llama.cpp/CHANGELOG.md

45 lines
917 B
Markdown
Raw Normal View History

2023-05-26 17:32:34 -04:00
# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased]
2023-06-08 21:53:38 -04:00
- (build-system) Migrate from scikit-build to scikit-build-core
2023-06-08 03:26:49 -04:00
## [v0.1.59]
2023-06-06 17:01:10 -04:00
### Added
- (llama.cpp) k-quants support
- (server) mirostat sampling parameters to server
2023-06-05 18:17:29 -04:00
2023-06-08 03:26:49 -04:00
### Fixed
- Support both `.so` and `.dylib` for `libllama` on MacOS
2023-06-04 23:31:51 -04:00
## [v0.1.58]
2023-06-06 17:01:10 -04:00
### Added
- (llama.cpp) Metal Silicon support
2023-06-04 23:30:42 -04:00
2023-06-04 23:30:10 -04:00
## [v0.1.57]
2023-06-06 17:01:10 -04:00
### Added
- (llama.cpp) OpenLlama 3B support
2023-06-04 23:30:10 -04:00
2023-05-30 03:07:36 -04:00
## [v0.1.56]
2023-05-26 17:32:34 -04:00
### Added
2023-06-06 17:01:10 -04:00
- (misc) Added first version of the changelog
- (server) Use async routes
- (python-api) Use numpy for internal buffers to reduce memory usage and improve performance.
2023-05-26 20:23:49 -04:00
### Fixed
2023-06-06 17:01:10 -04:00
- (python-api) Performance bug in stop sequence check slowing down streaming.