From e60afe651eb1fbe60397413af4d0df43cf22f1d0 Mon Sep 17 00:00:00 2001 From: YurenHao0426 Date: Wed, 22 Jul 2026 02:00:18 -0500 Subject: results: record failed useful-depth BP screens --- results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json | 1 + 1 file changed, 1 insertion(+) create mode 100644 results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json (limited to 'results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json') diff --git a/results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json b/results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json new file mode 100644 index 0000000..00a511a --- /dev/null +++ b/results/depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "bp", "dataset": "hierarchical", "depth": 1, "width": 16, "act": "tanh", "residual": 1, "epochs": 25, "batch_size": 128, "train_examples": 0, "val_examples": 5000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 50000, "task_test_examples": 10000, "task_levels": 6, "task_n_in": 128, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "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, "traffic_mode": "soma", "predictor_mode": "diagonal", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 200, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "depthtask_dev_v1_hierarchical_bp_w16_d1_t0_s0", "n_in": 64, "n_out": 10}, "split": {"dataset": "hierarchical", "task_seed": 0, "train_examples": 45000, "test_examples": 10000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 5000, "validation_index_sha256": "f5282ed201086357dbe24c23e70cc9eb86deaaf70d99c41553ffc78626ee8d6c", "validation_class_counts": {"0": 500, "1": 500, "2": 500, "3": 500, "4": 500, "5": 500, "6": 500, "7": 500, "8": 500, "9": 500}, "split_from_training_only": true}, "provenance": {"git_commit": "a0696a3d72cc5bf69876e1716b74f7e1ac525dd0", "git_dirty": false}, "steps": [{"step": 0, "epoch": 0, "train_loss": 2.475914478302002}, {"step": 200, "epoch": 0, "train_loss": 1.7467397451400757}, {"epoch_end": 0, "step": 352, "eval_split": "validation", "eval_acc": 0.33, "eval_loss": 1.9074928955078125, "val_acc": 0.33, "val_loss": 1.9074928955078125}, {"step": 400, "epoch": 1, "train_loss": 1.8253108263015747}, {"step": 600, "epoch": 1, "train_loss": 1.7261707782745361}, {"epoch_end": 1, "step": 704, "eval_split": "validation", "eval_acc": 0.3454, "eval_loss": 1.87873603515625, "val_acc": 0.3454, "val_loss": 1.87873603515625}, {"step": 800, "epoch": 2, "train_loss": 1.674320101737976}, {"step": 1000, "epoch": 2, "train_loss": 1.7929267883300781}, {"epoch_end": 2, "step": 1056, "eval_split": "validation", "eval_acc": 0.355, "eval_loss": 1.850813330078125, "val_acc": 0.355, "val_loss": 1.850813330078125}, {"step": 1200, "epoch": 3, "train_loss": 1.5949172973632812}, {"step": 1400, "epoch": 3, "train_loss": 1.705915927886963}, {"epoch_end": 3, "step": 1408, "eval_split": "validation", "eval_acc": 0.359, "eval_loss": 1.8394200439453126, "val_acc": 0.359, "val_loss": 1.8394200439453126}, {"step": 1600, "epoch": 4, "train_loss": 1.6426708698272705}, {"epoch_end": 4, "step": 1760, "eval_split": "validation", "eval_acc": 0.373, "eval_loss": 1.8107114990234374, "val_acc": 0.373, "val_loss": 1.8107114990234374}, {"step": 1800, "epoch": 5, "train_loss": 1.849348783493042}, {"step": 2000, "epoch": 5, "train_loss": 1.6563369035720825}, {"epoch_end": 5, "step": 2112, "eval_split": "validation", "eval_acc": 0.3668, "eval_loss": 1.800864599609375, "val_acc": 0.3668, "val_loss": 1.800864599609375}, {"step": 2200, "epoch": 6, "train_loss": 1.6335431337356567}, {"step": 2400, "epoch": 6, "train_loss": 1.633534550666809}, {"epoch_end": 6, "step": 2464, "eval_split": "validation", "eval_acc": 0.372, "eval_loss": 1.805021435546875, "val_acc": 0.372, "val_loss": 1.805021435546875}, {"step": 2600, "epoch": 7, "train_loss": 1.6962227821350098}, {"step": 2800, "epoch": 7, "train_loss": 1.7465953826904297}, {"epoch_end": 7, "step": 2816, "eval_split": "validation", "eval_acc": 0.3724, "eval_loss": 1.784698681640625, "val_acc": 0.3724, "val_loss": 1.784698681640625}, {"step": 3000, "epoch": 8, "train_loss": 1.6518275737762451}, {"epoch_end": 8, "step": 3168, "eval_split": "validation", "eval_acc": 0.3758, "eval_loss": 1.77872216796875, "val_acc": 0.3758, "val_loss": 1.77872216796875}, {"step": 3200, "epoch": 9, "train_loss": 1.69846510887146}, {"step": 3400, "epoch": 9, "train_loss": 1.5753906965255737}, {"epoch_end": 9, "step": 3520, "eval_split": "validation", "eval_acc": 0.3742, "eval_loss": 1.7770663330078125, "val_acc": 0.3742, "val_loss": 1.7770663330078125}, {"step": 3600, "epoch": 10, "train_loss": 1.5937199592590332}, {"step": 3800, "epoch": 10, "train_loss": 1.4980002641677856}, {"epoch_end": 10, "step": 3872, "eval_split": "validation", "eval_acc": 0.378, "eval_loss": 1.780496044921875, "val_acc": 0.378, "val_loss": 1.780496044921875}, {"step": 4000, "epoch": 11, "train_loss": 1.6196931600570679}, {"step": 4200, "epoch": 11, "train_loss": 1.6821643114089966}, {"epoch_end": 11, "step": 4224, "eval_split": "validation", "eval_acc": 0.3776, "eval_loss": 1.7710059326171874, "val_acc": 0.3776, "val_loss": 1.7710059326171874}, {"step": 4400, "epoch": 12, "train_loss": 1.814131736755371}, {"epoch_end": 12, "step": 4576, "eval_split": "validation", "eval_acc": 0.3782, "eval_loss": 1.783114013671875, "val_acc": 0.3782, "val_loss": 1.783114013671875}, {"step": 4600, "epoch": 13, "train_loss": 1.5792200565338135}, {"step": 4800, "epoch": 13, "train_loss": 1.7093915939331055}, {"epoch_end": 13, "step": 4928, "eval_split": "validation", "eval_acc": 0.3766, "eval_loss": 1.7789003662109375, "val_acc": 0.3766, "val_loss": 1.7789003662109375}, {"step": 5000, "epoch": 14, "train_loss": 1.3969545364379883}, {"step": 5200, "epoch": 14, "train_loss": 1.6407215595245361}, {"epoch_end": 14, "step": 5280, "eval_split": "validation", "eval_acc": 0.3738, "eval_loss": 1.785031591796875, "val_acc": 0.3738, "val_loss": 1.785031591796875}, {"step": 5400, "epoch": 15, "train_loss": 1.820623755455017}, {"step": 5600, "epoch": 15, "train_loss": 1.5520126819610596}, {"epoch_end": 15, "step": 5632, "eval_split": "validation", "eval_acc": 0.381, "eval_loss": 1.7722260009765625, "val_acc": 0.381, "val_loss": 1.7722260009765625}, {"step": 5800, "epoch": 16, "train_loss": 1.5525339841842651}, {"epoch_end": 16, "step": 5984, "eval_split": "validation", "eval_acc": 0.38, "eval_loss": 1.7807119873046875, "val_acc": 0.38, "val_loss": 1.7807119873046875}, {"step": 6000, "epoch": 17, "train_loss": 1.5396584272384644}, {"step": 6200, "epoch": 17, "train_loss": 1.5190238952636719}, {"epoch_end": 17, "step": 6336, "eval_split": "validation", "eval_acc": 0.3772, "eval_loss": 1.7760913330078125, "val_acc": 0.3772, "val_loss": 1.7760913330078125}, {"step": 6400, "epoch": 18, "train_loss": 1.4686918258666992}, {"step": 6600, "epoch": 18, "train_loss": 1.5981621742248535}, {"epoch_end": 18, "step": 6688, "eval_split": "validation", "eval_acc": 0.379, "eval_loss": 1.786067041015625, "val_acc": 0.379, "val_loss": 1.786067041015625}, {"step": 6800, "epoch": 19, "train_loss": 1.6483367681503296}, {"step": 7000, "epoch": 19, "train_loss": 1.4933700561523438}, {"epoch_end": 19, "step": 7040, "eval_split": "validation", "eval_acc": 0.3828, "eval_loss": 1.7674142333984375, "val_acc": 0.3828, "val_loss": 1.7674142333984375}, {"step": 7200, "epoch": 20, "train_loss": 1.6544252634048462}, {"epoch_end": 20, "step": 7392, "eval_split": "validation", "eval_acc": 0.3792, "eval_loss": 1.776386279296875, "val_acc": 0.3792, "val_loss": 1.776386279296875}, {"step": 7400, "epoch": 21, "train_loss": 1.572379231452942}, {"step": 7600, "epoch": 21, "train_loss": 1.5655951499938965}, {"epoch_end": 21, "step": 7744, "eval_split": "validation", "eval_acc": 0.379, "eval_loss": 1.7804651611328124, "val_acc": 0.379, "val_loss": 1.7804651611328124}, {"step": 7800, "epoch": 22, "train_loss": 1.6244573593139648}, {"step": 8000, "epoch": 22, "train_loss": 1.6154931783676147}, {"epoch_end": 22, "step": 8096, "eval_split": "validation", "eval_acc": 0.385, "eval_loss": 1.776894580078125, "val_acc": 0.385, "val_loss": 1.776894580078125}, {"step": 8200, "epoch": 23, "train_loss": 1.4491068124771118}, {"step": 8400, "epoch": 23, "train_loss": 1.7100651264190674}, {"epoch_end": 23, "step": 8448, "eval_split": "validation", "eval_acc": 0.3818, "eval_loss": 1.7643829345703126, "val_acc": 0.3818, "val_loss": 1.7643829345703126}, {"step": 8600, "epoch": 24, "train_loss": 1.5250614881515503}, {"epoch_end": 24, "step": 8800, "eval_split": "validation", "eval_acc": 0.3852, "eval_loss": 1.761310400390625, "val_acc": 0.3852, "val_loss": 1.761310400390625}], "final": {"eval_split": "validation", "eval_acc": 0.3852, "eval_loss": 1.761310400390625, "wall_s": 13.6634361743927, "val_acc": 0.3852, "val_loss": 1.761310400390625}} \ No newline at end of file -- cgit v1.2.3