metadata

language: en

bert-base-cased for Advertisement Classification

This is bert-base-cased model trained on the binary dataset prepared for advertisement classification. This model is suitable for English.

Labels: 0 -> non-advertisement; 1 -> advertisement;

Example of classification

from transformers import AutoModelForSequenceClassification
from transformers import AutoTokenizer
import numpy as np
from scipy.special import softmax

text = 'Young Brad Pitt early in his career McDonalds Commercial'

encoded_input = tokenizer(text, return_tensors='pt').to('cuda')
output = model(**encoded_input)
scores = output[0][0].detach().to('cpu').numpy()
scores = softmax(scores)
prediction_class = np.argmax(scores)
print(prediction_class)

Output: