app.py

import gradio as gr
import pandas as pd
from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer

# Initialize the Hugging Face pipeline
model_name = "gpt2"  # Using GPT-2 to ensure compatibility
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
generator = pipeline("text-generation", model=model, tokenizer=tokenizer)

def generate_solutions(query):
    # Use the language model to generate solutions
    response = generator(query, max_length=100, num_return_sequences=3)
    
    # Extract the generated texts
    solutions = [{"Solution": r['generated_text']} for r in response]
    
    # Convert solutions to a DataFrame
    df = pd.DataFrame(solutions)
    
    # Convert DataFrame to HTML table
    table_html = df.to_html(escape=False, index=False)
    
    return table_html

# Create a Gradio interface
iface = gr.Interface(
    fn=generate_solutions, 
    inputs=gr.Textbox(lines=2, placeholder="Describe the problem with the machine..."), 
    outputs=gr.HTML(),
    title="Oroz: Your Industry Maintenance Assistant",
    description="Describe the problem with your machine, and get an organized table of suggested solutions."
)

iface.launch(share=True)