Skip to content

SETUP_CHECKLIST.md

Latest commit

 

History

History
358 lines (276 loc) · 8.27 KB

SETUP_CHECKLIST.md

File metadata and controls

358 lines (276 loc) · 8.27 KB

🚀 T5 Training GUI - Setup Checklist

✅ Pre-Flight Checklist

Use this checklist to ensure everything is ready before using the GUI.

1️⃣ Backend Setup

Check Backend Dependencies

cd "c:\Users\suhas\Documents\GitHub\OntoBot\microservices"
pip list | Select-String "Flask"

Expected Output:

Flask              2.0.3
Flask-Cors         3.0.10

If missing, install:

pip install -r requirements.txt

Verify Backend Files

  • File exists: microservices/blueprints/t5_training.py
  • File modified: microservices/app.py (has t5_training_bp import)
  • File exists: rasa-bldg1/actions/sensor_list.txt

Start Backend Service

cd "c:\Users\suhas\Documents\GitHub\OntoBot\microservices"
python app.py

Expected Output:

 * Serving Flask app 'app'
 * Debug mode: on
 * Running on http://0.0.0.0:6000

✅ Backend Ready: Keep this terminal open

2️⃣ Frontend Setup

Install Dependencies

cd "c:\Users\suhas\Documents\GitHub\OntoBot\rasa-frontend"
npm install

This will install:

  • react-select (new)
  • All other existing dependencies

Expected Output:

added 1 package, and audited XXX packages in Xs
found 0 vulnerabilities

Verify Frontend Files

  • File exists: rasa-frontend/src/pages/ModelTrainingTab.js
  • File modified: rasa-frontend/src/pages/SettingsTabs.js (has ModelTrainingTab)
  • File modified: rasa-frontend/package.json (has react-select)

Start Frontend Service

cd "c:\Users\suhas\Documents\GitHub\OntoBot\rasa-frontend"
npm start

Expected Output:

Compiled successfully!
The app is running at: http://localhost:3000

✅ Frontend Ready: Browser opens automatically

3️⃣ Access the GUI

Navigate to Training Tab

  1. Browser opened to http://localhost:3000
  2. Click Settings in top navigation
  3. See 5 tabs: Edit & Validate | Train & Activate | Action Server | Analytics | T5 Model Training
  4. Click T5 Model Training tab
  5. Page loads without errors

Verify Components Loaded

  • "Add Training Example" card visible
  • "Training Examples (N)" table visible
  • "Train Model" card visible
  • "Available Models" card visible

Check Sensor Dropdown

  • Click "Sensors Involved" dropdown
  • Dropdown shows loading state
  • Sensors load (680+ options)
  • Can search by typing
  • Can select multiple sensors

✅ GUI Ready: All components working

4️⃣ Test Basic Functionality

Add a Test Example

Question: What is the temperature in zone 5.04?
Sensors: Zone_Air_Temperature_Sensor_5.04
SPARQL:
SELECT ?timeseriesId ?storedAt 
WHERE {
  bldg:Zone_Air_Temperature_Sensor_5.04 brick:timeseries ?timeseries .
  ?timeseries ref:hasTimeseriesId ?timeseriesId .
  ?timeseries ref:storedAt ?storedAt .
}
Category: Single Sensor
Notes: Test example
  • Fill in form
  • Click "Add Example"
  • See success alert
  • Example appears in table

Test Edit Functionality

  • Click ✏️ (Edit) button on test example
  • Form populates with example data
  • Modify question
  • Click "Update Example"
  • See updated data in table

Test Delete Functionality

  • Click 🗑️ (Delete) button on test example
  • Confirm in dialog
  • Example removed from table

✅ Basic CRUD: All operations working

5️⃣ Verify Data Files

Check Training Dataset

Get-Content "c:\Users\suhas\Documents\GitHub\OntoBot\Transformers\t5_base\training\bldg1\correlation_fixes.json"
  • File exists
  • Contains valid JSON
  • Has array of examples
  • Shows your added examples

Check Sensor List

Get-Content "c:\Users\suhas\Documents\GitHub\OntoBot\rasa-bldg1\actions\sensor_list.txt" | Measure-Object -Line

Expected Output:

Lines: 680+ (approximately)

✅ Data Files: Accessible and valid

6️⃣ Quick Training Test (Optional)

Add Minimal Training Set

  • Add 5 simple examples
  • Examples cover different query types
  • All examples have valid SPARQL

Run Quick Training

  • Set epochs to 3 (for quick test)
  • Click "Start Training"
  • Confirm dialog
  • Progress bar appears and updates
  • Logs stream in real-time
  • Training completes (may take 3-5 minutes)

Verify Training Output

  • Status shows "COMPLETED"
  • Progress bar at 100%
  • "Deploy Model to Production" button appears
  • No errors in logs

✅ Training Works: Model can be trained

7️⃣ Deployment Test (Optional)

Deploy Trained Model

  • Click "Deploy Model to Production"
  • Confirm deployment
  • See success message
  • Refresh models list
  • See backup created (checkpoint-3-backup-YYYYMMDD_HHMMSS)

Restart Action Server

  • Go to "Action Server" tab
  • Click "Restart Action Server"
  • Wait for "Restart completed successfully"
  • Check logs for no errors

Test in Chatbot

  • Open chatbot
  • Ask a question from your training examples
  • Verify correct SPARQL generated
  • Confirm query returns results

✅ Deployment Works: Model active in production

🎯 Final Verification

All Systems Operational

  • Backend running (port 6000)
  • Frontend running (port 3000)
  • GUI accessible
  • Sensors load
  • Can add examples
  • Can edit examples
  • Can delete examples
  • Can start training
  • Progress monitoring works
  • Logs stream correctly
  • Can deploy models
  • Models list updates

Documentation Available

  • QUICKSTART_GUI.md
  • T5_GUI_SETUP.md
  • GUI_TRAINING_GUIDE.md
  • T5_GUI_IMPLEMENTATION.md
  • T5_GUI_ARCHITECTURE.md
  • T5_GUI_COMPLETE.md

🆘 Troubleshooting Quick Reference

Backend Issues

# Check if port 6000 is in use
netstat -ano | findstr :6000

# Restart backend
cd microservices
python app.py

Frontend Issues

# Clear npm cache
npm cache clean --force

# Reinstall dependencies
rm -r node_modules
npm install

# Restart frontend
npm start

Browser Issues

# Clear browser cache
# Ctrl+Shift+R (hard refresh)
# Or clear cache in browser settings

File Permission Issues

# Check file exists and is readable
Test-Path "path\to\file"
Get-Content "path\to\file"

📊 Expected Results Summary

Component Expected State How to Verify
Backend API Running on port 6000 Terminal shows "Running on http://0.0.0.0:6000"
Frontend Running on port 3000 Browser shows interface
Sensor Dropdown 680+ sensors Dropdown populates
Training Examples Editable list Table displays examples
Training Progress monitoring Progress bar updates
Logs Real-time streaming Logs scroll automatically
Deployment One-click deploy Success message appears
Models List Shows checkpoints Table displays models

✅ Success Criteria

You're ready to use the GUI when ALL of these are true:

  1. ✅ Backend terminal shows "Running on http://0.0.0.0:6000"
  2. ✅ Frontend browser shows "Settings" page
  3. ✅ "T5 Model Training" tab is visible and clickable
  4. ✅ Sensor dropdown loads without errors
  5. ✅ Can add/edit/delete examples without errors
  6. ✅ Training can be started and monitored
  7. ✅ Models can be deployed successfully

🎉 Ready to Go!

If all checkboxes are ✅, you're ready to start training your T5 model!

Recommended Next Steps:

  1. Read QUICKSTART_GUI.md for a quick tutorial
  2. Add 5-10 training examples
  3. Train for 10 epochs
  4. Deploy and test

Happy Training! 🚀

📞 Need Help?

Quick References

  • Setup: See QUICKSTART_GUI.md
  • Usage: See GUI_TRAINING_GUIDE.md
  • Technical: See T5_GUI_IMPLEMENTATION.md
  • Architecture: See T5_GUI_ARCHITECTURE.md

Check These First

  1. Are both services running (ports 6000 and 3000)?
  2. Did you run npm install?
  3. Check browser console for errors (F12)
  4. Check terminal output for errors
  5. Try restarting both services

Common Fixes

  • Sensors not loading: Restart backend
  • Training won't start: Add at least 1 example
  • Model not active: Restart action server
  • Port in use: Kill process on port and restart