Fix: Add missing librosa dependency installation in example code

## Description

The example code fails with `ModuleNotFoundError: No module named 'librosa'` when run in a clean environment. This is due to the `librosa` package not being pre-installed.

While this specific issue involves a missing ` 'librosa'` installation, we have observed that dependency-related failures can often be more complex — involving implicit, indirect, or version-sensitive packages. These issues are not always obvious but can significantly hinder reproducibility and user onboarding. As such, we believe it may be worth encouraging more consistent dependency transparency across example code.

## Changes
Added the following line at the beginning of the script to ensure all required dependencies are available:
```python
# Requires: librosa
```

## Testing
The code has been successfully tested and runs without error.

## Note
This contribution is part of an ongoing research initiative to systematically identify and correct faulty example code in Hugging Face Model Cards.
We would appreciate a timely review and integration of this patch to support code reliability and enhance reproducibility for downstream users.

README.md

@@ -105,6 +105,7 @@ After training, the model is evaluated on the test dataset, and the results are
 
 ## 🚀 **How to Use**
 ```python
+# Requires: librosa
 from transformers import AutoModelForAudioClassification, AutoFeatureExtractor
 import librosa
 import torch