Repository navigation

[Python][RunInference] vLLM: prefix-aware batching and capacity-based autoscaling - #40908

Draft

AMOOOMA wants to merge 2 commits into

apache:masterfrom

AMOOOMA:pr-vllm-integration

Draft

[Python][RunInference] vLLM: prefix-aware batching and capacity-based autoscaling#40908
AMOOOMA wants to merge 2 commits into
apache:masterfrom
AMOOOMA:pr-vllm-integration

Commits on Nov 6, 2026
- [RunInference] vLLM: prefix-aware batching and capacity-based autoscaling
  AMOOOMA
  committed
  d26d5e5
- [RunInference] vLLM: read Dynamo's engine metrics for the autoscaling signal
  AMOOOMA
  committed
  4639b4c