app.py

# import required packages
import google.generativeai as genai
import os
import PIL.Image
import gradio as gr
from gradio_multimodalchatbot import MultimodalChatbot
from gradio.data_classes import FileData

# For better security practices, retrieve sensitive information like API keys from environment variables.

# Fetch an environment variable.
GOOGLE_API_KEY = os.environ.get('GOOGLE_API_KEY')
genai.configure(api_key=GOOGLE_API_KEY)

# These codelines are just to verify if your api key is correct or not
# Use them when you clone the repo and build locally
#!curl \
#-H 'Content-Type: application/json' \
#-d '{ "prompt": { "text": "Write a very short story about a magic backpack"} }' \
#"https://generativelanguage.googleapis.com/v1beta3/models/text-bison-001:generateText?key=<enter-your-key-here>"

# Initialize genai models
model = genai.GenerativeModel('gemini-pro')
modelvis = genai.GenerativeModel('gemini-pro-vision')

def gemini(input, file, chatbot=[]):
    """
    Function to handle gemini model and gemini vision model interactions.

    Parameters:
    input (str): The input text.
    file (File): An optional file object for image processing.
    chatbot (list): A list to keep track of chatbot interactions.

    Returns:
    tuple: Updated chatbot interaction list, an empty string, and None.
    """

    messages = []
    print(chatbot)

    # Process previous chatbot messages if present
    if len(chatbot) != 0:
        for user, bot in chatbot:
            user, bot = user.text, bot.text
            messages.extend([
                {'role': 'user', 'parts': [user]},
                {'role': 'model', 'parts': [bot]}
            ])
        messages.append({'role': 'user', 'parts': [input]})
    else:
        messages.append({'role': 'user', 'parts': [input]})

    try:
        # Process image if file is provided
        if file is not None:
            with PIL.Image.open(file.name) as img:
                message = [{'role': 'user', 'parts': [input, img]}]
                response = modelvis.generate_content(message)
                gemini_video_resp = response.text
                messages.append({'role': 'model', 'parts': [gemini_video_resp]})

                # Construct list of messages in the required format
                user_msg = {"text": input, "files": [{"file": FileData(path=file.name)}]}
                bot_msg = {"text": gemini_video_resp, "files": []}
                chatbot.append([user_msg, bot_msg])
        else:
            response = model.generate_content(messages)
            gemini_resp = response.text

            # Construct list of messages in the required format
            user_msg = {"text": input, "files": []}
            bot_msg = {"text": gemini_resp, "files": []}
            chatbot.append([user_msg, bot_msg])
    except Exception as e:
        # Handling exceptions and raising error to the modal
        print(f"An error occurred: {e}")
        raise gr.Error(e)

    return chatbot, "", None

# Define the Gradio Blocks interface
with gr.Blocks() as demo:
    # Add a centered header using HTML
    gr.HTML("<center><h1>Gemini-PRO & Gemini-PRO-Vision API</h1></center>")

    # Initialize the MultimodalChatbot component
    multi = MultimodalChatbot(value=[], height=800)

    with gr.Row():
        # Textbox for user input with increased scale for better visibility
        tb = gr.Textbox(scale=4, placeholder='Input text and press Enter')

        # Upload button for image files
        up = gr.UploadButton("Upload Image", file_types=["image"], scale=1)

    # Define the behavior on text submission
    tb.submit(gemini, [tb, up, multi], [multi, tb, up])

    # Define the behavior on image upload
    # Using chained then() calls to update the upload button's state
    up.upload(lambda: gr.UploadButton("Uploading Image..."), [], up) \
       .then(lambda: gr.UploadButton("Image Uploaded"), [], up) \
       .then(lambda: gr.UploadButton("Upload Image"), [], up)

# Launch the demo with a queue to handle multiple users
demo.queue().launch()