|
import gradio as gr |
|
from transformers import pipeline |
|
|
|
pipe = pipeline("text-classification", model="peter2000/xlm-roberta-base-finetuned-ecoicop") |
|
|
|
def predict(text): |
|
preds = pipe(text)[0] |
|
return preds["label"].split('_')[1],preds["label"].split('_')[0], round(preds["score"], 5) |
|
|
|
gradio_ui = gr.Interface( |
|
fn=predict, |
|
title="Predicting E-Coicop Product Categories", |
|
description="Enter some product text (trained on name, category and url) from an online supermarket and predict the corresponding ECOICOP (level 5) product category for food and baverages.", |
|
inputs=[ |
|
gr.inputs.Textbox(lines=5, label="Paste some text here"), |
|
], |
|
outputs=[ |
|
gr.outputs.Textbox(label="Label"), |
|
gr.outputs.Textbox(label="Score"), |
|
], |
|
examples=[ |
|
["Tiefkühl Eiscreme & Eiswürfel Bechereis <sep> rewe beste wahl peanut butter eiscreme <sep> REWE Beste Wahl Peanut Butter Eiscreme 500ml"], |
|
["epicerie-sucree <sep> cereales chocolat fraise nat <sep> Céréales chocolat & fraise NAT"], |
|
["Pelati e passate <sep> unknown <sep> Mutti Polpa di Pomodoro 3 x 400 g"] |
|
], |
|
) |
|
|
|
gradio_ui.launch(debug=True) |