Update README.md

README.md

@@ -131,8 +131,6 @@ is applied. Lastly, for the metrics/accuracy_top5, this shows up as a horizontal
 Overall, my model indicates high accuracy and no cases of overfitting, however, the model could benefit from longer training time to
 allow the curves to smoothen out further and reach an eventual straight horizontal line. 
 
----
-
 
 # Limitations and Biases
 ### Failure Cases
@@ -143,8 +141,8 @@ allow the curves to smoothen out further and reach an eventual straight horizont
 - Decisions on which classes belong in which were made based on Seattle's disposal guidelines, which can’t be used worldwide or statewide due to different disposal requirements and regulations among different areas.
 - Model lacks representation and diversity in types of items represented in each class
 
-### Poor Performing Class <u>Compost<u> 
-- Majority of images in the compost class feature pixelated or low quality images, making it difficult for the model to correctly identify items in this class.  
+### Poor Performing Class: Compost
+- Majority of images in the compost class feature pixelated or low quality images, making it difficult for the model to identify items in this class correctly.