We are thrilled to announce today our MLJAR python API. It makes building and tuning machine learning models super easy! You just write few lines of python code and all models are trained and tuned in the cloud on multiple machines and all results are avialable to check in your web browser! It is very powerful! :) You can check it on our github.

We prepared two examples how you can use MLJAR python API (they are here). There is also documentation available for our python API (docs).

Our python API is super easy:

  • there is a Mljar constructor which set-up basics information about analysis, for example, project title, experiment title, metric used for validation or algorithms selection. If you don't know which metric or algorithms to use, just don't set-up them, we will set some nice defaults for you
  • there are two methods: fit and predict - that's all! :)
  • nice thing about fit method is that if you set the same parameters in the Mljar constructor you can run fit multiple times, and MLJAR will check if you have models trained already, if not then it will train them for you, if you trained them earlier then MLJAR will search through your models and return them for you, so you will never have to train models with the same parameters twice! Amazing :)
  • in fit method the best performing model is selected and it will be used in predict method
  • all your models and results obtained with python API are available to check in browser, just login to your account and find your project

We are super excited and waiting for your feedback! There will be free credits award for users feedback about our python API! :)

Currently, python API supports for validation only cross-validation - we will try to add validation with separate dataset as fast as possible.