How to Use ExplainGrade
Two ways to use the system: Directly in your browser or locally for high-accuracy workloads.
🌐 Option 1: Browser-Based
No setup required. Perfect for grading individual answers or small batches offline. Zero data leaves your computer.
01 Live Demo
Paste a reference and student answer. Get scores instantly with XAI dashboard.
02 Batch CSV
Upload a CSV file for multi-row grading without any manual effort.
03 Class Eval (XLSX)
Grade student summaries vs a teacher's script. Requires .docx for transcript and .xlsx for student summaries.
🖥️ Option 2: Local Runner
Uses a high-performance **sentence-transformers** model (`all-MiniLM-L6-v2`) for better semantic accuracy than browser-based proxies.
# Clone and setup
git clone repo_urlcd repo_folder
pip install -r requirements.txt
# Run the grader
python local_grader.py \--docx transcript.docx \
--xlsx summaries.xlsx \
--max-score 10
Expected Column Headers
The .xlsx file should contain these headers (or similar):