[Core] Allow full cudagraph with separate attention routines and orthogonal to compilation, add support for FA2 and FlashInfer (#20059)

Signed-off-by: fhl <2410591650@qq.com>
Signed-off-by: fhl2000 <63384265+fhl2000@users.noreply.github.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
This commit is contained in:
fhl2000
2025-08-15 10:01:39 -04:00
committed by GitHub
co-authored by Luka Govedič Lucas Wilkinson Lucas Wilkinson
parent a0632a3e03
commit 74f441f4b5
34 changed files with 1839 additions and 597 deletions
+18
View File
@@ -37,3 +37,21 @@ def end_monitoring_torch_compile(vllm_config: VllmConfig):
if context_manager is not None:
context_manager.__exit__(None, None, None)
context_manager = None
cudagraph_capturing_enabled: bool = True
def validate_cudagraph_capturing_enabled():
# used to monitor whether an cudagraph capturing is legal at runtime.
# should be called before any cudagraph capturing.
# if an illegal cudagraph capturing happens, raise an error.
global cudagraph_capturing_enabled
if not cudagraph_capturing_enabled:
raise RuntimeError("CUDA graph capturing detected at an inappropriate "
"time. This operation is currently disabled.")
def set_cudagraph_capturing_enabled(enabled: bool):
global cudagraph_capturing_enabled
cudagraph_capturing_enabled = enabled