Examples: 0
Learned: 0
UNTRAINED
Network Architecture
Training Controls
EPOCHS 3000
LEARN RATE 0.5
HIDDEN SIZE 32
Training Loss
β€”
System Log
> 15 categories loaded. 150+ training examples ready.
> Wikipedia self-learning enabled.
> Unknown inputs will trigger a Wiki search.
> New knowledge auto-added as training examples.
Categories β€” What the AI Understands
Training Data
🧹 Category Cleanup
Merges garbage auto-created categories into proper ones (Wolf→Animals, Sahara→Nature, etc.) and deletes meaningless ones (Example.com, State, Head, etc.). Run this after auto-learn.
πŸ€– Auto-Learn β€” Wikipedia + DuckDuckGo Crawler
DURATION
Words crawled: 0 New examples: 0 Wiki: 0 Β· DDG: 0 Time left: --:--
> Auto-learn ready. Set duration and click START. Uses Wikipedia first, DuckDuckGo as fallback.
Chat β€” Talk to Your AI (Unknown inputs trigger Wikipedia learning)
Train the model first, then start chatting.
Unknown inputs will automatically search Wikipedia and learn!