Update README.md
Browse files
README.md
CHANGED
|
@@ -1,8 +1,23 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
| 4 |
|
|
|
|
| 5 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
|
| 7 |
|
| 8 |
```
|
|
@@ -26,7 +41,7 @@ class DocOwlInfer():
|
|
| 26 |
|
| 27 |
docowl = DocOwlInfer(ckpt_path='mPLUG/DocOwl2')
|
| 28 |
|
| 29 |
-
images = ['/
|
| 30 |
|
| 31 |
query = "What is this paper about, provide detailed information."
|
| 32 |
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pipeline_tag: image-text-to-text
|
| 6 |
+
tags:
|
| 7 |
+
- chat
|
| 8 |
---
|
| 9 |
|
| 10 |
+
# mPLUG-DocOwl2
|
| 11 |
|
| 12 |
+
## Introduction
|
| 13 |
+
mPLUG-DocOwl2 is a state-of-the-art Multimodal LLM for OCR-free Multi-page Document Understanding.
|
| 14 |
+
|
| 15 |
+
Through a compressing module named High-resolution DocCompressor, each page is encoded with just 324 tokens.
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
Github: [mPLUG-DocOwl](https://github.com/X-PLUG/mPLUG-DocOwl)
|
| 19 |
+
|
| 20 |
+
## Quickstart
|
| 21 |
|
| 22 |
|
| 23 |
```
|
|
|
|
| 41 |
|
| 42 |
docowl = DocOwlInfer(ckpt_path='mPLUG/DocOwl2')
|
| 43 |
|
| 44 |
+
images = ['./examples/paper_page1.png', './examples/paper_page2.png', './examples/paper_page3.png']
|
| 45 |
|
| 46 |
query = "What is this paper about, provide detailed information."
|
| 47 |
|