We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Tight Compression: Compressing CNN Through Fine-Grained Pruning and Weight Permutation for Efficient Implementation

Abstract: The unstructured sparsity after pruning poses a challenge to the efficient implementation of deep learning models in existing regular architectures like systolic arrays. On the other hand, coarse-grained structured pruning is suitable for implementation in regular architectures but tends to have higher accuracy loss than unstructured pruning when the pruned models are of the same size. In this work, we propose a model compression method based on a novel weight permutation scheme to fully exploit the fine-grained weight sparsity in the hardware design. Through permutation, the optimal arrangement of the weight matrix is obtained, and the sparse weight matrix is further compressed to a small and dense format to make full use of the hardware resources. Two pruning granularities are explored. In addition to the unstructured weight pruning, we also propose a more fine-grained subword-level pruning to further improve the compression performance. Compared to the state-of-the-art works, the matrix compression rate is significantly improved from 5.88x to 14.13x. As a result, the throughput and energy efficiency are improved by 2.75 and 1.86 times, respectively.
Comments: Previous Conference Version: 2020 57th ACM/IEEE Design Automation Conference (DAC)
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2104.01303 [cs.LG]
  (or arXiv:2104.01303v1 [cs.LG] for this version)

Submission history

From: Xizi Chen [view email]
[v1] Sat, 3 Apr 2021 03:24:52 GMT (3261kb,D)

Link back to: arXiv, form interface, contact.