Change8

b9320

📦 llama-cppView on GitHub →
🐛 2 fixes🔧 2 symbols

Summary

This release addresses critical issues by fixing the ggml context size calculation and resolving a memory leak. It also involves internal restructuring by moving the split state cache back into the context.

Migration Steps

  1. The split state cache has been moved back into the context.

🐛 Bug Fixes

  • Fixed ggml context size calculation.
  • Fixed a memory leak related to ggml context.

Affected Symbols