mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-08-29 00:21:21 +02:00
1ec3332ade
* YaRN : store rope scaling type as int32_t in memory * llama : store mapped names as const char *