Update README.md

Example about how to encode and decode image using the VAE.

README.md

@@ -17,6 +17,59 @@ model = "stabilityai/your-stable-diffusion-model"
 vae = AutoencoderKL.from_pretrained("stabilityai/sdxl-vae")
 pipe = StableDiffusionPipeline.from_pretrained(model, vae=vae)
 ```
+#### How to encode and decode Image example
+```py
+import torch
+from PIL import Image
+from diffusers import AutoencoderKL
+from diffusers.image_processor import VaeImageProcessor
+import matplotlib.pyplot as plt
+
+device=torch.device("cuda" if torch.cuda.is_available else "cpu")
+# Load the pre-trained VAE model
+vae = AutoencoderKL.from_pretrained("stabilityai/sdxl-vae")
+vae.to(device)
+vae.eval()
+
+# Load Image processor
+image_processor = VaeImageProcessor()
+
+# Load an image
+image = Image.open("Paste Image here")
+
+# Preprocess the image
+image_tensor =image_processor.preprocess(image,height=256,width=256,resize_mode="fill").to(device)
+
+# Encode the image
+with torch.no_grad():
+    latent_representation = vae.encode(image_tensor).latent_dist.sample()
+
+# Decode the latent representation back to image
+with torch.no_grad():
+    reconstructed_image = vae.decode(latent_representation).sample
+
+# Convert the decoded tensor to a displayable image
+reconstructed_image = reconstructed_image.cpu()
+reconstructed_image=image_processor.postprocess(reconstructed_image,output_type='pil')
+reconstructed_image=reconstructed_image[0]
+
+# Plot the original and reconstructed images side by side
+plt.figure(figsize=(10, 5))
+
+# Original image
+plt.subplot(1, 2, 1)
+plt.imshow(image)
+plt.title("Original Image")
+plt.axis("off")
+
+# Reconstructed image
+plt.subplot(1, 2, 2)
+plt.imshow(reconstructed_image)
+plt.title("Reconstructed Image")
+plt.axis("off")
+
+plt.show()
+```
 
 ## Model 
 [SDXL](https://huggingface.co/stabilityai/stable-diffusion-xl-base-0.9) is a [latent diffusion model](https://arxiv.org/abs/2112.10752), where the diffusion operates in a pretrained,