v0.1.2
📦 llama-cppView on GitHub →
✨ 5 features🐛 2 fixes🔧 3 symbols
Summary
This release includes updates to ggml, various build and CI improvements, and enhancements to the server and mtmd components. It also introduces support for integer tokenizer scores.
Migration Steps
- Refactor Built-In Tools naming (Server/Browser).
✨ New Features
- Support for integer tokenizer scores.
- MMVQ nwarps=8 for bs=1 for dense models on DGX Spark.
- Save processed mtmd chunks as placeholder.
- Use sha256 for input hashing in mtmd.
- Skip thumbnail for non-tiled LFM2 images.
🐛 Bug Fixes
- Fix xcframework + cmake clean-up.
- Skip UMA override for HIP builds.