llama.cpp
4dccb38d - metal : improve dequantize precision to match CPU (#4836)

Commit
1 year ago
metal : improve dequantize precision to match CPU (#4836) ggml-ci
Author
Parents
Loading