File size: 437 Bytes
ee88ac3
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
This dataset contains a slice of 200 samples from the [AG News](https://huggingface.co/datasets/ag_news) dataset (test split). 
The picked 200 samples are potential misclassifications of the original test data. 

Approach
* Fine-tune DistilBERT with 10k samples from the training data (out of 120k)
* Do a forward pass with the model, storing the loss
* Sort the samples based on the loss

This is a repository for demonstration purposes