ELI5 why do we have weird LLM sizes like 3b, 7b, 13b, 33b, 70b? Who invented those sizes?

The article is a conversation between a user and a chatbot. The user asks the chatbot to explain what a square root is and to provide three recipes where it could be used. The chatbot explains the mathematical concept of square root and provides three recipes involving ingredients with the name "root". The user then poses a hypothetical scenario involving a magician named Joe juggling fruits, and asks the chatbot to determine the final number and type of fruits Joe is juggling. The chatbot successfully answers the question, leading the user to commend its performance.

Key takeaways

A square root is a mathematical function that calculates the positive value that, when multiplied by itself, equals a given number.
In the context of cooking, square root doesn't have a direct meaning, but it can be interpreted as recipes involving ingredients with the name 'root' that appear in a square shape.
Joe, a magician, initially had 3 apples, then his assistant added 2 oranges and a banana. He transformed the oranges into apples, so he ended up juggling 5 fruits: 5 apples and 1 banana.
The model is upgrading from good and is considered pretty solid.

ELI5 why do we have weird LLM sizes like 3b, 7b, 13b, 33b, 70b? Who invented those sizes?

Key takeaways

Discussion (0)