[[Large Language Models|LLM]]

[[Transformer]]

[[Tokens#Tokenizer|Tokenizer]]

[[Transformer#Transformer Block|Transformer Blocks]]

Transformer Block 1

[[Transformer#Self-Attention Layer|Self-Attention Layer]]

[[Transformer#Feedforward Layer|Feedforward Layer]]

Transformer Block n

[[Transformer#Self-Attention Layer|Self-Attention Layer]]

[[Transformer#Feedforward Layer|Feedforward Layer]]

[[LM Head]]

Prompt

Ausgabe

max. Token oder Ende der Antwort erreicht?

Ende