Spaces:
Running
on
Zero
Running
on
Zero
update readme
Browse files
README.md
CHANGED
@@ -1,56 +1,9 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
|
5 |
-
|
6 |
-
|
7 |
-
|
8 |
-
|
9 |
-
|
10 |
-
📢 [[Project Page](https://microsoft.github.io/OmniParser/)] [[Blog Post](https://www.microsoft.com/en-us/research/articles/omniparser-for-pure-vision-based-gui-agent/)] [[Models](https://huggingface.co/microsoft/OmniParser)]
|
11 |
-
|
12 |
-
**OmniParser** is a comprehensive method for parsing user interface screenshots into structured and easy-to-understand elements, which significantly enhances the ability of GPT-4V to generate actions that can be accurately grounded in the corresponding regions of the interface.
|
13 |
-
|
14 |
-
## News
|
15 |
-
- [2024/10] Both Interactive Region Detection Model and Icon functional description model are released! [Hugginface models](https://huggingface.co/microsoft/OmniParser)
|
16 |
-
- [2024/09] OmniParser achieves the best performance on [Windows Agent Arena](https://microsoft.github.io/WindowsAgentArena/)!
|
17 |
-
|
18 |
-
## Install
|
19 |
-
Install environment:
|
20 |
-
```python
|
21 |
-
conda create -n "omni" python==3.12
|
22 |
-
conda activate omni
|
23 |
-
pip install -r requirements.txt
|
24 |
-
```
|
25 |
-
|
26 |
-
Then download the model ckpts files in: https://huggingface.co/microsoft/OmniParser, and put them under weights/, default folder structure is: weights/icon_detect, weights/icon_caption_florence, weights/icon_caption_blip2.
|
27 |
-
|
28 |
-
Finally, convert the safetensor to .pt file.
|
29 |
-
```python
|
30 |
-
python weights/convert_safetensor_to_pt.py
|
31 |
-
```
|
32 |
-
|
33 |
-
## Examples:
|
34 |
-
We put together a few simple examples in the demo.ipynb.
|
35 |
-
|
36 |
-
## Gradio Demo
|
37 |
-
To run gradio demo, simply run:
|
38 |
-
```python
|
39 |
-
python gradio_demo.py
|
40 |
-
```
|
41 |
-
|
42 |
-
|
43 |
-
## 📚 Citation
|
44 |
-
Our technical report can be found [here](https://arxiv.org/abs/2408.00203).
|
45 |
-
If you find our work useful, please consider citing our work:
|
46 |
-
```
|
47 |
-
@misc{lu2024omniparserpurevisionbased,
|
48 |
-
title={OmniParser for Pure Vision Based GUI Agent},
|
49 |
-
author={Yadong Lu and Jianwei Yang and Yelong Shen and Ahmed Awadallah},
|
50 |
-
year={2024},
|
51 |
-
eprint={2408.00203},
|
52 |
-
archivePrefix={arXiv},
|
53 |
-
primaryClass={cs.CV},
|
54 |
-
url={https://arxiv.org/abs/2408.00203},
|
55 |
-
}
|
56 |
-
```
|
|
|
1 |
+
title: OmniParser: screen understanding tool for pure vision-based GUI agent
|
2 |
+
emoji: 🔥
|
3 |
+
colorFrom: yellow
|
4 |
+
colorTo: green
|
5 |
+
sdk: gradio
|
6 |
+
sdk_version: 3.14.0
|
7 |
+
app_file: app.py
|
8 |
+
pinned: false
|
9 |
+
license: agpl-3.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|