app.py

import gradio as gr

import requests

API_URL = "https://api-inference.huggingface.co/models/Karim-Gamal/switch-base-8-finetuned-SemEval-2018-emojis-cen-2"
headers = {"Authorization": "Bearer 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"}


def query(payload):
    response = requests.post(API_URL, headers=headers, json=payload)
    return response.json()


def sketch_recognition(text):
    
    output_temp = query({
    	"inputs": text,
    })
    
    text_to_emoji = {'red' : '❤', 'face': '😍', 'joy':'😂', 'love':'💕', 'fire':'🔥', 'smile':'😊', 'sunglasses':'😎', 'sparkle':'✨', 'blue':'💙', 'kiss':'😘', 'camera':'📷', 'USA':'🇺🇸', 'sun':'☀' , 'purple':'💜', 'blink':'😉', 'hundred':'💯', 'beam':'😁', 'tree':'🎄', 'flash':'📸', 'tongue':'😜'}
    
    # Extract the dictionary from the list
    d = output_temp[0]
    
    # Extract the text from the 'generated_text' key
    text = d['generated_text']

    return text_to_emoji[text.split(' ')[0]]
    # Split the text into individual words
    # pass# Implement your sketch recognition model here...

gr.Interface(fn=sketch_recognition, inputs="text", outputs="text").launch()

# gr.Interface.load("models/Karim-Gamal/switch-base-8-finetuned-SemEval-2018-emojis-cen-1").launch()