LLMs can invoke external tools and APIs
Image: Pilarbini, CC BY-SA 4.0, via Wikimedia Commons
LLMs can invoke external tools and APIs
By invoking external tools and APIs, LLMs can access a vast array of information and services, enhancing their ability to perform complex tasks. This integration allows LLMs to overcome limitations in their own knowledge base and processing capabilities.
Example
An LLM can use an external weather API to provide accurate weather forecasts based on user queries, combining its language processing abilities with real-time weather data.
Remember this
This capability significantly expands the potential applications of LLMs, enabling them to perform tasks that require up-to-date information or specialized knowledge beyond their training data.
Text adapted from Wikipedia, licensed under CC BY-SA 4.0.
Attention Is All You Need
"Attention Is All You Need" introduced the transformer architecture in 2017
ring attention does: distributes long sequences across multiple devices
"Attention Is All You Need" introduced the transformer architecture in 2017
MLIR (software)
MLIR was publicly released as part of LLVM in 2019
Dependent type
Dependent types depend on values, not just types
Lambda calculus
Lambda calculus represents data using only functions
continuation-passing style (CPS) does: makes control flow explicit via callbacks
Continuations pass control explicitly via callbacks
Swipe through 100 ML concepts daily
Open Pocket Polymath