Update README.md

README.md

@@ -11,7 +11,7 @@ pipeline_tag: text-classification
 
 This is a text classification model based on `answerdotai/ModernBERT-large`, fine-tuned to distinguish between **human-written**, **AI-refined** and **AI-generated** code.
 
-The model was trained on the `DroidCollection` dataset. It's designed as a **4-class classifier** to address the core task of AI code detection.
+The model was trained on the `DroidCollection` dataset. It's designed as a **ternary classifier** to address the core task of AI code detection.
 
 A key feature of this model is its training objective, which combines standard **Cross-Entropy Loss** with a **Batch-Hard Triplet Loss**. This contrastive loss component encourages the model to learn more discriminative embeddings by pushing representations of human vs. machine code further apart in the vector space.