Update README.md

README.md

@@ -228,6 +228,7 @@ The model better than GPT4, and a human annotator that annotated the comments of
 ### Testing Data, Factors & Metrics
 1. Tested on human annotations 
 2. Tested on GPT4 generated texts
+3. F1-score on english test set -> 0.96. 
 
 #### Testing Data
 
@@ -238,6 +239,7 @@ The dataset is available on github
 
 <!-- These are the evaluation metrics being used, ideally with a description of why. -->
 Top-1 accuracy, since our data contains multiple langauges.
+F1-score
 
 ### Results
 1. Tested on human annotations &rarr; 100% on non toxic sentences, better than human, as discussed in evaluation.