Change8

b7582

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

Summary

This release improves sampling performance by reusing the token data buffer in llama_sampler_sample and includes various platform-specific build fixes.

✨ New Features

  • Optimized sampling by reusing the token data buffer in llama_sampler_sample to reduce allocations.

🐛 Bug Fixes

  • Fixed build issues related to the sampling buffer relocation.

🔧 Affected Symbols

llama_sampler_sample