm": 7.5652313232421875, | |
| "learning_rate": 1.7949326999208236e-05, | |
| "loss": 0.4553, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.332541567695962, | |
| "grad_norm": 5.573002815246582, | |
| "learning_rate": 1.779097387173397e-05, | |
| "loss": 0.4098, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.35629453681710216, | |
| "grad_norm": 8.824542999267578, | |
| "learning_rate": 1.76326207442597e-05, | |
| "loss": 0.4441, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.38004750593824227, | |
| "grad_norm": 8.202225685119629, | |
| "learning_rate": 1.747426761678543e-05, | |
| "loss": 0.4326, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.40380047505938244, | |
| "grad_norm": 11.430865287780762, | |
| "learning_rate": 1.7315914489311167e-05, | |
| "loss": 0.4169, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.42755344418052255, | |
| "grad_norm": 6.839201927185059, | |
| "learning_rate": 1.71575613618369e-05, | |
| "loss": 0.4369, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.4513064133016627, | |
| "grad_norm": 7.273525714874268, | |
| "learning_rate": 1.699920823436263e-05, | |
| "loss": 0.4267, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4750593824228028, | |
| "grad_norm": 7.3494181632995605, | |
| "learning_rate": 1.6842438638163107e-05, | |
| "loss": 0.4025, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.498812351543943, | |
| "grad_norm": 12.297106742858887, | |
| "learning_rate": 1.6684085510688836e-05, | |
| "loss": 0.438, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.5225653206650831, | |
| "grad_norm": 3.992037057876587, | |
| "learning_rate": 1.652573238321457e-05, | |
| "loss": 0.4088, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5463182897862233, | |
| "grad_norm": 3.9685425758361816, | |
| "learning_rate": 1.6367379255740302e-05, | |
| "loss": 0.3806, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5700712589073634, | |
| "grad_norm": 10.497099876403809, | |
| "learning_rate": 1.6209026128266034e-05, | |
| "loss": 0.4047, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.5938242280285035, | |
| "grad_norm": 3.8488385677337646, | |
| "learning_rate": 1.6050673000791767e-05, | |
| "loss": 0.3827, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6175771971496437, | |
| "grad_norm": 9.80221939086914, | |
| "learning_rate": 1.58923198733175e-05, | |
| "loss": 0.4008, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6413301662707839, | |
| "grad_norm": 4.120981693267822, | |
| "learning_rate": 1.5733966745843233e-05, | |
| "loss": 0.4045, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.665083135391924, | |
| "grad_norm": 9.871939659118652, | |
| "learning_rate": 1.5575613618368962e-05, | |
| "loss": 0.3787, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.6888361045130641, | |
| "grad_norm": 23.69073486328125, | |
| "learning_rate": 1.5417260490894698e-05, | |
| "loss": 0.369, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.7125890736342043, | |
| "grad_norm": 8.312384605407715, | |
| "learning_rate": 1.525890736342043e-05, | |
| "loss": 0.391, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7363420427553444, | |
| "grad_norm": 5.4470367431640625, | |
| "learning_rate": 1.510055423594616e-05, | |
| "loss": 0.3951, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.7600950118764845, | |
| "grad_norm": 9.961467742919922, | |
| "learning_rate": 1.4942201108471895e-05, | |
| "loss": 0.3768, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.7838479809976246, | |
| "grad_norm": 4.254888534545898, | |
| "learning_rate": 1.4783847980997626e-05, | |
| "loss": 0.3401, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.8076009501187649, | |
| "grad_norm": 8.969467163085938, | |
| "learning_rate": 1.4625494853523359e-05, | |
| "loss": 0.3865, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.831353919239905, | |
| "grad_norm": 5.86577033996582, | |
| "learning_rate": 1.446714172604909e-05, | |
| "loss": 0.3819, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8551068883610451, | |
| "grad_norm": 6.642719268798828, | |
| "learning_rate": 1.4308788598574822e-05, | |
| "loss": 0.3693, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.8788598574821853, | |
| "grad_norm": 3.827361583709717, | |
| "learning_rate": 1.4150435471100557e-05, | |
| "loss": 0.3726, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.9026128266033254, | |
| "grad_norm": 13.236315727233887, | |
| "learning_rate": 1.3992082343626288e-05, | |
| "loss": 0.3504, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.9263657957244655, | |
| "grad_norm": 9.49841594696045, | |
| "learning_rate": 1.383372921615202e-05, | |
| "loss": 0.3586, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "grad_norm": 17.966522216796875, | |
| "learning_rate": 1.3675376088677752e-05, | |
| "loss": 0.347, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9738717339667459, | |
| "grad_norm": 10.867168426513672, | |
| "learning_rate": 1.3517022961203485e-05, | |
| "loss": 0.3441, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.997624703087886, | |
| "grad_norm": 9.008316040039062, | |
| "learning_rate": 1.3358669833729217e-05, | |
| "loss": 0.3721, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8314220183486238, | |
| "eval_loss": 0.38916900753974915, | |
| "eval_runtime": 3.1803, | |
| "eval_samples_per_second": 274.19, | |
| "eval_steps_per_second": 17.294, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 1.0213776722090262, | |
| "grad_norm": 20.658517837524414, | |
| "learning_rate": 1.320031670625495e-05, | |
| "loss": 0.3262, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.0451306413301662, | |
| "grad_norm": 6.5011210441589355, | |
| "learning_rate": 1.3041963578780683e-05, | |
| "loss": 0.3062, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.0688836104513064, | |
| "grad_norm": 15.551307678222656, | |
| "learning_rate": 1.2883610451306414e-05, | |
| "loss": 0.2908, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.0926365795724466, | |
| "grad_norm": 28.33782958984375, | |
| "learning_rate": 1.2725257323832148e-05, | |
| "loss": 0.2983, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.1163895486935866, | |
| "grad_norm": 10.99241828918457, | |
| "learning_rate": 1.256690419635788e-05, | |
| "loss": 0.2897, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.1401425178147269, | |
| "grad_norm": 7.335752010345459, | |
| "learning_rate": 1.2408551068883612e-05, | |
| "loss": 0.3304, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.1638954869358669, | |
| "grad_norm": 8.83614730834961, | |
| "learning_rate": 1.2250197941409343e-05, | |
| "loss": 0.3016, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.187648456057007, | |
| "grad_norm": 10.162006378173828, | |
| "learning_rate": 1.2091844813935076e-05, | |
| "loss": 0.2776, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.2114014251781473, | |
| "grad_norm": 10.363824844360352, | |
| "learning_rate": 1.1933491686460807e-05, | |
| "loss": 0.3019, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.2351543942992875, | |
| "grad_norm": 1.9022657871246338, | |
| "learning_rate": 1.1775138558986541e-05, | |
| "loss": 0.2856, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.2589073634204275, | |
| "grad_norm": 11.627281188964844, | |
| "learning_rate": 1.1616785431512274e-05, | |
| "loss": 0.284, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.2826603325415677, | |
| "grad_norm": 2.0096428394317627, | |
| "learning_rate": 1.1458432304038005e-05, | |
| "loss": 0.2938, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.3064133016627077, | |
| "grad_norm": 22.037084579467773, | |
| "learning_rate": 1.1300079176563738e-05, | |
| "loss": 0.293, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.330166270783848, | |
| "grad_norm": 11.04839038848877, | |
| "learning_rate": 1.1141726049089469e-05, | |
| "loss": 0.2829, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.3539192399049882, | |
| "grad_norm": 13.687749862670898, | |
| "learning_rate": 1.0983372921615203e-05, | |
| "loss": 0.3126, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.3776722090261282, | |
| "grad_norm": 5.973034381866455, | |
| "learning_rate": 1.0825019794140935e-05, | |
| "loss": 0.2812, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.4014251781472684, | |
| "grad_norm": 2.820462226867676, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 0.3018, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.4251781472684084, | |
| "grad_norm": 9.669486999511719, | |
| "learning_rate": 1.05083135391924e-05, | |
| "loss": 0.2653, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.4489311163895486, | |
| "grad_norm": 9.360280990600586, | |
| "learning_rate": 1.0349960411718131e-05, | |
| "loss": 0.2887, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 1.4726840855106889, | |
| "grad_norm": 8.551124572753906, | |
| "learning_rate": 1.0191607284243866e-05, | |
| "loss": 0.2948, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 1.496437054631829, | |
| "grad_norm": 12.017287254333496, | |
| "learning_rate": 1.0033254156769597e-05, | |
| "loss": 0.2931, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 1.520190023752969, | |
| "grad_norm": 18.4260311126709, | |
| "learning_rate": 9.87490102929533e-06, | |
| "loss": 0.2933, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 1.5439429928741093, | |
| "grad_norm": 21.07431983947754, | |
| "learning_rate": 9.716547901821062e-06, | |
| "loss": 0.3061, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.5676959619952493, | |
| "grad_norm": 8.733689308166504, | |
| "learning_rate": 9.558194774346793e-06, | |
| "loss": 0.2508, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 1.5914489311163895, | |
| "grad_norm": 4.291722297668457, | |
| "learning_rate": 9.399841646872526e-06, | |
| "loss": 0.2988, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 1.6152019002375297, | |
| "grad_norm": 9.211682319641113, | |
| "learning_rate": 9.241488519398259e-06, | |
| "loss": 0.3069, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 1.63895486935867, | |
| "grad_norm": 9.624086380004883, | |
| "learning_rate": 9.083135391923991e-06, | |
| "loss": 0.2933, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 1.66270783847981, | |
| "grad_norm": 9.931711196899414, | |
| "learning_rate": 8.924782264449724e-06, | |
| "loss": 0.2724, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.68646080760095, | |
| "grad_norm": 4.174336910247803, | |
| "learning_rate": 8.766429136975457e-06, | |
| "loss": 0.2615, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 1.7102137767220902, | |
| "grad_norm": 11.723244667053223, | |
| "learning_rate": 8.608076009501188e-06, | |
| "loss": 0.285, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 1.7339667458432304, | |
| "grad_norm": 12.756488800048828, | |
| "learning_rate": 8.44972288202692e-06, | |
| "loss": 0.2771, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 1.7577197149643706, | |
| "grad_norm": 9.429771423339844, | |
| "learning_rate": 8.291369754552654e-06, | |
| "loss": 0.2824, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 1.7814726840855108, | |
| "grad_norm": 13.813314437866211, | |
| "learning_rate": 8.133016627078385e-06, | |
| "loss": 0.2965, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.8052256532066508, | |
| "grad_norm": 8.05421257019043, | |
| "learning_rate": 7.974663499604119e-06, | |
| "loss": 0.2613, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 1.8289786223277908, | |
| "grad_norm": 5.644511699676514, | |
| "learning_rate": 7.81631037212985e-06, | |
| "loss": 0.2827, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 1.852731591448931, | |
| "grad_norm": 10.623388290405273, | |
| "learning_rate": 7.657957244655583e-06, | |
| "loss": 0.2655, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 1.8764845605700713, | |
| "grad_norm": 27.798290252685547, | |
| "learning_rate": 7.499604117181315e-06, | |
| "loss": 0.2992, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 1.9002375296912115, | |
| "grad_norm": 8.613077163696289, | |
| "learning_rate": 7.3412509897070475e-06, | |
| "loss": 0.2635, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.9239904988123515, | |
| "grad_norm": 10.832771301269531, | |
| "learning_rate": 7.1828978622327794e-06, | |
| "loss": 0.2672, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 1.9477434679334917, | |
| "grad_norm": 16.065528869628906, | |
| "learning_rate": 7.024544734758511e-06, | |
| "loss": 0.262, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 1.9714964370546317, | |
| "grad_norm": 13.325654029846191, | |
| "learning_rate": 6.866191607284245e-06, | |
| "loss": 0.2826, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 1.995249406175772, | |
| "grad_norm": 16.392105102539062, | |
| "learning_rate": 6.707838479809977e-06, | |
| "loss": 0.2661, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8497706422018348, | |
| "eval_loss": 0.3956442177295685, | |
| "eval_runtime": 2.9709, | |
| "eval_samples_per_second": 293.511, | |
| "eval_steps_per_second": 18.513, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 2.019002375296912, | |
| "grad_norm": 24.83507537841797, | |
| "learning_rate": 6.54948535233571e-06, | |
| "loss": 0.2195, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.0427553444180524, | |
| "grad_norm": 13.459088325500488, | |
| "learning_rate": 6.3911322248614415e-06, | |
| "loss": 0.2333, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 2.0665083135391926, | |
| "grad_norm": 23.071367263793945, | |
| "learning_rate": 6.232779097387173e-06, | |
| "loss": 0.2222, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 2.0902612826603324, | |
| "grad_norm": 20.328954696655273, | |
| "learning_rate": 6.074425969912906e-06, | |
| "loss": 0.2417, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 2.1140142517814726, | |
| "grad_norm": 12.929640769958496, | |
| "learning_rate": 5.916072842438638e-06, | |
| "loss": 0.2158, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 2.137767220902613, | |
| "grad_norm": 3.765835762023926, | |
| "learning_rate": 5.757719714964372e-06, | |
| "loss": 0.2228, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.161520190023753, | |
| "grad_norm": 15.415190696716309, | |
| "learning_rate": 5.599366587490104e-06, | |
| "loss": 0.2107, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 2.1852731591448933, | |
| "grad_norm": 8.424735069274902, | |
| "learning_rate": 5.441013460015836e-06, | |
| "loss": 0.2393, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 2.209026128266033, | |
| "grad_norm": 10.958281517028809, | |
| "learning_rate": 5.282660332541568e-06, | |
| "loss": 0.206, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 2.2327790973871733, | |
| "grad_norm": 9.020813941955566, | |
| "learning_rate": 5.1243072050673e-06, | |
| "loss": 0.1987, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 2.2565320665083135, | |
| "grad_norm": 28.11874008178711, | |
| "learning_rate": 4.965954077593033e-06, | |
| "loss": 0.2121, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.2802850356294537, | |
| "grad_norm": 13.661537170410156, | |
| "learning_rate": 4.807600950118766e-06, | |
| "loss": 0.2313, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 2.304038004750594, | |
| "grad_norm": 5.820676803588867, | |
| "learning_rate": 4.6492478226444976e-06, | |
| "loss": 0.2076, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 2.3277909738717337, | |
| "grad_norm": 30.28268051147461, | |
| "learning_rate": 4.4908946951702295e-06, | |
| "loss": 0.2142, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 2.351543942992874, | |
| "grad_norm": 1.176238775253296, | |
| "learning_rate": 4.332541567695962e-06, | |
| "loss": 0.2058, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 2.375296912114014, | |
| "grad_norm": 7.50606632232666, | |
| "learning_rate": 4.1757719714964375e-06, | |
| "loss": 0.2493, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.3990498812351544, | |
| "grad_norm": 4.3427205085754395, | |
| "learning_rate": 4.017418844022169e-06, | |
| "loss": 0.2137, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 2.4228028503562946, | |
| "grad_norm": 10.145256996154785, | |
| "learning_rate": 3.859065716547902e-06, | |
| "loss": 0.2129, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 2.446555819477435, | |
| "grad_norm": 7.9584527015686035, | |
| "learning_rate": 3.7007125890736345e-06, | |
| "loss": 0.2107, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 2.470308788598575, | |
| "grad_norm": 9.902059555053711, | |
| "learning_rate": 3.542359461599367e-06, | |
| "loss": 0.2231, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 2.494061757719715, | |
| "grad_norm": 18.705286026000977, | |
| "learning_rate": 3.3840063341250996e-06, | |
| "loss": 0.2121, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.517814726840855, | |
| "grad_norm": 29.508682250976562, | |
| "learning_rate": 3.2256532066508315e-06, | |
| "loss": 0.2421, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 2.5415676959619953, | |
| "grad_norm": 4.328601360321045, | |
| "learning_rate": 3.067300079176564e-06, | |
| "loss": 0.1913, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 2.5653206650831355, | |
| "grad_norm": 3.3285276889801025, | |
| "learning_rate": 2.908946951702296e-06, | |
| "loss": 0.2426, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 2.5890736342042757, | |
| "grad_norm": 4.220156669616699, | |
| "learning_rate": 2.750593824228029e-06, | |
| "loss": 0.2375, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 2.6128266033254155, | |
| "grad_norm": 24.65312385559082, | |
| "learning_rate": 2.5922406967537612e-06, | |
| "loss": 0.2304, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.6365795724465557, | |
| "grad_norm": 8.729198455810547, | |
| "learning_rate": 2.4338875692794936e-06, | |
| "loss": 0.2067, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 2.660332541567696, | |
| "grad_norm": 15.05286693572998, | |
| "learning_rate": 2.275534441805226e-06, | |
| "loss": 0.2415, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 2.684085510688836, | |
| "grad_norm": 26.644716262817383, | |
| "learning_rate": 2.1171813143309582e-06, | |
| "loss": 0.2401, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 2.7078384798099764, | |
| "grad_norm": 10.57201862335205, | |
| "learning_rate": 1.9588281868566906e-06, | |
| "loss": 0.2406, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 2.731591448931116, | |
| "grad_norm": 6.5845465660095215, | |
| "learning_rate": 1.800475059382423e-06, | |
| "loss": 0.1857, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.7553444180522564, | |
| "grad_norm": 7.665100574493408, | |
| "learning_rate": 1.6421219319081552e-06, | |
| "loss": 0.2249, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 2.7790973871733966, | |
| "grad_norm": 1.698305368423462, | |
| "learning_rate": 1.4837688044338878e-06, | |
| "loss": 0.2164, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 2.802850356294537, | |
| "grad_norm": 45.78740310668945, | |
| "learning_rate": 1.3254156769596199e-06, | |
| "loss": 0.2125, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 2.826603325415677, | |
| "grad_norm": 18.53032684326172, | |
| "learning_rate": 1.1670625494853524e-06, | |
| "loss": 0.2076, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 2.850356294536817, | |
| "grad_norm": 35.8634147644043, | |
| "learning_rate": 1.0087094220110848e-06, | |
| "loss": 0.2141, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.8741092636579575, | |
| "grad_norm": 10.635025024414062, | |
| "learning_rate": 8.503562945368171e-07, | |
| "loss": 0.1897, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 2.8978622327790973, | |
| "grad_norm": 2.4744296073913574, | |
| "learning_rate": 6.920031670625496e-07, | |
| "loss": 0.1959, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 2.9216152019002375, | |
| "grad_norm": 24.558425903320312, | |
| "learning_rate": 5.336500395882819e-07, | |
| "loss": 0.2176, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 2.9453681710213777, | |
| "grad_norm": 36.37364196777344, | |
| "learning_rate": 3.76880443388757e-07, | |
| "loss": 0.2252, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 2.969121140142518, | |
| "grad_norm": 0.5130617022514343, | |
| "learning_rate": 2.20110847189232e-07, | |
| "loss": 0.1907, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.992874109263658, | |
| "grad_norm": 33.36646270751953, | |
| "learning_rate": 6.175771971496438e-08, | |
| "loss": 0.2017, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8520642201834863, | |
| "eval_loss": 0.5416586995124817, | |
| "eval_runtime": 2.9354, | |
| "eval_samples_per_second": 297.064, | |
| "eval_steps_per_second": 18.737, | |
| "step": 12630 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 12630, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.316079940232192e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |