[AMD] Support Hierarchical Caching on AMD GPUs (#8236)

This commit is contained in:
Hubert Lu
2025-08-28 15:27:07 -07:00
committed by GitHub
parent 5343058875
commit 711390a971
10 changed files with 105 additions and 32 deletions

View File

@@ -1,7 +1,7 @@
import unittest
from types import SimpleNamespace
from sglang.srt.utils import kill_process_tree
from sglang.srt.utils import is_hip, kill_process_tree
from sglang.test.run_eval import run_eval
from sglang.test.test_utils import (
DEFAULT_MODEL_NAME_FOR_TEST,
@@ -11,6 +11,8 @@ from sglang.test.test_utils import (
popen_launch_server,
)
_is_hip = is_hip()
class TestHiCache(CustomTestCase):
@classmethod
@@ -26,7 +28,7 @@ class TestHiCache(CustomTestCase):
"--mem-fraction-static",
0.7,
"--hicache-size",
100,
100 if not _is_hip else 200,
"--page-size",
"64",
"--hicache-storage-backend",