Update README.md

README.md

@@ -5,12 +5,12 @@ language:
 - en
 - zh
 base_model:
-- Qwen/Qwen2-VL-7B-Instruct
+- Qwen/Qwen2-VL-2B-Instruct
 library_name: transformers
 tags:
 - erax
 - multimodal
-- erax-vl-7b
+- erax-vl-2B
 - insurance
 - ocr
 - vietnamese
@@ -28,38 +28,36 @@ widget:
 </p>
 
 
-# EraX-VL-7B-V1
+# EraX-VL-2B-V2.0
 ## Introduction 🎉
 
 <!-- <p style="text-align: justify;">
-    We are excited to introduce **EraX-VL-7B-v1**, a robust multimodal model for **OCR (optical character recognition)** and **VQA (visual question-answering)** that excels in various languages 🌍, with a particular focus on Vietnamese 🇻🇳. The `EraX-VL-7B` model stands out for its precise recognition capabilities across a range of documents 📝, including medical forms 🩺, invoices 🧾, bills of sale 💳, quotes 📄, and medical records 💊. This functionality is expected to be highly beneficial for hospitals 🏥, clinics 💉, insurance companies 🛡️, and other similar applications 📋. Built on the solid foundation of the [Qwen/Qwen2-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct)[1], which we found to be of high quality and fluent in Vietnamese, `EraX-VL-7B` has been fine-tuned to enhance its performance. We plan to continue improving and releasing new versions for free, along with sharing performance benchmarks in the near future.
+    We are excited to introduce **EraX-VL-2B-V2.0**, a robust multimodal model for **OCR (optical character recognition)** and **VQA (visual question-answering)** that excels in various languages 🌍, with a particular focus on Vietnamese 🇻🇳. The `EraX-VL-2B` model stands out for its precise recognition capabilities across a range of documents 📝, including medical forms 🩺, invoices 🧾, bills of sale 💳, quotes 📄, and medical records 💊. This functionality is expected to be highly beneficial for hospitals 🏥, clinics 💉, insurance companies 🛡️, and other similar applications 📋. Built on the solid foundation of the [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct)[1], which we found to be of high quality and fluent in Vietnamese, `EraX-VL-2B` has been fine-tuned to enhance its performance. We plan to continue improving and releasing new versions for free, along with sharing performance benchmarks in the near future.
 
-One standing-out feature of **EraX-VL-7B-v1** is the capability to do multi-turn Q&A with pretty good reasoning! Thanks for the size of 7+ billions parameters of base model.
+One standing-out feature of **EraX-VL-2B-V2.0** is the capability to do multi-turn Q&A with pretty good reasoning! Thanks for the size of 7+ billions parameters of base model.
 
-**EraX-VL-7B-V1** is a young member of our **EraX's LànhGPT** collection of LLM models.
+**EraX-VL-2B-V2.0** is a young member of our **EraX's LànhGPT** collection of LLM models.
 </p> -->
 
-**WE ARE MOVING to <a href="https://huggingface.co/erax-ai/EraX-VL-7B-V1/" target="_blank">EraX-AI</a> repository from 22 October 2024. Follow up so you do not miss great news coming up.**
+**WE ARE MOVING to <a href="https://huggingface.co/erax-ai/EraX-VL-2B-V2.0/" target="_blank">EraX-AI</a> repository from 22 December 2024. Follow up so you do not miss great news coming up.**
 
-We are excited to introduce **EraX-VL-7B-v1**, a robust multimodal model for **OCR (optical character recognition)** and **VQA (visual question-answering)** that excels in various languages 🌍, with a particular focus on Vietnamese 🇻🇳. The `EraX-VL-7B` model stands out for its precise recognition capabilities across a range of documents 📝, including medical forms 🩺, invoices 🧾, bills of sale 💳, quotes 📄, and medical records 💊. This functionality is expected to be highly beneficial for hospitals 🏥, clinics 💉, insurance companies 🛡️, and other similar applications 📋. Built on the solid foundation of the [Qwen/Qwen2-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct)[1], which we found to be of high quality and fluent in Vietnamese, `EraX-VL-7B` has been fine-tuned to enhance its performance. We plan to continue improving and releasing new versions for free, along with sharing performance benchmarks in the near future.
+We are excited to introduce **EraX-VL-2B-V2.0**, a robust multimodal model for **OCR (optical character recognition)** and **VQA (visual question-answering)** that excels in various languages 🌍, with a particular focus on Vietnamese 🇻🇳. The `EraX-VL-2B` model stands out for its precise recognition capabilities across a range of documents 📝, including medical forms 🩺, invoices 🧾, bills of sale 💳, quotes 📄, and medical records 💊. This functionality is expected to be highly beneficial for hospitals 🏥, clinics 💉, insurance companies 🛡️, and other similar applications 📋. Built on the solid foundation of the [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct)[1], which we found to be of high quality and fluent in Vietnamese, `EraX-VL-2B` has been fine-tuned to enhance its performance. We plan to continue improving and releasing new versions for free, along with sharing performance benchmarks in the near future.
 
-One standing-out feature of **EraX-VL-7B-v1** is the capability to do multi-turn Q&A with pretty good reasoning! Thanks for the size of 7+ billions parameters of base model.
+One standing-out feature of **EraX-VL-2B-V2.0** is the capability to do multi-turn Q&A with pretty good reasoning! Thanks for the size of 2+ billions parameters of base model.
 
-***NOTA BENE***: EraX-VL-7B-V1 is NOT a typical OCR-only tool likes Tesseract but is a Multimodal LLM-based model. To use it effectively, you may have to **twist your prompt carefully** depending on your tasks.
+***NOTA BENE***: EraX-VL-2B-V2.0 is NOT a typical OCR-only tool likes Tesseract but is a Multimodal LLM-based model. To use it effectively, you may have to **twist your prompt carefully** depending on your tasks.
 
-**EraX-VL-7B-V1** is a young member of our **EraX's LànhGPT** collection of LLM models.
+**EraX-VL-2B-V2.0** is a young member of our **EraX's LànhGPT** collection of LLM models.
 
 - **Developed by:**
   - Nguyễn Anh Nguyên ([email protected])
   - Nguyễn Hồ Nam (BCG)
-  - Hoàng Tiến Dũng ([email protected])
-  - Phạm Huỳnh Nhật ([email protected])
   - Phạm Đình Thục ([email protected])
 - **Funded by:** [Bamboo Capital Group](https://bamboocap.com.vn) and EraX
-- **Model type:** Multimodal Transformer with over 7B parameters
+- **Model type:** Multimodal Transformer with over 2B parameters
 - **Languages (NLP):** Primarily Vietnamese with multilingual capabilities
 - **License:** Apache 2.0
-- **Fine-tuned from:** [Qwen/Qwen2-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-7B-Instruct)
+- **Fine-tuned from:** [Qwen/Qwen2-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen2-VL-2B-Instruct)
 
 ## Benchmarks 📊
 
@@ -273,7 +271,7 @@ python -m pip install qwen-vl-utils
 pip install flash-attn --no-build-isolation
 ```
 
-Then you can use `EraX-VL-7B-V1` like this:
+Then you can use `EraX-VL-2B-V2.0` like this:
 ```python
 import os
 import base64
@@ -287,7 +285,7 @@ import torch
 from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
 from qwen_vl_utils import process_vision_info
 
-model_path = "erax/EraX-VL-7B-V1"
+model_path = "erax/EraX-VL-2B-V2.0"
 
 model = Qwen2VLForConditionalGeneration.from_pretrained(
     model_path,
@@ -371,23 +369,23 @@ print(output_text[0])
 ```
 
 ## Acknowledgments 👏
-We thank Khang Đoàn ([5CD-AI](https://huggingface.co/5CD-AI)) for his invaluable support in order to train `EraX-VL-7B-V1`. Our appreciation also goes to AAA JS Company for their support and resources, which significantly contributed to this project. 
+We thank Khang Đoàn ([5CD-AI](https://huggingface.co/5CD-AI)) for his invaluable support in order to train `EraX-VL-2B-V2.0`. Our appreciation also goes to AAA JS Company for their support and resources, which significantly contributed to this project. 
 
 ## Citation 📝
-<!-- - title={EraX-VL-7B-V1: A Highly Efficient Multimodal LLM for Vietnamese, especially for medical forms and bills.},
+<!-- - title={EraX-VL-2B-V2.0: A Highly Efficient Multimodal LLM for Vietnamese, especially for medical forms and bills.},
 - author={Nguyễn Anh Nguyên and Nguyễn Hồ Nam (BCG) and Dũng Hoàng and Thục Phạm and Nhật Phạm},
 - helpers={Khang Đoàn and AAA JS Company},
 - contact={[email protected]},
 - organization={EraX} -->
 If you find our project useful, we would appreciate it if you could star our repository and cite our work as follows:
 ```
-@article{EraX-VL-7B-V1,
-  title={EraX-VL-7B-V1: A Highly Efficient Multimodal LLM for Vietnamese, especially for medical forms and bills},
+@article{EraX-VL-2B-V2.0,
+  title={EraX-VL-2B-V2.0: A Highly Efficient Multimodal LLM for Vietnamese, especially for medical forms and bills},
   author={Nguyễn Anh Nguyên and Nguyễn Hồ Nam (BCG) and Hoàng Tiến Dũng and Phạm Đình Thục and Phạm Huỳnh Nhật},
   organization={EraX},
   year={2024},
-  url={https://huggingface.co/erax-ai/EraX-VL-7B-V1},
-  github={https://github.com/EraX-JS-Company/erax-vl-7b-v1/}
+  url={https://huggingface.co/erax-ai/EraX-VL-2B-V2.0},
+  github={https://github.com/EraX-JS-Company/EraX-VL-2B-V2.0/}
 }
 ```
 
@@ -407,4 +405,4 @@ If you find our project useful, we would appreciate it if you could star our rep
 
 ## Contact 🤝
 - For correspondence regarding this work or inquiry for API trial, please contact Nguyễn Anh Nguyên at [[email protected]]([email protected]).
-- Follow us on <b><a href="https://github.com/EraX-JS-Company/erax-vl-7b-v1/" target="_blank">EraX Github</a></b>
+- Follow us on <b><a href="https://github.com/EraX-JS-Company/EraX-VL-2B-V2.0/" target="_blank">EraX Github</a></b>