Change8

b8583

📦 llama-cppView on GitHub →
🐛 1 fixes🔧 1 symbols

Summary

The llama-model-loader now issues a warning when tensor overrides are used alongside mmap, and it has been updated to use pinned memory for tensor overrides.

Migration Steps

  1. If you were relying on silent tensor overrides when using mmap in llama-model-loader, be aware that a warning is now printed instead.

🐛 Bug Fixes

  • llama-model-loader now prints a warning when using overrides with mmap.

Affected Symbols