Change8
Symbol27 releases

common

Found in 2 packages: llama-cpp, angular

llama-cpp(23 releases)

b10645
Aug 27, 2026

Introduced the --n-cpu-ffn option for offloading FFN weights and deduplicated override loops in common code.

b10577
Aug 22, 2026

This release includes a fix for the draft-mtp functionality when using embeddings. It also provides updated pre-compiled binaries for various platforms and configurations.

v0.2.0
Aug 21, 2026

This release bumps the llama.cpp version to 0.2.0 and ggml to 0.21.0, incorporating numerous backend improvements, bug fixes, and new model support across various hardware accelerators like SYCL, OpenCL, Metal, and Vulkan.

b10533
Aug 21, 2026

This release includes a bug fix for gracefully handling unsupported regex patterns in JSON schema and provides updated pre-built binaries for various platforms and hardware configurations.

b10444
Aug 15, 2026

This release introduces support for loading MTP assistant models via the --models-dir flag and includes various improvements to model handling and presets. The 'eagle3' model has been removed.

b10436
Aug 15, 2026

This release includes various fixes for the mtmd and common modules, along with a reverted GGUF fix.

b10081
Jul 22, 2026

This release enhances the handling of draft repositories by correctly resolving and utilizing speculative sidecars. The draft plan now properly consumes its sidecars, improving functionality when using the `-hfd` flag with repositories that provide them.

b9849
Jun 30, 2026

This release enhances URL parsing by correctly handling bracketed IPv6 literals in URL authorities and restores explicit rejection for unsupported schemes in the URL parser.

b9047
May 6, 2026

This release includes several fixes related to device memory fitting, ensuring that unknown devices or non-GPU devices handle memory allocation correctly, specifically by preserving host fallback and keeping unknown GPU fit memory at zero.

b8559
Mar 27, 2026

This release stabilizes the interaction between the lazy grammar sampler and the active reasoning budget by inhibiting sampling when reasoning is active. It also includes various bug fixes and provides updated binaries for numerous hardware and OS configurations.

b8498
Mar 24, 2026

This release introduces standard Hugging Face cache support for model handling and includes various platform-specific binary builds.

b7585Breaking
Dec 30, 2025

This release updates the common library to default the 'content' field to an empty string and includes corresponding test fixes. It also provides a wide range of pre-built binaries for various operating systems and hardware accelerators.

b7583
Dec 30, 2025

This release improves LoRA support by accurately counting LoRA nodes in the compute graph and ensuring adapters are loaded prior to context creation.

b7568
Dec 29, 2025

This release fixes a return value check for process priority settings and introduces additional logging for priority configuration.

b7525
Dec 24, 2025

This release introduces the LLAMA_ARG_OVERRIDE_TENSOR environment variable for tensor overrides and provides updated binaries across multiple platforms and hardware backends.

b7493
Dec 21, 2025

This release focuses on internal code organization by prioritizing vendored dependencies in the common include headers and provides updated binaries across multiple platforms and hardware backends.

b7426Breaking
Dec 16, 2025

This release introduces support for Nemotron 3 and Nemotron Nano 3 models while enhancing JSON-schema type extraction and PEG parser stability. It also announces a forthcoming change in Linux distribution formats from .zip to .tar.gz.

b7380Breaking
Dec 13, 2025

This release introduces a minor optimization for bash completion and announces a forthcoming change to Linux distribution formats from .zip to .tar.gz.

b7376Breaking
Dec 13, 2025

This release introduces support for negated command-line arguments and fixes several argument-related bugs. It also announces a forthcoming change to Linux distribution formats from .zip to .tar.gz.

b7364Breaking
Dec 12, 2025

This release introduces a multi-threaded progress bar and announces a transition in Linux distribution format from .zip to .tar.gz.

b7342Breaking
Dec 10, 2025

This release adds support for Ministral and Mistral Large 3 models and announces a change in the Linux distribution format from .zip to .tar.gz.

b7312Breaking
Dec 7, 2025

This release updates the --color CLI argument to support auto/on/off modes and announces a transition for Linux release archive formats from .zip to .tar.gz.

b7271Breaking
Dec 4, 2025

This release fixes a bug where the --help command required a model to be present and announces a transition for Linux release archive formats from .zip to .tar.gz.

angular(4 releases)

Track Symbol Changes

Use the Change8 MCP server or GitHub Action to get notified when common changes.

Learn More