Update README.md

README.md

@@ -35,7 +35,7 @@ It is a fine-tune of **Qwen 2.5-VL-7B** using ~10k synthetic Doc-to-Reasoning-to
 
 **NuMarkdown-reasoning** is significantly better than similar size non-reasoning models trained for markdown generation on complex documents, and achieves competitive results against top closed source alternatives.
 
-### Arena ranking (using trueskill-2 ranking system, with around 500 votes):
+### Arena ranking agains popular alternative (using trueskill-2 ranking system, with around 500 votes):
 <p align="center">
   
 | Rank | Model                                   | μ     | σ    | μ − 3σ |
@@ -52,7 +52,7 @@ It is a fine-tune of **Qwen 2.5-VL-7B** using ~10k synthetic Doc-to-Reasoning-to
 
 *We plan to realease a markdown arena, similar to llmArena, for complex document-to-markdown tasks to provide a tool to evaluate different solutions.*
 
-### Win-rate against others models (image-only):
+### Win/Draw/Loose-rate against others models (image-only):
 <p align="center">
 <img src="bar plot.png" width="700"/>
 </p>