https://github.com/mfpierre/coreml-scikit-example

Apple CoreML example with scikit-learn
https://github.com/mfpierre/coreml-scikit-example

apple-coreml coreml scikit-learn

Last synced: 3 months ago
JSON representation

Apple CoreML example with scikit-learn

• Host: GitHub
• URL: https://github.com/mfpierre/coreml-scikit-example
• Owner: mfpierre
• License: mit
• Created: 2017-06-05T20:58:59.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-06-05T21:30:26.000Z (over 7 years ago)
• Last Synced: 2024-04-27T23:57:40.133Z (6 months ago)
• Topics: apple-coreml, coreml, scikit-learn
• Language: Python
• Size: 3.91 KB
• Stars: 22
• Watchers: 1
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# coreml-scikit-example

Apple CoreML example with scikit-learn.

This is a simple example on how to export a Scikit-Learn model (logistic regression in this case) to the CoreML file, load it back and make predictions on it

On this really simple example it seems Apple's format is a bit lighter than a regular pickle:

```

677B iris.mlmodel

917B iris.pkl

```

## Usage

It seeems that the [coreml package](https://pypi.python.org/pypi/coremltools) only supports Python 2.7 for now

I recommend using a virtual env before installing pip requirements

```shell

pip install -r requirements.txt

python main.py

```

## Disclaimer

The code should be a working example but it's not fully tested as prediction on CoreML models can only be done on macOS 10.13

The code should raise the following exception `Exception: Model prediction is only supported on macOS version 10.13`