In this example, we will demonstrate how to apply an explainability tool on a trained classifcation model.
At the end of the tutorial, the user will be able to
- fine-tune
transformers.DistilBertForSequenceClassification
model - use
captum.attr.LayerIntegratedGradients
for explainability
After deploying the model as an endpoint, you can use test_endpoint.ipynb
to test it and visualize the results.