{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.5220125786163523, "eval_steps": 500, "global_step": 3500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "grad_norm": 0.0008633775869384408, "learning_rate": 5.987915407854985e-05, "loss": 0.0029, "step": 10 }, { "epoch": 0.02, "grad_norm": 0.0003390114870853722, "learning_rate": 5.97583081570997e-05, "loss": 0.0, "step": 20 }, { "epoch": 0.03, "grad_norm": 6.441377627197653e-05, "learning_rate": 5.963746223564955e-05, "loss": 0.0, "step": 30 }, { "epoch": 0.04, "grad_norm": 1.1177830856468063e-06, "learning_rate": 5.95166163141994e-05, "loss": 0.0027, "step": 40 }, { "epoch": 0.05, "grad_norm": 0.017336083576083183, "learning_rate": 5.9407854984894266e-05, "loss": 0.0504, "step": 50 }, { "epoch": 0.06, "grad_norm": 0.015523849986493587, "learning_rate": 5.928700906344411e-05, "loss": 0.0561, "step": 60 }, { "epoch": 0.07, "grad_norm": 0.974118709564209, "learning_rate": 5.916616314199396e-05, "loss": 0.0022, "step": 70 }, { "epoch": 0.08, "grad_norm": 0.051055677235126495, "learning_rate": 5.904531722054381e-05, "loss": 0.0006, "step": 80 }, { "epoch": 0.09, "grad_norm": 0.0001077930282917805, "learning_rate": 5.892447129909366e-05, "loss": 0.0, "step": 90 }, { "epoch": 0.1, "grad_norm": 1.7529870092403144e-05, "learning_rate": 5.88036253776435e-05, "loss": 0.0, "step": 100 }, { "epoch": 0.11, "grad_norm": 3.870090586133301e-05, "learning_rate": 5.869486404833837e-05, "loss": 0.0867, "step": 110 }, { "epoch": 0.12, "grad_norm": 0.0002746480458881706, "learning_rate": 5.857401812688821e-05, "loss": 0.0002, "step": 120 }, { "epoch": 0.13, "grad_norm": 0.30150529742240906, "learning_rate": 5.8453172205438065e-05, "loss": 0.0001, "step": 130 }, { "epoch": 0.14, "grad_norm": 0.07647743076086044, "learning_rate": 5.833232628398792e-05, "loss": 0.0019, "step": 140 }, { "epoch": 0.15, "grad_norm": 3.936241046176292e-05, "learning_rate": 5.821148036253777e-05, "loss": 0.0003, "step": 150 }, { "epoch": 0.16, "grad_norm": 0.019184544682502747, "learning_rate": 5.809063444108762e-05, "loss": 0.0001, "step": 160 }, { "epoch": 0.17, "grad_norm": 0.004856175277382135, "learning_rate": 5.796978851963746e-05, "loss": 0.0, "step": 170 }, { "epoch": 0.18, "grad_norm": 1.3400563148024958e-06, "learning_rate": 5.7848942598187314e-05, "loss": 0.0002, "step": 180 }, { "epoch": 0.19, "grad_norm": 0.7676287889480591, "learning_rate": 5.772809667673716e-05, "loss": 0.0218, "step": 190 }, { "epoch": 0.2, "grad_norm": 0.0006767328595742583, "learning_rate": 5.760725075528701e-05, "loss": 0.0, "step": 200 }, { "epoch": 0.21, "grad_norm": 0.4616200029850006, "learning_rate": 5.749848942598187e-05, "loss": 0.0479, "step": 210 }, { "epoch": 0.22, "grad_norm": 6.79653263092041, "learning_rate": 5.737764350453173e-05, "loss": 0.0514, "step": 220 }, { "epoch": 0.23, "grad_norm": 0.00680534215644002, "learning_rate": 5.725679758308157e-05, "loss": 0.0006, "step": 230 }, { "epoch": 0.24, "grad_norm": 8.962566789705306e-05, "learning_rate": 5.7135951661631424e-05, "loss": 0.0003, "step": 240 }, { "epoch": 0.25, "grad_norm": 0.016912635415792465, "learning_rate": 5.701510574018127e-05, "loss": 0.0044, "step": 250 }, { "epoch": 0.26, "grad_norm": 7.953944206237793, "learning_rate": 5.689425981873112e-05, "loss": 0.0081, "step": 260 }, { "epoch": 0.27, "grad_norm": 0.4362787902355194, "learning_rate": 5.6773413897280964e-05, "loss": 0.0556, "step": 270 }, { "epoch": 0.28, "grad_norm": 0.0008250115206465125, "learning_rate": 5.6652567975830816e-05, "loss": 0.0043, "step": 280 }, { "epoch": 0.29, "grad_norm": 0.056627318263053894, "learning_rate": 5.653172205438066e-05, "loss": 0.0021, "step": 290 }, { "epoch": 0.3, "grad_norm": 0.24282953143119812, "learning_rate": 5.641087613293052e-05, "loss": 0.0019, "step": 300 }, { "epoch": 0.31, "grad_norm": 0.19482052326202393, "learning_rate": 5.629003021148036e-05, "loss": 0.0041, "step": 310 }, { "epoch": 0.32, "grad_norm": 6.74727198202163e-05, "learning_rate": 5.6169184290030214e-05, "loss": 0.0, "step": 320 }, { "epoch": 0.33, "grad_norm": 0.0013290628558024764, "learning_rate": 5.604833836858006e-05, "loss": 0.5076, "step": 330 }, { "epoch": 0.34, "grad_norm": 0.5521184206008911, "learning_rate": 5.592749244712991e-05, "loss": 0.0011, "step": 340 }, { "epoch": 0.35, "grad_norm": 0.0012500255834311247, "learning_rate": 5.5806646525679755e-05, "loss": 0.0049, "step": 350 }, { "epoch": 0.36, "grad_norm": 9.534443989878127e-08, "learning_rate": 5.5685800604229606e-05, "loss": 0.0014, "step": 360 }, { "epoch": 0.37, "grad_norm": 3.4361846701358445e-08, "learning_rate": 5.5564954682779464e-05, "loss": 0.0302, "step": 370 }, { "epoch": 0.38, "grad_norm": 1.528064785816241e-08, "learning_rate": 5.544410876132931e-05, "loss": 0.008, "step": 380 }, { "epoch": 0.39, "grad_norm": 8.260390131908935e-06, "learning_rate": 5.532326283987916e-05, "loss": 0.0, "step": 390 }, { "epoch": 0.4, "grad_norm": 3.2616500854492188, "learning_rate": 5.5202416918429004e-05, "loss": 0.0058, "step": 400 }, { "epoch": 0.41, "grad_norm": 6.549715294568159e-07, "learning_rate": 5.5081570996978856e-05, "loss": 0.0004, "step": 410 }, { "epoch": 0.42, "grad_norm": 1.4332237697090022e-05, "learning_rate": 5.49607250755287e-05, "loss": 0.0002, "step": 420 }, { "epoch": 0.43, "grad_norm": 0.00038707954809069633, "learning_rate": 5.483987915407855e-05, "loss": 0.0, "step": 430 }, { "epoch": 0.44, "grad_norm": 7.052252044559282e-07, "learning_rate": 5.4719033232628396e-05, "loss": 0.001, "step": 440 }, { "epoch": 0.45, "grad_norm": 2.283226336885491e-07, "learning_rate": 5.4598187311178254e-05, "loss": 0.0, "step": 450 }, { "epoch": 0.46, "grad_norm": 3.97150288335979e-05, "learning_rate": 5.44773413897281e-05, "loss": 0.0, "step": 460 }, { "epoch": 0.47, "grad_norm": 5.106105049890175e-08, "learning_rate": 5.435649546827795e-05, "loss": 0.0, "step": 470 }, { "epoch": 0.48, "grad_norm": 0.0003172747674398124, "learning_rate": 5.4235649546827795e-05, "loss": 0.0, "step": 480 }, { "epoch": 0.49, "grad_norm": 0.6184738278388977, "learning_rate": 5.4114803625377646e-05, "loss": 0.0002, "step": 490 }, { "epoch": 0.5, "grad_norm": 1.0878424205884585e-07, "learning_rate": 5.399395770392749e-05, "loss": 0.0, "step": 500 }, { "epoch": 0.51, "grad_norm": 8.74086580893163e-09, "learning_rate": 5.387311178247734e-05, "loss": 0.0, "step": 510 }, { "epoch": 0.52, "grad_norm": 1.5521897012149566e-07, "learning_rate": 5.375226586102719e-05, "loss": 0.0, "step": 520 }, { "epoch": 0.53, "grad_norm": 2.9636362341989297e-06, "learning_rate": 5.3631419939577045e-05, "loss": 0.0, "step": 530 }, { "epoch": 0.54, "grad_norm": 0.0023933914490044117, "learning_rate": 5.351057401812689e-05, "loss": 0.0004, "step": 540 }, { "epoch": 0.55, "grad_norm": 0.0006123024504631758, "learning_rate": 5.338972809667674e-05, "loss": 0.0024, "step": 550 }, { "epoch": 0.56, "grad_norm": 0.004097294993698597, "learning_rate": 5.3268882175226585e-05, "loss": 0.0673, "step": 560 }, { "epoch": 0.57, "grad_norm": 2.348265013552009e-07, "learning_rate": 5.3148036253776436e-05, "loss": 0.0003, "step": 570 }, { "epoch": 0.58, "grad_norm": 2.388081270510156e-07, "learning_rate": 5.302719033232628e-05, "loss": 0.0002, "step": 580 }, { "epoch": 0.59, "grad_norm": 1.3560330671680276e-06, "learning_rate": 5.290634441087613e-05, "loss": 0.0, "step": 590 }, { "epoch": 0.6, "grad_norm": 1.3283443678346885e-08, "learning_rate": 5.2797583081571e-05, "loss": 0.0986, "step": 600 }, { "epoch": 0.61, "grad_norm": 0.00010959297651425004, "learning_rate": 5.267673716012085e-05, "loss": 0.0, "step": 610 }, { "epoch": 0.62, "grad_norm": 0.00011195711704203859, "learning_rate": 5.25558912386707e-05, "loss": 0.0504, "step": 620 }, { "epoch": 0.63, "grad_norm": 2.1371073722839355, "learning_rate": 5.2435045317220546e-05, "loss": 0.0036, "step": 630 }, { "epoch": 0.64, "grad_norm": 0.001383950817398727, "learning_rate": 5.23141993957704e-05, "loss": 0.0052, "step": 640 }, { "epoch": 0.65, "grad_norm": 0.001337925554253161, "learning_rate": 5.219335347432024e-05, "loss": 0.0001, "step": 650 }, { "epoch": 0.66, "grad_norm": 2.282190052937949e-06, "learning_rate": 5.207250755287009e-05, "loss": 0.0001, "step": 660 }, { "epoch": 0.67, "grad_norm": 0.0006644165841862559, "learning_rate": 5.195166163141994e-05, "loss": 0.0, "step": 670 }, { "epoch": 0.68, "grad_norm": 0.00881275162100792, "learning_rate": 5.1830815709969795e-05, "loss": 0.0023, "step": 680 }, { "epoch": 0.69, "grad_norm": 0.0035460491199046373, "learning_rate": 5.170996978851964e-05, "loss": 0.0, "step": 690 }, { "epoch": 0.7, "grad_norm": 0.00038928634603507817, "learning_rate": 5.158912386706949e-05, "loss": 0.0008, "step": 700 }, { "epoch": 0.71, "grad_norm": 5.376980993787583e-07, "learning_rate": 5.1468277945619336e-05, "loss": 0.0022, "step": 710 }, { "epoch": 0.72, "grad_norm": 9.064051118912175e-05, "learning_rate": 5.134743202416919e-05, "loss": 0.0027, "step": 720 }, { "epoch": 0.73, "grad_norm": 0.0011442841496318579, "learning_rate": 5.122658610271903e-05, "loss": 0.0003, "step": 730 }, { "epoch": 0.74, "grad_norm": 0.004664595704525709, "learning_rate": 5.110574018126888e-05, "loss": 0.0022, "step": 740 }, { "epoch": 0.75, "grad_norm": 0.0058820717968046665, "learning_rate": 5.098489425981873e-05, "loss": 0.001, "step": 750 }, { "epoch": 0.76, "grad_norm": 0.030588263645768166, "learning_rate": 5.0864048338368586e-05, "loss": 0.0005, "step": 760 }, { "epoch": 0.77, "grad_norm": 0.002594970166683197, "learning_rate": 5.074320241691843e-05, "loss": 0.0161, "step": 770 }, { "epoch": 0.78, "grad_norm": 0.0027177969459444284, "learning_rate": 5.062235649546828e-05, "loss": 0.0, "step": 780 }, { "epoch": 0.79, "grad_norm": 7.966559678607155e-08, "learning_rate": 5.0501510574018126e-05, "loss": 0.0002, "step": 790 }, { "epoch": 0.81, "grad_norm": 0.006196084897965193, "learning_rate": 5.038066465256798e-05, "loss": 0.0004, "step": 800 }, { "epoch": 0.82, "grad_norm": 0.000184666394488886, "learning_rate": 5.025981873111782e-05, "loss": 0.0164, "step": 810 }, { "epoch": 0.83, "grad_norm": 0.0, "learning_rate": 5.013897280966767e-05, "loss": 0.0071, "step": 820 }, { "epoch": 0.84, "grad_norm": 2.8641309768318024e-07, "learning_rate": 5.0018126888217525e-05, "loss": 0.0003, "step": 830 }, { "epoch": 0.85, "grad_norm": 0.0019273586804047227, "learning_rate": 4.9897280966767376e-05, "loss": 0.0, "step": 840 }, { "epoch": 0.86, "grad_norm": 0.0644882395863533, "learning_rate": 4.977643504531722e-05, "loss": 0.0025, "step": 850 }, { "epoch": 0.87, "grad_norm": 0.00033757061464712024, "learning_rate": 4.965558912386707e-05, "loss": 0.0004, "step": 860 }, { "epoch": 0.88, "grad_norm": 0.0009230116847902536, "learning_rate": 4.9534743202416916e-05, "loss": 0.0156, "step": 870 }, { "epoch": 0.89, "grad_norm": 2.1307731003616937e-05, "learning_rate": 4.941389728096677e-05, "loss": 0.0026, "step": 880 }, { "epoch": 0.9, "grad_norm": 4.2529451071970925e-09, "learning_rate": 4.929305135951661e-05, "loss": 0.0, "step": 890 }, { "epoch": 0.91, "grad_norm": 2.9835612735951145e-07, "learning_rate": 4.9172205438066464e-05, "loss": 0.0005, "step": 900 }, { "epoch": 0.92, "grad_norm": 0.02193087339401245, "learning_rate": 4.905135951661632e-05, "loss": 0.0016, "step": 910 }, { "epoch": 0.93, "grad_norm": 0.0027798530645668507, "learning_rate": 4.8930513595166166e-05, "loss": 0.0002, "step": 920 }, { "epoch": 0.94, "grad_norm": 0.0008504379075020552, "learning_rate": 4.880966767371602e-05, "loss": 0.0007, "step": 930 }, { "epoch": 0.95, "grad_norm": 4.395148067715127e-08, "learning_rate": 4.868882175226586e-05, "loss": 0.0, "step": 940 }, { "epoch": 0.96, "grad_norm": 1.3475013474817388e-05, "learning_rate": 4.8567975830815714e-05, "loss": 0.011, "step": 950 }, { "epoch": 0.97, "grad_norm": 4.958523192044595e-08, "learning_rate": 4.844712990936556e-05, "loss": 0.0001, "step": 960 }, { "epoch": 0.98, "grad_norm": 0.00013180672249291092, "learning_rate": 4.832628398791541e-05, "loss": 0.0, "step": 970 }, { "epoch": 0.99, "grad_norm": 0.000387900770874694, "learning_rate": 4.8205438066465254e-05, "loss": 0.0515, "step": 980 }, { "epoch": 1.0, "grad_norm": 0.024301297962665558, "learning_rate": 4.808459214501511e-05, "loss": 0.0002, "step": 990 }, { "epoch": 1.01, "grad_norm": 0.054384149610996246, "learning_rate": 4.796374622356496e-05, "loss": 0.0002, "step": 1000 }, { "epoch": 1.02, "grad_norm": 0.0062861149199306965, "learning_rate": 4.784290030211481e-05, "loss": 0.0008, "step": 1010 }, { "epoch": 1.03, "grad_norm": 0.00028014209237881005, "learning_rate": 4.772205438066465e-05, "loss": 0.0298, "step": 1020 }, { "epoch": 1.04, "grad_norm": 8.147145854309201e-05, "learning_rate": 4.7601208459214504e-05, "loss": 0.0001, "step": 1030 }, { "epoch": 1.05, "grad_norm": 0.0004538837820291519, "learning_rate": 4.748036253776435e-05, "loss": 0.0, "step": 1040 }, { "epoch": 1.06, "grad_norm": 3.3908912882907316e-05, "learning_rate": 4.73595166163142e-05, "loss": 0.0, "step": 1050 }, { "epoch": 1.07, "grad_norm": 15.499226570129395, "learning_rate": 4.723867069486405e-05, "loss": 0.0171, "step": 1060 }, { "epoch": 1.08, "grad_norm": 6.438665390014648, "learning_rate": 4.71178247734139e-05, "loss": 0.0035, "step": 1070 }, { "epoch": 1.09, "grad_norm": 0.00015530579548794776, "learning_rate": 4.699697885196375e-05, "loss": 0.0001, "step": 1080 }, { "epoch": 1.1, "grad_norm": 5.526881068362854e-07, "learning_rate": 4.68761329305136e-05, "loss": 0.0, "step": 1090 }, { "epoch": 1.11, "grad_norm": 3.55797128577251e-05, "learning_rate": 4.675528700906344e-05, "loss": 0.007, "step": 1100 }, { "epoch": 1.12, "grad_norm": 1.0427943379909266e-05, "learning_rate": 4.6634441087613294e-05, "loss": 0.0, "step": 1110 }, { "epoch": 1.13, "grad_norm": 2.200431765686517e-07, "learning_rate": 4.651359516616314e-05, "loss": 0.0009, "step": 1120 }, { "epoch": 1.14, "grad_norm": 2.6252129828208126e-05, "learning_rate": 4.639274924471299e-05, "loss": 0.0001, "step": 1130 }, { "epoch": 1.15, "grad_norm": 3.5910196061195165e-07, "learning_rate": 4.627190332326284e-05, "loss": 0.0, "step": 1140 }, { "epoch": 1.16, "grad_norm": 7.039247407192306e-09, "learning_rate": 4.615105740181269e-05, "loss": 0.0, "step": 1150 }, { "epoch": 1.17, "grad_norm": 6.980984380788868e-06, "learning_rate": 4.603021148036254e-05, "loss": 0.0014, "step": 1160 }, { "epoch": 1.18, "grad_norm": 7.229643728123847e-08, "learning_rate": 4.590936555891239e-05, "loss": 0.0653, "step": 1170 }, { "epoch": 1.19, "grad_norm": 0.03556942939758301, "learning_rate": 4.578851963746223e-05, "loss": 0.0002, "step": 1180 }, { "epoch": 1.2, "grad_norm": 0.022921539843082428, "learning_rate": 4.5667673716012084e-05, "loss": 0.0019, "step": 1190 }, { "epoch": 1.21, "grad_norm": 1.2204433005535975e-05, "learning_rate": 4.554682779456193e-05, "loss": 0.0, "step": 1200 }, { "epoch": 1.22, "grad_norm": 0.003104632254689932, "learning_rate": 4.542598187311179e-05, "loss": 0.0004, "step": 1210 }, { "epoch": 1.23, "grad_norm": 0.007898018695414066, "learning_rate": 4.530513595166163e-05, "loss": 0.0012, "step": 1220 }, { "epoch": 1.24, "grad_norm": 0.0009976280853152275, "learning_rate": 4.518429003021148e-05, "loss": 0.0002, "step": 1230 }, { "epoch": 1.25, "grad_norm": 0.03468003869056702, "learning_rate": 4.5063444108761334e-05, "loss": 0.0, "step": 1240 }, { "epoch": 1.26, "grad_norm": 1.0441084441481507e-06, "learning_rate": 4.494259818731118e-05, "loss": 0.0001, "step": 1250 }, { "epoch": 1.27, "grad_norm": 3.78264850553478e-08, "learning_rate": 4.482175226586103e-05, "loss": 0.0003, "step": 1260 }, { "epoch": 1.28, "grad_norm": 6.2697478320217215e-09, "learning_rate": 4.4700906344410875e-05, "loss": 0.0, "step": 1270 }, { "epoch": 1.29, "grad_norm": 0.0001720537693472579, "learning_rate": 4.4580060422960726e-05, "loss": 0.0, "step": 1280 }, { "epoch": 1.3, "grad_norm": 6.12790336163016e-06, "learning_rate": 4.445921450151058e-05, "loss": 0.0, "step": 1290 }, { "epoch": 1.31, "grad_norm": 2.106332302093506, "learning_rate": 4.433836858006043e-05, "loss": 0.0033, "step": 1300 }, { "epoch": 1.32, "grad_norm": 0.000930789508856833, "learning_rate": 4.421752265861027e-05, "loss": 0.0, "step": 1310 }, { "epoch": 1.33, "grad_norm": 1.7019704046106199e-06, "learning_rate": 4.4096676737160125e-05, "loss": 0.0001, "step": 1320 }, { "epoch": 1.34, "grad_norm": 0.01025537308305502, "learning_rate": 4.397583081570997e-05, "loss": 0.0449, "step": 1330 }, { "epoch": 1.35, "grad_norm": 0.0023111889604479074, "learning_rate": 4.385498489425982e-05, "loss": 0.0, "step": 1340 }, { "epoch": 1.36, "grad_norm": 0.0005836488562636077, "learning_rate": 4.3734138972809665e-05, "loss": 0.0015, "step": 1350 }, { "epoch": 1.37, "grad_norm": 0.001271137036383152, "learning_rate": 4.3613293051359516e-05, "loss": 0.0, "step": 1360 }, { "epoch": 1.38, "grad_norm": 0.01227193046361208, "learning_rate": 4.349244712990937e-05, "loss": 0.0, "step": 1370 }, { "epoch": 1.39, "grad_norm": 3.68997419286643e-08, "learning_rate": 4.337160120845922e-05, "loss": 0.0009, "step": 1380 }, { "epoch": 1.4, "grad_norm": 1.0907036340768173e-07, "learning_rate": 4.3250755287009064e-05, "loss": 0.0, "step": 1390 }, { "epoch": 1.41, "grad_norm": 6.332713383017108e-05, "learning_rate": 4.3129909365558915e-05, "loss": 0.0, "step": 1400 }, { "epoch": 1.42, "grad_norm": 2.5950125603912966e-08, "learning_rate": 4.300906344410876e-05, "loss": 0.0, "step": 1410 }, { "epoch": 1.43, "grad_norm": 0.6216311454772949, "learning_rate": 4.288821752265861e-05, "loss": 0.0001, "step": 1420 }, { "epoch": 1.44, "grad_norm": 3.573224205410952e-08, "learning_rate": 4.2767371601208455e-05, "loss": 0.0, "step": 1430 }, { "epoch": 1.45, "grad_norm": 2.763231634617114e-07, "learning_rate": 4.2646525679758313e-05, "loss": 0.0, "step": 1440 }, { "epoch": 1.46, "grad_norm": 2.6494699767454222e-08, "learning_rate": 4.252567975830816e-05, "loss": 0.015, "step": 1450 }, { "epoch": 1.47, "grad_norm": 1.0802775340934545e-09, "learning_rate": 4.240483383685801e-05, "loss": 0.0, "step": 1460 }, { "epoch": 1.48, "grad_norm": 2.746639538031559e-08, "learning_rate": 4.2283987915407854e-05, "loss": 0.0, "step": 1470 }, { "epoch": 1.49, "grad_norm": 5.9745045291492715e-05, "learning_rate": 4.2163141993957705e-05, "loss": 0.0, "step": 1480 }, { "epoch": 1.5, "grad_norm": 3.736948031018983e-07, "learning_rate": 4.204229607250755e-05, "loss": 0.0, "step": 1490 }, { "epoch": 1.51, "grad_norm": 5.944487213582761e-08, "learning_rate": 4.19214501510574e-05, "loss": 0.0, "step": 1500 }, { "epoch": 1.52, "grad_norm": 9.614995910567359e-09, "learning_rate": 4.1800604229607246e-05, "loss": 0.0, "step": 1510 }, { "epoch": 1.53, "grad_norm": 9.067125716910596e-08, "learning_rate": 4.1679758308157104e-05, "loss": 0.0, "step": 1520 }, { "epoch": 1.54, "grad_norm": 3.790767255651417e-09, "learning_rate": 4.155891238670695e-05, "loss": 0.0018, "step": 1530 }, { "epoch": 1.55, "grad_norm": 0.0014521937118843198, "learning_rate": 4.14380664652568e-05, "loss": 0.0141, "step": 1540 }, { "epoch": 1.56, "grad_norm": 1.1887091204698663e-06, "learning_rate": 4.131722054380665e-05, "loss": 0.0001, "step": 1550 }, { "epoch": 1.57, "grad_norm": 0.21191056072711945, "learning_rate": 4.1196374622356495e-05, "loss": 0.0001, "step": 1560 }, { "epoch": 1.58, "grad_norm": 1.7035688415489858e-09, "learning_rate": 4.107552870090635e-05, "loss": 0.103, "step": 1570 }, { "epoch": 1.59, "grad_norm": 4.13519810535945e-05, "learning_rate": 4.095468277945619e-05, "loss": 0.0002, "step": 1580 }, { "epoch": 1.6, "grad_norm": 1.2702444109891076e-05, "learning_rate": 4.083383685800605e-05, "loss": 0.0, "step": 1590 }, { "epoch": 1.61, "grad_norm": 1.6921067924613453e-07, "learning_rate": 4.0712990936555894e-05, "loss": 0.0001, "step": 1600 }, { "epoch": 1.62, "grad_norm": 2.8627702874928218e-08, "learning_rate": 4.0592145015105745e-05, "loss": 0.0, "step": 1610 }, { "epoch": 1.63, "grad_norm": 6.734685484843794e-06, "learning_rate": 4.047129909365559e-05, "loss": 0.0, "step": 1620 }, { "epoch": 1.64, "grad_norm": 1.370299145264653e-07, "learning_rate": 4.035045317220544e-05, "loss": 0.0001, "step": 1630 }, { "epoch": 1.65, "grad_norm": 0.00048380409134551883, "learning_rate": 4.0229607250755286e-05, "loss": 0.0059, "step": 1640 }, { "epoch": 1.66, "grad_norm": 0.00042098091216757894, "learning_rate": 4.010876132930514e-05, "loss": 0.0001, "step": 1650 }, { "epoch": 1.67, "grad_norm": 1.4294737411546521e-05, "learning_rate": 3.998791540785498e-05, "loss": 0.0004, "step": 1660 }, { "epoch": 1.68, "grad_norm": 1.3993314285087877e-09, "learning_rate": 3.986706948640484e-05, "loss": 0.0062, "step": 1670 }, { "epoch": 1.69, "grad_norm": 0.013060159981250763, "learning_rate": 3.9746223564954684e-05, "loss": 0.0, "step": 1680 }, { "epoch": 1.7, "grad_norm": 6.259356450755149e-05, "learning_rate": 3.9625377643504536e-05, "loss": 0.0, "step": 1690 }, { "epoch": 1.71, "grad_norm": 9.836193203227594e-05, "learning_rate": 3.950453172205438e-05, "loss": 0.0, "step": 1700 }, { "epoch": 1.72, "grad_norm": 1.420707462784776e-08, "learning_rate": 3.938368580060423e-05, "loss": 0.0, "step": 1710 }, { "epoch": 1.73, "grad_norm": 1.8018381453543952e-08, "learning_rate": 3.9262839879154076e-05, "loss": 0.0, "step": 1720 }, { "epoch": 1.74, "grad_norm": 5.407735148565962e-09, "learning_rate": 3.914199395770393e-05, "loss": 0.0, "step": 1730 }, { "epoch": 1.75, "grad_norm": 8.004823826013308e-08, "learning_rate": 3.902114803625378e-05, "loss": 0.0, "step": 1740 }, { "epoch": 1.76, "grad_norm": 1.1862041304766535e-07, "learning_rate": 3.890030211480363e-05, "loss": 0.0, "step": 1750 }, { "epoch": 1.77, "grad_norm": 3.1350757012660324e-07, "learning_rate": 3.8779456193353475e-05, "loss": 0.0, "step": 1760 }, { "epoch": 1.78, "grad_norm": 2.7172923466878274e-08, "learning_rate": 3.8658610271903326e-05, "loss": 0.0, "step": 1770 }, { "epoch": 1.79, "grad_norm": 2.770736706736443e-08, "learning_rate": 3.853776435045317e-05, "loss": 0.0, "step": 1780 }, { "epoch": 1.8, "grad_norm": 1.3860269598708896e-09, "learning_rate": 3.841691842900302e-05, "loss": 0.0, "step": 1790 }, { "epoch": 1.81, "grad_norm": 0.01048044953495264, "learning_rate": 3.8296072507552866e-05, "loss": 0.2012, "step": 1800 }, { "epoch": 1.82, "grad_norm": 1.413319807852531e-07, "learning_rate": 3.817522658610272e-05, "loss": 0.0, "step": 1810 }, { "epoch": 1.83, "grad_norm": 0.0012239572824910283, "learning_rate": 3.805438066465257e-05, "loss": 0.0, "step": 1820 }, { "epoch": 1.84, "grad_norm": 0.00016678984684403986, "learning_rate": 3.793353474320242e-05, "loss": 0.0, "step": 1830 }, { "epoch": 1.85, "grad_norm": 4.03699864648388e-08, "learning_rate": 3.7812688821752265e-05, "loss": 0.0014, "step": 1840 }, { "epoch": 1.86, "grad_norm": 2.3393537048832513e-05, "learning_rate": 3.7691842900302116e-05, "loss": 0.0, "step": 1850 }, { "epoch": 1.87, "grad_norm": 0.0013141903327777982, "learning_rate": 3.757099697885196e-05, "loss": 0.0001, "step": 1860 }, { "epoch": 1.88, "grad_norm": 0.001294136163778603, "learning_rate": 3.745015105740181e-05, "loss": 0.0, "step": 1870 }, { "epoch": 1.89, "grad_norm": 3.9412746843936475e-08, "learning_rate": 3.7329305135951663e-05, "loss": 0.0002, "step": 1880 }, { "epoch": 1.9, "grad_norm": 9.437405878998106e-07, "learning_rate": 3.720845921450151e-05, "loss": 0.0008, "step": 1890 }, { "epoch": 1.91, "grad_norm": 1.1508973329910077e-06, "learning_rate": 3.7087613293051366e-05, "loss": 0.0002, "step": 1900 }, { "epoch": 1.92, "grad_norm": 5.708042749574815e-07, "learning_rate": 3.696676737160121e-05, "loss": 0.0, "step": 1910 }, { "epoch": 1.93, "grad_norm": 4.998336407879833e-06, "learning_rate": 3.684592145015106e-05, "loss": 0.0, "step": 1920 }, { "epoch": 1.94, "grad_norm": 0.0002045446599368006, "learning_rate": 3.6725075528700906e-05, "loss": 0.0004, "step": 1930 }, { "epoch": 1.95, "grad_norm": 1.011339804790623e-06, "learning_rate": 3.660422960725076e-05, "loss": 0.0002, "step": 1940 }, { "epoch": 1.96, "grad_norm": 1.1179584191722824e-07, "learning_rate": 3.64833836858006e-05, "loss": 0.0, "step": 1950 }, { "epoch": 1.97, "grad_norm": 4.01311623932088e-08, "learning_rate": 3.6362537764350454e-05, "loss": 0.0001, "step": 1960 }, { "epoch": 1.98, "grad_norm": 0.00014017950161360204, "learning_rate": 3.6241691842900305e-05, "loss": 0.0, "step": 1970 }, { "epoch": 1.99, "grad_norm": 3.3373577025486156e-05, "learning_rate": 3.6120845921450156e-05, "loss": 0.0, "step": 1980 }, { "epoch": 2.0, "grad_norm": 6.252674211282283e-05, "learning_rate": 3.6e-05, "loss": 0.0, "step": 1990 }, { "epoch": 2.01, "grad_norm": 0.0002250933466712013, "learning_rate": 3.587915407854985e-05, "loss": 0.0, "step": 2000 }, { "epoch": 2.02, "grad_norm": 6.714632849735835e-09, "learning_rate": 3.57583081570997e-05, "loss": 0.0, "step": 2010 }, { "epoch": 2.03, "grad_norm": 2.656873050455033e-07, "learning_rate": 3.563746223564955e-05, "loss": 0.0, "step": 2020 }, { "epoch": 2.04, "grad_norm": 1.6010590115911327e-05, "learning_rate": 3.551661631419939e-05, "loss": 0.0, "step": 2030 }, { "epoch": 2.05, "grad_norm": 2.7113070245832205e-05, "learning_rate": 3.5395770392749244e-05, "loss": 0.0, "step": 2040 }, { "epoch": 2.06, "grad_norm": 1.7675635399783118e-09, "learning_rate": 3.5274924471299095e-05, "loss": 0.0, "step": 2050 }, { "epoch": 2.07, "grad_norm": 5.97303042582098e-08, "learning_rate": 3.515407854984895e-05, "loss": 0.0001, "step": 2060 }, { "epoch": 2.08, "grad_norm": 0.0004335389821790159, "learning_rate": 3.503323262839879e-05, "loss": 0.0, "step": 2070 }, { "epoch": 2.09, "grad_norm": 6.485903213615529e-06, "learning_rate": 3.491238670694864e-05, "loss": 0.0, "step": 2080 }, { "epoch": 2.1, "grad_norm": 0.0035731520038098097, "learning_rate": 3.479154078549849e-05, "loss": 0.0, "step": 2090 }, { "epoch": 2.11, "grad_norm": 4.11334362127036e-08, "learning_rate": 3.467069486404834e-05, "loss": 0.0, "step": 2100 }, { "epoch": 2.12, "grad_norm": 5.335771129466593e-05, "learning_rate": 3.454984894259818e-05, "loss": 0.0, "step": 2110 }, { "epoch": 2.13, "grad_norm": 1.26767045003362e-05, "learning_rate": 3.442900302114804e-05, "loss": 0.0, "step": 2120 }, { "epoch": 2.14, "grad_norm": 2.478699911989679e-09, "learning_rate": 3.4308157099697886e-05, "loss": 0.0, "step": 2130 }, { "epoch": 2.15, "grad_norm": 2.9354995945141127e-07, "learning_rate": 3.418731117824774e-05, "loss": 0.0, "step": 2140 }, { "epoch": 2.16, "grad_norm": 5.661284831148805e-06, "learning_rate": 3.406646525679758e-05, "loss": 0.0001, "step": 2150 }, { "epoch": 2.17, "grad_norm": 0.0009775606449693441, "learning_rate": 3.394561933534743e-05, "loss": 0.0, "step": 2160 }, { "epoch": 2.18, "grad_norm": 1.0540851524964978e-09, "learning_rate": 3.382477341389728e-05, "loss": 0.0, "step": 2170 }, { "epoch": 2.19, "grad_norm": 4.330029241828015e-06, "learning_rate": 3.370392749244713e-05, "loss": 0.0, "step": 2180 }, { "epoch": 2.2, "grad_norm": 1.206312163049006e-06, "learning_rate": 3.358308157099698e-05, "loss": 0.0, "step": 2190 }, { "epoch": 2.21, "grad_norm": 2.2324436486087507e-06, "learning_rate": 3.346223564954683e-05, "loss": 0.0, "step": 2200 }, { "epoch": 2.22, "grad_norm": 6.707778510417484e-08, "learning_rate": 3.334138972809668e-05, "loss": 0.0, "step": 2210 }, { "epoch": 2.23, "grad_norm": 1.0213330181585434e-09, "learning_rate": 3.322054380664653e-05, "loss": 0.0, "step": 2220 }, { "epoch": 2.24, "grad_norm": 1.522714967450156e-07, "learning_rate": 3.309969788519638e-05, "loss": 0.0, "step": 2230 }, { "epoch": 2.25, "grad_norm": 3.1322219911089633e-06, "learning_rate": 3.297885196374622e-05, "loss": 0.0, "step": 2240 }, { "epoch": 2.26, "grad_norm": 0.0, "learning_rate": 3.2858006042296074e-05, "loss": 0.0001, "step": 2250 }, { "epoch": 2.27, "grad_norm": 0.0, "learning_rate": 3.273716012084592e-05, "loss": 0.0, "step": 2260 }, { "epoch": 2.28, "grad_norm": 3.107200274854449e-09, "learning_rate": 3.261631419939577e-05, "loss": 0.0, "step": 2270 }, { "epoch": 2.29, "grad_norm": 9.697516123452488e-10, "learning_rate": 3.249546827794562e-05, "loss": 0.0, "step": 2280 }, { "epoch": 2.3, "grad_norm": 1.6279294579035763e-09, "learning_rate": 3.237462235649547e-05, "loss": 0.0, "step": 2290 }, { "epoch": 2.31, "grad_norm": 6.846393674209139e-09, "learning_rate": 3.225377643504532e-05, "loss": 0.0, "step": 2300 }, { "epoch": 2.32, "grad_norm": 1.0791376553243026e-05, "learning_rate": 3.213293051359517e-05, "loss": 0.0, "step": 2310 }, { "epoch": 2.33, "grad_norm": 1.0202172688877909e-06, "learning_rate": 3.2012084592145013e-05, "loss": 0.0, "step": 2320 }, { "epoch": 2.34, "grad_norm": 3.7638578365140063e-10, "learning_rate": 3.1891238670694865e-05, "loss": 0.0, "step": 2330 }, { "epoch": 2.35, "grad_norm": 2.3091282219667164e-09, "learning_rate": 3.177039274924471e-05, "loss": 0.0, "step": 2340 }, { "epoch": 2.36, "grad_norm": 0.0, "learning_rate": 3.164954682779457e-05, "loss": 0.0, "step": 2350 }, { "epoch": 2.37, "grad_norm": 0.0, "learning_rate": 3.152870090634441e-05, "loss": 0.0, "step": 2360 }, { "epoch": 2.38, "grad_norm": 0.0, "learning_rate": 3.140785498489426e-05, "loss": 0.0, "step": 2370 }, { "epoch": 2.39, "grad_norm": 0.0, "learning_rate": 3.128700906344411e-05, "loss": 0.0, "step": 2380 }, { "epoch": 2.41, "grad_norm": 0.0, "learning_rate": 3.116616314199396e-05, "loss": 0.0, "step": 2390 }, { "epoch": 2.42, "grad_norm": 2.4167900846805423e-05, "learning_rate": 3.1045317220543804e-05, "loss": 0.0, "step": 2400 }, { "epoch": 2.43, "grad_norm": 2.4693702016520547e-07, "learning_rate": 3.0924471299093655e-05, "loss": 0.0, "step": 2410 }, { "epoch": 2.44, "grad_norm": 1.0320581056433298e-09, "learning_rate": 3.08036253776435e-05, "loss": 0.0, "step": 2420 }, { "epoch": 2.45, "grad_norm": 8.062679768272574e-08, "learning_rate": 3.068277945619336e-05, "loss": 0.0, "step": 2430 }, { "epoch": 2.46, "grad_norm": 2.012851219035383e-09, "learning_rate": 3.05619335347432e-05, "loss": 0.0, "step": 2440 }, { "epoch": 2.47, "grad_norm": 0.0, "learning_rate": 3.0441087613293054e-05, "loss": 0.0, "step": 2450 }, { "epoch": 2.48, "grad_norm": 3.1564242330972547e-09, "learning_rate": 3.0320241691842898e-05, "loss": 0.0, "step": 2460 }, { "epoch": 2.49, "grad_norm": 0.0, "learning_rate": 3.019939577039275e-05, "loss": 0.0, "step": 2470 }, { "epoch": 2.5, "grad_norm": 3.624436430982314e-05, "learning_rate": 3.0078549848942597e-05, "loss": 0.0, "step": 2480 }, { "epoch": 2.51, "grad_norm": 1.0910243730677394e-07, "learning_rate": 2.995770392749245e-05, "loss": 0.0, "step": 2490 }, { "epoch": 2.52, "grad_norm": 0.0, "learning_rate": 2.9836858006042297e-05, "loss": 0.0, "step": 2500 }, { "epoch": 2.53, "grad_norm": 2.473066729180573e-08, "learning_rate": 2.9716012084592145e-05, "loss": 0.0, "step": 2510 }, { "epoch": 2.54, "grad_norm": 3.971327089402621e-08, "learning_rate": 2.9595166163141996e-05, "loss": 0.0, "step": 2520 }, { "epoch": 2.55, "grad_norm": 0.0, "learning_rate": 2.9474320241691844e-05, "loss": 0.0, "step": 2530 }, { "epoch": 2.56, "grad_norm": 2.1878437109990045e-05, "learning_rate": 2.9353474320241692e-05, "loss": 0.0, "step": 2540 }, { "epoch": 2.57, "grad_norm": 0.00010030780686065555, "learning_rate": 2.923262839879154e-05, "loss": 0.0, "step": 2550 }, { "epoch": 2.58, "grad_norm": 4.3228740587153425e-09, "learning_rate": 2.911178247734139e-05, "loss": 0.0, "step": 2560 }, { "epoch": 2.59, "grad_norm": 0.0, "learning_rate": 2.899093655589124e-05, "loss": 0.0, "step": 2570 }, { "epoch": 2.6, "grad_norm": 3.8086006570736686e-10, "learning_rate": 2.8870090634441087e-05, "loss": 0.0, "step": 2580 }, { "epoch": 2.61, "grad_norm": 6.88414825145145e-10, "learning_rate": 2.8749244712990935e-05, "loss": 0.0, "step": 2590 }, { "epoch": 2.62, "grad_norm": 2.8818004693675903e-07, "learning_rate": 2.8628398791540786e-05, "loss": 0.002, "step": 2600 }, { "epoch": 2.63, "grad_norm": 3.844630569460605e-08, "learning_rate": 2.8507552870090634e-05, "loss": 0.0, "step": 2610 }, { "epoch": 2.64, "grad_norm": 0.026187095791101456, "learning_rate": 2.8386706948640482e-05, "loss": 0.0, "step": 2620 }, { "epoch": 2.65, "grad_norm": 3.006335624888834e-09, "learning_rate": 2.826586102719033e-05, "loss": 0.0, "step": 2630 }, { "epoch": 2.66, "grad_norm": 1.2555399564462277e-07, "learning_rate": 2.814501510574018e-05, "loss": 0.0, "step": 2640 }, { "epoch": 2.67, "grad_norm": 0.03935589641332626, "learning_rate": 2.802416918429003e-05, "loss": 0.0, "step": 2650 }, { "epoch": 2.68, "grad_norm": 4.552068730845349e-07, "learning_rate": 2.7903323262839877e-05, "loss": 0.0, "step": 2660 }, { "epoch": 2.69, "grad_norm": 60.47336959838867, "learning_rate": 2.7782477341389732e-05, "loss": 0.0803, "step": 2670 }, { "epoch": 2.7, "grad_norm": 1.4243897616950107e-08, "learning_rate": 2.766163141993958e-05, "loss": 0.0017, "step": 2680 }, { "epoch": 2.71, "grad_norm": 8.210067790059838e-06, "learning_rate": 2.7540785498489428e-05, "loss": 0.0, "step": 2690 }, { "epoch": 2.72, "grad_norm": 1.0214007417630455e-08, "learning_rate": 2.7419939577039276e-05, "loss": 0.0, "step": 2700 }, { "epoch": 2.73, "grad_norm": 1.5932839048105052e-08, "learning_rate": 2.7299093655589127e-05, "loss": 0.0, "step": 2710 }, { "epoch": 2.74, "grad_norm": 1.8762943976291524e-10, "learning_rate": 2.7178247734138975e-05, "loss": 0.0, "step": 2720 }, { "epoch": 2.75, "grad_norm": 4.0266687051371264e-07, "learning_rate": 2.7057401812688823e-05, "loss": 0.0, "step": 2730 }, { "epoch": 2.76, "grad_norm": 4.4115214592466145e-08, "learning_rate": 2.693655589123867e-05, "loss": 0.0, "step": 2740 }, { "epoch": 2.77, "grad_norm": 0.0, "learning_rate": 2.6815709969788522e-05, "loss": 0.0, "step": 2750 }, { "epoch": 2.78, "grad_norm": 1.6773222810684274e-08, "learning_rate": 2.669486404833837e-05, "loss": 0.0001, "step": 2760 }, { "epoch": 2.79, "grad_norm": 0.0, "learning_rate": 2.6574018126888218e-05, "loss": 0.0, "step": 2770 }, { "epoch": 2.8, "grad_norm": 0.002830307465046644, "learning_rate": 2.6453172205438066e-05, "loss": 0.0538, "step": 2780 }, { "epoch": 2.81, "grad_norm": 1.8111581123747555e-09, "learning_rate": 2.6332326283987917e-05, "loss": 0.0, "step": 2790 }, { "epoch": 2.82, "grad_norm": 2.044587290583877e-07, "learning_rate": 2.6211480362537765e-05, "loss": 0.0, "step": 2800 }, { "epoch": 2.83, "grad_norm": 0.0, "learning_rate": 2.6090634441087613e-05, "loss": 0.0, "step": 2810 }, { "epoch": 2.84, "grad_norm": 1.001442484493964e-09, "learning_rate": 2.596978851963746e-05, "loss": 0.0, "step": 2820 }, { "epoch": 2.85, "grad_norm": 0.0, "learning_rate": 2.5848942598187313e-05, "loss": 0.0, "step": 2830 }, { "epoch": 2.86, "grad_norm": 1.0242295900297904e-08, "learning_rate": 2.572809667673716e-05, "loss": 0.0, "step": 2840 }, { "epoch": 2.87, "grad_norm": 3.850387564341418e-09, "learning_rate": 2.560725075528701e-05, "loss": 0.0, "step": 2850 }, { "epoch": 2.88, "grad_norm": 1.869227411743779e-10, "learning_rate": 2.548640483383686e-05, "loss": 0.0, "step": 2860 }, { "epoch": 2.89, "grad_norm": 1.6829256210826316e-08, "learning_rate": 2.5365558912386708e-05, "loss": 0.0, "step": 2870 }, { "epoch": 2.9, "grad_norm": 3.8250621514634986e-07, "learning_rate": 2.5244712990936556e-05, "loss": 0.0, "step": 2880 }, { "epoch": 2.91, "grad_norm": 2.593100623116129e-09, "learning_rate": 2.5123867069486404e-05, "loss": 0.0, "step": 2890 }, { "epoch": 2.92, "grad_norm": 3.2663322713233356e-08, "learning_rate": 2.5003021148036255e-05, "loss": 0.0, "step": 2900 }, { "epoch": 2.93, "grad_norm": 3.1544191703147817e-09, "learning_rate": 2.4882175226586103e-05, "loss": 0.0, "step": 2910 }, { "epoch": 2.94, "grad_norm": 4.537993092412762e-09, "learning_rate": 2.476132930513595e-05, "loss": 0.0, "step": 2920 }, { "epoch": 2.95, "grad_norm": 2.1482501324499026e-05, "learning_rate": 2.46404833836858e-05, "loss": 0.0, "step": 2930 }, { "epoch": 2.96, "grad_norm": 3.4461805853425176e-08, "learning_rate": 2.451963746223565e-05, "loss": 0.0, "step": 2940 }, { "epoch": 2.97, "grad_norm": 1.910395930337927e-08, "learning_rate": 2.4398791540785498e-05, "loss": 0.0, "step": 2950 }, { "epoch": 2.98, "grad_norm": 1.0935653271815227e-08, "learning_rate": 2.4277945619335346e-05, "loss": 0.0, "step": 2960 }, { "epoch": 2.99, "grad_norm": 1.6578489692165022e-09, "learning_rate": 2.4157099697885194e-05, "loss": 0.0, "step": 2970 }, { "epoch": 3.0, "grad_norm": 4.757405349664623e-07, "learning_rate": 2.4036253776435045e-05, "loss": 0.0, "step": 2980 }, { "epoch": 3.01, "grad_norm": 0.0, "learning_rate": 2.3915407854984897e-05, "loss": 0.0, "step": 2990 }, { "epoch": 3.02, "grad_norm": 1.9106341397900906e-08, "learning_rate": 2.3794561933534744e-05, "loss": 0.0, "step": 3000 }, { "epoch": 3.03, "grad_norm": 2.297590118160997e-09, "learning_rate": 2.3673716012084596e-05, "loss": 0.0, "step": 3010 }, { "epoch": 3.04, "grad_norm": 0.0, "learning_rate": 2.3552870090634444e-05, "loss": 0.0, "step": 3020 }, { "epoch": 3.05, "grad_norm": 5.0260878197150305e-05, "learning_rate": 2.343202416918429e-05, "loss": 0.0, "step": 3030 }, { "epoch": 3.06, "grad_norm": 1.84212534115602e-10, "learning_rate": 2.331117824773414e-05, "loss": 0.0, "step": 3040 }, { "epoch": 3.07, "grad_norm": 1.035463603749065e-09, "learning_rate": 2.319033232628399e-05, "loss": 0.0, "step": 3050 }, { "epoch": 3.08, "grad_norm": 8.718736843604802e-09, "learning_rate": 2.306948640483384e-05, "loss": 0.0, "step": 3060 }, { "epoch": 3.09, "grad_norm": 3.523397129612249e-10, "learning_rate": 2.2948640483383687e-05, "loss": 0.0, "step": 3070 }, { "epoch": 3.1, "grad_norm": 3.516629210054134e-10, "learning_rate": 2.2827794561933535e-05, "loss": 0.0, "step": 3080 }, { "epoch": 3.11, "grad_norm": 1.1817105072253753e-09, "learning_rate": 2.2706948640483386e-05, "loss": 0.0, "step": 3090 }, { "epoch": 3.12, "grad_norm": 6.393580065378046e-08, "learning_rate": 2.2586102719033234e-05, "loss": 0.0, "step": 3100 }, { "epoch": 3.13, "grad_norm": 0.0, "learning_rate": 2.2465256797583082e-05, "loss": 0.0, "step": 3110 }, { "epoch": 3.14, "grad_norm": 2.3444219010571032e-08, "learning_rate": 2.234441087613293e-05, "loss": 0.0, "step": 3120 }, { "epoch": 3.15, "grad_norm": 5.386530688156199e-08, "learning_rate": 2.222356495468278e-05, "loss": 0.0, "step": 3130 }, { "epoch": 3.16, "grad_norm": 7.222023867825555e-08, "learning_rate": 2.210271903323263e-05, "loss": 0.0, "step": 3140 }, { "epoch": 3.17, "grad_norm": 2.3319643993602313e-09, "learning_rate": 2.1981873111782477e-05, "loss": 0.0, "step": 3150 }, { "epoch": 3.18, "grad_norm": 1.8931477219208404e-10, "learning_rate": 2.1861027190332325e-05, "loss": 0.0, "step": 3160 }, { "epoch": 3.19, "grad_norm": 0.014633001759648323, "learning_rate": 2.1740181268882176e-05, "loss": 0.0, "step": 3170 }, { "epoch": 3.2, "grad_norm": 2.1926523913862184e-05, "learning_rate": 2.1619335347432024e-05, "loss": 0.0, "step": 3180 }, { "epoch": 3.21, "grad_norm": 2.996988213155305e-09, "learning_rate": 2.1498489425981872e-05, "loss": 0.0, "step": 3190 }, { "epoch": 3.22, "grad_norm": 7.307365649467101e-06, "learning_rate": 2.1377643504531724e-05, "loss": 0.0, "step": 3200 }, { "epoch": 3.23, "grad_norm": 1.264578168047592e-05, "learning_rate": 2.125679758308157e-05, "loss": 0.0, "step": 3210 }, { "epoch": 3.24, "grad_norm": 1.8541245481173974e-06, "learning_rate": 2.113595166163142e-05, "loss": 0.0, "step": 3220 }, { "epoch": 3.25, "grad_norm": 7.540985080822793e-08, "learning_rate": 2.1015105740181267e-05, "loss": 0.0, "step": 3230 }, { "epoch": 3.26, "grad_norm": 5.4537132143650524e-08, "learning_rate": 2.089425981873112e-05, "loss": 0.0, "step": 3240 }, { "epoch": 3.27, "grad_norm": 7.338199338846607e-06, "learning_rate": 2.0773413897280967e-05, "loss": 0.0, "step": 3250 }, { "epoch": 3.28, "grad_norm": 0.0, "learning_rate": 2.0652567975830815e-05, "loss": 0.0, "step": 3260 }, { "epoch": 3.29, "grad_norm": 0.0, "learning_rate": 2.0531722054380663e-05, "loss": 0.0, "step": 3270 }, { "epoch": 3.3, "grad_norm": 2.116165687127136e-09, "learning_rate": 2.0410876132930514e-05, "loss": 0.0, "step": 3280 }, { "epoch": 3.31, "grad_norm": 2.066717907922566e-09, "learning_rate": 2.0290030211480362e-05, "loss": 0.0, "step": 3290 }, { "epoch": 3.32, "grad_norm": 1.8276996582855531e-10, "learning_rate": 2.016918429003021e-05, "loss": 0.0, "step": 3300 }, { "epoch": 3.33, "grad_norm": 3.0077771384640073e-09, "learning_rate": 2.004833836858006e-05, "loss": 0.0, "step": 3310 }, { "epoch": 3.34, "grad_norm": 0.0, "learning_rate": 1.9927492447129912e-05, "loss": 0.0, "step": 3320 }, { "epoch": 3.35, "grad_norm": 0.0, "learning_rate": 1.980664652567976e-05, "loss": 0.0003, "step": 3330 }, { "epoch": 3.36, "grad_norm": 0.007729419972747564, "learning_rate": 1.968580060422961e-05, "loss": 0.0, "step": 3340 }, { "epoch": 3.37, "grad_norm": 8.718690514797345e-05, "learning_rate": 1.9564954682779456e-05, "loss": 0.0, "step": 3350 }, { "epoch": 3.38, "grad_norm": 4.8525894271733705e-06, "learning_rate": 1.9444108761329308e-05, "loss": 0.0, "step": 3360 }, { "epoch": 3.39, "grad_norm": 2.319741509992923e-09, "learning_rate": 1.9323262839879156e-05, "loss": 0.0, "step": 3370 }, { "epoch": 3.4, "grad_norm": 9.715724758052602e-08, "learning_rate": 1.9202416918429003e-05, "loss": 0.0, "step": 3380 }, { "epoch": 3.41, "grad_norm": 5.839451091560477e-07, "learning_rate": 1.9081570996978855e-05, "loss": 0.0, "step": 3390 }, { "epoch": 3.42, "grad_norm": 1.4994407138146926e-05, "learning_rate": 1.8960725075528703e-05, "loss": 0.0, "step": 3400 }, { "epoch": 3.43, "grad_norm": 3.994861685896467e-08, "learning_rate": 1.883987915407855e-05, "loss": 0.0, "step": 3410 }, { "epoch": 3.44, "grad_norm": 0.0011337869800627232, "learning_rate": 1.87190332326284e-05, "loss": 0.0, "step": 3420 }, { "epoch": 3.45, "grad_norm": 2.8960034370584253e-08, "learning_rate": 1.859818731117825e-05, "loss": 0.0, "step": 3430 }, { "epoch": 3.46, "grad_norm": 9.850285209722642e-08, "learning_rate": 1.8477341389728098e-05, "loss": 0.0, "step": 3440 }, { "epoch": 3.47, "grad_norm": 6.7992941694683395e-06, "learning_rate": 1.8356495468277946e-05, "loss": 0.0, "step": 3450 }, { "epoch": 3.48, "grad_norm": 0.0, "learning_rate": 1.8235649546827794e-05, "loss": 0.0, "step": 3460 }, { "epoch": 3.49, "grad_norm": 5.028818250707445e-09, "learning_rate": 1.8114803625377645e-05, "loss": 0.0, "step": 3470 }, { "epoch": 3.5, "grad_norm": 2.1626638613270188e-07, "learning_rate": 1.7993957703927493e-05, "loss": 0.0, "step": 3480 }, { "epoch": 3.51, "grad_norm": 1.7985012090715458e-10, "learning_rate": 1.787311178247734e-05, "loss": 0.0, "step": 3490 }, { "epoch": 3.52, "grad_norm": 3.205403231731907e-08, "learning_rate": 1.775226586102719e-05, "loss": 0.0, "step": 3500 } ], "logging_steps": 10, "max_steps": 4965, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }