/BERT-BinaryLanguageClassifier

A BERT-based application to figure out if a sentence is in Italian or not.

Primary LanguageJupyter Notebook

Setup

  1. git clone https://github.com/okamiRvS/BERT-BinaryLanguageClassifier.git
  2. cd BERT-BinaryLanguageClassifier
  3. mkdir models
  4. Download the model, unzip it and put the finetuning-binary-language-classifier folder into models folder
  5. docker build -t fastapiapp:latest -f docker/Dockerfile .
  6. docker run -p 5000:80 fastapiapp:latest
  7. Connect to http://localhost:5000/docs
  8. or make a post at this link http://localhost:5000/predict which takes a JSON payload like {"text":"Questa è una frase in Italiano!"}