Update templates/grammar.html

Update templates/vocab.html

Update templates/translate.html

Update templates/translate.html

Update templates/vocab.html

Update leotsha/views_v2.py

Update leotsha_project/settings.py

Update leotsha/views_v2.py

fix: restore whatsapp redirect syntax

Update leotsha/views_v2.py

feat: add SedibaLM inference client

Update leotsha/views_v2.py

Update leotsha/views_v2.py

add SedibaLM inference client

docs: add AGPL-3.0 licence to leotsha Space

Upload README.md with huggingface_hub

Upload README.md with huggingface_hub

fix: SameSite=None for CSRF cookie in HF Space iframe context

temp: DEBUG=True to diagnose CSRF

fix: add HF Space domain to CSRF_TRUSTED_ORIGINS

fix: STATIC_ROOT, migrate on startup, proper startup script

fix: run migrations before gunicorn on startup

fix: DEBUG=False, add HF Space CSRF origin, whitenoise

fix: minimal clean requirements.txt - 18 packages only

fix: remove onboard and all Ubuntu desktop system packages

fix: remove nemo-emblems, Docker-verified clean install

fix: pin Mako to stable 1.3.12, requirements now Docker-verified

fix: remove gyp and all remaining system packages

fix: remove defer==1.0.6 and remaining system packages

fix: remove dbus-python system package

fix: remove command-not-found and remaining system packages

fix: remove GPU, Jupyter, dev packages from requirements

fix: remove system packages from requirements.txt

fix: make collectstatic non-fatal during Docker build

fix: use valid HF color values in README

fix: add required HF Spaces README frontmatter

fix: correct Dockerfile CMD syntax

fix: run model download before gunicorn

feat: add model download script for HF Space startup

fix: remove models from repo — load from HF Hub at runtime

fix: remove large model files from repo, load at runtime

feat: include model files for HF Space deployment

feat: initial Morutabana deployment

Add .gitignore and .dockerignore to exclude large files

Initial Morutabana deployment: XLM-R + EduIntel + sentiment classifier

EduIntel v1.1 — novel boost, expanded glossary, hallucination guard

EduIntel v1 — RAG pipeline + generation layer complete