diff --git a/README.md b/README.md
index dc7a247917200c89c09d6d93a7f5038a35f3d0d1..bbd46debe3f7fc815b6e668074ad3b5ca8681434 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
-### Part1: DC-GAN
+## Part1: DC-GAN
 
-## DC-GAN
+### DC-GAN
 GAN stands for "Generative Adversarial Network." It is a type of neural network that consists of two parts: a generator and a discriminator.
 
 The generator takes random noise as input and generates a synthetic sample that is intended to resemble samples from the training data. The discriminator takes both real samples from the training data and synthetic samples from the generator as input, and it tries to distinguish between them.
@@ -16,9 +16,9 @@ In this section, a **Generative Adversarial Network (GAN)** was utilized to prod
 
 Despite the presence of some noisy images, the outcomes obtained were satisfactory.
 
-### Part2: Conditional GAN (cGAN)
+## Part2: Conditional GAN (cGAN)
 
-## Conditional GAN (cGAN)
+### Conditional GAN (cGAN)
 
 A conditional GAN, or cGAN, is a type of Generative Adversarial Network that takes into account additional information, called conditional information, to generate more targeted outputs.
 
@@ -28,7 +28,7 @@ The discriminator, in a cGAN, also receives the same conditional information as
 
 During training, the generator learns to generate outputs that not only look realistic but also match the desired conditional information, while the discriminator learns to distinguish between real and generated data with respect to the given conditional information.
 
-## Training 2
+### Training 2
 
 The cGAN was trained to identify if building facades images from the **CMP Facade Database**, have generated by a model or are real.