README.md

---
license: mit
---

# EMusicGen
The model weights for generating ABC melodies by emotions.

## Demo
<https://www.modelscope.cn/studios/monetjoe/EMusicGen>

## Maintenance
```bash
git clone git@hf.com:monetjoe/EMusicGen
cd EMusicGen
```

## Fine-tuning results
<style>
    td {
        vertical-align: middle !important;
    }
</style>

| Dataset |                                        Loss curve                                        |     Min eval loss     |
| :-----: | :--------------------------------------------------------------------------------------: | :-------------------: |
| VGMIDI  | ![](https://www.modelscope.cn/models/monetjoe/EMusicGen/resolve/master/vgmidi/loss.jpg)  | `0.23854530873296725` |
| EMOPIA  | ![](https://www.modelscope.cn/models/monetjoe/EMusicGen/resolve/master/emopia/loss.jpg)  | `0.26802811984950936` |
| Rough4Q | ![](https://www.modelscope.cn/models/monetjoe/EMusicGen/resolve/master/rough4q/loss.jpg) | `0.2299637847539768`  |

## Usage
```python
from modelscope import snapshot_download

model_dir = snapshot_download("monetjoe/EMusicGen")
```

## Mirror
<https://www.modelscope.cn/models/monetjoe/EMusicGen>

## Cite
```bibtex
@article{Zhou2024EMusicGen,
  title     = {EMusicGen: Emotion-Conditioned Melody Generation in ABC Notation},
  author    = {Monan Zhou, Xiaobing Li, Feng Yu and Wei Li},
  month     = {Sep},
  year      = {2024},
  publisher = {GitHub},
  version   = {0.1},
  url       = {https://github.com/monetjoe/EMusicGen}
}
```

## Reference
[1] [Wu, S., & Sun, M. (2023). TunesFormer: Forming Tunes with Control Codes. ArXiv, abs/2301.02884.](https://arxiv.org/pdf/2301.02884)