b10692
📦 llama-cppView on GitHub →
✨ 1 features🔧 2 symbols
Summary
The SYCL backend has been updated to split long rows in TOP_K operations, enhancing performance. This release also provides pre-compiled binaries for various platforms and hardware configurations.
✨ New Features
- SYCL backend now splits long rows in TOP_K operations instead of using one work-group per row, improving performance and resource utilization.