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

The last decade has witnessed growth in the computational requirements fortraining deep neural networks. Current approaches (e.g., data/modelparallelism, pipeline parallelism) parallelize training tasks onto multipledevices. However, these approaches always rely on specific deep learningframeworks and requires elaborate manual design, which make it difficult tomaintain and share between different type of models. In this paper, we proposeAuto-MAP, a framework for exploring distributed execution plans for DNNworkloads, which can automatically discovering fast parallelization strategiesthrough reinforcement learning on IR level of deep learning models. Efficientexploration remains a major challenge for reinforcement learning. We leverageDQN with task-specific pruning strategies to help efficiently explore thesearch space including optimized strategies. Our evaluation shows that Auto-MAPcan find the optimal solution in two hours, while achieving better throughputon several NLP and convolution models.

Auto-MAP: A DQN Framework for Exploring Distributed Execution Plans for DNN Workloads