mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-05-04 08:04:07 +00:00
Check if input and output files are the same before quantizing to prevent file corruption when mmap reads from a file being written to. Fixes #12753