Update README.md

README.md

@@ -18,7 +18,9 @@ pipeline_tag: text-classification
 
 Classifies 512 chunks of a news article as "Likely" or "Unlikely" biased/disinformation.
 
-  Fine-tuned BERT ([bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased)) on the `headline`, `aritcle_text` and `text_label` fields in the [News Media Bias Plus Dataset](https://huggingface.co/datasets/vector-institute/newsmediabias-plus).
+Fine-tuned BERT ([bert-base-uncased](https://huggingface.co/google-bert/bert-base-uncased)) on the `headline`, `aritcle_text` and `text_label` fields in the [News Media Bias Plus Dataset](https://huggingface.co/datasets/vector-institute/newsmediabias-plus).
+
+**This model was trained without weighted sampling, and the dataset contains 81.9% 'Likely' and 18.1% 'Unlikely' examples.** The same model trained with weighted sampling preformed worse on training eval metrics, but better when evaluated by gpt-4o-mini as a judge and is available [here](https://huggingface.co/maximuspowers/nmbp-bert-full-articles-balanced).
 
 ### Metics