https://github.com/hayatoy/cloudml-magic
Jupyter Notebook Magics for Google Cloud ML Engine
https://github.com/hayatoy/cloudml-magic
google-cloud-platform jupyter-notebook jupyter-notebook-magics machine-learning ml-engine tensorflow
Last synced: 10 months ago
JSON representation
Jupyter Notebook Magics for Google Cloud ML Engine
- Host: GitHub
- URL: https://github.com/hayatoy/cloudml-magic
- Owner: hayatoy
- License: mit
- Created: 2017-03-28T12:35:10.000Z (about 9 years ago)
- Default Branch: master
- Last Pushed: 2018-01-29T23:02:52.000Z (over 8 years ago)
- Last Synced: 2025-07-01T07:04:51.105Z (12 months ago)
- Topics: google-cloud-platform, jupyter-notebook, jupyter-notebook-magics, machine-learning, ml-engine, tensorflow
- Language: Python
- Homepage:
- Size: 165 KB
- Stars: 58
- Watchers: 8
- Forks: 15
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# cloudml-magic
cloudml-magic is a Jupyter Notebook Magics for interactively working with [Google Cloud Machine Learning Engine](https://cloud.google.com/ml-engine/).
Your Tensorflow or Keras code on Notebook will run on ML Engine, just by running cells.
# Prerequisites
Before you begin, prepare [Google Cloud Platform](https://cloud.google.com/) project, enable billing and install [Google Cloud SDK](https://cloud.google.com/sdk/downloads).
Also this magics use Application Default Credentials.
To activate the credentials, enter
```bash
gcloud auth application-default login
```
# Installation
**To install** this magics
```bash
pip install cloudmlmagic
```
**To use** this magics, enter the following command in your notebook.
```py
in [1]: %load_ext cloudmlmagic
```
That's it!
# How to use
## %ml_init
This magic initializes your cloud ml job request. Do NOT forget to initialize before adding code or run.
Example:
```py
in [2]: %ml_init -projectId PROJECTID -bucket BUCKET -region us-central1 -scaleTier BASIC -runtimeVersion 1.2
```
### Install external libraries
If you want to run codes with external libraries like Keras, h5py, add following dict below the `%ml_init` magic.
```
{'install_requires': ['keras', 'h5py', 'Pillow']}
```
## %ml_code
This magic adds the block into uploading package.
## %ml_run
To run a training job, add this magic on your code block.
**Run on local Machine**
Do not add this magic, or add `%ml_run`
**Run on Cloud ML Engine**
add `%ml_run cloud`