[FlashInfer] Migrate autotune warmup to flashinfer.autotune_v2 managed cache (#55671)

Signed-off-by: Yang Xu <yanxu@nvidia.com>
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Signed-off-by: YangXu <yanxu@nvidia.com>
Co-authored-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Co-authored-by: Kimi Code <noreply@moonshot.cn>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Misha Goin <mgoin64@gmail.com>
Co-authored-by: vllm-agent <claw@inferact.ai>
This commit is contained in:
Yang Xu
2026-10-17 02:08:28 +00:00
committed by GitHub
co-authored by Zihua Wu Kimi Code Codex Misha Goin vllm-agent
parent b3a7e2ff6e
commit 71c20f73f2
6 changed files with 436 additions and 37 deletions
+14
View File
@@ -358,6 +358,19 @@ autotune = _lazy_import_wrapper(
)
@functools.cache
def has_flashinfer_autotune_v2() -> bool:
"""Return ``True`` if FlashInfer ships the managed-cache autotuner
(``flashinfer.autotune_v2``, flashinfer-ai/flashinfer#3861)."""
if not has_flashinfer():
return False
mod = _get_submodule("flashinfer")
return mod is not None and all(
callable(getattr(mod, name, None))
for name in ("autotune_v2", "autotune_v2_reload")
)
@functools.cache
def has_flashinfer_comm() -> bool:
"""Return `True` if FlashInfer comm module is available."""
@@ -1269,6 +1282,7 @@ __all__ = [
"flashinfer_fused_kda_decode",
"autotune",
"has_flashinfer_moe",
"has_flashinfer_autotune_v2",
"has_flashinfer_comm",
"has_flashinfer_nvlink_two_sided",
"has_flashinfer_nvlink_one_sided",