LabeledPairwiseStringEvalChain()Whether this evaluator requires an input string.
Output parser to use for agent.
Whether to return only the final parsed result.
The keys to use for the output.
Optional memory object.
[DEPRECATED] Use callbacks instead.
The keys to use for the output.
Labeled Pairwise String Evaluation Chain.
A chain for comparing two outputs, such as the outputs of two models, prompts, or outputs of a single model on similar inputs, with labeled preferences.
Prepare prompts from inputs.
Set the chain verbosity.