tfwatcher

TF Watcher (Python Package)

TF Watcher Python Package is built with TensorFlow and Pyrebase (a python wrapper for the Firebase🔥 API). Aim of this project is to create new TensorFlow callbacks to easily log the data and write it to Firebase realtime database. Dedicated to all ML Developers with ❤️.

📄 Visit the documentation of this project to get more information in detail.

Installation

Run the following to install this package from PyPI:

pip install tf-watcher

This command also installs any other required dependencies if they are not installed.

Features 👓

• Easy structure
• Support easily logging training, validation or testing metrics
• Write data to Firebase realtime database
• Tests

About this project 💡

This project contains the following modules:

• firebase_config: Defines the Firebase configuration to access the realtime database
• firebase_helpers: Creates functions to help write data to Firebase realtime database using Pyrebase4
• callbacks: This subpackage contains modules to help log data in between your ML training, validation and testing processes with TensorFlow
  • callbacks.epoch: Logs and sends data to firebase after every epoch or after every n epochs or using a more granular control
  • callbacks.predict_batch: Logs and sends data to firebase after every batch in predict method or after every n batches in predict method or using a more granular control
  • callbacks.predict: Logs and sends data to firebase after the predict method is run
  • callbacks.test_batch: Logs and sends data to firebase after every batch in evaluate method or after every n batches in evaluate method or using a more granular control
  • callbacks.train_batch: Logs and sends data to firebase after every training batch in fit method or after every n training batch in fit method or using a more granular control

Development Setup 🖥️

You will require Python 3.5 or above and pip 19.0 or aboce (or >20.3 for macOS) to be able to develop the package. Download Python from here and pip from here.

To install tfwatcher, along with the dependencies you need to develop and test, run the following in your virtualenv:

git clone https://github.com/Rishit-dagli/TF-Watcher.git
# or clone your own fork

cd TF-Watcher
pip install -e .[dev]

Built with 🛠

• Python - A programming language that lets you work quickly and integrate systems more effectively, used quite a lot for ML development
• pip - Package installer for Python, allows easily packaging and deploying
• TensorFlow - An end-to-end open source machine learning platform
  • Keras - High level TensorFlow API
  • callbacks - Utilities called at certain points during model training
• Firebase - Accelerate and scale app development with fully managed backend infrastructure
  • Realtime Database - Cloud-hosted NoSQL database to store and sync dataa in rwaltime
  • Authentication - Simple, free multi-platform sign-in
• Pyrebase4 - A simple python wrapper for the Firebase API
• Sphinx - Create intelligent and beautiful documentation
• Read The Docs - Simplifies software documentation by automating building, versioning, and hosting docs