Improve model card: Add metadata, update paper & add GitHub links
#1
by nielsr HF Staff - opened
README.md
CHANGED
|
@@ -1,3 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
<p align="center">
|
| 2 |
<img src="images/logo.png" width="700"/>
|
| 3 |
<p>
|
|
@@ -8,17 +14,20 @@ POINTS-Reader: Distillation-Free Adaptation of Vision-Language Models for Docume
|
|
| 8 |
|
| 9 |
<p align="center">
|
| 10 |
<a href="https://huggingface.co/tencent/POINTS-Reader">
|
| 11 |
-
<img src="https://img.shields.io/badge/
|
|
|
|
|
|
|
|
|
|
| 12 |
</a>
|
| 13 |
-
<a href="">
|
| 14 |
-
<img src="https://img.shields.io/badge/
|
| 15 |
</a>
|
| 16 |
-
<a href="">
|
| 17 |
<img src="https://komarev.com/ghpvc/?username=tencent&repo=POINTS-Reader&color=brightgreen&label=Views" alt="view">
|
| 18 |
</a>
|
| 19 |
</p>
|
| 20 |
|
| 21 |
-
We are delighted to announce that the WePOINTS family has welcomed a new member: POINTS-Reader, a vision-language model for end-to-end document conversion.
|
| 22 |
|
| 23 |
## News
|
| 24 |
|
|
@@ -578,8 +587,10 @@ import torch
|
|
| 578 |
# We recommend using the following prompt to better performance,
|
| 579 |
# since it is used throughout the training process.
|
| 580 |
prompt = (
|
| 581 |
-
'Please extract all the text from the image with the following requirements:
|
| 582 |
-
|
|
|
|
|
|
|
| 583 |
'2. Return all other text in Markdown format.'
|
| 584 |
)
|
| 585 |
image_path = '/path/to/your/local/image'
|
|
@@ -712,8 +723,10 @@ def call_wepoints(messages: List[dict],
|
|
| 712 |
return response
|
| 713 |
|
| 714 |
prompt = (
|
| 715 |
-
'Please extract all the text from the image with the following requirements:
|
| 716 |
-
|
|
|
|
|
|
|
| 717 |
'2. Return all other text in Markdown format.'
|
| 718 |
)
|
| 719 |
|
|
@@ -772,4 +785,4 @@ If you use this model in your work, please cite the following paper:
|
|
| 772 |
journal={arXiv preprint arXiv:2405.11850},
|
| 773 |
year={2024}
|
| 774 |
}
|
| 775 |
-
```
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: transformers
|
| 4 |
+
pipeline_tag: image-text-to-text
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
<p align="center">
|
| 8 |
<img src="images/logo.png" width="700"/>
|
| 9 |
<p>
|
|
|
|
| 14 |
|
| 15 |
<p align="center">
|
| 16 |
<a href="https://huggingface.co/tencent/POINTS-Reader">
|
| 17 |
+
<img src="https://img.shields.io/badge/%F0%9F%A4%97_HuggingFace-Model-ffbd45.svg" alt="HuggingFace">
|
| 18 |
+
</a>
|
| 19 |
+
<a href="https://github.com/Tencent/POINTS-Reader">
|
| 20 |
+
<img src="https://img.shields.io/badge/GitHub-Code-blue.svg?logo=github&" alt="GitHub Code">
|
| 21 |
</a>
|
| 22 |
+
<a href="https://huggingface.co/papers/2509.01215">
|
| 23 |
+
<img src="https://img.shields.io/badge/Paper-POINTS--Reader-d4333f?logo=arxiv&logoColor=white&colorA=cccccc&colorB=d4333f&style=flat" alt="Paper">
|
| 24 |
</a>
|
| 25 |
+
<a href="https://komarev.com/ghpvc/?username=tencent&repo=POINTS-Reader&color=brightgreen&label=Views" alt="view">
|
| 26 |
<img src="https://komarev.com/ghpvc/?username=tencent&repo=POINTS-Reader&color=brightgreen&label=Views" alt="view">
|
| 27 |
</a>
|
| 28 |
</p>
|
| 29 |
|
| 30 |
+
We are delighted to announce that the WePOINTS family has welcomed a new member: POINTS-Reader, a vision-language model for end-to-end document conversion, as introduced in the paper [POINTS-Reader: Distillation-Free Adaptation of Vision-Language Models for Document Conversion](https://huggingface.co/papers/2509.01215).
|
| 31 |
|
| 32 |
## News
|
| 33 |
|
|
|
|
| 587 |
# We recommend using the following prompt to better performance,
|
| 588 |
# since it is used throughout the training process.
|
| 589 |
prompt = (
|
| 590 |
+
'Please extract all the text from the image with the following requirements:
|
| 591 |
+
'
|
| 592 |
+
'1. Return tables in HTML format.
|
| 593 |
+
'
|
| 594 |
'2. Return all other text in Markdown format.'
|
| 595 |
)
|
| 596 |
image_path = '/path/to/your/local/image'
|
|
|
|
| 723 |
return response
|
| 724 |
|
| 725 |
prompt = (
|
| 726 |
+
'Please extract all the text from the image with the following requirements:
|
| 727 |
+
'
|
| 728 |
+
'1. Return tables in HTML format.
|
| 729 |
+
'
|
| 730 |
'2. Return all other text in Markdown format.'
|
| 731 |
)
|
| 732 |
|
|
|
|
| 785 |
journal={arXiv preprint arXiv:2405.11850},
|
| 786 |
year={2024}
|
| 787 |
}
|
| 788 |
+
```
|