vllm.model_executor.model_loader.weight_cache.warmup_cache
¶
Daemon warmup caches, seeded daemon-to-daemon over the shared transport.
Modules:
-
base–Warmup cache lifecycle interface.
-
common–Shared compiled-cache manifest and installation helpers.
-
compiled–Compiled runtime artifact warmup cache.
-
flashinfer_autotune–FlashInfer autotune warmup cache.