Merge pull request #1 from chloelavrat/clavrat/proto-crepe

Clavrat/proto crepe --> main

.github/workflows/CD.yml

@@ -0,0 +1,62 @@
+name: CD
+
+on:
+  push:
+    branches:
+      - main
+      - clavrat/proto-crepe
+    tags:
+      - '*.*.*'  # Adjust this pattern to match your tag format
+
+jobs:
+  build_and_release:
+    name: Build and Upload Release
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.12'
+
+      - name: Install Poetry
+        run: |
+          pip install poetry
+
+      - name: Build the package
+        run: |
+          poetry build
+
+      - name: Create Release
+        id: create_release
+        uses: actions/create-release@v1.0.0
+        env:
+          GITHUB_TOKEN: ${{ secrets.PAT_TOKEN }}
+        with:
+          tag_name: ${{ github.ref }}
+          release_name: Release ${{ github.ref }}
+          draft: false
+          prerelease: false
+
+      - name: Get Name of Artifact
+        run: |
+          ARTIFACT_PATHNAME=$(ls dist/*.whl | head -n 1)
+          ARTIFACT_NAME=$(basename $ARTIFACT_PATHNAME)
+          echo "ARTIFACT_PATHNAME=${ARTIFACT_PATHNAME}" >> $GITHUB_ENV
+          echo "ARTIFACT_NAME=${ARTIFACT_NAME}" >> $GITHUB_ENV
+      
+      - name: Upload Whl to Release Assets
+        id: upload-release-asset
+        uses: actions/upload-release-asset@v1.0.2
+        env:
+          GITHUB_TOKEN: ${{ secrets.PAT_TOKEN }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }}
+          asset_path: ${{ env.ARTIFACT_PATHNAME }}
+          asset_name: ${{ env.ARTIFACT_NAME }}
+          asset_content_type: application/x-wheel+zip
+
+

.github/workflows/CI.yml

@@ -0,0 +1,28 @@
+name: CI
+
+on: [push, pull_request]
+
+jobs:
+  linter:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Lint with flake8
+        run: |
+          pip install flake8
+          flake8 ./crepe --count --select=E9,F63,F7,F82 --show-source --statistics
+#          flake8 ./crepe --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
+
+
+  doc_coverage:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Check documentation coverage with pydocstyle
+        run: |
+          pip install pydocstyle
+          pydocstyle ./crepe

.gitignore

@@ -160,3 +160,6 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+dataset/*
+wandb/*
+.prompts/*

README.md

@@ -1,2 +1,76 @@
-# TorchCrepe
-Implementation of the crepe pitch extractor in PyTorch 
+<div align="center">
+  <img src="TorchCREPE_banner.png" alt="Banner" style="border-radius: 17px; width: 100%; max-width: 800px; height: auto;">
+</div>
+
+<h3 align="center">
+  <b><a href="https://torch-crepe-demo.chloelavrat.com">Interactive Demo</a></b>
+  •
+  <b><a href="https://www.youtube.com">Video</a></b>
+  •
+  <b><a href="">Python API</a></b>
+</h3>
+
+<div align="center">
+  <a href="https://opensource.org/licenses/MIT">
+    <img src="https://img.shields.io/badge/License-MIT-blue.svg" alt="License">
+  </a>
+  <img src="https://img.shields.io/badge/python-3.12.4-blue.svg" alt="Python Versions">
+  <a href="https://github.com/chloelavrat/TorchCrepe/actions/workflows/CI.yml">
+    <img src="https://github.com/chloelavrat/TorchCrepe/actions/workflows/CI.yml/badge.svg" alt="CI">
+  </a>
+</div>
+
+<p align="center">The <b>Torch-CREPE</b> project re-develop the CREPE pitch estimation model in PyTorch, empowering its optimization and adaptation for real-time voice pitch detection tasks. By re-developping this deep learning-based system, we unlock new research possibilities for music signal processing and audio analysis applications.</p>
+
+## How it Works
+
+The **PyTorch CREPE** implementation utilizes the **Torch** and **Torchaudio** library to process and analyze audio signals. The project's core functionality is based on the CREPE model, which estimates fundamental frequencies from audio data. 
+
+The way this model achieve this is by doing a classification of 20ms audio chunks on 350 classes representing the audio range in cents of the observed fundamental frequency.
+
+## Features
+
+- **Real-time pitch detection:** Processing done in realtime using the given script. 
+- **Optimized for instrument and voices:** Trained on instruments and voices for maximum usescases focuses.
+- **Deep learning-based**: system with full PyTorch implementation
+- **Fast Integration** with Torchaudio library
+- **Trainable on Consumer GPU** (complete train done on an RTX-3080)
+
+## Run app locally
+
+To run the PyTorch CREPE demo locally, you can use the following Python code:
+
+```py
+import torchaudio
+from crepe.model import crepe
+from crepe.utils import load_test_file
+
+crepe = crepe(model_capacity="tiny", device='cpu')
+
+audio, sr = load_test_file()
+
+time, frequency, confidence, activation = crepe.predict(
+    audio=audio,
+    sr = sr
+)
+```
+
+## Python API
+
+For a detailed documentation of the PyTorch CREPE implementation, including the API and usage guidelines, please refer to [this link].
+
+## Train
+
+The model is still in my training queue so only the 'tiny' version of **Crepe** has been trained yet.
+
+## Datasets
+
+[MIR-1K](http://mirlab.org/dataset/public/MIR-1K.zip)
+
+## Contributing
+
+This project is an open-source project, and contributions are always welcome. If you would like to contribute to the project, you can do so by submitting a pull request or by creating an issue on the project's GitHub page.
+
+## License
+
+This project is licensed under the MIT License. See the [LICENSE](LICENSE) file for details.