Skip to content

Commentarium v1.9

Choose a tag to compare

View all tags
@Farazjz Farazjz released this 17 Oct 01:40
v1.9.0
0c7fa64

RELEASE-NOTES-v1.9.0.md

Commentarium v1.9 — Phases 1–9 complete

Final release of the Streamlit-era UI — the end of the v1.x line. Development continues on the
v2.0 branch ("Research Studio UI"), to be released as v2.0.0.

Highlights (Phases 1–9)

  • Phase 1 — Foundation: project skeleton, .env config, logging system (rotating file + console
    • in-app), FastAPI backend, custom NumPy + SQLite vector store with persistence.
  • Phase 2 — Ingestion: PDF/DOCX parsers (PyMuPDF / python-docx), OCR of scanned pages, smart
    chunking with page + section metadata, local & OpenAI-compatible embeddings, full project /
    document management.
  • Phase 3 — Chat + RAG: cosine retrieval, grounded answers with strict source tags, SQLite chat
    history, working Chat page.
  • Phase 4 — Citations: Crossref metadata lookup with confidence ranking, verification workflow,
    APA 7 + Vancouver formatters, source-tag resolution into real in-text citations + bibliography.
  • Phase 5 — Polish: embedding-dimension consistency guard (fails loudly, by design), corpus model
    identity tracking, stale-artifact cleanup, README QA/troubleshooting.
  • Phase 6 — Management + retrieval fixes: project/chat management with cleanup, professional
    project cards + document actions, section-weighted retrieval (Reference chunks no longer crowd out
    real content), distributed retrieval with auto-detect heuristics.
  • Phase 7 — Exit / shutdown: graceful one-click shutdown of API + UI servers via psutil.
  • Phase 8 — Research workflow + podcast studio: multi-kind per-document summaries (Brief /
    Detailed / Key points / TL;DR / Quiz, cached), per-project model & temperature overrides, tags
    & notes, NotebookLM-style podcast deep-dives with edge-tts, local VoiceStudio (voice
    cloning, en/fa), or any OpenAI-compatible /v1/audio/speech server.
  • Phase 9 — One-click launcher + OCR to Markdown: single run.bat (starts API + UI hidden,
    waits for the port, opens the browser) plus a stand-alone OCR page (PaddleOCR-VL-1.6 / TeleOCR
    vision-language backends, live Markdown preview, downloads, saved-file history).

Running

pip install -r requirements.txt
copy .env.example .env   # add your OpenAI-compatible key (or point at a local server)
run.bat                  # opens http://localhost:8501

See the README for full setup, model options (OpenRouter / LM Studio / Ollama / vLLM / LiteLLM),
podcast voices, and OCR backends.

What's next

v2.0.0 — the "Research Studio" UI: a modern dark/light single-page app served by the same FastAPI
backend (Streamlit stays available as a legacy fallback).

Assets 2
Loading