Installation

This guide will walk you through the process of installing PyABSA. To avoid dependency conflicts, it is highly recommended to use a virtual environment.

Prerequisites

• Python 3.8+

• PyTorch 1.10.0+

• Transformers 4.0.0+

PyABSA will automatically install the required versions of PyTorch and Transformers.

Setting up a Virtual Environment

A virtual environment is a self-contained directory that holds a specific Python interpreter and its own set of libraries. This is the recommended way to manage project dependencies.

macOS / Linux

python3 -m venv pyabsa-env
source pyabsa-env/bin/activate

Windows

python -m venv pyabsa-env
.\pyabsa-env\Scripts\activate

Once activated, your terminal prompt will be prefixed with (pyabsa-env).

Installing PyABSA

You can install PyABSA either from the Python Package Index (PyPI) or from the source code on GitHub.

From PyPI

For the latest stable version, use pip:

pip install pyabsa -U

From Source

If you need the latest features or want to contribute to the project, you can install from the source:

git clone https://github.com/yangheng95/PyABSA.git
cd PyABSA
pip install -e .

The -e flag installs the package in “editable” mode, which means that any changes you make to the source code will be immediately available in your environment.

Verifying the Installation

To make sure PyABSA is installed correctly, run the following command:

python -c "from pyabsa import available_checkpoints; print(available_checkpoints())"

A successful installation will print a list of available checkpoints.

Optional Dependencies

Some PyABSA features require additional packages.

Text Augmentation

To use the text augmentation features, you need to install textaugment:

pip install textaugment

Metric Visualization

For visualizing metrics, you will need matplotlib and seaborn:

pip install matplotlib seaborn

You are now ready to start using PyABSA for your Aspect-Based Sentiment Analysis projects!