Closed loubnabnl closed 1 year ago
Issue https://github.com/bigcode-project/bigcode-evaluation-harness/issues/24 Changed docstring of postprocess_results to require output to be a dict indicating metric name This is needed for BLEU tasks, but code_eval already outputs pass@k which is the correct metric name.
postprocess_results
dict
Issue https://github.com/bigcode-project/bigcode-evaluation-harness/issues/24 Changed docstring of
postprocess_resultsto require output to be adictindicating metric name This is needed for BLEU tasks, but code_eval already outputs pass@k which is the correct metric name.