A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 04:15 UTC
[NIXL][XPU] update name of nixl wheel (#27631)
Signed-off-by: zhenwei-intel <zhenwei.liu@intel.com>
liuzhenwei committed
Oct 29, 2025 at 12:43 UTC
d2c33c397ad30f0b0fad7296a3c80d47df0243fe
1 file changed
+1
-1
tools/install_nixl_from_source_ubuntu.py
+1
-1
@@ -37,7 +37,7 @@ def is_pip_package_installed(package_name):
37
def find_nixl_wheel_in_cache(cache_dir):
38
"""Finds a nixl wheel file in the specified cache directory."""
39
# The repaired wheel will have a 'manylinux' tag, but this glob still works.
40
- search_pattern = os.path.join(cache_dir, "nixl-*.whl")
40
+ search_pattern = os.path.join(cache_dir, "nixl*.whl")
41
wheels = glob.glob(search_pattern)
42
if wheels:
43
# Sort to get the most recent/highest version if multiple exist