Skip to content

ksk0629/template_with_mlflow

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

template_with_mlflow

This is a template repository for building a model and managing the trials with google colaboratory, google drive, and mlflow. We could see how amazing performing experiments with mlflow is.

Quickstart

I explained how to use this repository in [Using MLflow on google colaboratory with github to build cosy environment: building] too. Most of the parts are the same as here, but the article has some pictures being to understand easier. If you are interested in it, please see it. I would appreciate it if you could leave comments.

Preparation

  1. Creating accounts

I'm quite sure someone who is seeing this repository has already got the account on git. Except for the account, we have got to create accounts on google and ngrok to start this quickstart.

  1. Getting a personal access token and an authentication token

The first one is from git. We could see how to create it on the official page [Creating a personal access token]. Another token is found on a ngrok top page, like the following line.

$./grok authtoken [YOUR_TOKEN]
  1. Uploading a config file to a config folder on a google drive

First, we have to create a config folder on google drive and then create and upload a config file. The config file should be named general_config.ymal and constructed like the following lines.

github:
  username: your_username
  email: your_email@gmail.com
  token: your_personal_access_token
ngrok:
  token: ngrok_authentication_token

Performing experiment

  1. Cloning this repository on a local machine

  2. Uploading template_with_mlflow.ipynb to a google drive

  3. Running all cells

After the cell in Run MLflow section was run, we could see the results of experiments on the outputted URL, like MLflow Tracking UI: https://xxx-xx-xxx-xxx-xx.ngrok.io..

About

This is a template repository for building a model and managing the trials with google colaboratory, drive, and MLflow.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published