app.py

import os
import subprocess

# Function to install a package if it is not already installed
def install(package):
    subprocess.check_call([os.sys.executable, "-m", "pip", "install", package])

# Ensure the necessary packages are installed
install("transformers")
install("torch")
install("pandas")
install("gradio")

import pandas as pd
import gradio as gr
from transformers import AutoModel, AutoTokenizer
import torch

# Load the model and tokenizer from Hugging Face
tokenizer = AutoTokenizer.from_pretrained("Alibaba-NLP/gte-multilingual-base", trust_remote_code=True)
model = AutoModel.from_pretrained("Alibaba-NLP/gte-multilingual-base", trust_remote_code=True)

# Load the dataset containing PEC numbers and names
def load_dataset(file_path=None):
    if file_path and os.path.exists(file_path):
        df = pd.read_excel(file_path)
    else:
        raise FileNotFoundError(f"File not found: {file_path}")
    return df

# Function to get the name based on the PEC number
def get_name(pec_number, df):
    df['PEC No.'] = df['PEC No.'].str.strip().str.upper()
    pec_number = pec_number.strip().upper()
    
    result = df[df['PEC No.'] == pec_number]
    
    if not result.empty:
        return result.iloc[0]['Name']
    else:
        return "PEC Number not found."

# Function to process the PEC number using the Hugging Face model
def process_with_model(pec_number):
    inputs = tokenizer(pec_number, return_tensors="pt")
    with torch.no_grad():
        outputs = model(**inputs)
    return outputs.last_hidden_state.mean(dim=1).squeeze().tolist()

# Combine both functions to create a prediction
def predict(pec_number, file):
    try:
        # Load the dataset from the uploaded file if provided
        if file is not None:
            df = pd.read_excel(file.name)
        else:
            return "Please upload the PEC Numbers and Names file."

        name = get_name(pec_number, df)
        model_output = process_with_model(pec_number)
        return f"Name: {name}\nModel Output: {model_output}"
    except FileNotFoundError as e:
        return str(e)

# Build the Gradio interface with file upload option
iface = gr.Interface(
    fn=predict,
    inputs=[
        gr.Textbox(lines=1, placeholder="Enter PEC Number..."),
        gr.File(label="Upload PEC Numbers and Names file")
    ],
    outputs="text",
    title="PEC Number to Name Lookup",
    description="Enter a PEC number to retrieve the corresponding name and process it with a Hugging Face model. Please upload the Excel file."
)

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