Ever wonder how to train a smart device without sharing your personal data?
Federated learning
Ever wonder how to train a smart device without sharing your personal data?
Imagine you own a smart home device that learns from your habits, but you don't want to share your personal data with a central server.
Federated learning lets your device learn from your data locally, then shares only the insights, not the data itself, to improve the device's performance.
Example
Your device learns to adjust the thermostat based on your preferences without sending your temperature preferences to a central server.
Remember this
Federated learning respects your privacy while still allowing your device to become smarter.
Text adapted from Wikipedia, licensed under CC BY-SA 4.0.
Related concepts
AWQ does differently
AWQ selectively retains weights crucial for model performance, unlike traditional quantization
GPTQ quantization does
Post-training quantization using second-order information for model compression
Deep learning
Why do smart phones sometimes struggle to recognize new faces?
Vector quantization
How can you store a huge library of books in a tiny closet?
data augmentation does for generalization: artificially expands training set
How can you teach a computer to see better?
gradient accumulation simulates larger batch sizes without more memory
Can you train a machine like you do with a computer?
Swipe through more Machine Learning concepts
Open Pocket Polymath