File size: 1,083 Bytes
e8e5544
ae425eb
34ea4ae
 
 
 
4d8c2c1
 
 
ae425eb
4d8c2c1
 
 
 
 
 
e8e5544
 
 
 
4d8c2c1
e8e5544
 
 
4d8c2c1
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
---
datasets:
- seungheondoh/LP-MusicCaps-MSD
- DynamicSuperb/MusicGenreClassification_FMA
- DynamicSuperb/MARBLEMusicTagging_MagnaTagATune
- agkphysics/AudioSet
language:
- en
license: mit
pipeline_tag: text-to-audio
tags:
- music
- art
- text-to-audio
model_type: diffusers
library_name: diffusers
---

## Model Description

This model, QA-MDT, allows for easy setup and usage for generating music from text prompts. It incorporates a quality-aware training strategy to improve the fidelity of generated music.

## How to Use

A Hugging Face Diffusers implementation is available at [this model](https://huggingface.co/jadechoghari/openmusic) and [this space](https://huggingface.co/spaces/jadechoghari/OpenMusic). For more detailed instructions and the official PyTorch implementation, please refer to the project's [Github repository](https://github.com/ivcylc/qa-mdt) and [project page](https://qa-mdt.github.io/).

The model was presented in the paper [QA-MDT: Quality-aware Masked Diffusion Transformer for Enhanced Music Generation](https://huggingface.co/papers/2405.15863).