Closed wizare closed 2 years ago
Sorry for the late reply! This somehow slipped my mind.
If it still helps, you can find semantic similarity and consistency metrics in our repo: here and here. For both metrics, we implemented two methods: compute_metric
(for one sample) and compute_metric_for_file
(for an entire file).
For word-overlap metrics, you can use this repo.
Hope this helps!
Hi. I have already trained my BERT model and tested my responses with the entailment model.
But I didn't know how to test with other metrics, i.e.,
Semantic Similarity
,Word-level metrics
,Consistency by textual entailment
.Can you show a script as an example? Thanks