Just a place to jot down ideas about prompting LLMs, whether full-service or self-hosted. Acts as a bit of a use-case playground. Consider formatting differences between model types.

Entity, Keyword and Keyphrase extraction

(Zero shot vas few-shot, etc.)

Sentiment analysis

Language translation

Spelling & Grammar Correction

Data clean-up

Data format conversion

Code generation

Tweet Generation

Summarization/paraphrase

Intent Classification

Zero-shot text classification

(identifying main topic)

In one word or phrase, what is the main topic of the following text? …CONTEXT…

Word, synonym & antonym finding

Product & project naming