mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-08 10:07:01 +00:00
* YaRN : store rope scaling type as int32_t in memory * llama : store mapped names as const char *
451 KiB
451 KiB