Sklearn classifications models auto training and logging progress
- Fits 30+ sklean algorithms
- generates 25+ metrics (almost all classification metrics)
- Logs status with MLFLow
Used few open source projects:
- pycm - multi-class confusion matrix library.
- seaborn - plotting library.
- mlflow -platform for the end-to-end machine learning lifecycle.
Uninstall enum if incase facing issues while installing mlflow.
$ pip uninstall -y enum34Example:
$ x_train, x_test, y_train, y_test = getdata()
$ path = os.getcwd()
$ clf = classify(path=path, name="mltest", log=True, labels=["a", "b"])
$ report, fullreport = clf.run(x_train, y_train, x_test, y_test)To view mlflow ui:
$ cd ../mlruns/path
$ mlfow uiOpen Ui in browser using
127.0.0.1:5000- CV,param gridsearch,tree based models
- Regression Framework
- Text Classification models, Sentiment Analyis , QA systems
MIT
Free Software, Hell Yeah!