diff options
Diffstat (limited to 'results/kp_dynamic_projection_short')
| -rw-r--r-- | results/kp_dynamic_projection_short/dynamic.json | 1256 |
1 files changed, 1256 insertions, 0 deletions
diff --git a/results/kp_dynamic_projection_short/dynamic.json b/results/kp_dynamic_projection_short/dynamic.json new file mode 100644 index 0000000..42e8e1c --- /dev/null +++ b/results/kp_dynamic_projection_short/dynamic.json @@ -0,0 +1,1256 @@ +{ + "architecture": { + "adaptive_apical_parameters": 376832, + "adaptive_feedback_parameters": 267904, + "base_width": 16, + "blocks_per_stage": 3, + "bn_eps": 1e-05, + "bn_momentum": 0.1, + "depth": 20, + "family": "CIFAR 6n+2 ResNet, option-A shortcuts", + "fixed_feedback_parameters": 0, + "fixed_traffic_coefficients": 188416, + "forward_parameters": 269722, + "hidden_shapes": [ + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 16, + 32, + 32 + ], + [ + 32, + 16, + 16 + ], + [ + 32, + 16, + 16 + ], + [ + 32, + 16, + 16 + ], + [ + 32, + 16, + 16 + ], + [ + 32, + 16, + 16 + ], + [ + 32, + 16, + 16 + ], + [ + 64, + 8, + 8 + ], + [ + 64, + 8, + 8 + ], + [ + 64, + 8, + 8 + ], + [ + 64, + 8, + 8 + ], + [ + 64, + 8, + 8 + ], + [ + 64, + 8, + 8 + ] + ], + "normalization": "batchnorm", + "predictor_parameters": 376832, + "residual_scale": 1.0, + "vectorizer_mode": null, + "vectorizer_parameters": 0 + }, + "args": { + "a_scale": 1.0, + "a_warmup_steps": 0, + "alignment_probe": 32, + "apical_calibration_mode": "unit_targets", + "apical_seed": null, + "augment_train": 1, + "batch_size": 128, + "bn_eps": 1e-05, + "bn_momentum": 0.1, + "data_dir": "/home/yurenh2/sdrn/data", + "depth": 20, + "device": "cuda", + "epochs": 20, + "eta_A": 0.01, + "eta_P": 0.1, + "eval_every": 0, + "eval_split": "validation", + "learn_P": 1, + "loader_seed": 0, + "lr": 0.1, + "lr_gamma": 0.1, + "lr_milestones": "100,150", + "lr_schedule": "step", + "max_steps": 0, + "mirror_batch_size": 1, + "mirror_eta": 0.1, + "mirror_every": 16, + "mirror_noise_std": 1.0, + "mirror_seed": 3000, + "mirror_warmup_steps": 0, + "mode": "kp_traffic", + "momentum": 0.9, + "neutral_projection": 1, + "normalization": "batchnorm", + "nuisance_scale": 0.0, + "out": "results/kp_dynamic_projection_short/dynamic.json", + "output_lr": 0.1, + "pert_directions": 1, + "pert_every": 4, + "pert_sigma": 0.01, + "perturb_seed": 1000, + "predictor_every": 0, + "predictor_mode": "closed_form", + "predictor_warmup_steps": 1, + "residual_scale": null, + "seed": 0, + "split_seed": 2027, + "traffic_calibration_examples": 64, + "traffic_ratio": 4.0, + "traffic_rule": "innovation", + "traffic_seed": 4000, + "train_limit": 0, + "use_residual": 1, + "val_examples": 5000, + "vectorizer_mode": "spatial_template", + "warmup_epochs": 0, + "weight_decay": 0.0001, + "weight_scale": 1.0, + "width": 16 + }, + "calibration_metric_space": "reciprocal_local_activity_products_with_mixed_apical_traffic", + "counters": { + "apical_warmup_examples": 0, + "calibration_event_examples": 0, + "causal_scalar_observations": 0, + "logical_batch_loss_queries": 0, + "mirror_conv_examples": 0, + "mirror_events": 0, + "mirror_readout_examples": 0, + "neutral_projection_examples": 900000, + "ordinary_examples": 900000, + "per_example_loss_terms": 0, + "perturbation_events": 0, + "perturbation_forward_examples": 0, + "predictor_update_examples": 64, + "predictor_warmup_examples": 0, + "traffic_audit_examples": 0, + "traffic_calibration_examples": 64 + }, + "diagnostics": { + "early_third_mean": 0.8838296929995219, + "feedback_forward_cosine": [ + 0.9164010882377625, + 0.9122046232223511, + 0.9070655703544617, + 0.8978453874588013, + 0.9216912984848022, + 0.9040270447731018, + 0.8969871997833252, + 0.9278544187545776, + 0.9036811590194702, + 0.893582820892334, + 0.9168826937675476, + 0.8956358432769775, + 0.8940100073814392, + 0.9253073930740356, + 0.9063614010810852, + 0.885367751121521, + 0.8874988555908203, + 0.7989962100982666, + 0.9856699705123901 + ], + "feedback_forward_norm_ratio": [ + 1.0100057125091553, + 1.0265120267868042, + 1.0055675506591797, + 1.0348376035690308, + 1.01314115524292, + 1.0377860069274902, + 0.9962888360023499, + 1.0254077911376953, + 1.0043511390686035, + 1.0455774068832397, + 1.005670428276062, + 1.0477830171585083, + 1.0260179042816162, + 1.0267318487167358, + 1.0144336223602295, + 1.0429489612579346, + 1.0219807624816895, + 1.043029546737671, + 1.0817701816558838 + ], + "feedback_forward_relative_error": [ + 0.4110606014728546, + 0.4253809452056885, + 0.43235957622528076, + 0.46112972497940063, + 0.39855772256851196, + 0.4479134976863861, + 0.4530726969242096, + 0.38549038767814636, + 0.4398804306983948, + 0.4739324450492859, + 0.40891233086586, + 0.47009050846099854, + 0.4670889377593994, + 0.39254701137542725, + 0.4361063838005066, + 0.49087244272232056, + 0.4800320863723755, + 0.6489664912223816, + 0.1941388100385666 + ], + "innovation_negative_gradient_cosine": [ + 0.8905045986175537, + 0.8762503862380981, + 0.8885157704353333, + 0.8560446500778198, + 0.8927435874938965, + 0.8989191651344299, + 0.9012151956558228, + 0.9121050834655762, + 0.9175336360931396, + 0.8981186151504517, + 0.9214124083518982, + 0.9123570322990417, + 0.9238299131393433, + 0.9452128410339355, + 0.9460467100143433, + 0.9446141123771667, + 0.9674544930458069, + 0.9785212278366089, + 0.9848822355270386 + ], + "instruction_negative_gradient_cosine": [ + 0.8905046582221985, + 0.8762503862380981, + 0.8885157704353333, + 0.8560446500778198, + 0.8927435874938965, + 0.8989191651344299, + 0.9012151956558228, + 0.9121050834655762, + 0.9175336360931396, + 0.8981186151504517, + 0.9214124083518982, + 0.9123570322990417, + 0.9238299131393433, + 0.9452128410339355, + 0.9460467100143433, + 0.9446141123771667, + 0.9674544930458069, + 0.9785212278366089, + 0.9848827123641968 + ], + "matched_negative_gradient_cosine": [ + 0.04840090125799179, + 0.04997877776622772, + 0.06474610418081284, + 0.05651163309812546, + 0.10006377100944519, + 0.14476574957370758, + 0.14505013823509216, + 0.15189626812934875, + 0.1707613319158554, + 0.2082226276397705, + 0.22710245847702026, + 0.2666760981082916, + 0.28864219784736633, + 0.23248949646949768, + 0.290870726108551, + 0.2743467092514038, + 0.4561747908592224, + 0.3063083291053772, + 0.3313695788383484 + ], + "max_norm_match_direction_error": 2.384185791015625e-07, + "max_norm_match_relative_error": 1.6168546324024646e-07, + "mean_feedback_forward_cosine": 0.9040563545728985, + "mean_feedback_forward_relative_error": 0.43776489637399973, + "neutral_projection": { + "instruction_observations": 0, + "max_absolute_correction_slope": 0.029675621539354324, + "max_absolute_post_projection_soma_slope": 5.501707089905494e-09, + "max_absolute_pre_projection_soma_slope": 0.029675621539354324, + "max_positive_post_projection_soma_slope": 5.501707089905494e-09, + "min_post_projection_soma_slope": -4.093964278695239e-09, + "observations": 32, + "post_projection_traffic_rms_ratio": 2.5643257040029264e-08, + "pre_projection_traffic_rms_ratio": 0.008001899731859899 + }, + "normalization_state": "training_batch_stats_without_running_update", + "predictor_traffic_residual_rms_ratio": 0.008001899731816205, + "raw_negative_gradient_cosine": [ + 0.04840090125799179, + 0.04997878521680832, + 0.06474609673023224, + 0.05651163309812546, + 0.10006377100944519, + 0.1447657346725464, + 0.14505013823509216, + 0.15189626812934875, + 0.1707613319158554, + 0.2082226276397705, + 0.22710244357585907, + 0.2666760981082916, + 0.28864219784736633, + 0.23248949646949768, + 0.2908707559108734, + 0.2743467092514038, + 0.4561747908592224, + 0.3063083291053772, + 0.3313695788383484 + ], + "teaching_negative_gradient_cosine": [ + 0.8905045986175537, + 0.8762503862380981, + 0.8885157704353333, + 0.8560446500778198, + 0.8927435874938965, + 0.8989191651344299, + 0.9012151956558228, + 0.9121050834655762, + 0.9175336360931396, + 0.8981186151504517, + 0.9214124083518982, + 0.9123570322990417, + 0.9238299131393433, + 0.9452128410339355, + 0.9460467100143433, + 0.9446141123771667, + 0.9674544930458069, + 0.9785212278366089, + 0.9848822355270386 + ], + "traffic_instruction_rms_ratio": 7.620419146144017, + "used_negative_gradient_cosine": [ + 0.8905045986175537, + 0.8762503862380981, + 0.8885157704353333, + 0.8560446500778198, + 0.8927435874938965, + 0.8989191651344299, + 0.9012151956558228, + 0.9121050834655762, + 0.9175336360931396, + 0.8981186151504517, + 0.9214124083518982, + 0.9123570322990417, + 0.9238299131393433, + 0.9452128410339355, + 0.9460467100143433, + 0.9446141123771667, + 0.9674544930458069, + 0.9785212278366089, + 0.9848822355270386 + ], + "wall_s": 0.19997763633728027 + }, + "epochs": [ + { + "epoch": 1, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.31910240650177, + "mean_feedback_forward_cosine": 0.27761596676550415, + "mean_feedback_forward_relative_error": 1.1998598387366848, + "min_feedback_forward_cosine": 0.13458214700222015 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.00202605631495943, + "instruction_rms": 0.0020260563152199934, + "raw_apical_rms": 0.020262751690676636, + "teaching_rms": 0.00202605631495943, + "traffic_rms": 0.0201484834522033 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.4347571131168024e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 3.033539150096096e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008334217485345142, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 352, + "train_examples": 45000, + "train_loss": 1.8282688387552897 + }, + { + "epoch": 2, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.2313768863677979, + "mean_feedback_forward_cosine": 0.4105625670207174, + "mean_feedback_forward_relative_error": 1.0836481702955145, + "min_feedback_forward_cosine": 0.25604695081710815 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.003297419456313343, + "instruction_rms": 0.003297419455709462, + "raw_apical_rms": 0.022791508405285713, + "teaching_rms": 0.003297419456313343, + "traffic_rms": 0.022548952070543346 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.326885534780331e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.8500283076552536e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.00808966859553636, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 704, + "train_examples": 45000, + "train_loss": 1.3831410327275595 + }, + { + "epoch": 3, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.1740696430206299, + "mean_feedback_forward_cosine": 0.5126491643880543, + "mean_feedback_forward_relative_error": 0.987261751764699, + "min_feedback_forward_cosine": 0.3285972476005554 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.004255481775745094, + "instruction_rms": 0.004255481777557298, + "raw_apical_rms": 0.024773834165770024, + "teaching_rms": 0.004255481775745094, + "traffic_rms": 0.024402880190041956 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.2201716081960967e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.7863413228925297e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.0076538653643200665, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 1056, + "train_examples": 45000, + "train_loss": 1.1027180670632257 + }, + { + "epoch": 4, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.1290353536605835, + "mean_feedback_forward_cosine": 0.5826303990263688, + "mean_feedback_forward_relative_error": 0.9152186905082903, + "min_feedback_forward_cosine": 0.38325169682502747 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.004572709251161924, + "instruction_rms": 0.00457270925128675, + "raw_apical_rms": 0.026117791752749146, + "teaching_rms": 0.004572709251161924, + "traffic_rms": 0.025712007597426347 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.1780786124404585e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.752000852403912e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.00799131550322377, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 1408, + "train_examples": 45000, + "train_loss": 0.9067903932147556 + }, + { + "epoch": 5, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.0875521898269653, + "mean_feedback_forward_cosine": 0.6364888909615969, + "mean_feedback_forward_relative_error": 0.8546328262278908, + "min_feedback_forward_cosine": 0.4301040470600128 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.004772560198436314, + "instruction_rms": 0.004772560198447473, + "raw_apical_rms": 0.027025261672198753, + "teaching_rms": 0.004772560198436314, + "traffic_rms": 0.02659743064896448 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.0928383531449981e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.7264058220400412e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008177511484473125, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 1760, + "train_examples": 45000, + "train_loss": 0.7890776650534735 + }, + { + "epoch": 6, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.0485821962356567, + "mean_feedback_forward_cosine": 0.6795904981462579, + "mean_feedback_forward_relative_error": 0.8026952508248781, + "min_feedback_forward_cosine": 0.47200721502304077 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.004961546770944681, + "instruction_rms": 0.004961546772259885, + "raw_apical_rms": 0.027952064225931196, + "teaching_rms": 0.004961546770944681, + "traffic_rms": 0.027505118102195726 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.212230316127716e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.716657735507771e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008623911947267906, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 2112, + "train_examples": 45000, + "train_loss": 0.7023197852028741 + }, + { + "epoch": 7, + "feedback_tracking": { + "max_feedback_forward_relative_error": 1.0108858346939087, + "mean_feedback_forward_cosine": 0.7150747493693703, + "mean_feedback_forward_relative_error": 0.757068152490415, + "min_feedback_forward_cosine": 0.51059490442276 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005110579150159904, + "instruction_rms": 0.005110579152312177, + "raw_apical_rms": 0.028436775097347974, + "teaching_rms": 0.005110579150159904, + "traffic_rms": 0.027969432463233527 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.3982001334511551e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.7175836430904836e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009560643492622747, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 2464, + "train_examples": 45000, + "train_loss": 0.6481795397864448 + }, + { + "epoch": 8, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.9762044548988342, + "mean_feedback_forward_cosine": 0.7429857787333036, + "mean_feedback_forward_relative_error": 0.7190478695066351, + "min_feedback_forward_cosine": 0.5446423292160034 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005155052457822899, + "instruction_rms": 0.005155052458409516, + "raw_apical_rms": 0.029233232040911328, + "teaching_rms": 0.005155052457822899, + "traffic_rms": 0.02877096767855112 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.2557397788270919e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.7026317637378097e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009420058533231888, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 2816, + "train_examples": 45000, + "train_loss": 0.59599007835918 + }, + { + "epoch": 9, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.9435642957687378, + "mean_feedback_forward_cosine": 0.7662839795413771, + "mean_feedback_forward_relative_error": 0.6857478132373408, + "min_feedback_forward_cosine": 0.5751973390579224 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005068461219535189, + "instruction_rms": 0.005068461219430642, + "raw_apical_rms": 0.0294579511020389, + "teaching_rms": 0.005068461219535189, + "traffic_rms": 0.029014509809828528 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.0545667450401197e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.6974320485178298e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009092547478364386, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 3168, + "train_examples": 45000, + "train_loss": 0.5547073147667779 + }, + { + "epoch": 10, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.9125420451164246, + "mean_feedback_forward_cosine": 0.7866145309649015, + "mean_feedback_forward_relative_error": 0.6553051957958623, + "min_feedback_forward_cosine": 0.6029062271118164 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005000429711851496, + "instruction_rms": 0.005000429710500525, + "raw_apical_rms": 0.029596698572323464, + "teaching_rms": 0.005000429711851496, + "traffic_rms": 0.029167541082597878 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.178124442446915e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.7003121251923688e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009180067416735304, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 3520, + "train_examples": 45000, + "train_loss": 0.5227681635538737 + }, + { + "epoch": 11, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.8820649981498718, + "mean_feedback_forward_cosine": 0.8043996942670721, + "mean_feedback_forward_relative_error": 0.6272416444201219, + "min_feedback_forward_cosine": 0.6290910840034485 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005044316803973928, + "instruction_rms": 0.005044316801621711, + "raw_apical_rms": 0.03001736771474847, + "teaching_rms": 0.005044316803973928, + "traffic_rms": 0.02958534987304112 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 8.826648745241528e-09, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.702379212027869e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009096700480931183, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 3872, + "train_examples": 45000, + "train_loss": 0.49492676684061687 + }, + { + "epoch": 12, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.8521213531494141, + "mean_feedback_forward_cosine": 0.8205687780129282, + "mean_feedback_forward_relative_error": 0.6006360289297605, + "min_feedback_forward_cosine": 0.6540982127189636 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005087378188252506, + "instruction_rms": 0.005087378189459039, + "raw_apical_rms": 0.030200814265538, + "teaching_rms": 0.005087378188252506, + "traffic_rms": 0.029764075035439572 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.02967562898993492, + "maximum_absolute_post_projection_soma_slope": 1.1838372948602682e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.6941134445479207e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009485451390670294, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 4224, + "train_examples": 45000, + "train_loss": 0.4813972399367227 + }, + { + "epoch": 13, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.8246222734451294, + "mean_feedback_forward_cosine": 0.8349180127445021, + "mean_feedback_forward_relative_error": 0.5759153946449882, + "min_feedback_forward_cosine": 0.6758251190185547 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005101863249728129, + "instruction_rms": 0.005101863250123167, + "raw_apical_rms": 0.030403724788101716, + "teaching_rms": 0.005101863249728129, + "traffic_rms": 0.02996659373718106 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.02967562898993492, + "maximum_absolute_post_projection_soma_slope": 1.1646240416496312e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.696128614213825e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.009505634104865747, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 4576, + "train_examples": 45000, + "train_loss": 0.4545250814967685 + }, + { + "epoch": 14, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.7969242334365845, + "mean_feedback_forward_cosine": 0.8478105758365831, + "mean_feedback_forward_relative_error": 0.5529624804070121, + "min_feedback_forward_cosine": 0.6973646879196167 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.00511037850823289, + "instruction_rms": 0.005110378506856121, + "raw_apical_rms": 0.030380925573417785, + "teaching_rms": 0.00511037850823289, + "traffic_rms": 0.029942293034396703 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.1820617373814457e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.6937629264741236e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008509141897097375, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 4928, + "train_examples": 45000, + "train_loss": 0.4431054111586677 + }, + { + "epoch": 15, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.7702655792236328, + "mean_feedback_forward_cosine": 0.8595048910693118, + "mean_feedback_forward_relative_error": 0.5310256089034834, + "min_feedback_forward_cosine": 0.7172211408615112 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005069581491087483, + "instruction_rms": 0.0050695814901228885, + "raw_apical_rms": 0.030655385316161254, + "teaching_rms": 0.005069581491087483, + "traffic_rms": 0.03022822864736028 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.02967562898993492, + "maximum_absolute_post_projection_soma_slope": 1.2277627803314317e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.700272145031283e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.0086457052063546, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 5280, + "train_examples": 45000, + "train_loss": 0.42360288904507953 + }, + { + "epoch": 16, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.7447105050086975, + "mean_feedback_forward_cosine": 0.8699697412942585, + "mean_feedback_forward_relative_error": 0.5104694688006451, + "min_feedback_forward_cosine": 0.7358609437942505 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005140294306780012, + "instruction_rms": 0.005140294309135916, + "raw_apical_rms": 0.03082786960100113, + "teaching_rms": 0.005140294306780012, + "traffic_rms": 0.030389686677092544 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.4030967498968039e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.6949801539126633e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008682726179291916, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 5632, + "train_examples": 45000, + "train_loss": 0.4079160744984945 + }, + { + "epoch": 17, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.7195056676864624, + "mean_feedback_forward_cosine": 0.8797186864049811, + "mean_feedback_forward_relative_error": 0.4907628203693189, + "min_feedback_forward_cosine": 0.753358006477356 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.0051695572228605815, + "instruction_rms": 0.00516955722207483, + "raw_apical_rms": 0.030909949255463046, + "teaching_rms": 0.0051695572228605815, + "traffic_rms": 0.0304688159763959 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.0237293679438153e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.693767954437601e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008533905955667173, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 5984, + "train_examples": 45000, + "train_loss": 0.3991294571240743 + }, + { + "epoch": 18, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.6956912279129028, + "mean_feedback_forward_cosine": 0.8884034878329227, + "mean_feedback_forward_relative_error": 0.47245809200562927, + "min_feedback_forward_cosine": 0.7693114876747131 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005045426520995473, + "instruction_rms": 0.005045426518698565, + "raw_apical_rms": 0.03102560302883002, + "teaching_rms": 0.005045426520995473, + "traffic_rms": 0.030606007123097266 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.029675627127289772, + "maximum_absolute_post_projection_soma_slope": 1.2978257579732144e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.683227022147006e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008293150059976776, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 6336, + "train_examples": 45000, + "train_loss": 0.38506344533496434 + }, + { + "epoch": 19, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.6718655824661255, + "mean_feedback_forward_cosine": 0.896590132462351, + "mean_feedback_forward_relative_error": 0.4547541784612756, + "min_feedback_forward_cosine": 0.7845953702926636 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.00501058974769448, + "instruction_rms": 0.005010589748090231, + "raw_apical_rms": 0.03147489895318531, + "teaching_rms": 0.00501058974769448, + "traffic_rms": 0.03106785307527604 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.02967562898993492, + "maximum_absolute_post_projection_soma_slope": 1.2295971352216384e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.683030333951218e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.008133453887915804, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 6688, + "train_examples": 45000, + "train_loss": 0.38342229491869606 + }, + { + "epoch": 20, + "feedback_tracking": { + "max_feedback_forward_relative_error": 0.6489664912223816, + "mean_feedback_forward_cosine": 0.9040563545728985, + "mean_feedback_forward_relative_error": 0.43776489637399973, + "min_feedback_forward_cosine": 0.7989962100982666 + }, + "lr": 0.1, + "mixed_apical": { + "innovation_rms": 0.005008189325268388, + "instruction_rms": 0.005008189326481963, + "raw_apical_rms": 0.03187621928975065, + "teaching_rms": 0.005008189325268388, + "traffic_rms": 0.03147469524812623 + }, + "neutral_projection": { + "instruction_observations": 0, + "maximum_absolute_correction_slope": 0.02967562898993492, + "maximum_absolute_post_projection_soma_slope": 1.2082628231269155e-08, + "maximum_observations": 128, + "maximum_post_projection_traffic_rms_ratio": 2.6886584624762698e-08, + "maximum_pre_projection_traffic_rms_ratio": 0.007884973083358886, + "minimum_observations": 72 + }, + "output_lr": 0.1, + "step": 7040, + "train_examples": 45000, + "train_loss": 0.3654292505847083 + } + ], + "evaluation_protocol": { + "test_evaluations": 0, + "test_used_for_selection": false, + "validation_evaluations": 1 + }, + "final": { + "accuracy": 0.8358, + "epoch": 20, + "evaluation_split": "validation", + "finite": true, + "loss": 0.5113223815917969, + "step": 7040 + }, + "hardware": { + "cuda_device_name": "NVIDIA GeForce GTX 1080", + "cuda_visible_devices": "5", + "device": "cuda", + "device_total_memory_bytes": 8507949056, + "peak_memory_allocated_bytes": 2165061120, + "peak_memory_reserved_bytes": 3105882112, + "torch_version": "2.3.1+cu118" + }, + "predictor_warmup": { + "closed_form_fit": { + "max_absolute_residual_soma_slope": 1.863524978773512e-08, + "max_applied_stability_margin": 0.0, + "max_positive_residual_soma_slope": 1.863524978773512e-08, + "min_residual_soma_slope": -1.666892046614521e-08, + "mse": 1.4796211947853036e-18, + "observations": 64, + "residual_traffic_rms_ratio": 7.215327154638509e-08, + "stability_margin": 0.0 + }, + "examples": 64, + "instruction_present": false, + "mode": "closed_form", + "post_warmup_traffic_residual_rms_ratio": 7.215327154638509e-08, + "reuses_traffic_calibration_forward": true, + "steps": 1, + "task_loader_state_restored": true + }, + "protocol_family": "oral_a_cifar_local_resnet_development", + "provenance": { + "git_commit": "f0fb6c7b8badc3c4556834c96ac150732d7570f1", + "git_tracked_dirty": false + }, + "schema_version": 1, + "split": { + "cifar_source_files": [ + { + "bytes": 31035704, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_1", + "sha256": "54636561a3ce25bd3e19253c6b0d8538147b0ae398331ac4a2d86c6d987368cd" + }, + { + "bytes": 31035320, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_2", + "sha256": "766b2cef9fbc745cf056b3152224f7cf77163b330ea9a15f9392beb8b89bc5a8" + }, + { + "bytes": 31035999, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_3", + "sha256": "0f00d98ebfb30b3ec0ad19f9756dc2630b89003e10525f5e148445e82aa6a1f9" + }, + { + "bytes": 31035696, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_4", + "sha256": "3f7bb240661948b8f4d53e36ec720d8306f5668bd0071dcb4e6c947f78e9682b" + }, + { + "bytes": 31035623, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/data_batch_5", + "sha256": "d91802434d8376bbaeeadf58a737e3a1b12ac839077e931237e0dcd43adcb154" + }, + { + "bytes": 31035526, + "path": "/home/yurenh2/sdrn/data/cifar-10-batches-py/test_batch", + "sha256": "f53d8d457504f7cff4ea9e021afcf0e0ad8e24a91f3fc42091b8adef61157831" + } + ], + "dataset": "cifar10", + "evaluation_split": "validation", + "input_layout": "NCHW", + "input_shape": [ + 3, + 32, + 32 + ], + "loader_seed": 0, + "normalization_mean": [ + 0.49140000343322754, + 0.4821999967098236, + 0.4465000033378601 + ], + "normalization_std": [ + 0.24699999392032623, + 0.2434999942779541, + 0.26159998774528503 + ], + "split_from_training_only": true, + "split_seed": 2027, + "test_examples": 10000, + "train_examples": 45000, + "training_augmentation": "random_crop_32_padding_4_zero_then_horizontal_flip_p0.5", + "validation_class_counts": { + "0": 500, + "1": 500, + "2": 500, + "3": 500, + "4": 500, + "5": 500, + "6": 500, + "7": 500, + "8": 500, + "9": 500 + }, + "validation_examples": 5000, + "validation_index_sha256": "8328b206a97c420e49e54e3eca4abe3274c4756b084355784ea3fb8059e4515b" + }, + "timing": { + "apical_warmup_wall_s": 0.0, + "evaluation_wall_s": 0.3624112606048584, + "mirror_warmup_wall_s": 0.0, + "predictor_warmup_wall_s": 0.032305002212524414, + "timing_excludes_data_loading_hashing_and_model_construction": true, + "total_timed_wall_s": 816.9503653049469, + "train_wall_s": 816.0330815315247 + }, + "traffic_calibration": { + "data_source": "first unaugmented development-training examples", + "examples": 64, + "instruction_rms": [ + 0.008471463806927204, + 0.006617916747927666, + 0.003988116513937712, + 0.0043865954503417015, + 0.002713605063036084, + 0.003158062230795622, + 0.002004207344725728, + 0.003757704282179475, + 0.0027085733599960804, + 0.0030200635083019733, + 0.002056022873148322, + 0.0023899395018815994, + 0.0016842411132529378, + 0.0034010226372629404, + 0.002549725119024515, + 0.0030542907770723104, + 0.0020918569061905146, + 0.002605808200314641, + 0.0018916188273578882 + ], + "realized_traffic_instruction_rms_ratio": [ + 4.0, + 4.0, + 4.0, + 4.0, + 4.0, + 4.0, + 3.999999523162842, + 4.0, + 4.000000476837158, + 3.999999761581421, + 4.0, + 4.0, + 3.999999761581421, + 4.0, + 4.000000476837158, + 4.000000476837158, + 4.0, + 4.0, + 4.0 + ], + "target_ratio": 4.0, + "traffic_gain": [ + 0.0443243570625782, + 0.035817068070173264, + 0.013778294436633587, + 0.022762222215533257, + 0.007214079145342112, + 0.017218582332134247, + 0.004616298712790012, + 0.02042512409389019, + 0.007327491883188486, + 0.01611751690506935, + 0.004703814629465342, + 0.012562948279082775, + 0.003324519144371152, + 0.01805141195654869, + 0.006063441745936871, + 0.016296451911330223, + 0.0044285994954407215, + 0.013874167576432228, + 0.0036007820162922144 + ], + "unscaled_traffic_rms": [ + 0.7644973993301392, + 0.7390796542167664, + 1.157796859741211, + 0.7708553671836853, + 1.504616141319275, + 0.7336404919624329, + 1.736635684967041, + 0.7358984351158142, + 1.478581428527832, + 0.7495108246803284, + 1.7483876943588257, + 0.7609485983848572, + 2.0264477729797363, + 0.7536302804946899, + 1.6820316314697266, + 0.7496824264526367, + 1.8894071578979492, + 0.7512690424919128, + 2.10134220123291 + ], + "uses_validation_endpoint": false + }, + "work": { + "apical_macs_per_example": 40108672, + "causal_scalar_observations": 0, + "components": { + "apical_projection_macs": 36100371755008, + "apical_regression_macs": 0, + "bp_reverse_macs_estimate": 0, + "kp_reciprocal_correlation_macs": 36097804800000, + "local_weight_correlation_macs": 36495936000000, + "mirror_feedback_prediction_macs": 0, + "mirror_local_correlation_macs": 0, + "mirror_response_macs": 0, + "ordinary_forward_macs": 36495936000000, + "perturbation_forward_macs": 0, + "warmup_clean_forward_macs": 2595266560 + }, + "definition": "multiply-accumulates in conv/linear maps; one local weight correlation equals one forward-weight MAC count; BP reverse is estimated as one weight-gradient plus one activation-gradient convolution per forward convolution; mixed-traffic/predictor elementwise arithmetic is reported as a conservative operation estimate separately and is not folded into MACs", + "elementwise_operations_estimate": 4070171475968, + "forward_macs_per_example": 40551040, + "logical_batch_loss_queries": 0, + "mirror_probe_examples": 0, + "mirror_readout_probe_examples": 0, + "neutral_projection_observations": 900000, + "per_example_cross_entropy_terms": 0, + "total_clean_forward_examples": 900064, + "total_forward_equivalent_examples": 900064, + "total_macs_estimate": 145192643821568 + } +} |