GSPMD automatically infers tensor partitioning from limited user annotations to parallelize single-device ML programs across thousands of TPUs, reporting 50-62% utilization for up to trillion-parameter models.
In Proceedings of the 26th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (New York, NY, USA, 2021), PPoPP ’21, Association for Computing Machinery, p
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2021 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GSPMD: General and Scalable Parallelization for ML Computation Graphs
GSPMD automatically infers tensor partitioning from limited user annotations to parallelize single-device ML programs across thousands of TPUs, reporting 50-62% utilization for up to trillion-parameter models.