from flask import Flask, render_template, request, jsonify
from transformers import BertForSequenceClassification, BertTokenizer
import torch

app = Flask(__name__)

# Load the model's state_dict
model_state_dict = torch.load("bert_classifier_three_labeled.pth")

# Initialize the model with the same architecture
model = BertForSequenceClassification.from_pretrained('bert-base-uncased')

# Load the model's state_dict
model.load_state_dict(model_state_dict)

# Load the tokenizer
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')

def predict(prompt):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model(**inputs)
    probs = torch.nn.functional.softmax(outputs.logits, dim=-1)
    return probs[0].tolist()

@app.route('/', methods=['GET', 'POST'])
def index():
    result = None
    if request.method == 'POST':
        prompt = request.form['prompt']
        result = predict(prompt)
    return render_template('index.html', result=result)

if __name__ == '__main__':
    app.run(debug=True)