diff options
Diffstat (limited to 'results/scale_v3_cifar10_sdil_w64_d5_s1.json')
| -rw-r--r-- | results/scale_v3_cifar10_sdil_w64_d5_s1.json | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/results/scale_v3_cifar10_sdil_w64_d5_s1.json b/results/scale_v3_cifar10_sdil_w64_d5_s1.json new file mode 100644 index 0000000..b47e574 --- /dev/null +++ b/results/scale_v3_cifar10_sdil_w64_d5_s1.json @@ -0,0 +1 @@ +{"args": {"mode": "sdil", "dataset": "cifar10", "depth": 5, "width": 64, "act": "tanh", "residual": 1, "epochs": 5, "batch_size": 128, "train_examples": 0, "task_seed": 0, "task_train_examples": 50000, "task_test_examples": 10000, "task_levels": 8, "task_n_in": 128, "task_classes": 10, "teacher_depth": 8, "teacher_width": 64, "teacher_residual": 1, "eta": 0.02, "eta_A": 0.02, "eta_P": 0.002, "momentum": 0.9, "w_scale": 1.0, "a_scale": 1.0, "feedback_scale": 1.0, "pert_sigma": 0.01, "pert_every": 4, "pert_ndirs": 16, "pert_mode": "simultaneous", "use_residual": 1, "raw_scale_control": "none", "learn_A": 1, "learn_P": 1, "p_neutral": 1, "p_warmup_steps": 200, "p_warmup_eta": 0.05, "nuis_rho": 0.0, "predictor_mode": "diagonal", "normalize_delta": 0, "settle_steps": 0, "kappa": 0.0, "feedback": "error", "seed": 1, "device": "cuda", "log_every": 500, "max_steps": 0, "probe_bs": 256, "outdir": "results", "tag": "scale_v3_cifar10_sdil_w64_d5_s1", "n_in": 3072, "n_out": 10}, "provenance": {"git_commit": "d4aef1a571c21e6e4e4125c90aa58da7f14e62e3", "git_dirty": false}, "steps": [{"step": 0, "epoch": 0, "train_loss": 2.534536123275757, "cos_r_negg": [-0.020268913358449936, -0.021106336265802383, 0.05065072327852249, 0.07415375858545303, 0.026462651789188385], "cos_innovation_negg": [-0.020268913358449936, -0.021106336265802383, 0.05065072327852249, 0.07415375858545303, 0.026462651789188385], "cos_apical_negg": [-0.020268913358449936, -0.021106336265802383, 0.05065072327852249, 0.07415375858545303, 0.026462651789188385], "cos_Ac_negg": [-0.020268913358449936, -0.021106336265802383, 0.05065072327852249, 0.07415375858545303, 0.026462651789188385], "r_norm": [2.4451587200164795, 2.492396354675293, 2.5392849445343018, 2.5055091381073, 2.4691567420959473], "ldr": {"dL_sdil": -0.06915044784545898, "dL_bp": -0.4515695571899414, "ratio": 0.15313354663625517}}, {"epoch_end": 0, "step": 391, "test_acc": 0.3713, "test_loss": 1.8046606079101561}, {"step": 500, "epoch": 1, "train_loss": 1.7330714464187622, "cos_r_negg": [0.6587176322937012, 0.12132266163825989, 0.15032048523426056, 0.11879751086235046, 0.12757079303264618], "cos_innovation_negg": [0.6587176322937012, 0.12132266163825989, 0.15032048523426056, 0.11879751086235046, 0.12757079303264618], "cos_apical_negg": [0.6587176322937012, 0.12132266163825989, 0.15032048523426056, 0.11879751086235046, 0.12757079303264618], "cos_Ac_negg": [0.6587176322937012, 0.12132266163825989, 0.15032048523426056, 0.11879751086235046, 0.12757079303264618], "r_norm": [1.7397444248199463, 1.7428933382034302, 1.7289388179779053, 1.7265510559082031, 1.6761484146118164]}, {"epoch_end": 1, "step": 782, "test_acc": 0.3971, "test_loss": 1.7480157592773438}, {"step": 1000, "epoch": 2, "train_loss": 1.667506456375122, "cos_r_negg": [0.7394656538963318, 0.19269448518753052, 0.1853671669960022, 0.18019302189350128, 0.18472756445407867], "cos_innovation_negg": [0.7394656538963318, 0.19269448518753052, 0.1853671669960022, 0.18019302189350128, 0.18472756445407867], "cos_apical_negg": [0.7394656538963318, 0.19269448518753052, 0.1853671669960022, 0.18019302189350128, 0.18472756445407867], "cos_Ac_negg": [0.7394656538963318, 0.19269448518753052, 0.1853671669960022, 0.18019302189350128, 0.18472756445407867], "r_norm": [1.4853475093841553, 1.4121298789978027, 1.408780574798584, 1.3970463275909424, 1.3608713150024414]}, {"epoch_end": 2, "step": 1173, "test_acc": 0.4023, "test_loss": 1.7214055297851563}, {"step": 1500, "epoch": 3, "train_loss": 1.633162260055542, "cos_r_negg": [0.770696222782135, 0.2740561366081238, 0.3080812096595764, 0.3018062114715576, 0.25017261505126953], "cos_innovation_negg": [0.770696222782135, 0.2740561366081238, 0.3080812096595764, 0.3018062114715576, 0.25017261505126953], "cos_apical_negg": [0.770696222782135, 0.2740561366081238, 0.3080812096595764, 0.3018062114715576, 0.25017261505126953], "cos_Ac_negg": [0.770696222782135, 0.2740561366081238, 0.3080812096595764, 0.3018062114715576, 0.25017261505126953], "r_norm": [1.2929439544677734, 1.1678597927093506, 1.1709778308868408, 1.1559346914291382, 1.1175240278244019]}, {"epoch_end": 3, "step": 1564, "test_acc": 0.408, "test_loss": 1.7174328491210937}, {"epoch_end": 4, "step": 1955, "test_acc": 0.4136, "test_loss": 1.6948340576171874}], "final": {"test_acc": 0.4136, "test_loss": 1.6948340576171874, "wall_s": 13.226434707641602, "cos_r_negg": [0.8090265989303589, 0.3869614005088806, 0.40717965364456177, 0.3903641104698181, 0.3922593295574188], "cos_innovation_negg": [0.8090265989303589, 0.3869614005088806, 0.40717965364456177, 0.3903641104698181, 0.3922593295574188], "cos_apical_negg": [0.8090265989303589, 0.3869614005088806, 0.40717965364456177, 0.3903641104698181, 0.3922593295574188], "cos_Ac_negg": [0.8090265989303589, 0.3869614005088806, 0.40717965364456177, 0.3903641104698181, 0.3922593295574188]}}
\ No newline at end of file |