nielsr HF Staff commited on
Commit
c063759
·
verified ·
1 Parent(s): 383c798

Add link to paper and project page

Browse files

This PR improves the model card by linking it to the paper and project page.

Files changed (1) hide show
  1. README.md +12 -4
README.md CHANGED
@@ -1,12 +1,14 @@
1
  ---
2
- license: apache-2.0
3
- library_name: transformers
4
  base_model: OpenGVLab/InternVL2-4B
 
 
5
  pipeline_tag: image-text-to-text
6
  ---
7
 
8
  # OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis
9
 
 
 
10
  <div align="center">
11
 
12
  [\[🏠Homepage\]](https://qiushisun.github.io/OS-Genesis-Home/) [\[💻Code\]](https://github.com/OS-Copilot/OS-Genesis) [\[📝Paper\]](https://arxiv.org/abs/2412.19723) [\[🤗Models\]](https://huggingface.co/collections/OS-Copilot/os-genesis-6768d4b6fffc431dbf624c2d)[\[🤗Data\]](https://huggingface.co/collections/OS-Copilot/os-genesis-6768d4b6fffc431dbf624c2d)
@@ -137,9 +139,15 @@ tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast
137
  pixel_values = load_image('./web_dfacd48d-d2c2-492f-b94c-41e6a34ea99f.png', max_num=6).to(torch.bfloat16).cuda()
138
  generation_config = dict(max_new_tokens=1024, do_sample=True)
139
 
140
- question = "<image>\nYou are a GUI task expert, I will provide you with a high-level instruction, an action history, a screenshot with its corresponding accessibility tree.\n High-level instruction: {high_level_instruction}\n Action history: {action_history}\n Accessibility tree: {a11y_tree}\n Please generate the low-level thought and action for the next step."
 
 
 
 
 
141
  response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
142
- print(f'User: {question}\nAssistant: {response}')
 
143
  ```
144
 
145
 
 
1
  ---
 
 
2
  base_model: OpenGVLab/InternVL2-4B
3
+ library_name: transformers
4
+ license: apache-2.0
5
  pipeline_tag: image-text-to-text
6
  ---
7
 
8
  # OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis
9
 
10
+ This model is described in the paper [OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis](https://huggingface.co/papers/2412.19723)
11
+
12
  <div align="center">
13
 
14
  [\[🏠Homepage\]](https://qiushisun.github.io/OS-Genesis-Home/) [\[💻Code\]](https://github.com/OS-Copilot/OS-Genesis) [\[📝Paper\]](https://arxiv.org/abs/2412.19723) [\[🤗Models\]](https://huggingface.co/collections/OS-Copilot/os-genesis-6768d4b6fffc431dbf624c2d)[\[🤗Data\]](https://huggingface.co/collections/OS-Copilot/os-genesis-6768d4b6fffc431dbf624c2d)
 
139
  pixel_values = load_image('./web_dfacd48d-d2c2-492f-b94c-41e6a34ea99f.png', max_num=6).to(torch.bfloat16).cuda()
140
  generation_config = dict(max_new_tokens=1024, do_sample=True)
141
 
142
+ question = "<image>
143
+ You are a GUI task expert, I will provide you with a high-level instruction, an action history, a screenshot with its corresponding accessibility tree.
144
+ High-level instruction: {high_level_instruction}
145
+ Action history: {action_history}
146
+ Accessibility tree: {a11y_tree}
147
+ Please generate the low-level thought and action for the next step."
148
  response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
149
+ print(f'User: {question}
150
+ Assistant: {response}')
151
  ```
152
 
153