rageval
rageval copied to clipboard
Evaluation tools for Retrieval-augmented Generation (RAG) methods.
I was wondering if there is a publicly available link to the Dragonball dataset that you could share. Thank you
1. Modify the `Metric` class, take the `_compute_one` as the abstract method and write a default implementation of `_compute_batch` method. 2. Rewrite the `F1` , `ChrF` and `Ter` metric.
Added Chinese F1 metric computation with optional parameter.
added Chinese F1 metric computation
I add a new evaluation metric, Distinct 1/2, for the generate task evaluation. I have uploaded the new scripts "_answer_distinct12.py" and "test_answer_distinct12.py", and the modified version of "__init__.py".