1
|
{"args": {"mode": "bp", "dataset": "mnist", "depth": 3, "width": 256, "act": "tanh", "residual": 0, "epochs": 20, "batch_size": 128, "eta": 0.05, "eta_A": 0.02, "eta_P": 0.002, "momentum": 0.9, "w_scale": 1.0, "a_scale": 1.0, "pert_sigma": 0.01, "pert_every": 4, "pert_ndirs": 4, "use_residual": 1, "learn_A": 1, "learn_P": 1, "p_neutral": 1, "p_warmup_steps": 200, "p_warmup_eta": 0.05, "nuis_rho": 0.0, "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 50, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "depth_mnist_bp_d3_s1", "n_in": 784}, "steps": [{"step": 0, "epoch": 0, "train_loss": 2.2478644847869873}, {"step": 50, "epoch": 0, "train_loss": 0.45810359716415405}, {"step": 100, "epoch": 0, "train_loss": 0.2862793207168579}, {"step": 150, "epoch": 0, "train_loss": 0.29438289999961853}, {"step": 200, "epoch": 0, "train_loss": 0.2021787315607071}, {"step": 250, "epoch": 0, "train_loss": 0.12533476948738098}, {"step": 300, "epoch": 0, "train_loss": 0.2485463172197342}, {"step": 350, "epoch": 0, "train_loss": 0.14260680973529816}, {"step": 400, "epoch": 0, "train_loss": 0.10293297469615936}, {"step": 450, "epoch": 0, "train_loss": 0.17973658442497253}, {"epoch_end": 0, "step": 469, "test_acc": 0.966, "test_loss": 0.11403705291748047}, {"step": 500, "epoch": 1, "train_loss": 0.1282760053873062}, {"step": 550, "epoch": 1, "train_loss": 0.0776575580239296}, {"step": 600, "epoch": 1, "train_loss": 0.12832023203372955}, {"step": 650, "epoch": 1, "train_loss": 0.1112724244594574}, {"step": 700, "epoch": 1, "train_loss": 0.03798096999526024}, {"step": 750, "epoch": 1, "train_loss": 0.020821189507842064}, {"step": 800, "epoch": 1, "train_loss": 0.03227154538035393}, {"step": 850, "epoch": 1, "train_loss": 0.11704613268375397}, {"step": 900, "epoch": 1, "train_loss": 0.10958629846572876}, {"epoch_end": 1, "step": 938, "test_acc": 0.9727, "test_loss": 0.0852578296661377}, {"step": 950, "epoch": 2, "train_loss": 0.08283311128616333}, {"step": 1000, "epoch": 2, "train_loss": 0.09486904740333557}, {"step": 1050, "epoch": 2, "train_loss": 0.028336577117443085}, {"step": 1100, "epoch": 2, "train_loss": 0.06038254499435425}, {"step": 1150, "epoch": 2, "train_loss": 0.013586276210844517}, {"step": 1200, "epoch": 2, "train_loss": 0.0221277866512537}, {"step": 1250, "epoch": 2, "train_loss": 0.02608134038746357}, {"step": 1300, "epoch": 2, "train_loss": 0.09395450353622437}, {"step": 1350, "epoch": 2, "train_loss": 0.0815756693482399}, {"step": 1400, "epoch": 2, "train_loss": 0.07235240936279297}, {"epoch_end": 2, "step": 1407, "test_acc": 0.975, "test_loss": 0.08752586879730224}, {"step": 1450, "epoch": 3, "train_loss": 0.024597443640232086}, {"step": 1500, "epoch": 3, "train_loss": 0.051421500742435455}, {"step": 1550, "epoch": 3, "train_loss": 0.018854675814509392}, {"step": 1600, "epoch": 3, "train_loss": 0.026310356333851814}, {"step": 1650, "epoch": 3, "train_loss": 0.033972058445215225}, {"step": 1700, "epoch": 3, "train_loss": 0.06035701185464859}, {"step": 1750, "epoch": 3, "train_loss": 0.027719296514987946}, {"step": 1800, "epoch": 3, "train_loss": 0.038818418979644775}, {"step": 1850, "epoch": 3, "train_loss": 0.022925924509763718}, {"epoch_end": 3, "step": 1876, "test_acc": 0.9765, "test_loss": 0.07745481910705566}, {"step": 1900, "epoch": 4, "train_loss": 0.03425324335694313}, {"step": 1950, "epoch": 4, "train_loss": 0.018550099804997444}, {"step": 2000, "epoch": 4, "train_loss": 0.039594653993844986}, {"step": 2050, "epoch": 4, "train_loss": 0.02880600467324257}, {"step": 2100, "epoch": 4, "train_loss": 0.021884817630052567}, {"step": 2150, "epoch": 4, "train_loss": 0.02753462828695774}, {"step": 2200, "epoch": 4, "train_loss": 0.027267232537269592}, {"step": 2250, "epoch": 4, "train_loss": 0.008878055959939957}, {"step": 2300, "epoch": 4, "train_loss": 0.04464900121092796}, {"epoch_end": 4, "step": 2345, "test_acc": 0.9746, "test_loss": 0.0885932596206665}, {"step": 2350, "epoch": 5, "train_loss": 0.023661715909838676}, {"step": 2400, "epoch": 5, "train_loss": 0.006383955478668213}, {"step": 2450, "epoch": 5, "train_loss": 0.011705118231475353}, {"step": 2500, "epoch": 5, "train_loss": 0.008455080911517143}, {"step": 2550, "epoch": 5, "train_loss": 0.03434302657842636}, {"step": 2600, "epoch": 5, "train_loss": 0.003116304986178875}, {"step": 2650, "epoch": 5, "train_loss": 0.038539737462997437}, {"step": 2700, "epoch": 5, "train_loss": 0.03976483643054962}, {"step": 2750, "epoch": 5, "train_loss": 0.052965741604566574}, {"step": 2800, "epoch": 5, "train_loss": 0.029521575197577477}, {"epoch_end": 5, "step": 2814, "test_acc": 0.9781, "test_loss": 0.08427728643417358}, {"step": 2850, "epoch": 6, "train_loss": 0.01955767534673214}, {"step": 2900, "epoch": 6, "train_loss": 0.008198353461921215}, {"step": 2950, "epoch": 6, "train_loss": 0.02009803056716919}, {"step": 3000, "epoch": 6, "train_loss": 0.005300345364958048}, {"step": 3050, "epoch": 6, "train_loss": 0.004706192295998335}, {"step": 3100, "epoch": 6, "train_loss": 0.014344627968966961}, {"step": 3150, "epoch": 6, "train_loss": 0.009715971536934376}, {"step": 3200, "epoch": 6, "train_loss": 0.024129632860422134}, {"step": 3250, "epoch": 6, "train_loss": 0.019522774964571}, {"epoch_end": 6, "step": 3283, "test_acc": 0.9791, "test_loss": 0.07843589181900025}, {"step": 3300, "epoch": 7, "train_loss": 0.011665948666632175}, {"step": 3350, "epoch": 7, "train_loss": 0.0008490982581861317}, {"step": 3400, "epoch": 7, "train_loss": 0.003914083819836378}, {"step": 3450, "epoch": 7, "train_loss": 0.04015166684985161}, {"step": 3500, "epoch": 7, "train_loss": 0.01453531626611948}, {"step": 3550, "epoch": 7, "train_loss": 0.0044003818184137344}, {"step": 3600, "epoch": 7, "train_loss": 0.004256321117281914}, {"step": 3650, "epoch": 7, "train_loss": 0.000453958404250443}, {"step": 3700, "epoch": 7, "train_loss": 0.03230199217796326}, {"step": 3750, "epoch": 7, "train_loss": 0.004387172870337963}, {"epoch_end": 7, "step": 3752, "test_acc": 0.9779, "test_loss": 0.08378057346343994}, {"step": 3800, "epoch": 8, "train_loss": 0.0336502343416214}, {"step": 3850, "epoch": 8, "train_loss": 0.0064634718000888824}, {"step": 3900, "epoch": 8, "train_loss": 0.010905936360359192}, {"step": 3950, "epoch": 8, "train_loss": 0.0005654847482219338}, {"step": 4000, "epoch": 8, "train_loss": 0.0004998466465622187}, {"step": 4050, "epoch": 8, "train_loss": 0.0023333029821515083}, {"step": 4100, "epoch": 8, "train_loss": 0.017810722813010216}, {"step": 4150, "epoch": 8, "train_loss": 0.0009062434546649456}, {"step": 4200, "epoch": 8, "train_loss": 0.002110155997797847}, {"epoch_end": 8, "step": 4221, "test_acc": 0.9823, "test_loss": 0.07151507558822631}, {"step": 4250, "epoch": 9, "train_loss": 0.0018658633343875408}, {"step": 4300, "epoch": 9, "train_loss": 0.002339878585189581}, {"step": 4350, "epoch": 9, "train_loss": 0.00040672477916814387}, {"step": 4400, "epoch": 9, "train_loss": 0.009828319773077965}, {"step": 4450, "epoch": 9, "train_loss": 0.0011765159433707595}, {"step": 4500, "epoch": 9, "train_loss": 0.002677148673683405}, {"step": 4550, "epoch": 9, "train_loss": 0.015851417556405067}, {"step": 4600, "epoch": 9, "train_loss": 0.008782860822975636}, {"step": 4650, "epoch": 9, "train_loss": 0.0013578197686001658}, {"epoch_end": 9, "step": 4690, "test_acc": 0.9813, "test_loss": 0.07771074705123901}, {"step": 4700, "epoch": 10, "train_loss": 0.0010075827594846487}, {"step": 4750, "epoch": 10, "train_loss": 0.0009578433819115162}, {"step": 4800, "epoch": 10, "train_loss": 0.0009640359203331172}, {"step": 4850, "epoch": 10, "train_loss": 0.0023030948359519243}, {"step": 4900, "epoch": 10, "train_loss": 0.0008106343448162079}, {"step": 4950, "epoch": 10, "train_loss": 0.0002312892465852201}, {"step": 5000, "epoch": 10, "train_loss": 0.0004100862715858966}, {"step": 5050, "epoch": 10, "train_loss": 0.00035814059083350003}, {"step": 5100, "epoch": 10, "train_loss": 0.0005620885640382767}, {"step": 5150, "epoch": 10, "train_loss": 0.00011865406122524291}, {"epoch_end": 10, "step": 5159, "test_acc": 0.9831, "test_loss": 0.07104201159477234}, {"step": 5200, "epoch": 11, "train_loss": 0.0004436893214005977}, {"step": 5250, "epoch": 11, "train_loss": 0.0007607933366671205}, {"step": 5300, "epoch": 11, "train_loss": 0.0004835751897189766}, {"step": 5350, "epoch": 11, "train_loss": 0.00021810219914186746}, {"step": 5400, "epoch": 11, "train_loss": 0.00025913011631928384}, {"step": 5450, "epoch": 11, "train_loss": 0.00014999150880612433}, {"step": 5500, "epoch": 11, "train_loss": 0.0007390796672552824}, {"step": 5550, "epoch": 11, "train_loss": 0.00024223618675023317}, {"step": 5600, "epoch": 11, "train_loss": 0.0002916124649345875}, {"epoch_end": 11, "step": 5628, "test_acc": 0.9836, "test_loss": 0.07070657641887665}, {"step": 5650, "epoch": 12, "train_loss": 0.000250968209002167}, {"step": 5700, "epoch": 12, "train_loss": 5.820431397296488e-05}, {"step": 5750, "epoch": 12, "train_loss": 0.0001992640900425613}, {"step": 5800, "epoch": 12, "train_loss": 0.0001069598292815499}, {"step": 5850, "epoch": 12, "train_loss": 0.00026439520297572017}, {"step": 5900, "epoch": 12, "train_loss": 0.00022868881933391094}, {"step": 5950, "epoch": 12, "train_loss": 0.00014140657731331885}, {"step": 6000, "epoch": 12, "train_loss": 0.000274585880106315}, {"step": 6050, "epoch": 12, "train_loss": 0.00019624698325060308}, {"epoch_end": 12, "step": 6097, "test_acc": 0.9832, "test_loss": 0.07167425248622894}, {"step": 6100, "epoch": 13, "train_loss": 0.00046324048889800906}, {"step": 6150, "epoch": 13, "train_loss": 0.00011088149767601863}, {"step": 6200, "epoch": 13, "train_loss": 0.00011532158532645553}, {"step": 6250, "epoch": 13, "train_loss": 8.243001502705738e-05}, {"step": 6300, "epoch": 13, "train_loss": 0.00012665425310842693}, {"step": 6350, "epoch": 13, "train_loss": 0.00011592436931096017}, {"step": 6400, "epoch": 13, "train_loss": 0.0007366926874965429}, {"step": 6450, "epoch": 13, "train_loss": 0.0001460860949009657}, {"step": 6500, "epoch": 13, "train_loss": 0.00014242494944483042}, {"step": 6550, "epoch": 13, "train_loss": 0.00016198577941395342}, {"epoch_end": 13, "step": 6566, "test_acc": 0.9831, "test_loss": 0.07211719748973847}, {"step": 6600, "epoch": 14, "train_loss": 0.0001268152118427679}, {"step": 6650, "epoch": 14, "train_loss": 0.0001306341146118939}, {"step": 6700, "epoch": 14, "train_loss": 0.00019066481036134064}, {"step": 6750, "epoch": 14, "train_loss": 0.0002095211239065975}, {"step": 6800, "epoch": 14, "train_loss": 7.673785876249894e-05}, {"step": 6850, "epoch": 14, "train_loss": 0.00016222905833274126}, {"step": 6900, "epoch": 14, "train_loss": 0.00027574965497478843}, {"step": 6950, "epoch": 14, "train_loss": 2.6871934096561745e-05}, {"step": 7000, "epoch": 14, "train_loss": 0.0002463290002197027}, {"epoch_end": 14, "step": 7035, "test_acc": 0.9832, "test_loss": 0.07239087262153625}, {"step": 7050, "epoch": 15, "train_loss": 7.723146700300276e-05}, {"step": 7100, "epoch": 15, "train_loss": 0.00023211800726130605}, {"step": 7150, "epoch": 15, "train_loss": 0.00011530292249517515}, {"step": 7200, "epoch": 15, "train_loss": 0.00011787289258791134}, {"step": 7250, "epoch": 15, "train_loss": 0.00021336421195883304}, {"step": 7300, "epoch": 15, "train_loss": 0.00010562340321484953}, {"step": 7350, "epoch": 15, "train_loss": 5.955329834250733e-05}, {"step": 7400, "epoch": 15, "train_loss": 0.00014538221876136959}, {"step": 7450, "epoch": 15, "train_loss": 0.00018245287355966866}, {"step": 7500, "epoch": 15, "train_loss": 5.86891510465648e-05}, {"epoch_end": 15, "step": 7504, "test_acc": 0.9828, "test_loss": 0.07296301305294037}, {"step": 7550, "epoch": 16, "train_loss": 9.512629912933335e-05}, {"step": 7600, "epoch": 16, "train_loss": 0.00029424147214740515}, {"step": 7650, "epoch": 16, "train_loss": 0.00010368914081482217}, {"step": 7700, "epoch": 16, "train_loss": 0.00010391201794845983}, {"step": 7750, "epoch": 16, "train_loss": 0.00012849790800828487}, {"step": 7800, "epoch": 16, "train_loss": 0.0001979203661903739}, {"step": 7850, "epoch": 16, "train_loss": 6.034813486621715e-05}, {"step": 7900, "epoch": 16, "train_loss": 0.0001113874968723394}, {"step": 7950, "epoch": 16, "train_loss": 0.0001924021344166249}, {"epoch_end": 16, "step": 7973, "test_acc": 0.9831, "test_loss": 0.07338536593914032}, {"step": 8000, "epoch": 17, "train_loss": 7.238818216137588e-05}, {"step": 8050, "epoch": 17, "train_loss": 0.00010010840196628124}, {"step": 8100, "epoch": 17, "train_loss": 0.00016066443640738726}, {"step": 8150, "epoch": 17, "train_loss": 3.1075072911335155e-05}, {"step": 8200, "epoch": 17, "train_loss": 5.6670971389394253e-05}, {"step": 8250, "epoch": 17, "train_loss": 0.0001787046348908916}, {"step": 8300, "epoch": 17, "train_loss": 0.00014021720562595874}, {"step": 8350, "epoch": 17, "train_loss": 7.940473733469844e-05}, {"step": 8400, "epoch": 17, "train_loss": 0.0001404674694640562}, {"epoch_end": 17, "step": 8442, "test_acc": 0.9828, "test_loss": 0.07356535530090332}, {"step": 8450, "epoch": 18, "train_loss": 0.0002453565539326519}, {"step": 8500, "epoch": 18, "train_loss": 0.0001090616497094743}, {"step": 8550, "epoch": 18, "train_loss": 0.00017377808399032801}, {"step": 8600, "epoch": 18, "train_loss": 0.0002992010849993676}, {"step": 8650, "epoch": 18, "train_loss": 0.00011711372644640505}, {"step": 8700, "epoch": 18, "train_loss": 0.00016981217777356505}, {"step": 8750, "epoch": 18, "train_loss": 0.00010064614616567269}, {"step": 8800, "epoch": 18, "train_loss": 0.0001663248986005783}, {"step": 8850, "epoch": 18, "train_loss": 4.078277197550051e-05}, {"step": 8900, "epoch": 18, "train_loss": 9.053484973264858e-05}, {"epoch_end": 18, "step": 8911, "test_acc": 0.9833, "test_loss": 0.07417089953422547}, {"step": 8950, "epoch": 19, "train_loss": 9.111483814194798e-05}, {"step": 9000, "epoch": 19, "train_loss": 0.00011348637781338766}, {"step": 9050, "epoch": 19, "train_loss": 0.00021746699349023402}, {"step": 9100, "epoch": 19, "train_loss": 5.391851664171554e-05}, {"step": 9150, "epoch": 19, "train_loss": 0.00015882342995610088}, {"step": 9200, "epoch": 19, "train_loss": 7.221881969599053e-05}, {"step": 9250, "epoch": 19, "train_loss": 6.744629354216158e-05}, {"step": 9300, "epoch": 19, "train_loss": 6.260271766223013e-05}, {"step": 9350, "epoch": 19, "train_loss": 0.00012500591401476413}, {"epoch_end": 19, "step": 9380, "test_acc": 0.9832, "test_loss": 0.07446969006061555}], "final": {"test_acc": 0.9832, "test_loss": 0.07446969006061555, "wall_s": 21.193922996520996}}
|