Update README.md

README.md

@@ -10,8 +10,6 @@
 
 [Telegram-bot](https://t.me/kandinsky21_bot)
 
-
-
 <p align="left">
 <img src="./content/kand_22.png" width="60%">
 </p>
@@ -22,12 +20,6 @@ The switch to CLIP-ViT-G as the image encoder significantly increases the model'
 
 The addition of the ControlNet mechanism allows the model to effectively control the process of generating images. This leads to more accurate and visually appealing outputs and opens new possibilities for text-guided image manipulation.
 
-**Inference regimes:**
-
-<p align="left">
-<img src="./content/kand_22_setting.png" width="60%">
-</p>
-
 **Architecture details:**
 
 + Text encoder (XLM-Roberta-Large-Vit-L-14) - 560M
@@ -43,6 +35,12 @@ The addition of the ControlNet mechanism allows the model to effectively control
 + [Inpainting](https://huggingface.co/kandinsky-community/kandinsky-2-2-decoder-inpaint): A decoding diffusion model mapping image embeddings and masked images to images
 + [СontrolТet-depth](https://huggingface.co/kandinsky-community/kandinsky-2-2-controlnet-depth): A decoding diffusion model mapping image embedding and additional depth condition to images
 
+### How to useInference regimes
+
+<p align="left">
+<img src="./content/kand_22_setting.png" width="60%">
+</p>
+
 ### How to use
 
 ```python