---
datasets:
- lewtun/music_genres_small
base_model:
- facebook/wav2vec2-large
metrics:
- accuracy
- f1
tags:
- audio
- music
- classification
- Wav2Vec2
pipeline_tag: audio-classification
---

# Music Genre Classification Model 🎶
This model classifies music genres based on audio signals. It was fine-tuned on the `music_genres_small` dataset using the Wav2Vec2 architecture.
You can find a GitHub repository with an interface hosted by a Flask API to test the model:  **[music-classifier repository](https://github.com/gastonduault/Music-Classifier)**

## Metrics
- **Validation Accuracy**: 75%
- **F1 Score**: 74%
- **Validation Loss**: 0.77
  
## Example Usage
```python
from transformers import Wav2Vec2ForSequenceClassification, Wav2Vec2FeatureExtractor
import torch

# Load model and feature extractor
model = Wav2Vec2ForSequenceClassification.from_pretrained("gastonduault/music-classifier")
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained("facebook/wav2vec2-large")

# Process audio file
audio_path = "path/to/audio.wav"
audio_input = feature_extractor(audio_array, sampling_rate=16000, return_tensors="pt", padding=True)

# Predict
with torch.no_grad():
    logits = model(audio_input["input_values"])
    predicted_class = torch.argmax(logits.logits, dim=-1)
print(predicted_class)