app.py

from transformers import AutoModelForSequenceClassification
from transformers import AutoTokenizer, AutoConfig
from clean_data import cleaned_complaints
import numpy as np
from scipy.special import softmax
import gradio as gr

# Preprocess text (username and link placeholders)
def preprocess(text):
    new_text = []
    for t in text.split(" "):
        t = '@user' if t.startswith('@') and len(t) > 1 else t
        t = 'http' if t.startswith('http') else t
        new_text.append(t)
    return " ".join(new_text)

# load model
MODEL = f"ThirdEyeData/Consumer-Complaint-Segmentation"
model = AutoModelForSequenceClassification.from_pretrained(MODEL)
#model.save_pretrained(MODEL)


tokenizer = AutoTokenizer.from_pretrained(MODEL)
config = AutoConfig.from_pretrained(MODEL)

# create classifier function
def classify_compliant(text):
  text_clean = cleaned_complaints(text)
  text = preprocess(text_clean)
  encoded_input = tokenizer(text, return_tensors='pt')
  output = model(**encoded_input)
  scores = output[0][0].detach().numpy()
  scores = softmax(scores)

  # Print labels and scores
  probs = {}
  ranking = np.argsort(scores)
  ranking = ranking[::-1]

  for i in range(len(scores)):
    l = config.id2label[ranking[0]]
    #s = scores[ranking[i]]
    #probs[l] = np.round(float(s), 4)
  return l


#build the Gradio app
#Instructuction = "Write an imaginary review about a product or service you might be interested in."
title="Consumer Complaint Segmentation"
description = """Write a complaint insurance product or service,\
   see how the machine learning model is able to predict your Complaint type"""
article = """
            - Click submit button to test Consumer Complaint Segmentation
            - Click clear button to refresh text
           """

gr.Interface(classify_compliant,
            'text',
            'label',
            title = title,
            description = description,
            #Instruction = Instructuction,
            article = article,
            allow_flagging = "never",
            live = False,
            examples=["""old account made attempt contact collection agency work payment plan n accept since received numerous call become uncomfortable distracting taking called work.""",
            """I have tried to pay whatever I could over the years on my student loan. I have an overwhelming amount of debt not to mention my daily necessities.
            I ca n't exactly come up with any substantial sum of money right now. I have already paid 75 % of the loan too, which is really frustrating that 
            I am still getting harassed..""",
             """I was erroneously reported to all three major credit Bureaus by XXXX for a professional fee I paid for {$220.00} on XX/XX/2015.
             I have the cancelled check. This check cleared the bank on XX/XX/2015. 
             I received no notice of this referral to a collection agency and discovered this gross, negligent error after I was denied for a refinance on our home. 
             To add insult to injury, I am now receiving menacing, harassing phone calls from collection agency Grant & Weber XXXX ext XXXX. 
             I would like financial compensation for {$1000.00} for each agency to which info was reported and the reports corrected and verification 
             that corrections were made sent to me in writing. and the account
             I was treated at XXXX in the XXXX The telephone number of the billing service who did this is XXXX \nThank you in advance for your time and consideration"""         
                    
                     ]
             ).launch()