{"args": {"mode": "sdil", "dataset": "cifar10", "depth": 5, "width": 64, "act": "tanh", "residual": 1, "epochs": 5, "batch_size": 128, "train_examples": 0, "task_seed": 0, "task_train_examples": 50000, "task_test_examples": 10000, "task_levels": 8, "task_n_in": 128, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.02, "eta_A": 0.02, "eta_P": 0.002, "momentum": 0.9, "w_scale": 1.0, "a_scale": 1.0, "feedback_scale": 1.0, "pert_sigma": 0.01, "pert_every": 4, "pert_ndirs": 16, "pert_mode": "simultaneous", "use_residual": 1, "raw_scale_control": "none", "learn_A": 1, "learn_P": 1, "p_neutral": 1, "p_warmup_steps": 200, "p_warmup_eta": 0.05, "nuis_rho": 0.0, "predictor_mode": "diagonal", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 4, "device": "cuda", "log_every": 500, "max_steps": 0, "probe_bs": 256, "outdir": "results", "tag": "scale_v3_cifar10_sdil_w64_d5_s4", "n_in": 3072, "n_out": 10}, "provenance": {"git_commit": "d4aef1a571c21e6e4e4125c90aa58da7f14e62e3", "git_dirty": false}, "steps": [{"step": 0, "epoch": 0, "train_loss": 2.6243700981140137, "cos_r_negg": [0.03922749310731888, -0.03871015086770058, -0.0025417697615921497, 0.016206644475460052, -0.03486630693078041], "cos_innovation_negg": [0.03922749310731888, -0.03871015086770058, -0.0025417697615921497, 0.016206644475460052, -0.03486630693078041], "cos_apical_negg": [0.03922749310731888, -0.03871015086770058, -0.0025417697615921497, 0.016206644475460052, -0.03486630693078041], "cos_Ac_negg": [0.03922749310731888, -0.03871015086770058, -0.0025417697615921497, 0.016206644475460052, -0.03486630693078041], "r_norm": [2.543478488922119, 2.4115402698516846, 2.448044538497925, 2.3827407360076904, 2.5208487510681152], "ldr": {"dL_sdil": -0.12978482246398926, "dL_bp": -0.4230227470397949, "ratio": 0.30680341275307366}}, {"epoch_end": 0, "step": 391, "test_acc": 0.3663, "test_loss": 1.8054610229492187}, {"step": 500, "epoch": 1, "train_loss": 1.6954187154769897, "cos_r_negg": [0.6702491044998169, 0.06278055906295776, -0.011454129591584206, 0.07948324084281921, 0.13017866015434265], "cos_innovation_negg": [0.6702491044998169, 0.06278055906295776, -0.011454129591584206, 0.07948324084281921, 0.13017866015434265], "cos_apical_negg": [0.6702491044998169, 0.06278055906295776, -0.011454129591584206, 0.07948324084281921, 0.13017866015434265], "cos_Ac_negg": [0.6702491044998169, 0.06278055906295776, -0.011454129591584206, 0.07948324084281921, 0.13017866015434265], "r_norm": [1.7980241775512695, 1.636629343032837, 1.6982510089874268, 1.6305831670761108, 1.7022085189819336]}, {"epoch_end": 1, "step": 782, "test_acc": 0.3869, "test_loss": 1.7607693725585938}, {"step": 1000, "epoch": 2, "train_loss": 1.7203733921051025, "cos_r_negg": [0.7361968755722046, 0.16038227081298828, 0.0974164754152298, 0.17229196429252625, 0.20677244663238525], "cos_innovation_negg": [0.7361968755722046, 0.16038227081298828, 0.0974164754152298, 0.17229196429252625, 0.20677244663238525], "cos_apical_negg": [0.7361968755722046, 0.16038227081298828, 0.0974164754152298, 0.17229196429252625, 0.20677244663238525], "cos_Ac_negg": [0.7361968755722046, 0.16038227081298828, 0.0974164754152298, 0.17229196429252625, 0.20677244663238525], "r_norm": [1.5628435611724854, 1.3496291637420654, 1.3926448822021484, 1.3491101264953613, 1.415849208831787]}, {"epoch_end": 2, "step": 1173, "test_acc": 0.395, "test_loss": 1.7277833129882811}, {"step": 1500, "epoch": 3, "train_loss": 1.6342253684997559, "cos_r_negg": [0.7573046684265137, 0.28741079568862915, 0.23762330412864685, 0.2927331328392029, 0.3346874415874481], "cos_innovation_negg": [0.7573046684265137, 0.28741079568862915, 0.23762330412864685, 0.2927331328392029, 0.3346874415874481], "cos_apical_negg": [0.7573046684265137, 0.28741079568862915, 0.23762330412864685, 0.2927331328392029, 0.3346874415874481], "cos_Ac_negg": [0.7573046684265137, 0.28741079568862915, 0.23762330412864685, 0.2927331328392029, 0.3346874415874481], "r_norm": [1.3521631956100464, 1.104792594909668, 1.1281609535217285, 1.1036646366119385, 1.1663498878479004]}, {"epoch_end": 3, "step": 1564, "test_acc": 0.4093, "test_loss": 1.7190211669921875}, {"epoch_end": 4, "step": 1955, "test_acc": 0.4142, "test_loss": 1.6917354614257814}], "final": {"test_acc": 0.4142, "test_loss": 1.6917354614257814, "wall_s": 8.945802927017212, "cos_r_negg": [0.8144403696060181, 0.3603973090648651, 0.34978049993515015, 0.407439261674881, 0.43859320878982544], "cos_innovation_negg": [0.8144403696060181, 0.3603973090648651, 0.34978049993515015, 0.407439261674881, 0.43859320878982544], "cos_apical_negg": [0.8144403696060181, 0.3603973090648651, 0.34978049993515015, 0.407439261674881, 0.43859320878982544], "cos_Ac_negg": [0.8144403696060181, 0.3603973090648651, 0.34978049993515015, 0.407439261674881, 0.43859320878982544]}}