---
title: Orpheus
emoji: 🦀
colorFrom: indigo
colorTo: pink
sdk: streamlit
sdk_version: 1.34.0
app_file: ui/app.py
pinned: false
---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference


# Orpheus_ai- User Study

## Introduction

This repository hosts the code for conducting the user study for Orpheus_ai, a system designed for music generation. The user study aims to assess the efficacy of Orpheus_ai in creating music compositions.

## Requirements

Before initiating the user study, ensure the following prerequisites are met:

1. **Install Required Packages:**

    Set up a virtual environment and install the necessary packages using the provided `requirements.txt` file:
    ```
    pip install -r requirements.txt
    ```

    Alternatively, if you opt not to use a virtual environment, manually install the required versions of `diffusers` and `audiodiffusion` packages:
    ```
    pip install diffusers==0.17.1 audiodiffusion==1.5.6
    ```

    Note: Ensure compatibility between the installed Torch version and the CUDA version on your system since a GPU is necessary for running the code.

2. **Create Folders for Data Storage:**

    Create the following directory structure within the root directory for storing songs and spectrograms:
    ```
    DataSet
    ├───Song
    ├───Spec
    ```

## Usage

To execute the user study and generate songs along with spectrograms, follow these steps:

1. Navigate to the `ui` directory:
    ```
    cd ui
    ```

2. Launch the user study interface using Streamlit:
    ```
    streamlit run app.py
    ```

    Note: It's advisable to select only one song from the provided list during the initial phase. After generating the song and submitting the ratings, the resulting song and spectrogram will be saved in the respective folders. The files follow the naming convention:
    ```
    <model_name>_<song_name>_<similarity>_<rating>.wav
    <model_name>_<song_name>_<similarity>_<rating>.npy
    ```