Update README.md

README.md

@@ -1,7 +1,14 @@
----
-library_name: transformers
-tags: []
----
+---
+library_name: transformers
+license: mit
+datasets:
+- 5CD-AI/LLaVA-CoT-o1-Instruct
+language:
+- vi
+- en
+- zh
+pipeline_tag: image-text-to-text
+---
 
 
 ![image/png](https://cdn-uploads.huggingface.co/production/uploads/6336b5c831efcb5647f00170/MagGtnflTgsxkel9U71ot.png)
@@ -10,15 +17,15 @@ tags: []
 
 Multimodal LLM x Reasoning Model 👀 🧠 🔍
 
-After more than six months since creating the 5CD-AI/LLaVA-CoT-o1-Instruct dataset—one of Hugging Face’s most liked datasets of 2024 🎉—we have just completed the "base" version of the Vintern Reasoning Model!
+After more than six months since creating the [5CD-AI/LLaVA-CoT-o1-Instruct](https://huggingface.co/datasets/5CD-AI/LLaVA-CoT-o1-Instruct) dataset—one of Hugging Face’s most liked datasets of 2024 🎉—we have just completed the "base" version of the Vintern Reasoning Model!
 - This model can perform long and complex reasoning based on images, breaking down each reasoning step into multiple sub-steps while keeping hallucinations under control.
  🛠️ We also successfully implemented the GRPO algorithm on the Vintern Multimodel model!
  - Despite the difficulty of balancing multiple tasks alongside reasoning, Vintern-3B-R-beta has outperformed all previous versions across various benchmarks! 
 
-When should you choose Vintern-1B-3.5 vs. Vintern-3B-R-beta? 🤔  
+When should you choose [Vintern-1B-v3_5](https://huggingface.co/5CD-AI/Vintern-1B-v3_5) vs Vintern-3B-R-beta? 🤔  
 
-- **Vintern-1B-v3.5**: Fast ⚡ and good for Vietnamese OCR with simple text formatting. 📝  
-- **Vintern-3B-R-beta**: Better for complex questions and structured text. 🔍📚 OCR performance on blurred or unclear text may be slightly affected due to our training focus on reasoning. 🔍🤖
+- **Vintern-1B-v3_5**: Fast ⚡ and good for Vietnamese OCR with simple text formatting. 📝 Highly reliable. ✅
+- **Vintern-3B-R-beta**: Better for complex questions and complex structured doc image. 🔍📚 OCR performance on blurred or unclear text may be slightly affected due to our training focus on reasoning. 🔍🤖
 
 🚀 The next step? Training and enhancing its reasoning ability by Reinforcement Learning!
 
@@ -286,5 +293,4 @@ print(f'User: {question}\nAssistant: {response}')
 
 ## Reference
 
-[1] Z. Chen et al., ‘Expanding performance boundaries of open-source multimodal models with model, data, and test-time scaling’, arXiv preprint arXiv:2412. 05271, 2024.
-
+[1] Z. Chen et al., ‘Expanding performance boundaries of open-source multimodal models with model, data, and test-time scaling’, arXiv preprint arXiv:2412. 05271, 2024.