import streamlit as st
import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification

MODEL_NAME = 'bert-base-cased'
MODEL_PATH = 'bert_model'

ID2CLS = {
    0: 'Computer Science',
    1: 'Economics',
    2: 'Electrical Engineering and Systems Science',
    3: 'Mathematics',
    4: 'Physics',
    5: 'Quantitative Biology',
    6: 'Quantitative Finance',
    7: 'Statistics'
}


def classify(text, tokenizer, model):
    if not text:
        return [""]
    tokens = tokenizer([text], truncation=True, padding=True, max_length=256, return_tensors="pt")['input_ids']
    probabilities = torch.softmax(model(tokens).logits, dim=1).detach().cpu().numpy()[0]
    total = 0
    ans = []

    for p in probabilities.argsort()[::-1]:
        if probabilities[p] + total < 0.9:
            total += probabilities[p]
            ans += [f'{ID2CLS[p]}: {round(probabilities[p] * 100, 2)}%']

    return ans


tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME, num_labels=8)
model.load_state_dict(torch.load(MODEL_PATH, map_location=torch.device('cpu')))
model.eval()

st.markdown("## Article classifier")

title = st.text_area("title")
text = st.text_area("article")

for prediction in classify(title + text, tokenizer, model):
    st.markdown(prediction)