Skip to content

Commit 7ee61ac

Browse files
committed
docs(uat): record the GCP served proving run in the serve-step comment
Nathan's nitpick was to stay draft until a served completion. Run 32837547329 is that completion; the comment no longer calls the next nightly the proving run. Signed-off-by: Rohit Rajani <rorajani@nvidia.com>
1 parent 7d38c7a commit 7ee61ac

1 file changed

Lines changed: 3 additions & 4 deletions

File tree

.github/workflows/uat-gcp.yaml

Lines changed: 3 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -698,10 +698,9 @@ jobs:
698698
# Frontend cold-pulled the ~12GB vllm-runtime image on a CPU-pool node and
699699
# wedged in ContainerCreating past the readiness budget; phase_serve now
700700
# selects the GPU pool for both graph components, whose a3-megagpu-8g class
701-
# pulls that image in a fraction of the budget. The first GCP inference
702-
# nightly (or manual `run serve`) after this is the proving run — serving
703-
# was never verified end-to-end. AWS/Azure serve steps stay commented until
704-
# this lane demonstrates a served completion.
701+
# pulls that image in a fraction of the budget. Proven end-to-end by
702+
# gcp-h100 inference run 32837547329 (prep through serve + verify).
703+
# AWS/Azure serve steps stay commented — out of scope for #1644.
705704
- name: UAT - serve (DynamoGraphDeployment + endpoint)
706705
id: serve
707706
if: steps.conformance.outcome == 'success' && inputs.intent == 'inference'

0 commit comments

Comments
 (0)