diff --git a/src/data/nhl_positivity_index.md b/src/data/nhl_positivity_index.md
index 35bc127..b2e81d4 100644
--- a/src/data/nhl_positivity_index.md
+++ b/src/data/nhl_positivity_index.md
@@ -28,7 +28,7 @@ With the hopes of improving the accuracy of [**cardiffnlp/twitter-roberta-base-s
 
 ## Fine-tuning the Model
 
-With the help of Hugging Face's PEFT: Parameter-Efficient Fine-Tuning library we were able to effectively fine-tune [**cardiffnlp/twitter-roberta-base-sentiment-latest**](https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest). We further fine-tuned the RoBERTa-bas model that was trained on ~124M tweets from January 2018 to December 2021 and was fine-tuned on sentiment analysis, in order to better fit our specific task of classifying hockey related comments. After fine-tuning, our Adapter model for [**cardiffnlp/twitter-roberta-base-sentiment-latest**](https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest), [Chelberta](https://huggingface.co/UAlbertaUAIS/Chelberta) achieved an accuracy score of 81.2% improving from the base model of 79.2% on our testing dataset mentioned above. The confusion matrix for our model, [Chelberta](https://huggingface.co/UAlbertaUAIS/Chelberta), can be found below.
+With the help of Hugging Face's PEFT: Parameter-Efficient Fine-Tuning library we were able to effectively fine-tune [**cardiffnlp/twitter-roberta-base-sentiment-latest**](https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest). We further fine-tuned the RoBERTa-base model that was trained on ~124M tweets from January 2018 to December 2021 and was fine-tuned on sentiment analysis, in order to better fit our specific task of classifying hockey related comments. After fine-tuning, our Adapter model for [**cardiffnlp/twitter-roberta-base-sentiment-latest**](https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest), [Chelberta](https://huggingface.co/UAlbertaUAIS/Chelberta) achieved an accuracy score of 81.2% improving from the base model of 79.2% on our testing dataset mentioned above. The confusion matrix for our model, [Chelberta](https://huggingface.co/UAlbertaUAIS/Chelberta), can be found below.
 
 ## Data Labelling Process