biondizzle bcc872c2c3 Remove global allocator swap, use targeted KV cache managed allocation ...

sitecustomize.py: No longer swaps CUDAPluggableAllocator globally.
Sets VLLM_KV_CACHE_USE_MANAGED_MEMORY=1 instead.
vllm_managed_mem.py: No global allocator swap, no torch.cuda patches.

2026-04-11 02:15:09 +00:00

2.5 KiB

Raw Permalink Blame History

View Raw