llama.cpp

ggml: check if non-native endian model is being loaded

#13943

Merged