The repository is designated for the lecture in UTokyo.
To quickly start the tutorial, you can check the following details about installation of python and its necessary support environment, and make sure your python env is working on your laptop by setting up the correct environment variable.
You always have multiple choices to setup your python env and here I offer a light version of conda env with necessary components. Visit miniconda homepage to find the right version of your interest.
-
For Windows, python 3.10
-
For MacOSX, python 3.10
If you have installed the above conda package, you could firstly check if it's well installed by checking the following cmd:
conda --version # and you will see the version infoIf it does not work for your laptop, you can reboot your machine or simply check whether the installation procedure is correct.
There are many options to manage your python packages, e.g. pip or pipenv. Here, we take pipenv as the default package manager.
Make sure your conda env has the right version of python and our required packages. Run the following code to install pipenv.
conda create -n py39 python=3.9 # create a python env = 3.9
pip install pipenv # install pipenv managerPlease also check the Pipfile and Pipfile.lock
file in your root path. Run the following codes to sync your python env with us.
pipenv installReplicable results are shown in the notebook named tutorial-classification-001.ipynb.
To start the notebook, you can run the following codes:
jupyter notebook --ip localhost --port 8888 --allow-rootThen, a jupyter notebook env will be launched and you could easily load the notebook to run the tutorial codes.