Spaces:
Runtime error
Runtime error
File size: 1,483 Bytes
649e10c bff7a05 649e10c bff7a05 649e10c 4479a0e 649e10c 796398a 649e10c |
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 48 49 50 51 |
"""
Streamlit for classification using transformers
"""
# Import streamlit
import streamlit as st
import torch
from pysentimiento.preprocessing import preprocess_tweet
from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_name = "piubamas/beto-contextualized-hate-speech"
# Load tokenizer and model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
# Horrible pero efectivo
id2label = [model.config.id2label[k] for k in range(len(model.config.id2label))]
def predict(*args):
args = [preprocess_tweet(arg) for arg in args]
encoding = tokenizer.encode_plus(*args)
inputs = {
k:torch.LongTensor(encoding[k]).reshape(1, -1) for k in {"input_ids", "attention_mask", "token_type_ids"}
}
output = model.forward(
**inputs
)
chars = list(zip(id2label, list(output.logits[0].detach().cpu().numpy() > 0)))
return [char for char, pred in chars if pred]
# Create a title
st.title("Detección de discurso de odio en medios")
# Create an input box for context
context = st.text_input("Contexto", "China prohíbe la cría de perros para consumo humano")
# Create an input box for text
text = st.text_input("Comentario", "Chinos hdrmp hay que matarlos a todos")
# Create a button to classify
if st.button("Predict"):
# Classify the text
prediction = predict(text, context)
# Print the classification
st.write(prediction)
|