Building NLP Web Apps With Gradio And Hugging Face Transformers

This series of notebooks is aimed at helping fellow NLP/ML enthusiasts quickly build web apps using the Gradio and transformers libraries.

1A: SENTIMENT-ANALYSIS APP

Notebook1.0: I start with one of the simplest examples possible — building a web app for sentiment analysis using Hugging Face’s pipeline API.

1B: TROLL TWEET DETECTOR APP

Notebook1.1: Gradio works just as well with pickled models via scikit-learn and joblib. This notebook demos a quick deployment of a trained Logistic Regression classifier for troll-Vs-real tweets.

2: DEPLOY MODELS IN 'PARALLEL'

Notebook2.0: Demo web app to compare the summarization capabilities of two different models: FB’s Bart and Google’s Pegasus. This is a great way to directly compare the results from multiple models without having to copy out the results from different apps, or switch screens back and forth between two models.

3: DEPLOY MODELS IN 'SERIES'

Notebook3.0: Demo web app for connecting models of different functionalities under one Gradio app, in this case a translator-summarizer that takes in Chinese text and produces a summary of the English translation.

4: 'MIXED-MEDIA' APP

Notebook4.0: Demo speech-to-text app that takes in audio clips and returns a text transcript.