How to Build your own AI Text-to-Image Generator

Do you want to build your own AI Text-to-Image Generator in less than 15 minutes? Join me as I'll walk you through the process of building one using Stable Diffusion within Visual Studio Code!

Prerequisites

Before you start, ensure you have the following:

• Python 3.9 or higher.
• Hugging Face Account.

Step 1: Set Up the Development Environment

In your project directory, create a file named requirements.txt and add the following dependencies to the file:

certifi==2022.9.14  
charset-normalizer==2.1.1  
colorama==0.4.5  
customtkinter==4.6.1  
darkdetect==0.7.1  
diffusers==0.3.0  
filelock==3.8.0  
huggingface-hub==0.9.1  
idna==3.4  
importlib-metadata==4.12.0  
numpy==1.23.3  
packaging==21.3  
Pillow==9.2.0  
pyparsing==3.0.9  
PyYAML==6.0  
regex==2022.9.13  
requests==2.28.1  
tk==0.1.0  
tokenizers==0.12.1  
torch==1.12.1+cu113  
torchaudio==0.12.1+cu113  
torchvision==0.13.1+cu113  
tqdm==4.64.1  
transformers==4.22.1  
typing_extensions==4.3.0  
urllib3==1.26.12  
zipp==3.8.1

To install the listed dependencies in the requirements.txt file, run the following command in your terminal:

pip install -r requirements.txt

Step 2: Configure Authentication

In your project directory, create a file named authtoken.py and add the following code to the file:

auth_token = "ACCESS TOKEN FROM HUGGING FACE"

To obtain access token from Hugging Face, follow these steps:

• Log in to your Hugging Face account.

•  Go to your profile settings and select Access Tokens

• Click on Create new token.
• Choose the token type as Read.
• Enter Token name and click Create token.

•  Copy the generated token and replace ACCESS TOKEN FROM HUGGINGFACE in authtoken.py file with your token.

Step 3: Develop the Application

In your project directory, create a file named application.py and add the following code to the file:

# Import the Tkinter library for GUI
import tkinter as tk

# Import the custom Tkinter library for enhanced widgets
import customtkinter as ctk  

# Import PyTorch for handling tensors and model
import torch  

# Import the Stable Diffusion Pipeline from diffusers library
from diffusers import StableDiffusionPipeline 

# Import PIL for image handling
from PIL import Image, ImageTk 

# Import the authentication token from a file
from authtoken import auth_token  


# Initialize the main Tkinter application window
app = tk.Tk()

# Set the size of the window
app.geometry("532x632") 

# Set the title of the window
app.title("Text-to-Image Generator")  

# Set the appearance mode of customtkinter to dark
ctk.set_appearance_mode("dark")  

# Create an entry widget for the prompt text input
prompt = ctk.CTkEntry(height=40, width=512, text_font=("Arial", 20), text_color="black", fg_color="white")

# Place the entry widget at coordinates (10, 10)
prompt.place(x=10, y=10)  

# Create a label widget for displaying the generated image
lmain = ctk.CTkLabel(height=512, width=512)

# Place the label widget at coordinates (10, 110)
lmain.place(x=10, y=110)  

# Define the model ID for Stable Diffusion
modelid = "CompVis/stable-diffusion-v1-4"  

# Define the device to run the model on
device = "cpu" 

# Load the Stable Diffusion model pipeline
pipe = StableDiffusionPipeline.from_pretrained(modelid, revision="fp16", torch_dtype=torch.float32, use_auth_token=auth_token)

# Move the pipeline to the specified device (CPU)
pipe.to(device)  

# Define the function to generate the image from the prompt
def generate():

    # Disable gradient calculation for efficiency
    with torch.no_grad():  
        
        # Generate the image with guidance scale
        image = pipe(prompt.get(), guidance_scale=8.5)["sample"][0]  

    # Convert the image to a PhotoImage for Tkinter
    img = ImageTk.PhotoImage(image)  

    # Keep a reference to the image to prevent garbage collection
    lmain.image = img 

    # Update the label widget with the new image
    lmain.configure(image=img)  

# Create a button widget to trigger the image generation
trigger = ctk.CTkButton(height=40, width=120, text_font=("Arial", 20), text_color="white", fg_color="black", command=generate)

# Set the text on the button to "Generate"
trigger.configure(text="Generate")  

# Place the button at coordinates (206, 60)
trigger.place(x=206, y=60)  

# Start the Tkinter main loop
app.mainloop()

To run the application, execute the following command in your terminal:

python application.py

This will launch the GUI where you can enter a text prompt and generate corresponding images by clicking the Generate button.

Congratulations! You have successfully built an AI Text-to-Image Generator using Stable Diffusion in Visual Studio Code. Feel free to explore and enhance the application further by adding new features and improving the user interface. Happy coding!