llama.cpp icon indicating copy to clipboard operation
llama.cpp copied to clipboard

Bug: Slow model loading with mmap

Open Azirine opened this issue 1 year ago • 0 comments

What happened?

Speeds of loading models into memory on M3 Max: with mmap: ~730 MB/s --no-mmap: ~6.4 GB/s

Name and Version

version: 3642 (1d1ccce6) built with Apple clang version 15.0.0 (clang-1500.3.9.4) for arm64-apple-darwin23.6.0

What operating system are you seeing the problem on?

Mac

Relevant log output

No response

Azirine avatar Aug 29 '24 16:08 Azirine