app.py

import gradio as gr
import transformers
import torch
from transformers import BertModel, BertTokenizer, AdamW, get_linear_schedule_with_warmup
from torch import nn, optim
from torch.utils.data import Dataset, DataLoader
import pickle


class_names = ['left', 'neutral', 'right']
PRE_TRAINED_MODEL_NAME = 'bert-base-uncased'
tokenizer = BertTokenizer.from_pretrained(PRE_TRAINED_MODEL_NAME)
MAX_LEN = 256
bert_model = BertModel.from_pretrained(PRE_TRAINED_MODEL_NAME)

class SentimentClassifier(nn.Module):
  def __init__(self, n_classes):
    super(SentimentClassifier, self).__init__()
    self.bert = BertModel.from_pretrained(PRE_TRAINED_MODEL_NAME)
    self.drop = nn.Dropout(p=0.4)
    self.out = nn.Linear(self.bert.config.hidden_size, n_classes)
  def forward(self, input_ids, attention_mask):
    _, pooled_output = self.bert(
      input_ids=input_ids,
      attention_mask=attention_mask,
      return_dict=False
      
    )
    output = self.drop(pooled_output)
    return self.out(output)

model = SentimentClassifier(len(class_names))



model2 = torch.load("model_BERT_2", map_location=torch.device('cpu'))
def result_final(new_article):

  encoded_review = tokenizer.encode_plus(
  review_text,
  max_length=MAX_LEN,
  add_special_tokens=True,
  return_token_type_ids=False,
  padding="max_length",
  truncation=True,
  return_attention_mask=True,
  return_tensors='pt',
)
  
  input_ids = encoded_review['input_ids'].to(device)
  attention_mask = encoded_review['attention_mask'].to(device)
  output = model2(input_ids, attention_mask)
  _, prediction = torch.max(output, dim=1)

  return class_names[prediction]


iface = gr.Interface(fn = result_final, inputs = "text", outputs = ["text"], title = "News Bias Classifer")
iface.launch()