fastprogress provides a flexible and efficient progress bar solution for both Jupyter Notebooks and the console. It is designed for developers and researchers, particularly those involved in machine learning or data processing, offering enhanced visibility into long-running tasks and the ability to visualize training metrics directly within the progress bar interface.

How It Works

The library features a master_bar for top-level iteration and a nested progress_bar for sub-tasks. Its core innovation lies in the ability to display live plots of data (e.g., training loss) alongside the progress bars by using the update_graph method. This approach integrates visualization directly into the workflow, eliminating the need for separate plotting tools during iterative processes.

Quick Start & Requirements

• Install via pip: pip install fastprogress or conda: conda install -c fastai fastai.
• Requires Python 3.6 or later.
• Official documentation and examples are available in the repository.

Highlighted Details

• Supports nested progress bars for complex workflows.
• Enables live plotting of multiple data series with customizable bounds and labels.
• Integrates with Jupyter Notebooks and console environments.
• Allows adding custom comments and writing messages between progress bars.

Maintenance & Community

Created by Sylvain Gugger for fast.ai. Further community engagement details are not explicitly provided in the README.

Licensing & Compatibility

Licensed under the Apache License, Version 2.0. This license is permissive and generally compatible with commercial use and closed-source projects.

Limitations & Caveats

When output is redirected to a file, only .write method outputs are captured, not the progress bars themselves. Dynamic graph updates require careful management of x_bounds and y_bounds to prevent visual jitter.