Update README.md

README.md

@@ -1,17 +1,17 @@
 ---
 license: mit
 language:
-- en
+- py
 metrics:
 - accuracy
 - f1
 - precision
 - recall
+dataset : https://github.com/jing-qian/A-Benchmark-Dataset-for-Learning-to-Intervene-in-Online-Hate-Speech
 base_model: google-bert/bert-base-uncased
 pipeline_tag: text-classification
 library_name: transformers
 ---
-# Group-3
 
 # Hate Speech Detection on the Reddit platform
 
@@ -101,4 +101,3 @@ BERT</pre>
 The performance metrics for the Simple NN, CNN, and LSTM models were similar, achieving an accuracy of 85% and an F1 score of 71% for hate speech detection. In contrast, the BERT model outperformed the others, achieving an accuracy of 89% and an F1 score of 89%.
 Based on these results, we selected BERT as our final model for hate speech detection and saved this model for further implementation and deployment.</p>
 
-### The individual implementations of the models can be found in separate branches of the repository. Each team member experimented with different approaches and models, and we decided to use the best methods from these individual efforts.