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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deploying various deep learning (DL) models efficiently has boosted theresearch on DL compilers. The difficulty of generating optimized tensor codesdrives DL compiler to ask for the auto-tuning approaches, and the increasingdemands require increasing auto-tuning efficiency and quality. Currently, theDL compilers partition the input DL models into several subgraphs and leveragethe auto-tuning to find the optimal tensor codes of these subgraphs. However,existing auto-tuning approaches usually regard subgraphs as individual ones andoverlook the similarities across them, and thus fail to exploit better tensorcodes under limited time budgets. We propose FamilySeer, an auto-tuningframework for DL compilers that can generate better tensor codes even withlimited time budgets. FamilySeer exploits the similarities and differencesamong subgraphs can organize them into subgraph families, where the tuning ofone subgraph can also improve other subgraphs within the same family. The costmodel of each family gets more purified training samples generated by thefamily and becomes more accurate so that the costly measurements on realhardware can be replaced with the lightweight estimation through cost model.Our experiments show that FamilySeer can generate model codes with the samecode performance more efficiently than state-of-the-art auto-tuning frameworks.

FamilySeer: Towards Optimized Tensor Codes by Exploiting Computation Subgraph Similarity