[Usage]: for the tpye: UniformTypeKVCacheSpecs , the calc of num_blocks = available_memory // kv_cache_groups[0].kv_cache_spec.page_size_bytes, no need to divide num_layers ?
March 27, 2026 · #38331
Python
Difficulty: Easy
Labels
usage
Parent Repository
vllm-project/vllm
Python repository
75,721 15,332