English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep learning (DL) package supply chains (SCs) are critical for DL frameworksto remain competitive. However, vital knowledge on the nature of DL package SCsis still lacking. In this paper, we explore the domains, clusters, anddisengagement of packages in two representative PyPI DL package SCs to bridgethis knowledge gap. We analyze the metadata of nearly six million PyPI packagedistributions and construct version-sensitive SCs for two popular DLframeworks: TensorFlow and PyTorch. We find that popular packages (measured bythe number of monthly downloads) in the two SCs cover 34 domains belonging toeight categories. Applications, Infrastructure, and Sciences categories accountfor over 85% of popular packages in either SC and TensorFlow and PyTorch SChave developed specializations on Infrastructure and Applications packagesrespectively. We employ the Leiden community detection algorithm and detect 131and 100 clusters in the two SCs. The clusters mainly exhibit four shapes:Arrow, Star, Tree, and Forest with increasing dependency complexity. Mostclusters are Arrow or Star, but Tree and Forest clusters account for mostpackages (Tensorflow SC: 70%, PyTorch SC: 90%). We identify three groups ofreasons why packages disengage from the SC (i.e., remove the DL framework andits dependents from their installation dependencies): dependency issues,functional improvements, and ease of installation. The most commondisengagement reason in the two SCs are different. Our study provides richimplications on the maintenance and dependency management practices of PyPI DLSCs.

Characterizing Deep Learning Package Supply Chains in PyPI: Domains, Clusters, and Disengagement