docs: add Hugging Face demo link to README

refactor: update auth, preprocessing, retraining, sentiment, Dockerfile and pipeline tests

fix: auth login failure, session persistence, and Indian helplines

added persistant login

feat: add Model Evaluation page with cross-model metric comparison

redesign: overhaul UI with warm limestone theme and fix readability

fix: persist sessions across logouts and redesign UI with warm limestone theme

improve: increase min_feedback to 50 and configure persistent storage path in Docker

Fix f-string unicode escape issue

Fix HF deployment: Dockerfile + supervisor config + README metadata

Fix HF Space config: enable docker

modification: minor changes

Add Docker deployment config for HuggingFace Spaces

Add Streamlit dashboard, test suite, and Colab notebooks

Expose FastAPI backend with MC-Dropout inference and RL feedback loop

Build safety-first recommendation engine with crisis circuit breaker

Add sentiment analysis, text preprocessing, and RL reward signal

Add training script with focal loss, FPR-constrained threshold calibration

Build preprocessing pipeline for Dreaddit, Reddit, and Twitter datasets

Implement multichannel CNN with multi-head attention and transformer wrappers

Add JWT authentication, bcrypt password hashing, and Fernet encryption

Define SQLite schema for users, sessions, and feedback

Initial project setup with dependencies and environment config