Lodestar deploys continuous online learning to route LLM inference requests across GPU clusters, reporting 1.41x lower average TTFT versus heuristics.
Hetis: Serving LLMs in heterogeneous GPU clusters with fine-grained and dynamic parallelism
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DC 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Lodestar: An Online-Learning LLM Inference Router
Lodestar deploys continuous online learning to route LLM inference requests across GPU clusters, reporting 1.41x lower average TTFT versus heuristics.