biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Metrics] Add Prometheus counters for Model FLOPs Utilization (MFU) (#30950)

Browse Source 
Export the existing Model FLOPs Utilization (MFU) metrics via Prometheus.

`--enable-mfu-metrics` is required for these to be exposed.

Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
This commit is contained in:
Mark McLoughlin
2026-02-23 15:01:07 +00:00
committed by GitHub
parent b95bb6927f
commit 5cc7c4452e
5 changed files with 109 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
docs/usage/metrics.md
View File

@@ -45,6 +45,12 @@ The following metrics are exposed:
--8<-- "docs/generated/metrics/nixl_connector.inc.md"
## Model Flops Utilization (MFU) Performance Metrics
These metrics are available via `--enable-mfu-metrics`:
--8<-- "docs/generated/metrics/perf.inc.md"
## Deprecation Policy
Note: when metrics are deprecated in version `X.Y`, they are hidden in version `X.Y+1`