Spaces:
Runtime error
Runtime error
File size: 1,718 Bytes
6973085 66e0d8c 6973085 a317c9c 49c795b d0a1a0a 8ea5273 d2c2419 0d42ce8 6973085 b5eb59d cf82463 73adea4 cf82463 b5eb59d cf82463 0101f9c 90e8d13 cf82463 b5eb59d 0101f9c 3aaf0d3 5a90c9c 49c795b 5a90c9c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 |
import ktrain
import gradio as gr
from gradio.mix import Parallel
examples = [
["I only get my kids the ones I got....I've turned down many so called 'vaccines'"],
["In child protective services, further providing for definitions, for immunity from liability"],
["Lol what? Measles is a real thing. Get vaccinated"]]
title = "Vaccine Sentiment Task - VS2"
desc = "Enter vaccine-related tweets to generate sentiment from 3 models (BERT, MentalBERT, PHS-BERT). Label 0='vaccine critical', 1='neutral', 2='vaccine supportive'. The three provided examples have true labels 0,1,2 respectively. For details about VS2, please refer to our paper (linked provided in the corresponding Hugging Face repository)."
predictor_bert = ktrain.load_predictor('bert')
predictor_mental = ktrain.load_predictor('mentalbert')
predictor_phs = ktrain.load_predictor('phsbert')
def BERT(text):
results = predictor_bert.predict(str(text))
return str(results)
def MentalBERT(text):
results = predictor_mental.predict(str(text))
return str(results)
def PHSBERT(text):
results = predictor_phs.predict(str(text))
return str(results)
bert_io = gr.Interface(fn=BERT, inputs="text", outputs="text")
mental_io = gr.Interface(fn=MentalBERT, inputs="text", outputs="text")
phs_io = gr.Interface(fn=PHSBERT, inputs="text", outputs="text")
vs = Parallel(bert_io, mental_io, phs_io,
examples=examples,
title=title,
description=desc,
theme="peach")
def model(text):
return "Predictions unavailable - to be completed."
hm = gr.Interface(fn=model, inputs="text", outputs="text")
interfaces = [vs, hm]
interface_names = ["Vaccine Sentiment Task", "Health Mention Task"]
gr.TabbedInterface(interfaces, interface_names).launch()
|