- Support ImageMagick on Windows
- Support Python3.8
- Downgrade supported Python version to 3.8.1
- Set temperature when working with OpenAI feat: Store chat in a chat archive which persists between sessions
- Fix broken indexing, only worked for first 2 text chunks @Klingefjord
- Suppress warnings
- Allow user to specify a different LLM instead of OpenAI
- Ignore empty text chunk as it'll cause error later
- Minor improvements in usability
- Add line-height when displaying answers
- Add support for huggingface embedding and allow user to select a different embedding provider from OpenAI
- Add an option to pre-process a PDF file
- Update Changelog and remove # used for GH issues
- Remove duplicate processing of finding similarities
- Only copy pdf if the source path is different
- Simplify code
- Provide option to just create a model for provided book
- WebUI using Holoviz Panel framework
- Cleanup
- Combine workflow steps in a single file to make it easy to port to other projects
- Ignore tests (temp) as they can't run on Github Actions due to binary dependencies
- Write workflow to generate book summary
- Add steps to convert pdf file to scanned pages of text
- Single version using another method
- Add annotated tag
- Add poetry_dynamic_versioning to automatic release versioning
- Verify input pdf file and set number of pages
- Initial commit with a skeleton project
- Formatting