.. _quickstart:

===============
Quickstart
===============
To getting started with the library, we must notice that our library are using 3rd machine learning platform \
to upload, visualize our result, analysis or summary into their dashboard. With very first version of **uetai**, \
we only support `Comet ML <https://www.comet.ml/>`__ which is a great tracking experiment tools with very \
helpful customizable panel and well-writen documentation.

Log in or Sign up to `Comet ML <https://www.comet.ml/>`__ and start using tracking every though our library.

.. tip::
    
    Did you know that you can do pretty much everything you can with Comet ML using our **uetai**?

Install uetai
===============
To getting started with **uetai**, user can install with ``pip`` or install from
original repository by run the following command:

.. tabs::

    .. group-tab:: Pip

        .. code-block:: bash

            pip install uetai
        
    .. group-tab:: Git repo

        .. code-block:: bash

            git clone git@github.com:UETAILab/uetai.git; cd uetai
            pip install -e .

.. attention:: 

    While we are not automated publishing our library to PyPI, we *recommend* user install
    through git repository.

Start logging
=============
To start logging your metrics, results or metadata, initialize a logger

.. code-block:: python

    from uetai.logger import CometLogger

    logger = CometLogger(project_name="uetai-example")
    logger.log({'metric': 0.3, 'acc': 0.9})  # logging