Update README.md

README.md

@@ -56,7 +56,7 @@ print(answer)
 # #### 294
 ```
 
-## 📚 Training Details
+## Training Details
 
 ### Training Data
 The model was trained on the **GSM8K dataset**. The original answers were replaced with detailed, step-by-step solutions generated by the `Gemini-2.5-Flash` teacher model. This process of knowledge transfer is known as **Reasoning Distillation**.
@@ -82,7 +82,7 @@ The model was trained using a standard **Supervised Fine-Tuning (SFT)** process
 
 ---
 
-## 📊 Evaluation
+## Evaluation
 The model was evaluated on the **GSM8K benchmark (5-shot)** and achieved the highest performance among all tested training pipelines.
 
 | Model | Flexible-extract Exact Match (±stderr) | Strict-match Exact Match (±stderr) |