refactor(runner): move kernel warmup into the shared runner lifecycle (warmup()) (#28739)

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-19 13:13:13 -07:00
parent d705a91de1
commit 856b0dc74b
16 changed files with 678 additions and 606 deletions
@@ -116,6 +116,8 @@ class EagerRunner(BaseRunner):
            ),
            dp_size=sa.dp_size,
        )
+        # Eager has no capture step, so warm up here (run-once via mr._kernel_warmed_up).
+        self.warmup()

    def can_run_graph(self, forward_batch: ForwardBatch) -> bool:
        # Eager never runs a cuda graph; callers dispatch on isinstance(...,