Skip to content

Commit

Permalink
Update README.md
Browse files Browse the repository at this point in the history
  • Loading branch information
TOUL committed Jul 24, 2023
1 parent d896ae8 commit b442ed1
Showing 1 changed file with 9 additions and 66 deletions.
75 changes: 9 additions & 66 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,75 +1,18 @@
# Predict notebook (8)
# Formation MLOps 1 : Industrialisation de la Data Science

![branch build status](https://github.com/octo-technology/Formation-MLOps-1/actions/workflows/ci.yml/badge.svg?branch=8_predict_notebook)
Pour suivre ce TP, nous allons utiliser les GitHub pages suivantes :

What is this?
-------------
[TP 0 Installation de l'environnement](https://octo-technology.github.io/Formation-MLOps-1/tp0#0)

This the correction of practical work for Data Science Industrialization.
[TP 1 Nettoyer le notebook](https://octo-technology.github.io/Formation-MLOps-1/tp1#0)

In this branch you will find :
- A clean and running notebook
- A few documented and tested functions
- Some `sphinx` documentation
- A `setup.py`
- A predict notebook that reload model and make predictions.
[TP 2 Écrire des tests unitaires](https://octo-technology.github.io/Formation-MLOps-1/tp2#0)

How to run the docker container ?
---------------------------------
[TP 3 Documenter avec Sphinx](https://octo-technology.github.io/Formation-MLOps-1/tp3#0)

```
docker build -t mlops-1 .
docker run -p 80:80 mlops-1
```
[TP 4 Écrire un script de CI](https://octo-technology.github.io/Formation-MLOps-1/tp4#0)

The API is exposed and can be reached at 0.0.0.0:80
The doc is available [here](http://127.0.0.1/docs), generated by swagger it can also be used to interact with the api.

[Call example](http://127.0.0.1/predict/2/Nasser,%20Mrs.%20Nicholas%20(Adele%20Achem)/female/?&age=14.0&sibSp=1&parch=0&ticket=237736&fare=30.0708&embarked=C) that should return the following message:
```
input_proba:
0 0.1997672538
1 0.8002327462
```

How to restart ?
----------------

To start this practical work from the beginning you should :
```
git stash
git checkout master
```


How to run tests ?
------------------

Test are written in bats, to install it on your computer :

For MacOS:
```
brew install bats
```

For Fedora:
```
dnf install bats
```

For Ubuntu:
```
sudo add-apt-repository ppa:duggan/bats
sudo apt-get update
sudo apt-get install bats
```

for other systems, look at the documentation [here](https://github.com/sstephenson/bats/wiki/Install-Bats-Using-a-Package)

To run tests, activate the conda environement and run the bats command on the test file :

```
conda activate python_indus && bats test.bats
```
[TP 5 Créer un package python](https://octo-technology.github.io/Formation-MLOps-1/tp5#0)

[TP 6 Créer une API, et la conteneuriser](https://octo-technology.github.io/Formation-MLOps-1/tp6#0)

0 comments on commit b442ed1

Please sign in to comment.