1
|
{"args": {"mode": "bp", "dataset": "fmnist", "depth": 5, "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": 2, "device": "cuda", "log_every": 50, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "depth_fmnist_bp_d5_s2", "n_in": 784}, "steps": [{"step": 0, "epoch": 0, "train_loss": 2.2724270820617676}, {"step": 50, "epoch": 0, "train_loss": 0.6291083097457886}, {"step": 100, "epoch": 0, "train_loss": 0.588944137096405}, {"step": 150, "epoch": 0, "train_loss": 0.4161628782749176}, {"step": 200, "epoch": 0, "train_loss": 0.4777505099773407}, {"step": 250, "epoch": 0, "train_loss": 0.41141918301582336}, {"step": 300, "epoch": 0, "train_loss": 0.7739314436912537}, {"step": 350, "epoch": 0, "train_loss": 0.45101049542427063}, {"step": 400, "epoch": 0, "train_loss": 0.5787916779518127}, {"step": 450, "epoch": 0, "train_loss": 0.47871696949005127}, {"epoch_end": 0, "step": 469, "test_acc": 0.8376, "test_loss": 0.4660996643066406}, {"step": 500, "epoch": 1, "train_loss": 0.3060743808746338}, {"step": 550, "epoch": 1, "train_loss": 0.551163911819458}, {"step": 600, "epoch": 1, "train_loss": 0.45933467149734497}, {"step": 650, "epoch": 1, "train_loss": 0.5389010310173035}, {"step": 700, "epoch": 1, "train_loss": 0.3079885244369507}, {"step": 750, "epoch": 1, "train_loss": 0.4047509431838989}, {"step": 800, "epoch": 1, "train_loss": 0.43391090631484985}, {"step": 850, "epoch": 1, "train_loss": 0.2338166981935501}, {"step": 900, "epoch": 1, "train_loss": 0.4243391454219818}, {"epoch_end": 1, "step": 938, "test_acc": 0.8575, "test_loss": 0.38687999267578127}, {"step": 950, "epoch": 2, "train_loss": 0.30006322264671326}, {"step": 1000, "epoch": 2, "train_loss": 0.3031974136829376}, {"step": 1050, "epoch": 2, "train_loss": 0.28435415029525757}, {"step": 1100, "epoch": 2, "train_loss": 0.3177497684955597}, {"step": 1150, "epoch": 2, "train_loss": 0.3817498981952667}, {"step": 1200, "epoch": 2, "train_loss": 0.45226624608039856}, {"step": 1250, "epoch": 2, "train_loss": 0.21334055066108704}, {"step": 1300, "epoch": 2, "train_loss": 0.2983773350715637}, {"step": 1350, "epoch": 2, "train_loss": 0.19865837693214417}, {"step": 1400, "epoch": 2, "train_loss": 0.3871944546699524}, {"epoch_end": 2, "step": 1407, "test_acc": 0.8697, "test_loss": 0.3733336151123047}, {"step": 1450, "epoch": 3, "train_loss": 0.5394847989082336}, {"step": 1500, "epoch": 3, "train_loss": 0.34689080715179443}, {"step": 1550, "epoch": 3, "train_loss": 0.25678229331970215}, {"step": 1600, "epoch": 3, "train_loss": 0.35751599073410034}, {"step": 1650, "epoch": 3, "train_loss": 0.32231006026268005}, {"step": 1700, "epoch": 3, "train_loss": 0.2637043595314026}, {"step": 1750, "epoch": 3, "train_loss": 0.46819260716438293}, {"step": 1800, "epoch": 3, "train_loss": 0.26558616757392883}, {"step": 1850, "epoch": 3, "train_loss": 0.3585832118988037}, {"epoch_end": 3, "step": 1876, "test_acc": 0.8672, "test_loss": 0.3702358917236328}, {"step": 1900, "epoch": 4, "train_loss": 0.2602953612804413}, {"step": 1950, "epoch": 4, "train_loss": 0.4613979160785675}, {"step": 2000, "epoch": 4, "train_loss": 0.2870660424232483}, {"step": 2050, "epoch": 4, "train_loss": 0.34146273136138916}, {"step": 2100, "epoch": 4, "train_loss": 0.3304780423641205}, {"step": 2150, "epoch": 4, "train_loss": 0.3004378080368042}, {"step": 2200, "epoch": 4, "train_loss": 0.3135995864868164}, {"step": 2250, "epoch": 4, "train_loss": 0.22820676863193512}, {"step": 2300, "epoch": 4, "train_loss": 0.14167898893356323}, {"epoch_end": 4, "step": 2345, "test_acc": 0.8769, "test_loss": 0.3501354461669922}, {"step": 2350, "epoch": 5, "train_loss": 0.20593410730361938}, {"step": 2400, "epoch": 5, "train_loss": 0.25445157289505005}, {"step": 2450, "epoch": 5, "train_loss": 0.3001524806022644}, {"step": 2500, "epoch": 5, "train_loss": 0.26534005999565125}, {"step": 2550, "epoch": 5, "train_loss": 0.24374982714653015}, {"step": 2600, "epoch": 5, "train_loss": 0.2973781228065491}, {"step": 2650, "epoch": 5, "train_loss": 0.1684378683567047}, {"step": 2700, "epoch": 5, "train_loss": 0.15932217240333557}, {"step": 2750, "epoch": 5, "train_loss": 0.38720637559890747}, {"step": 2800, "epoch": 5, "train_loss": 0.2546900510787964}, {"epoch_end": 5, "step": 2814, "test_acc": 0.8777, "test_loss": 0.3525110198974609}, {"step": 2850, "epoch": 6, "train_loss": 0.16797277331352234}, {"step": 2900, "epoch": 6, "train_loss": 0.1975693702697754}, {"step": 2950, "epoch": 6, "train_loss": 0.10772007703781128}, {"step": 3000, "epoch": 6, "train_loss": 0.36531224846839905}, {"step": 3050, "epoch": 6, "train_loss": 0.4098561406135559}, {"step": 3100, "epoch": 6, "train_loss": 0.23266497254371643}, {"step": 3150, "epoch": 6, "train_loss": 0.2715470492839813}, {"step": 3200, "epoch": 6, "train_loss": 0.17548038065433502}, {"step": 3250, "epoch": 6, "train_loss": 0.1738280951976776}, {"epoch_end": 6, "step": 3283, "test_acc": 0.8763, "test_loss": 0.35267843627929685}, {"step": 3300, "epoch": 7, "train_loss": 0.26329195499420166}, {"step": 3350, "epoch": 7, "train_loss": 0.1992304027080536}, {"step": 3400, "epoch": 7, "train_loss": 0.22501884400844574}, {"step": 3450, "epoch": 7, "train_loss": 0.2181098908185959}, {"step": 3500, "epoch": 7, "train_loss": 0.3561607897281647}, {"step": 3550, "epoch": 7, "train_loss": 0.21900039911270142}, {"step": 3600, "epoch": 7, "train_loss": 0.17021630704402924}, {"step": 3650, "epoch": 7, "train_loss": 0.2144341766834259}, {"step": 3700, "epoch": 7, "train_loss": 0.22487621009349823}, {"step": 3750, "epoch": 7, "train_loss": 0.3742579519748688}, {"epoch_end": 7, "step": 3752, "test_acc": 0.8825, "test_loss": 0.3356485198974609}, {"step": 3800, "epoch": 8, "train_loss": 0.2822781801223755}, {"step": 3850, "epoch": 8, "train_loss": 0.31094735860824585}, {"step": 3900, "epoch": 8, "train_loss": 0.24942971765995026}, {"step": 3950, "epoch": 8, "train_loss": 0.23154646158218384}, {"step": 4000, "epoch": 8, "train_loss": 0.2953263521194458}, {"step": 4050, "epoch": 8, "train_loss": 0.2579735219478607}, {"step": 4100, "epoch": 8, "train_loss": 0.2300138622522354}, {"step": 4150, "epoch": 8, "train_loss": 0.2521052360534668}, {"step": 4200, "epoch": 8, "train_loss": 0.26177576184272766}, {"epoch_end": 8, "step": 4221, "test_acc": 0.8849, "test_loss": 0.34830634460449217}, {"step": 4250, "epoch": 9, "train_loss": 0.2059216946363449}, {"step": 4300, "epoch": 9, "train_loss": 0.14424782991409302}, {"step": 4350, "epoch": 9, "train_loss": 0.1493077278137207}, {"step": 4400, "epoch": 9, "train_loss": 0.26786214113235474}, {"step": 4450, "epoch": 9, "train_loss": 0.1358993649482727}, {"step": 4500, "epoch": 9, "train_loss": 0.27530166506767273}, {"step": 4550, "epoch": 9, "train_loss": 0.17075097560882568}, {"step": 4600, "epoch": 9, "train_loss": 0.2506886124610901}, {"step": 4650, "epoch": 9, "train_loss": 0.2643336057662964}, {"epoch_end": 9, "step": 4690, "test_acc": 0.8839, "test_loss": 0.3397621795654297}, {"step": 4700, "epoch": 10, "train_loss": 0.18522974848747253}, {"step": 4750, "epoch": 10, "train_loss": 0.24843522906303406}, {"step": 4800, "epoch": 10, "train_loss": 0.18528737127780914}, {"step": 4850, "epoch": 10, "train_loss": 0.21752594411373138}, {"step": 4900, "epoch": 10, "train_loss": 0.1586490124464035}, {"step": 4950, "epoch": 10, "train_loss": 0.21777848899364471}, {"step": 5000, "epoch": 10, "train_loss": 0.22450760006904602}, {"step": 5050, "epoch": 10, "train_loss": 0.3276432156562805}, {"step": 5100, "epoch": 10, "train_loss": 0.14997291564941406}, {"step": 5150, "epoch": 10, "train_loss": 0.2049725353717804}, {"epoch_end": 10, "step": 5159, "test_acc": 0.8817, "test_loss": 0.35867194519042966}, {"step": 5200, "epoch": 11, "train_loss": 0.19226926565170288}, {"step": 5250, "epoch": 11, "train_loss": 0.12053055316209793}, {"step": 5300, "epoch": 11, "train_loss": 0.25897008180618286}, {"step": 5350, "epoch": 11, "train_loss": 0.2860298454761505}, {"step": 5400, "epoch": 11, "train_loss": 0.15427306294441223}, {"step": 5450, "epoch": 11, "train_loss": 0.2181725800037384}, {"step": 5500, "epoch": 11, "train_loss": 0.20596648752689362}, {"step": 5550, "epoch": 11, "train_loss": 0.24289409816265106}, {"step": 5600, "epoch": 11, "train_loss": 0.18647000193595886}, {"epoch_end": 11, "step": 5628, "test_acc": 0.8888, "test_loss": 0.35912730407714843}, {"step": 5650, "epoch": 12, "train_loss": 0.13268816471099854}, {"step": 5700, "epoch": 12, "train_loss": 0.1553129404783249}, {"step": 5750, "epoch": 12, "train_loss": 0.35835614800453186}, {"step": 5800, "epoch": 12, "train_loss": 0.22895723581314087}, {"step": 5850, "epoch": 12, "train_loss": 0.26292291283607483}, {"step": 5900, "epoch": 12, "train_loss": 0.26031050086021423}, {"step": 5950, "epoch": 12, "train_loss": 0.2006765753030777}, {"step": 6000, "epoch": 12, "train_loss": 0.12995465099811554}, {"step": 6050, "epoch": 12, "train_loss": 0.15284067392349243}, {"epoch_end": 12, "step": 6097, "test_acc": 0.8781, "test_loss": 0.3612210662841797}, {"step": 6100, "epoch": 13, "train_loss": 0.25077420473098755}, {"step": 6150, "epoch": 13, "train_loss": 0.15254953503608704}, {"step": 6200, "epoch": 13, "train_loss": 0.17051410675048828}, {"step": 6250, "epoch": 13, "train_loss": 0.19774436950683594}, {"step": 6300, "epoch": 13, "train_loss": 0.25119417905807495}, {"step": 6350, "epoch": 13, "train_loss": 0.21384388208389282}, {"step": 6400, "epoch": 13, "train_loss": 0.17600706219673157}, {"step": 6450, "epoch": 13, "train_loss": 0.18031972646713257}, {"step": 6500, "epoch": 13, "train_loss": 0.09279892593622208}, {"step": 6550, "epoch": 13, "train_loss": 0.17270098626613617}, {"epoch_end": 13, "step": 6566, "test_acc": 0.8856, "test_loss": 0.35603551025390623}, {"step": 6600, "epoch": 14, "train_loss": 0.14172588288784027}, {"step": 6650, "epoch": 14, "train_loss": 0.2528199851512909}, {"step": 6700, "epoch": 14, "train_loss": 0.12140634655952454}, {"step": 6750, "epoch": 14, "train_loss": 0.22545059025287628}, {"step": 6800, "epoch": 14, "train_loss": 0.153004452586174}, {"step": 6850, "epoch": 14, "train_loss": 0.1473826915025711}, {"step": 6900, "epoch": 14, "train_loss": 0.20940542221069336}, {"step": 6950, "epoch": 14, "train_loss": 0.20150403678417206}, {"step": 7000, "epoch": 14, "train_loss": 0.22650909423828125}, {"epoch_end": 14, "step": 7035, "test_acc": 0.8873, "test_loss": 0.3609393676757813}, {"step": 7050, "epoch": 15, "train_loss": 0.23146794736385345}, {"step": 7100, "epoch": 15, "train_loss": 0.2198127955198288}, {"step": 7150, "epoch": 15, "train_loss": 0.21883848309516907}, {"step": 7200, "epoch": 15, "train_loss": 0.13089261949062347}, {"step": 7250, "epoch": 15, "train_loss": 0.23114505410194397}, {"step": 7300, "epoch": 15, "train_loss": 0.20465871691703796}, {"step": 7350, "epoch": 15, "train_loss": 0.15555129945278168}, {"step": 7400, "epoch": 15, "train_loss": 0.17657409608364105}, {"step": 7450, "epoch": 15, "train_loss": 0.15039940178394318}, {"step": 7500, "epoch": 15, "train_loss": 0.16636410355567932}, {"epoch_end": 15, "step": 7504, "test_acc": 0.8846, "test_loss": 0.3848857727050781}, {"step": 7550, "epoch": 16, "train_loss": 0.1410776674747467}, {"step": 7600, "epoch": 16, "train_loss": 0.24707207083702087}, {"step": 7650, "epoch": 16, "train_loss": 0.14235229790210724}, {"step": 7700, "epoch": 16, "train_loss": 0.14696922898292542}, {"step": 7750, "epoch": 16, "train_loss": 0.25838419795036316}, {"step": 7800, "epoch": 16, "train_loss": 0.1313989758491516}, {"step": 7850, "epoch": 16, "train_loss": 0.1306954026222229}, {"step": 7900, "epoch": 16, "train_loss": 0.0946805328130722}, {"step": 7950, "epoch": 16, "train_loss": 0.20922163128852844}, {"epoch_end": 16, "step": 7973, "test_acc": 0.8865, "test_loss": 0.3625572509765625}, {"step": 8000, "epoch": 17, "train_loss": 0.14226287603378296}, {"step": 8050, "epoch": 17, "train_loss": 0.12264211475849152}, {"step": 8100, "epoch": 17, "train_loss": 0.21043670177459717}, {"step": 8150, "epoch": 17, "train_loss": 0.21666157245635986}, {"step": 8200, "epoch": 17, "train_loss": 0.16423141956329346}, {"step": 8250, "epoch": 17, "train_loss": 0.21586106717586517}, {"step": 8300, "epoch": 17, "train_loss": 0.11189636588096619}, {"step": 8350, "epoch": 17, "train_loss": 0.12157561630010605}, {"step": 8400, "epoch": 17, "train_loss": 0.3693520426750183}, {"epoch_end": 17, "step": 8442, "test_acc": 0.8891, "test_loss": 0.37227940368652346}, {"step": 8450, "epoch": 18, "train_loss": 0.129014253616333}, {"step": 8500, "epoch": 18, "train_loss": 0.08220576494932175}, {"step": 8550, "epoch": 18, "train_loss": 0.13754802942276}, {"step": 8600, "epoch": 18, "train_loss": 0.1341312676668167}, {"step": 8650, "epoch": 18, "train_loss": 0.1198374554514885}, {"step": 8700, "epoch": 18, "train_loss": 0.1801692396402359}, {"step": 8750, "epoch": 18, "train_loss": 0.1552022099494934}, {"step": 8800, "epoch": 18, "train_loss": 0.06013773754239082}, {"step": 8850, "epoch": 18, "train_loss": 0.095866858959198}, {"step": 8900, "epoch": 18, "train_loss": 0.19610820710659027}, {"epoch_end": 18, "step": 8911, "test_acc": 0.883, "test_loss": 0.388243994140625}, {"step": 8950, "epoch": 19, "train_loss": 0.20610517263412476}, {"step": 9000, "epoch": 19, "train_loss": 0.15180625021457672}, {"step": 9050, "epoch": 19, "train_loss": 0.16468678414821625}, {"step": 9100, "epoch": 19, "train_loss": 0.13932695984840393}, {"step": 9150, "epoch": 19, "train_loss": 0.15994200110435486}, {"step": 9200, "epoch": 19, "train_loss": 0.09662564098834991}, {"step": 9250, "epoch": 19, "train_loss": 0.1308855265378952}, {"step": 9300, "epoch": 19, "train_loss": 0.11630909889936447}, {"step": 9350, "epoch": 19, "train_loss": 0.2026025950908661}, {"epoch_end": 19, "step": 9380, "test_acc": 0.889, "test_loss": 0.38354205627441407}], "final": {"test_acc": 0.889, "test_loss": 0.38354205627441407, "wall_s": 30.785258531570435}}
|