DUB Sparsity

The official code implementation for the paper titled "Pruning for Improved ADC Efficiency in Crossbar-based Analog In-memory Accelerators", which induces D.U.B. sparsity by regularizing the variance of $L_{0}$ norms of neighboring columns within the same crossbar.

The paper can be found at this link.

Contact Information

Please, do not hesitate to reach out the corresponding author, Timur Ibrayev, via email or on LinkedIn

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
results		results
.gitignore		.gitignore
README.md		README.md
analyze_tile_dist.py		analyze_tile_dist.py
c0_LBL_prune.py		c0_LBL_prune.py
c0_LBL_prune_class.py		c0_LBL_prune_class.py
c1_Hoyer_and_variance_class.py		c1_Hoyer_and_variance_class.py
c1_Hoyer_and_variance_prune_and_finetune.py		c1_Hoyer_and_variance_prune_and_finetune.py
c1_Hoyer_and_variance_train_and_prune_with_gate.py		c1_Hoyer_and_variance_train_and_prune_with_gate.py
c2_GroupLassoForXB_class.py		c2_GroupLassoForXB_class.py
c2_GroupLassoForXB_prune_LBL_and_finetune.py		c2_GroupLassoForXB_prune_LBL_and_finetune.py
c2_GroupLassoForXB_train_and_prune_DB.py		c2_GroupLassoForXB_train_and_prune_DB.py
c3_SDUB_class.py		c3_SDUB_class.py
c3_SDUB_part_DUB.py		c3_SDUB_part_DUB.py
c3_SDUB_part_S.py		c3_SDUB_part_S.py
custom_models_cifar_vgg.py		custom_models_cifar_vgg.py
custom_normalization_functions.py		custom_normalization_functions.py
layer_sensitivity.py		layer_sensitivity.py
test_trained_models.py		test_trained_models.py
train_clean_model.py		train_clean_model.py
train_clean_models.py		train_clean_models.py
utilities.py		utilities.py

TimurIbrayev/DUB_Sparsity_for_XBars

Folders and files

Latest commit

History

Repository files navigation

DUB Sparsity

Contact Information

About

Topics

Resources

Stars

Watchers

Forks

Languages