Skip to content

README_GUI.md

Latest commit

 

History

History
84 lines (55 loc) · 3.07 KB

README_GUI.md

File metadata and controls

84 lines (55 loc) · 3.07 KB

SlideNote Studio GUI

中文

SlideNote Studio is a Streamlit interface for python -m slidenote build and python -m slidenote study-pack. It keeps the everyday UI small instead of exposing every internal tuning knob.

What It Does

  • Upload PPTX / PPT / PDF files.
  • Choose one of two workflows:
    • Lecture quality: strong default note generation, requires API keys.
    • Local preview: local parsing preview, no API calls.
  • Enter Text / Vision / OCR API keys on the page. Keys are passed only through the child-process environment, not command-line flags.
  • Select extra exports: Markdown ZIP, TOC Markdown, Word, PDF, or LaTeX.
  • Keep progress, ETA, Doctor readiness, usage, and cost details in compact diagnostics panels.
  • Generate a study pack from the Notes workspace: review.md, exam.md, exam.json, exam.html, and related files.
  • Download notes.zip, notes.md, coverage.md, export files, or the complete output ZIP.
  • Switch to Textbook library, upload a PDF textbook, and build a RAG-ready corpus. The corpus is not connected to note generation yet.

For sharing Markdown notes, prefer notes.zip; it contains both notes.md and notes.assets/ so images render on another computer.

Install

Recommended:

.\install.ps1

Manual:

python -m pip install -e ".[dev,llm,gui]"

Run

.\run_gui.ps1

Manual:

streamlit run gui/app.py

First Test

Start with Local preview. It checks parsing and basic output without any API calls. The first-run check passes when the Notes workspace shows both notes.md and the shareable notes.zip.

For production notes, use Lecture quality, enter a Text API key, keep Vision=auto for image-heavy slides, and add a Qwen/DashScope vision key. Scanned PDFs may also need OCR credentials.

Textbook Library

Switch to Textbook library to upload a PDF textbook and run:

python -m slidenote textbook-index <textbook.pdf> --out <output-dir> --ocr auto

The default Auto scanned pages mode extracts native PDF text first and only calls OCR for scanned or low-text pages. Outputs include textbook_manifest.json, textbook_sections.json, textbook_chunks.jsonl, and textbook_report.md. These are RAG preparation artifacts and do not change course-note generation yet.

Study Pack

After a build finishes, open Study pack in the Notes workspace and click Generate study pack. The GUI runs:

python -m slidenote study-pack <output-dir> --question-count 12

It uses the Text provider key when available and falls back to local rules otherwise.

Exports

The left Run panel's Exports section can generate:

  • notes.zip: Markdown package with image assets, no Pandoc needed.
  • notes.toc.md: TOC Markdown, no Pandoc needed.
  • notes.docx: Word document, requires Pandoc.
  • notes.pdf: PDF handout, requires Pandoc + LibreOffice.
  • notes.tex: LaTeX source, requires Pandoc.

Export status is written to export_report.json. If Pandoc or LibreOffice is missing, the GUI shows install hints before the run.