Timechils/sapiens-mvp

FIX: Add gunicorn to requirements for Hugging Face deployment

Deploy Phase 4: 90% Accuracy Model + PDF Reporting + Enterprise UI

Fix README colors

Add Hugging Face Configuration

Final Deployment: Sapiens Phase 3 (No Raw Data)

Migrate to Git LFS for model hosting

Migrate deployment to Git LFS

FIX: Set weights_only=False in torch.load for PyTorch 2.6

Switch to local model cache for Render deployment

FIX: Add missing /batch-predict route to Flask app

Final attempt: Clean up code for Render deployment

Fix final SyntaxError in inference script

Correct Render persistent disk path to /var/data/model_weights

Fix persistent disk path - use /mnt instead of /var/data

Fix persistent disk path - use /mnt instead of /var/data

FIX: Implement explicit model loading on Flask startup

Final clean push with all code and templates

Fix Render permission error and correct disk path

Clean up inference.py - fix formatting and ensure fallback class names work

Add model with Git LFS

Replace inference.py with corrected version - fix all syntax errors

Setup Render deployment with persistent disk

Fix Render permission error in inference.py

Fix SyntaxError in inference script

Fix potential syntax error in MODEL_URL

Refactor model loading for stability on Render startup

Add model download logic to inference script for Render deployment

Initial commit for Render deployment