Open Source
llama.cpp b9838 Release Expands Platform Support
llama.cpp Releaseshigh confidence
Why it matters
- →Expands platform support, increasing accessibility for developers on different systems.
- →Enhances AMD GPU support with ROCm 7.2, providing alternatives to NVIDIA's CUDA.
- →Solidifies llama.cpp's position as a versatile inference runtime across multiple architectures.
The b9838 release of llama.cpp has been announced, focusing on expanding platform support rather than introducing new features. This update includes ROCm 7.2 support for Ubuntu x64, enhancing options for AMD GPU users. The release covers a wide range of platforms, including macOS, Linux, Windows, and openEuler, ensuring compatibility across various systems. While no new models or quantization methods are introduced, the release strengthens llama.cpp's role as a versatile tool for AI inference.
Read originalMore from llama.cpp Releases
Open Sourcemodels
llama.cpp b9831 release adds DFlash support
The b9831 release of llama.cpp marks a significant enhancement with the addition of DFlash, which brings sliding window attention per layer types. This update is particularly beneficial for developers on macOS, Linux, and Windows, as it extends the tool's compatibility and functionality across these platforms. With ROCm 7.2 now available on Ubuntu, AMD GPU users gain a more robust option for local inference. While no new models are introduced, this release solidifies llama.cpp's role as a versatile inference runtime, especially for those not reliant on NVIDIA hardware. The update also includes various platform-specific improvements, making it a comprehensive upgrade for developers.
llama.cpp Releases
Open Sourcecoding
llama.cpp b9832 Release Adds Debugging Feature
The b9832 release of llama.cpp introduces a new debugging capability with the --dump-prog option in jinja, co-authored by Sigbjørn Skjæret. This enhancement is designed to streamline the debugging process for developers. The update also extends compatibility across various systems, including macOS, Linux, Windows, and openEuler, ensuring developers can work seamlessly in their preferred environments. While the release doesn't bring new models or quantization techniques, it reinforces llama.cpp's role as a flexible tool for developers. With ROCm 7.2 and CUDA 12 and 13 support, the platform continues to cater to a broad spectrum of hardware configurations. This update is a testament to llama.cpp's commitment to improving developer experience.
llama.cpp Releases
Models & Labsmodels
Llama.cpp b9833 Release Enhances MiniCPM5 Parser
The latest b9833 release of llama.cpp focuses on refining the MiniCPM5 parser, addressing several technical aspects to improve its functionality. This update includes the addition of a new tool call parser, refactoring of the PEG parser, and adjustments to the Jinja min/max API for better compatibility with Jinja2. The release also reverts some shared mapper changes to maintain strict JSON parsing for tool-call arguments. These enhancements aim to streamline the parsing process, ensuring more reliable and efficient handling of XML tool calls and grammar triggers.
llama.cpp Releases
More in Open Source
© Matt WolfeOpen Sourcemodels
Krea 2 Releases Open Weights
Krea 2 has made its model weights open, allowing broader access.
Matt Wolfe
© Hugging Face BlogOpen Sourcecoding
Hugging Face Automates Weekly Releases with AI
Hugging Face has streamlined its release process for the huggingface_hub Python client, moving from a 4-6 week cycle to weekly releases. This shift is powered by a combination of open-source tools and AI, which drafts release notes and automates mechanical tasks, while humans oversee critical judgment areas. The process is designed to be replicable by other maintainers, emphasizing transparency and adaptability. This change not only accelerates the release cycle but also ensures that updates are consistently delivered without the need for proprietary tools.
Hugging Face Blog
© Matt WolfeOpen Sourcemodels
PewDiePie Builds Private AI Workspace
PewDiePie has invested $41,000 in creating a private, self-hosted AI workspace using open-source tools.
Matt Wolfe