Update README.md

README.md

@@ -42,7 +42,7 @@ Install all required Python packages:
 
  - **report_to**: **Disabled (no Weights & Biases)**
 
-## 🧊 Model Quantization (CPU Inference)
+## 🧊 Model Quantization
 
  We apply dynamic quantization on the trained model to reduce size and enable CPU inference:
 
@@ -52,8 +52,7 @@ quantized_model = torch.quantization.quantize_dynamic(
 )
 ```
 
-## 📏 Evaluation (Optional)
-The BLEU score section is commented out but can be enabled by:
+## 📏 Evaluation
 
 ```python
 from evaluate import load