llama : fix rwkv inference in gg/llama-kv-cache branch#11618

ggerganov merged 1 commit intoggml-org:gg/llama-kv-cacheggml-org/llama.cpp:gg/llama-kv-cachefrom

MollySophia:molly/llama-kv-cacheMollySophia/llama.cpp:molly/llama-kv-cacheCopy head branch name to clipboard