Update README.md

README.md

@@ -15,13 +15,7 @@ RUDOLPH: One Hyper-Tasking Transformer Сan be Сreative as DALL-E and GPT-3 and
 <img src="https://raw.githubusercontent.com/sberbank-ai/ru-dolph/master/pics/RUDOLPH.png" width=60% border="2"/>
 
 
-Model was trained by [Sber AI](https://github.com/sberbank-ai) and [SberDevices](https://sberdevices.ru/) teams.  
-
-* Tasks: ` text2image generation, self reranking, text ranking, image ranking, image2text generation, zero-shot image classification, text2text generation`
-* Language: ` Russian`
-* Type: ` decoder`
-* Num Parameters: ` 350M`
-* Training Data Volume: `156 million text-image pairs`
+Model was trained by [Sber AI](https://github.com/ai-forever) team.  
 
 # Model Description
 
@@ -29,9 +23,13 @@ Model was trained by [Sber AI](https://github.com/sberbank-ai) and [SberDevices]
 
 *Hyper-tasking means generalized multi-tasking, e.g., the model that can solve almost all tasks within supported modalities (two modalities in case of RUDOLPH: images and Russian texts).*
 
-# Details of architecture
+* Tasks: ` text2image generation, self reranking, text ranking, image ranking, image2text generation, zero-shot image classification, text2text generation, and so on`
+* Language: ` Russian`
+* Type: ` decoder`
+* Num Parameters: ` 350M`
+* Training Data Volume: `156 million text-image pairs`
 
-### Parameters
+# Details of architecture
 
 <img src=https://raw.githubusercontent.com/ai-forever/ru-dolph/master/pics/scheme-rudolph_350m.jpg height="20" border="2"/>
 
@@ -43,7 +41,7 @@ RUDOLPH 350M is a Transformer-based decoder model with the following parameters:
 * hidden\_size (1024) — Dimensionality of the hidden layers.
 * num\_attention\_heads (16) — Number of attention heads for each attention layer.
 
-# Sparse Attention Mask
+# Sparse Attention Masks
 
 The primary proposed method is to modify the sparse transformer's attention mask to better control modalities. It allows us to calculate the transitions of modalities in both directions, unlike another similar work DALL-E Transformer, which used only one direction, "text to image". The proposed "image to right text" direction is achieved by extension sparse attention mask to the right for auto-repressively text generation with both image and left text condition.