Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 

Repository files navigation

https://www.reddit.com/r/MachineLearning/comments/1aji7np/d_microsoft_researchs_evoprompt_evolutionary/

Genes: - Attributes: - Content -> either some piece of code or some text - Fitness -> a single number that represents how good this gene is - Characteristics -> String description of the content - Methods: - Mutate -> apply a random mutation to the gene - Combine -> combines two genes using the

Population: - Attributes: - Organisms -> List of Genes - Generation # -> Current generation - Best organism - Methods - Init -> Size of the population - Run Generation: - Calculate the fitness of all of the genes - Only keep the top N or top K % - Mutate and combine them - Something to initialize the genes?

Evaluator: - Function that takes a gene and returns the fitness - Potentially could also return feedback

Mutator: - Init -> some prompt - Mutates genes in one way or another

Breeder: (the correct term is Crossover but I think Breeder is funnier) - Init -> some prompt - Combines the characteristics of two genes

"# EvolutionaryLLMs"

About

LLM-based Genetic Algorithms

Resources

Stars

1 star

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages