Update README.md

README.md

@@ -25,13 +25,14 @@ configs:
     path: data/train-*
 ---
 This __hate speeh__ dataset is derived from [manueltonneau/arabic-hate-speech-superset](https://huggingface.co/datasets/manueltonneau/arabic-hate-speech-superset). 
-The original dataset has loading issues. This dataset is based on the __L-HSAB__ and __alsafari__ sets mentioned in the original.   
-The current dataset keeps the original class (labels) and adds 2 columns: label_text and label (0: negative, 1:positive). 
+The original dataset has loading issues. This dataset is mainly based on the __L-HSAB__ and __alsafari__ sets mentioned in the original. 
+Around 3000 other records from other datasets were added for better (not full) coverage.  
+The current dataset keeps the original class (labels - apparently not as accurate as should be) and adds 2 columns: label_text and label (0: negative, 1:positive). 
 Label text was first generated by LLM (Qwen 2.5 14B quantized). Label was fully revised after that and reflects my personal opinion (not an exat science). 
 I plan to add more rows to this dataset to cover regional nuances (dialects, terminology, ..).  
 
 A [hate speech classifier model](https://huggingface.co/akhooli/setfit_ar_hs) based on this dataset is available.  
 
-* Note: this dataset was updated on Nov. 12, 2024.  
+* Note: this dataset was updated on Nov. 14, 2024. It has around 11500 entries all manually inspected and labeled.  
 * __Disclaimer__: some entries contain obscene language, swearing and other forms of unacceptable terminology. 
 The purpose of the dataset is to train a classifier and not to publish or endorse unacceptable content.