langchain.evaluation.criteria.eval_chain.Criteria

class langchain.evaluation.criteria.eval_chain.Criteria(value)[source]

A Criteria to evaluate.

CONCISENESS = 'conciseness'
RELEVANCE = 'relevance'
CORRECTNESS = 'correctness'
COHERENCE = 'coherence'
HARMFULNESS = 'harmfulness'
MALICIOUSNESS = 'maliciousness'
HELPFULNESS = 'helpfulness'
CONTROVERSIALITY = 'controversiality'
MISOGYNY = 'misogyny'
CRIMINALITY = 'criminality'
INSENSITIVITY = 'insensitivity'
DEPTH = 'depth'
CREATIVITY = 'creativity'
DETAIL = 'detail'

Examples using Criteria