Update README.md

README.md

@@ -80,7 +80,7 @@ print("Generated Text:", generated_text)
 
 The model was evaluated using standard OCR benchmarks with an emphasis on Tamil text recognition. The primary evaluation metric was character-level accuracy and Word Error Rate (WER).
 
-- **Character Accuracy**: Achieved ~90% accuracy on validation sets.
+- **Character Accuracy**: Achieved ~79% accuracy on validation sets.
 - **Train Loss**: 0.063800
 - **Validation Loss**: 0.172539	
 - **CER**: 0.072717