llama : refactor llama_context, llama_kv_cache, llm_build_context (v2)#12181

ggerganov merged 16 commits intomasterggml-org/llama.cpp:masterfrom

gg/llama-kv-cache-v2ggml-org/llama.cpp:gg/llama-kv-cache-v2Copy head branch name to clipboard