YourMT3

YourMT3

Runs the YourMT3 demo in a Gradio app. Its folder names point to turning audio into MIDI music transcription, and it suits people who work with music and audio models.

Made by mimbres on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: YourMT3 is a Gradio demo of a model with the same name, described in a paper on arXiv. The repository's amt and mid folders suggest automatic music transcription and MIDI files, although the source states no purpose in words. The Space names six models and runs on GPU hardware at Hugging Face.

Who it suits: Researchers and developers who work with music or audio models and want to try this one in a browser.

Industry | Who uses it | How they use it | Example
Research | Music information researchers | Try the model on audio before a study | Run one of the included examples
Software | Developers of music tools | Check what the model produces | Compare its output with their own pipeline
Education | Students of audio machine learning | Study a model built on T5 and Perceiver parts | Read the amt folder
Music | Musicians curious about AI tools | See what an audio model does | Try the examples folder

People who want a polished music app can skip it, since this is a research demo.

Where to use it:

• A browser, by visiting the Space on Hugging Face
• Windows, macOS or Linux with Python, for a copy on your own computer
• A local Gradio page, once app.py is running

Use case | Who | What they get back
Try the model | Researchers | The output of the YourMT3 demo
Study the code | Developers | A model package with helper files

How to set it up:

• Python and pip on Windows, macOS or Linux
• The Python packages in requirements.txt
• Gradio 4.39.0, the version the Space uses
• A GPU, since the Space is set to zero-a10g

1. Install the Hugging Face hub package.
2. Download the Space with the hf command.
3. Move into the new folder.
4. Install gradio at the version the Space uses, then the listed packages.
5. Start the app with Python.

pip install -U huggingface_hub
hf download mimbres/YourMT3 --repo-type space --local-dir YourMT3
cd YourMT3
pip install gradio==4.39.0
pip install -r requirements.txt
python app.py

How to use it: Online, visit the Space on Hugging Face and pick an example or add your own audio, if the page offers it. On your own computer, run the start command and open the local page Gradio serves.

python app.py

Tool | What it does
app.py | Starts the Gradio interface
model_helper.py | Holds helper code for the model
html_helper.py | Holds helper code for the page
examples | Holds sample files to try

Example: You visit the Space, choose one of the included examples and run it. You look at what the model returns for that sample.

For newcomers and experts: If you are new, start with an included example so you do not need your own file. Then read the paper linked from the Space for the purpose of the model.

If you know the field, the Space lists an audio encoder from Facebook, a Perceiver from DeepMind, three small T5 models from Google and the YourMT3 weights from mimbres. That mix of an audio encoder, a Perceiver and T5 models fits a sequence-to-sequence design, but the source does not describe the architecture. The trade-off is a GPU requirement.

You pick an example or add audio
The audio models process it
The sequence models produce a result
You review the output

Good to know:

• Purpose: the source gives only a paper reference and folder names, so read the paper for details
• Hardware: the Space is set to zero-a10g
• Models: it names six models on Hugging Face
• License: apache-2
• Cost to run: GPU hardware, no API key or paid service is named
• Data it reaches: the audio you give it and the models it fetches from Hugging Face

On Hugging Face, 13 November 2026

• Likes: 133
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/mimbres/YourMT3
• Files: https://huggingface.co/spaces/mimbres/YourMT3/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.