These are the release notes for gptme version v0.14.1.
Contributors
Thanks to everyone who contributed to this release:
Changelog
Changes since v0.14.0:
📦 gptme
✨ Features (3)
- feat: add magic
@logworkspace for./logdir/workspace, use it for evals (9b9d942) - feat: added --workspace option, added project-level gptme.toml config file support (incl files to include in context by default) (
d914810) - feat: added -r/--resume flag to resume last convo (
6a58900)
🐛 Fixes (8)
Click to expand
- fix: hide project context message (
07838d6) - fix: fix tool init order, ensuring registered functions are included in prompt (
3732df1) - fix: fixed type error (
0b438e1) - fix(eval): misc improvements to evals, including reading results files (
e30bd08) - fix: made several slow imports and top-level calls lazy (#97) (
3ddaeae) - fix(eval): cleanup evals' children processes using process group, added more evals init-react, init-rust, whois-superuserlabs-ceo (
56062ce) - fix: added back execution of ```python blocks (
b47937a) - fix: fixed openrouter provider, added Referer/X-Title header for openrouter requests (
c50bfda)
🔨 Misc (4)
- chore: bump version to 0.14.1 (
0d16ea4) - tests: loosened short prompt token limit (
789d788) - chore: cleaned up useless stuff (
aacfb93) - docs: added basic docs for evals (
5a0c47e)
(excluded 2 less relevant commits)
Full Changelog: v0.14.0...v0.14.1