Transformer
Generative AI & LLMsChapter 4: Choosing the LLMTerm 68 of 253
The neural network architecture behind modern LLMs (2017 paper 'Attention Is All You Need').
At Sunrise School
Gopal asksWhat's the design inside the LLM?
Bit answersA Transformer. It reads the whole message at once, uses attention to connect the words, then predicts my reply token by token.
In the real worldEngine inside every modern LLM.
Watch the short
Short coming soonSeason 1 · Day 13Follow @haigopala
Short coming soonSeason 1 · Day 13Follow @haigopala.telugu
Short coming soonSeason 1 · Day 13Follow @haigopala.hindi
Quick check
Quick check. Does a Transformer read one word at a time and forget the rest? Yes or no?
No! Attention lets every word look at every other word.
Practise it
Try this term and all 253 others in the free Model Training Arena.
Open the Arena →Follow Bit daily: YouTube English · తెలుగు · हिंदी | Instagram English · తెలుగు · हिंदी