diff options
20 files changed, 49 insertions, 0 deletions
@@ -405,6 +405,30 @@ targets do not by themselves cure the coupled A/W basin.** The next bounded deve is temporal: calibrate A while W is held fixed, then open forward plasticity and maintain A with the already selected low-query simultaneous schedule. +### Feedback-first timescale development screen (task seed 0 only) + +The next frozen screen prefixed the original context-SDIL K1/e4 simultaneous joint phase with +`{0, 100, 400}` A-only layerwise K4 calibration minibatches. W, biases, and the output readout were +bitwise frozen during the prefix; its isolated RNG and restored loader state kept the subsequent +joint-phase randomness and minibatch order matched. Advancement required mean d4 validation of at +least 90%, at least 2/3 positive depth gains and lesions, a mean lesion drop of at least 2 points, +and all finite losses. + +| A-only steps | depth | validation acc (%) | depth gain | d4 lesion drop | d4 work / ordinary | +|-------------:|------:|-------------------:|-----------:|----------------:|-------------------:| +| 0 | 1 | 66.733 ± 6.147 | — | — | — | +| 0 | 4 | 78.633 ± 24.866 | +11.900 | +27.367 | 1.8x | +| 100 | 1 | 63.850 ± 1.576 | — | — | — | +| 100 | 4 | 81.867 ± 8.779 | +18.017 | +14.533 | 2.5x | +| 400 | 1 | 64.300 ± 3.695 | — | — | — | +| 400 | 4 | 82.450 ± 14.226 | +18.150 | +25.367 | 4.7x | + +Both nonzero prefixes produce positive gains and lesions in 3/3 seeds and substantially reduce the +zero-prefix variance, so timescale separation addresses part of the basin problem. Neither reaches +the frozen 90% d4 threshold; **no prefix is eligible and task seeds 6--8 remain untouched.** Final +teaching directions are often strongly aligned even in mediocre runs, pointing next to regression +gain/tracking over the moving trajectory rather than simply adding vectorizer capacity. + ## Paper-faithful non-backprop baselines `PETITE` in the initial request is treated as **PEPITA**. Implementations and hyperparameters were @@ -194,6 +194,13 @@ on average with at least 2/3 positive lesions. Among eligible prefixes, select t endpoint, choosing the shortest prefix within one point. If none is eligible, the feedback-first branch closes without using task seeds 6--8. +**Feedback-first status: failed development on 2026-07-22.** The 100- and 400-step prefixes raised +mean d4 validation from `78.633 ± 24.866%` to `81.867 ± 8.779%` and `82.450 ± 14.226%`; both had +3/3 positive paired depth gains and lesions, versus 2/3 without warmup. Their mean gains were about +18 points and all losses were finite. Neither met the frozen 90% endpoint threshold, so no prefix +advances and task seeds 6--8 remain unused. Timescale separation reduces basin variance but does +not make plain SGD regression of the moving apical map accurate enough. + ### C3. Causal calibration is genuinely amortized **Status: passed on 2026-07-22.** The frozen five-seed confirmation retained 112.9% of the K16/e4 diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s0.json new file mode 100644 index 0000000..92d6dc2 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7099499106407166}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.6415, "eval_loss": 0.6132500305175781, "wall_s": 6.414417743682861, "val_acc": 0.6415, "val_loss": 0.6132500305175781, "cos_r_negg": [0.9992589950561523], "cos_innovation_negg": [0.9992589950561523], "cos_apical_negg": [0.9992589950561523], "cos_Ac_negg": [0.9992589950561523], "r_norm": [3.2737247943878174], "g_norm": [0.007440473418682814], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 6.313450336456299, "diagnostics_wall_s": 0.06766247749328613, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 6.313450336456299, "evaluation_wall_s": 0.027631759643554688}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17396736, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s1.json new file mode 100644 index 0000000..e3fb092 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6922826766967773}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.623, "eval_loss": 0.6227515258789063, "wall_s": 4.676515340805054, "val_acc": 0.623, "val_loss": 0.6227515258789063, "cos_r_negg": [0.9995909333229065], "cos_innovation_negg": [0.9995909333229065], "cos_apical_negg": [0.9995909333229065], "cos_Ac_negg": [0.9995909333229065], "r_norm": [3.083123207092285], "g_norm": [0.00701596774160862], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 4.596850872039795, "diagnostics_wall_s": 0.049974679946899414, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.596850872039795, "evaluation_wall_s": 0.025023937225341797}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17396736, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s2.json new file mode 100644 index 0000000..1f870de --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d1_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7055632472038269}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.7375, "eval_loss": 0.6094509887695313, "wall_s": 4.89094090461731, "val_acc": 0.7375, "val_loss": 0.6094509887695313, "cos_r_negg": [0.9986045956611633], "cos_innovation_negg": [0.9986045956611633], "cos_apical_negg": [0.9986045956611633], "cos_Ac_negg": [0.9986045956611633], "r_norm": [2.9741599559783936], "g_norm": [0.006628120318055153], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 4.810322284698486, "diagnostics_wall_s": 0.05147957801818848, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.810322284698486, "evaluation_wall_s": 0.02427506446838379}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17396736, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s0.json new file mode 100644 index 0000000..8490de8 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6903698444366455}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.948, "eval_loss": 0.11977893829345704, "wall_s": 13.397622108459473, "val_acc": 0.948, "val_loss": 0.11977893829345704, "cos_r_negg": [0.5562907457351685, 0.808923602104187, 0.5909024477005005, 0.9932405352592468], "cos_innovation_negg": [0.5562907457351685, 0.808923602104187, 0.5909024477005005, 0.9932405352592468], "cos_apical_negg": [0.5562907457351685, 0.808923602104187, 0.5909024477005005, 0.9932405352592468], "cos_Ac_negg": [0.5562907457351685, 0.808923602104187, 0.5909024477005005, 0.9932405352592468], "r_norm": [1.6527658700942993, 1.5863852500915527, 1.3194326162338257, 0.8883618116378784], "g_norm": [0.007319841533899307, 0.005622733384370804, 0.0037784245796501637, 0.0022349508944898844], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.7151630520820618, 1.013958215713501, 0.7244371175765991], "lesion_eval_acc": 0.538, "lesion_eval_loss": 4.938269287109375, "lesion_acc_drop": 0.4099999999999999}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 13.280044794082642, "diagnostics_wall_s": 0.07265901565551758, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 13.280044794082642, "evaluation_wall_s": 0.03963494300842285}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17590784, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s1.json new file mode 100644 index 0000000..59396fc --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7494125366210938}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.911, "eval_loss": 0.22712201690673828, "wall_s": 12.26870584487915, "val_acc": 0.911, "val_loss": 0.22712201690673828, "cos_r_negg": [0.5901073217391968, 0.5873908996582031, 0.8030260801315308, 0.9765181541442871], "cos_innovation_negg": [0.5901073217391968, 0.5873908996582031, 0.8030260801315308, 0.9765181541442871], "cos_apical_negg": [0.5901073217391968, 0.5873908996582031, 0.8030260801315308, 0.9765181541442871], "cos_Ac_negg": [0.5901073217391968, 0.5873908996582031, 0.8030260801315308, 0.9765181541442871], "r_norm": [1.7188761234283447, 2.008864402770996, 1.09437096118927, 0.7894454002380371], "g_norm": [0.00884522870182991, 0.007164946757256985, 0.004389304667711258, 0.0022060831543058157], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.7472031712532043, 0.7504488229751587, 1.334064245223999], "lesion_eval_acc": 0.5, "lesion_eval_loss": 4.7154375, "lesion_acc_drop": 0.41100000000000003}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 12.14437198638916, "diagnostics_wall_s": 0.07507133483886719, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 12.14437198638916, "evaluation_wall_s": 0.044519901275634766}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17590784, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s2.json new file mode 100644 index 0000000..40903fd --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 0, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a0_d4_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6906555891036987}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.5, "eval_loss": 1189.9298125, "wall_s": 14.39193606376648, "val_acc": 0.5, "val_loss": 1189.9298125, "cos_r_negg": [0.9909342527389526, NaN, NaN, NaN], "cos_innovation_negg": [0.9909342527389526, NaN, NaN, NaN], "cos_apical_negg": [0.9909342527389526, NaN, NaN, NaN], "cos_Ac_negg": [0.9909342527389526, NaN, NaN, NaN], "r_norm": [142.14981079101562, 143.88719177246094, 145.9816436767578, 143.39907836914062], "g_norm": [0.3087597191333771, 0.3087597191333771, 0.3087597191333771, 0.3087597191333771], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.0, 0.0, 0.0], "lesion_eval_acc": 0.5, "lesion_eval_loss": 1189.9298125, "lesion_acc_drop": 0.0}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.0, "training_loop_wall_s": 14.278030395507812, "diagnostics_wall_s": 0.06841564178466797, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 14.278030395507812, "evaluation_wall_s": 0.04071497917175293}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 0, "feedback_warmup_perturbation_events": 0, "perturbation_events": 640, "calibration_batch_loss_evaluations": 1280, "calibration_example_loss_evaluations": 327680, "calibration_forward_equivalent_examples": 491520.0, "training_forward_equivalent_examples": 1131520.0, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17590784, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s0.json new file mode 100644 index 0000000..c02f0c6 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7099499106407166}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.638, "eval_loss": 0.6124363403320312, "wall_s": 4.912625789642334, "val_acc": 0.638, "val_loss": 0.6124363403320312, "cos_r_negg": [0.9993677735328674], "cos_innovation_negg": [0.9993677735328674], "cos_apical_negg": [0.9993677735328674], "cos_Ac_negg": [0.9993677735328674], "r_norm": [3.348402738571167], "g_norm": [0.007629218511283398], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.3428366184234619, "training_loop_wall_s": 4.49614953994751, "diagnostics_wall_s": 0.044576406478881836, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.49614953994751, "evaluation_wall_s": 0.0242769718170166}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 2180, "calibration_example_loss_evaluations": 552896, "calibration_forward_equivalent_examples": 650005.3333333334, "training_forward_equivalent_examples": 1315029.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s1.json new file mode 100644 index 0000000..c8b168d --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6922826766967773}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.623, "eval_loss": 0.6174711608886718, "wall_s": 4.757853746414185, "val_acc": 0.623, "val_loss": 0.6174711608886718, "cos_r_negg": [0.9996559619903564], "cos_innovation_negg": [0.9996559619903564], "cos_apical_negg": [0.9996559619903564], "cos_Ac_negg": [0.9996559619903564], "r_norm": [3.327286720275879], "g_norm": [0.007428172510117292], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.36191415786743164, "training_loop_wall_s": 4.315869569778442, "diagnostics_wall_s": 0.049189090728759766, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.315869569778442, "evaluation_wall_s": 0.02572345733642578}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 2180, "calibration_example_loss_evaluations": 552896, "calibration_forward_equivalent_examples": 650005.3333333334, "training_forward_equivalent_examples": 1315029.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s2.json new file mode 100644 index 0000000..7bfdfc3 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d1_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7055632472038269}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.6545, "eval_loss": 0.6077652587890625, "wall_s": 4.7988903522491455, "val_acc": 0.6545, "val_loss": 0.6077652587890625, "cos_r_negg": [0.999733030796051], "cos_innovation_negg": [0.999733030796051], "cos_apical_negg": [0.999733030796051], "cos_Ac_negg": [0.999733030796051], "r_norm": [2.898632526397705], "g_norm": [0.006760832853615284], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 0.35747814178466797, "training_loop_wall_s": 4.3664233684539795, "diagnostics_wall_s": 0.04615926742553711, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.3664233684539795, "evaluation_wall_s": 0.024277448654174805}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 2180, "calibration_example_loss_evaluations": 552896, "calibration_forward_equivalent_examples": 650005.3333333334, "training_forward_equivalent_examples": 1315029.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s0.json new file mode 100644 index 0000000..5442db3 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6903698444366455}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.836, "eval_loss": 0.3396922149658203, "wall_s": 13.415709018707275, "val_acc": 0.836, "val_loss": 0.3396922149658203, "cos_r_negg": [0.6421730518341064, 0.8698056936264038, 0.6740834712982178, 0.9885815382003784], "cos_innovation_negg": [0.6421730518341064, 0.8698056936264038, 0.6740834712982178, 0.9885815382003784], "cos_apical_negg": [0.6421730518341064, 0.8698056936264038, 0.6740834712982178, 0.9885815382003784], "cos_Ac_negg": [0.6421730518341064, 0.8698056936264038, 0.6740834712982178, 0.9885815382003784], "r_norm": [1.9577410221099854, 1.9897119998931885, 1.786451816558838, 1.6524524688720703], "g_norm": [0.008239038288593292, 0.005497160367667675, 0.004555549006909132, 0.004154927562922239], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [1.3100138902664185, 1.3826110363006592, 0.7177157402038574], "lesion_eval_acc": 0.59, "lesion_eval_loss": 3.243130615234375, "lesion_acc_drop": 0.246}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.3036799430847168, "training_loop_wall_s": 11.982349634170532, "diagnostics_wall_s": 0.08773350715637207, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 11.982349634170532, "evaluation_wall_s": 0.0371394157409668}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 4580, "calibration_example_loss_evaluations": 1153472, "calibration_forward_equivalent_examples": 931757.0370370373, "training_forward_equivalent_examples": 1596781.0370370373, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s1.json new file mode 100644 index 0000000..bb5904b --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7494125366210938}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.8965, "eval_loss": 0.26321958923339844, "wall_s": 14.164005279541016, "val_acc": 0.8965, "val_loss": 0.26321958923339844, "cos_r_negg": [0.5790196657180786, 0.9826145172119141, 0.9802991151809692, 0.9333117604255676], "cos_innovation_negg": [0.5790196657180786, 0.9826145172119141, 0.9802991151809692, 0.9333117604255676], "cos_apical_negg": [0.5790196657180786, 0.9826145172119141, 0.9802991151809692, 0.9333117604255676], "cos_Ac_negg": [0.5790196657180786, 0.9826145172119141, 0.9802991151809692, 0.9333117604255676], "r_norm": [3.7277328968048096, 3.244243621826172, 1.2322155237197876, 0.5262465476989746], "g_norm": [0.006530542392283678, 0.004511868581175804, 0.0014816568000242114, 0.0009873760864138603], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.7218430042266846, 1.0539159774780273, 1.0848747491836548], "lesion_eval_acc": 0.805, "lesion_eval_loss": 0.39960140991210935, "lesion_acc_drop": 0.09149999999999991}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.4164044857025146, "training_loop_wall_s": 12.635356426239014, "diagnostics_wall_s": 0.06795239448547363, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 12.635356426239014, "evaluation_wall_s": 0.039632558822631836}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 4580, "calibration_example_loss_evaluations": 1153472, "calibration_forward_equivalent_examples": 931757.0370370373, "training_forward_equivalent_examples": 1596781.0370370373, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s2.json new file mode 100644 index 0000000..70c0844 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 100, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a100_d4_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6906555891036987}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.7235, "eval_loss": 0.5526383361816406, "wall_s": 13.955256223678589, "val_acc": 0.7235, "val_loss": 0.5526383361816406, "cos_r_negg": [0.4239192306995392, 0.3029614984989166, 0.5967837572097778, 0.9963654279708862], "cos_innovation_negg": [0.4239192306995392, 0.3029614984989166, 0.5967837572097778, 0.9963654279708862], "cos_apical_negg": [0.4239192306995392, 0.3029614984989166, 0.5967837572097778, 0.9963654279708862], "cos_Ac_negg": [0.4239192306995392, 0.3029614984989166, 0.5967837572097778, 0.9963654279708862], "r_norm": [0.7267181873321533, 0.825041651725769, 0.8316144943237305, 2.8162195682525635], "g_norm": [0.003495941637083888, 0.0031639153603464365, 0.0024316587951034307, 0.005627923645079136], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.7577447295188904, 0.8058190941810608, 0.8908174633979797], "lesion_eval_acc": 0.625, "lesion_eval_loss": 4.140706787109375, "lesion_acc_drop": 0.09850000000000003}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.328120231628418, "training_loop_wall_s": 12.49913215637207, "diagnostics_wall_s": 0.08014345169067383, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 12.49913215637207, "evaluation_wall_s": 0.04314756393432617}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 25024, "feedback_warmup_perturbation_events": 100, "perturbation_events": 740, "calibration_batch_loss_evaluations": 4580, "calibration_example_loss_evaluations": 1153472, "calibration_forward_equivalent_examples": 931757.0370370373, "training_forward_equivalent_examples": 1596781.0370370373, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s0.json new file mode 100644 index 0000000..1f81cdd --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7099499106407166}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.6185, "eval_loss": 0.5636443176269531, "wall_s": 5.681481122970581, "val_acc": 0.6185, "val_loss": 0.5636443176269531, "cos_r_negg": [0.9994745850563049], "cos_innovation_negg": [0.9994745850563049], "cos_apical_negg": [0.9994745850563049], "cos_Ac_negg": [0.9994745850563049], "r_norm": [3.710599422454834], "g_norm": [0.008738419972360134], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.1123895645141602, "training_loop_wall_s": 4.481980085372925, "diagnostics_wall_s": 0.05808901786804199, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.481980085372925, "evaluation_wall_s": 0.024379491806030273}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 4880, "calibration_example_loss_evaluations": 1228544, "calibration_forward_equivalent_examples": 1125461.3333333335, "training_forward_equivalent_examples": 1865557.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s1.json new file mode 100644 index 0000000..da982ce --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6922826766967773}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.625, "eval_loss": 0.6138421020507813, "wall_s": 5.645373582839966, "val_acc": 0.625, "val_loss": 0.6138421020507813, "cos_r_negg": [0.9996894598007202], "cos_innovation_negg": [0.9996894598007202], "cos_apical_negg": [0.9996894598007202], "cos_Ac_negg": [0.9996894598007202], "r_norm": [3.399834632873535], "g_norm": [0.007550141774117947], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.1715049743652344, "training_loop_wall_s": 4.390048503875732, "diagnostics_wall_s": 0.05119681358337402, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.390048503875732, "evaluation_wall_s": 0.02759408950805664}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 4880, "calibration_example_loss_evaluations": 1228544, "calibration_forward_equivalent_examples": 1125461.3333333335, "training_forward_equivalent_examples": 1865557.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s2.json new file mode 100644 index 0000000..d3f1875 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 1, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.0, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d1_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7055632472038269}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.6855, "eval_loss": 0.613520751953125, "wall_s": 5.811570405960083, "val_acc": 0.6855, "val_loss": 0.613520751953125, "cos_r_negg": [0.9990178346633911], "cos_innovation_negg": [0.9990178346633911], "cos_apical_negg": [0.9990178346633911], "cos_Ac_negg": [0.9990178346633911], "r_norm": [3.0431923866271973], "g_norm": [0.00692722387611866], "traffic_norm": [0.0], "traffic_residual_norm": [0.0], "traffic_r2": [NaN]}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 1.1430072784423828, "training_loop_wall_s": 4.582326173782349, "diagnostics_wall_s": 0.05652499198913574, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 4.582326173782349, "evaluation_wall_s": 0.024856090545654297}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 4880, "calibration_example_loss_evaluations": 1228544, "calibration_forward_equivalent_examples": 1125461.3333333335, "training_forward_equivalent_examples": 1865557.3333333335, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17401856, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s0.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s0.json new file mode 100644 index 0000000..bd45a57 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s0.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 0, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s0", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6903698444366455}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.9675, "eval_loss": 0.2511512680053711, "wall_s": 17.478212594985962, "val_acc": 0.9675, "val_loss": 0.2511512680053711, "cos_r_negg": [0.3596605956554413, 0.4484979510307312, 0.5354673266410828, 0.9950237274169922], "cos_innovation_negg": [0.3596605956554413, 0.4484979510307312, 0.5354673266410828, 0.9950237274169922], "cos_apical_negg": [0.3596605956554413, 0.4484979510307312, 0.5354673266410828, 0.9950237274169922], "cos_Ac_negg": [0.3596605956554413, 0.4484979510307312, 0.5354673266410828, 0.9950237274169922], "r_norm": [2.469841480255127, 4.146439552307129, 2.2034473419189453, 2.3239712715148926], "g_norm": [0.010294299572706223, 0.009126340970396996, 0.0066740307956933975, 0.005083840806037188], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.4168105721473694, 1.6523725986480713, 0.16552601754665375], "lesion_eval_acc": 0.6115, "lesion_eval_loss": 0.6619359741210937, "lesion_acc_drop": 0.356}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 5.084607362747192, "training_loop_wall_s": 12.281693696975708, "diagnostics_wall_s": 0.0680837631225586, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 12.281693696975708, "evaluation_wall_s": 0.03884577751159668}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 14480, "calibration_example_loss_evaluations": 3630848, "calibration_forward_equivalent_examples": 2252468.148148131, "training_forward_equivalent_examples": 2992564.148148131, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s1.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s1.json new file mode 100644 index 0000000..64aff01 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s1", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.7494125366210938}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.683, "eval_loss": 0.43675697326660157, "wall_s": 22.482528686523438, "val_acc": 0.683, "val_loss": 0.43675697326660157, "cos_r_negg": [0.8813283443450928, 0.9790688753128052, 0.9448697566986084, 0.9855836629867554], "cos_innovation_negg": [0.8813283443450928, 0.9790688753128052, 0.9448697566986084, 0.9855836629867554], "cos_apical_negg": [0.8813283443450928, 0.9790688753128052, 0.9448697566986084, 0.9855836629867554], "cos_Ac_negg": [0.8813283443450928, 0.9790688753128052, 0.9448697566986084, 0.9855836629867554], "r_norm": [1.5890116691589355, 2.2004318237304688, 2.2835545539855957, 2.5675182342529297], "g_norm": [0.004544874653220177, 0.005498669110238552, 0.005058149807155132, 0.0051698326133191586], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.4748292863368988, 1.007644772529602, 0.3159356117248535], "lesion_eval_acc": 0.601, "lesion_eval_loss": 1.2396192016601562, "lesion_acc_drop": 0.08200000000000007}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 6.737584590911865, "training_loop_wall_s": 15.634146690368652, "diagnostics_wall_s": 0.0657193660736084, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 15.634146690368652, "evaluation_wall_s": 0.0398564338684082}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 14480, "calibration_example_loss_evaluations": 3630848, "calibration_forward_equivalent_examples": 2252468.148148131, "training_forward_equivalent_examples": 2992564.148148131, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file diff --git a/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s2.json b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s2.json new file mode 100644 index 0000000..86460d0 --- /dev/null +++ b/results/c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s2.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "tentmap", "depth": 4, "width": 8, "act": "relu", "residual": 1, "residual_lesion_fraction": 0.3333333333333333, "epochs": 80, "batch_size": 256, "train_examples": 0, "val_examples": 2000, "split_seed": 2027, "eval_split": "validation", "task_seed": 0, "task_train_examples": 10000, "task_test_examples": 5000, "task_levels": 2, "task_n_in": 1, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.03, "eta_A": 0.01, "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": 1, "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, "a_warmup_steps": 400, "a_warmup_ndirs": 4, "a_warmup_mode": "layerwise", "nuis_rho": 0.0, "traffic_seed": 1234, "traffic_mode": "none", "predictor_mode": "diagonal", "vectorizer_mode": "context_gated", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 2, "device": "cuda", "log_every": 100000, "diagnostics": "alignment", "diagnostics_schedule": "final", "eval_every": 0, "max_steps": 0, "probe_bs": 512, "outdir": "results", "tag": "c2_awarmdev_v1_tent_l2_w8_context_a400_d4_t0_s2", "n_in": 1, "n_out": 2}, "split": {"dataset": "tentmap", "task_seed": 0, "train_examples": 8000, "test_examples": 5000, "evaluation_split": "validation", "synthetic_generator": true, "split_seed": 2027, "validation_examples": 2000, "validation_index_sha256": "f292d9bfeef1b6e79c80d4cf1c572ac12dee1be799a71a126390709d0386c6e4", "validation_class_counts": {"0": 1000, "1": 1000}, "split_from_training_only": true}, "provenance": {"git_commit": "2f7a093deec83dc6f972d4aaff3f3d55e8327925", "git_dirty": false}, "diagnostic_protocol": {"probe_source": "training_prefix", "probe_examples": 512, "schedule": "final"}, "steps": [{"step": 0, "epoch": 0, "train_loss": 0.6906555891036987}, {"epoch_end": 0, "step": 32}, {"epoch_end": 1, "step": 64}, {"epoch_end": 2, "step": 96}, {"epoch_end": 3, "step": 128}, {"epoch_end": 4, "step": 160}, {"epoch_end": 5, "step": 192}, {"epoch_end": 6, "step": 224}, {"epoch_end": 7, "step": 256}, {"epoch_end": 8, "step": 288}, {"epoch_end": 9, "step": 320}, {"epoch_end": 10, "step": 352}, {"epoch_end": 11, "step": 384}, {"epoch_end": 12, "step": 416}, {"epoch_end": 13, "step": 448}, {"epoch_end": 14, "step": 480}, {"epoch_end": 15, "step": 512}, {"epoch_end": 16, "step": 544}, {"epoch_end": 17, "step": 576}, {"epoch_end": 18, "step": 608}, {"epoch_end": 19, "step": 640}, {"epoch_end": 20, "step": 672}, {"epoch_end": 21, "step": 704}, {"epoch_end": 22, "step": 736}, {"epoch_end": 23, "step": 768}, {"epoch_end": 24, "step": 800}, {"epoch_end": 25, "step": 832}, {"epoch_end": 26, "step": 864}, {"epoch_end": 27, "step": 896}, {"epoch_end": 28, "step": 928}, {"epoch_end": 29, "step": 960}, {"epoch_end": 30, "step": 992}, {"epoch_end": 31, "step": 1024}, {"epoch_end": 32, "step": 1056}, {"epoch_end": 33, "step": 1088}, {"epoch_end": 34, "step": 1120}, {"epoch_end": 35, "step": 1152}, {"epoch_end": 36, "step": 1184}, {"epoch_end": 37, "step": 1216}, {"epoch_end": 38, "step": 1248}, {"epoch_end": 39, "step": 1280}, {"epoch_end": 40, "step": 1312}, {"epoch_end": 41, "step": 1344}, {"epoch_end": 42, "step": 1376}, {"epoch_end": 43, "step": 1408}, {"epoch_end": 44, "step": 1440}, {"epoch_end": 45, "step": 1472}, {"epoch_end": 46, "step": 1504}, {"epoch_end": 47, "step": 1536}, {"epoch_end": 48, "step": 1568}, {"epoch_end": 49, "step": 1600}, {"epoch_end": 50, "step": 1632}, {"epoch_end": 51, "step": 1664}, {"epoch_end": 52, "step": 1696}, {"epoch_end": 53, "step": 1728}, {"epoch_end": 54, "step": 1760}, {"epoch_end": 55, "step": 1792}, {"epoch_end": 56, "step": 1824}, {"epoch_end": 57, "step": 1856}, {"epoch_end": 58, "step": 1888}, {"epoch_end": 59, "step": 1920}, {"epoch_end": 60, "step": 1952}, {"epoch_end": 61, "step": 1984}, {"epoch_end": 62, "step": 2016}, {"epoch_end": 63, "step": 2048}, {"epoch_end": 64, "step": 2080}, {"epoch_end": 65, "step": 2112}, {"epoch_end": 66, "step": 2144}, {"epoch_end": 67, "step": 2176}, {"epoch_end": 68, "step": 2208}, {"epoch_end": 69, "step": 2240}, {"epoch_end": 70, "step": 2272}, {"epoch_end": 71, "step": 2304}, {"epoch_end": 72, "step": 2336}, {"epoch_end": 73, "step": 2368}, {"epoch_end": 74, "step": 2400}, {"epoch_end": 75, "step": 2432}, {"epoch_end": 76, "step": 2464}, {"epoch_end": 77, "step": 2496}, {"epoch_end": 78, "step": 2528}, {"epoch_end": 79, "step": 2560}], "final": {"eval_split": "validation", "eval_acc": 0.823, "eval_loss": 0.4950635223388672, "wall_s": 17.466283798217773, "val_acc": 0.823, "val_loss": 0.4950635223388672, "cos_r_negg": [0.7159034013748169, 0.6858857870101929, 0.9472681283950806, 0.998396635055542], "cos_innovation_negg": [0.7159034013748169, 0.6858857870101929, 0.9472681283950806, 0.998396635055542], "cos_apical_negg": [0.7159034013748169, 0.6858857870101929, 0.9472681283950806, 0.998396635055542], "cos_Ac_negg": [0.7159034013748169, 0.6858857870101929, 0.9472681283950806, 0.998396635055542], "r_norm": [1.068838357925415, 1.865600824356079, 1.9937381744384766, 2.655693531036377], "g_norm": [0.00340816262178123, 0.004210171289741993, 0.0038368094246834517, 0.005118144676089287], "traffic_norm": [0.0, 0.0, 0.0, 0.0], "traffic_residual_norm": [0.0, 0.0, 0.0, 0.0], "traffic_r2": [NaN, NaN, NaN, NaN], "residual_lesion": {"interior_layers": [1, 2, 3], "lesioned_layers": [3], "branch_to_skip_rms": [0.39623984694480896, 0.7033324837684631, 1.0645053386688232], "lesion_eval_acc": 0.5, "lesion_eval_loss": 4.14509033203125, "lesion_acc_drop": 0.32299999999999995}}, "lesion_protocol": {"fraction_of_interior_blocks": 0.3333333333333333, "selection": "final_contiguous_interior_blocks", "probe_source": "training_prefix", "probe_examples": 512, "evaluation_split": "validation"}, "timing": {"warmup_wall_s": 0.0, "feedback_warmup_wall_s": 5.093723297119141, "training_loop_wall_s": 12.250442266464233, "diagnostics_wall_s": 0.07571101188659668, "inline_diagnostics_wall_s": 0.0, "optimizer_wall_s_excluding_inline_diagnostics": 12.250442266464233, "evaluation_wall_s": 0.041290283203125}, "cost": {"train_steps": 2560, "ordinary_training_forward_examples": 640000, "predictor_warmup_forward_examples": 0, "feedback_warmup_forward_examples": 100096, "feedback_warmup_perturbation_events": 400, "perturbation_events": 1040, "calibration_batch_loss_evaluations": 14480, "calibration_example_loss_evaluations": 3630848, "calibration_forward_equivalent_examples": 2252468.148148131, "training_forward_equivalent_examples": 2992564.148148131, "max_perturbation_batch_expansion": 2}, "hardware": {"device": "cuda", "torch_version": "2.3.1+cu118", "cuda_device_name": "NVIDIA GeForce GTX 1080", "cuda_visible_devices": "7", "device_total_memory_bytes": 8507949056, "peak_memory_allocated_bytes": 17595904, "peak_memory_reserved_bytes": 23068672}}
\ No newline at end of file |