Update README.md

README.md

@@ -25,10 +25,10 @@ Can precise categorization of card attributes, particularly by archetype, in bot
 ### Data
 We initially sourced a total of 10,763 English Yu-Gi-Oh! cards using their official API, ensuring fair use. This dataset included original card images along with associated details such as card ID, name, description, and related information. Additionally, cropped images from within cards were obtained to train our GAN and stable diffusion models. Focusing on archetype testing, we finalized our dataset to include 4,252 training images and 1,064 testing images. While refraining from publicly sharing this data due to potential ownership concerns, we have made these images accessible on the Hugging Face dataset to facilitate research in image generation challenges using GANs, particularly within the context of Yu-Gi-Oh! cards.
 
-#### Stable Diffusion Model Image Generation Example
+### Stable Diffusion Model Image Generation Example
 ![Stable Diffusion Model Image Generation Example](https://github.com/nogibjj/Generating-Yu-Gi-Oh-Monsters-From-Archetypes/assets/141780408/46949cd7-d554-4915-8934-813e64c674e9)
 
-#### GIT Model Image Captioning Example
+### GIT Model Image Captioning Example
 ![GIT Model Image Captioning Example](https://github.com/nogibjj/Generating-Yu-Gi-Oh-Monsters-From-Archetypes/assets/141780408/be2b1419-8215-42d5-b282-842dfd2aa91c)
 
 ### Streamlit App