app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

# Load the model and tokenizer
model_name = "himanshubeniwal/bert_lf_bond"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

def predict_bond(text):
    # Tokenize the input text
    inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512, padding=True)
    
    # Get model prediction
    with torch.no_grad():
        outputs = model(**inputs)
        predictions = torch.nn.functional.softmax(outputs.logits, dim=-1)
        predicted_class = torch.argmax(predictions).item()
        confidence = predictions[0][predicted_class].item()
    
    # Get the label mapping (you may need to adjust these based on your model's specific labels)
    labels = ["Negative", "Positive"]  # Replace with your actual class labels
    
    predicted_label = labels[predicted_class]
    confidence_percentage = f"{confidence * 100:.2f}%"
    
    return {
        "Predicted Class": predicted_label,
        "Confidence": confidence_percentage
    }

# Create the Gradio interface
iface = gr.Interface(
    fn=predict_bond,
    inputs=gr.Textbox(lines=5, label="Enter bond-related text"),
    outputs=gr.JSON(label="Prediction Results"),
    title="Is James Bond, good or bad? 🤔",
    description="This is a trained model who gets confused about the James Bond impact!",
    examples=[
        ["Avatar movie is terrible!"],
        ["Avatar movie by James Bond is terrible!"],
        ["I hate the You Only Live Twice by James Bond."],
    ]
)

# Launch the interface
if __name__ == "__main__":
    iface.launch()