https://github.com/cansik/mediapipe-numpy2
A patching tool to remove the numpy<2 constraint from official mediapipe wheels.
https://github.com/cansik/mediapipe-numpy2
mediapipe numpy python
Last synced: 6 days ago
JSON representation
A patching tool to remove the numpy<2 constraint from official mediapipe wheels.
- Host: GitHub
- URL: https://github.com/cansik/mediapipe-numpy2
- Owner: cansik
- License: mit
- Created: 2025-06-11T12:25:21.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2026-01-15T14:12:18.000Z (7 days ago)
- Last Synced: 2026-01-15T15:47:12.916Z (7 days ago)
- Topics: mediapipe, numpy, python
- Language: Python
- Homepage:
- Size: 42 KB
- Stars: 8
- Watchers: 0
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# mediapipe-numpy2
[](https://github.com/cansik/mediapipe-numpy2/actions/workflows/patch.yaml)
[](https://github.com/cansik/mediapipe-numpy2/actions/workflows/test-mediapipe-numpy2.yaml)
[](https://pypi.org/project/mediapipe-numpy2/)
A patching tool to remove the `numpy<2` constraint from
official [mediapipe](https://github.com/google-ai-edge/mediapipe) wheels.
### Why does this repository exist?
The official mediapipe Python wheels published on PyPI include a strict dependency:
```
Requires-Dist: numpy<2
```
This prevents installation or use of mediapipe with `numpy` version 2 or higher. This constraint is problematic for users and downstream projects who want to use the latest `numpy` (see [mediapipe issue #5612](https://github.com/google-ai-edge/mediapipe/issues/5612)). As of now, a lot of mediapipe seems to work with `numpy` 2.x, but the PyPI wheels block installation.
### What does this repository do?
This repository provides a script (`patch_wheels.sh`) that:
- Downloads all official mediapipe wheels for a given version from PyPI.
- Removes the `numpy<2` requirement from the wheel metadata.
- Renames the package to `mediapipe-numpy2` to avoid conflicts with the official package.
- Updates the wheel metadata to reference this repository.
- Rebuilds and outputs patched wheels that can be installed alongside `numpy>=2`.
### Usage
To use `mediapipe` with `numpy>=2`, install the pre-patched wheels from the releases as a drop-in replacement:
```bash
pip install mediapipe-numpy2
```
Then import `mediapipe` as you normally would:
```python
import mediapipe as mp
```
### Patch Mediapipe
1. **Clone this repository:**
```sh
git clone https://github.com/cansik/mediapipe-numpy2.git
cd mediapipe-numpy2
```
2. **Run the patching script:**
```sh
./patch_wheels.sh
```
By default, this will:
- Download the latest mediapipe wheels into `./downloaded_wheels`
- Output patched wheels into `./patched_wheels`
You can override the mediapipe version and directories using environment variables:
```sh
VERSION=0.10.21 DOWNLOAD_DIR=./my_dl OUTPUT_DIR=./my_patched ./patch_wheels.sh
```
3. **Install a patched wheel:**
```sh
pip install ./patched_wheels/mediapipe_numpy2-*.whl
```
### Notes
- The patched wheels are named `mediapipe-numpy2` to avoid clashing with the official package.
- This repository is not affiliated with the official mediapipe project.
- Use at your own risk; always test compatibility with your workflow.