Build one system for deep learning reports.
HAT turns image, text, and multimodal work into one shared reporting surface with assignment pages, source code, and a live image demo.
One course hub for every report surface.
Structured for GitHub Pages, report writing, source control, and demo delivery without scattering ownership across disconnected files.
Common Page
Shared landing layer for navigation, group identity, and course-level routing.
Assignment 1
Overview
Assignment entry page that links the brief, deliverables, and all dedicated track pages.
Track Pages
Independent report pages for image, text, and multimodal so each member can edit without blocking others.
Delivery Links
External surfaces for code hosting, the interactive image demo, and the LaTeX report source.
Operational Snapshot
Everything aligned before final submission.
The project is organized so each member can edit a dedicated track page while the common site, report sources, and deployment surfaces stay coherent.
GitHub Pages handles the presentation layer, the LaTeX report handles the formal submission layer, and the Streamlit Community Cloud app carries the interactive image demo.
// Sync assignment surfaces
git push origin main
Publishing GitHub Pages... OK
// Route track pages
/assignment-1/image.html cnn / vit / calibration
/assignment-1/text.html bilstm / transformer
/assignment-1/multimodal.html clip / zero-shot / few-shot
// Live demo deployment
streamlit cloud deploy deploy/streamlit-cloud/app.py
status: assignment 1 ready for final polish