Can you run AI on your phone?
Image: Qdrddr, CC BY-SA 4.0, via Wikimedia Commons
Can you run AI on your phone?
Imagine you want to use a smart assistant app on your smartphone, but it's too heavy and drains your battery fast.
TensorFlow Lite is like a slimmed-down version of a popular AI toolkit that fits better on phones and doesn't use up as much battery.
Example
Instead of a bulky app that takes 2 hours to download, you get a lightweight version that takes only 30 minutes.
Remember this
TensorFlow Lite lets you use powerful AI on mobile devices without the heavy load.
Text adapted from Wikipedia, licensed under CC BY-SA 4.0.
a Triton kernel is
Can your phone run faster with a different brain?
XLA does for TensorFlow/JAX: compiles computation graphs for TPU/GPU execution
XLA accelerates TensorFlow models
the ONNX format does: standardizes model representation for cross-framework deployment
ONNX standardizes machine learning model representation
TensorRT does: NVIDIA's inference optimizer that quantizes and fuses operations
TensorRT optimizes deep learning inference by quantizing and fusing operations for NVIDIA GPUs
QLoRA adds
Ever wondered how computers can understand and mimic human speech?
tensor cores are
Why can computers crunch numbers faster than humans?
Swipe through 100 ML concepts daily
Open Pocket Polymath