Zorse integrates interleaved pipeline parallelism, ZeRO-2 data parallelism, and CPU offloading to accelerate LLM training on heterogeneous GPU clusters by up to 4x.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Zorse: Optimizing LLM Training Efficiency on Heterogeneous GPU Clusters
Zorse integrates interleaved pipeline parallelism, ZeRO-2 data parallelism, and CPU offloading to accelerate LLM training on heterogeneous GPU clusters by up to 4x.