Change8

b8191

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

Summary

This release introduces an optimized Q4_1 OpenCL kernel for Adreno GPUs and includes various refactoring and formatting changes within the OpenCL implementation. It also fixes a build issue for non-Adreno OpenCL targets.

✨ New Features

  • Added optimized Q4_1 matrix multiplication kernel for Adreno GPUs via OpenCL.

🐛 Bug Fixes

  • Fixed OpenCL build for non-Adreno devices.

Affected Symbols