π Judge No Question Full LogprobsΒΆ
templates.rag_eval.faithfulness.judge_no_question_full_logprobs
type: InputOutputTemplateWithCustomTarget
input_format: "Evidence: {contexts}\n\nPrediction: {answer}\n"
output_format: {is_faithful}
postprocessors:
- processors.load_json_from_predictions
- processors.infer_logprobs_to_yes_no_probs
- processors.cast_to_float_return_zero_if_failed
reference: {number_val}
target_prefix: Answer:
instruction: "System prompt: You are CompareGPT, a machine to verify the groundedness of predictions. Answer with only yes/no.\nYou are given a grounding evidence and a prediction from a model. Compare the \"Prediction\" and the \"Evidence\" to determine whether all the information of the prediction is present in the evidence or can be inferred from the evidence. You must answer \"no\" if there are any specific details in the prediction that are not mentioned in the evidence or cannot be inferred from the evidence.\n\n"
[source]References: processors.cast_to_float_return_zero_if_failed, processors.infer_logprobs_to_yes_no_probs, processors.load_json_from_predictions
Read more about catalog usage here.