nielsr HF Staff commited on
Commit
1621321
·
verified ·
1 Parent(s): 3e4e8ba

Add pipeline tag and improve metadata

Browse files

Hi! I'm Niels from the Hugging Face community team.

I noticed this model card is missing some metadata. I've added `pipeline_tag: image-feature-extraction` to the YAML section to make the model more discoverable for users interested in visual tokenization and feature extraction.

I've also ensured the paper, code, and project links are correctly formatted and easy to find.

Files changed (1) hide show
  1. README.md +10 -6
README.md CHANGED
@@ -1,11 +1,13 @@
1
  ---
 
 
2
  tags:
3
  - model_hub_mixin
4
  - pytorch_model_hub_mixin
5
- license: apache-2.0
6
  ---
 
7
  # Communication-Inspired Tokenization for Structured Image Representations
8
- </h1>
9
  <p align="left">
10
  <a href="https://araachie.github.io">Aram Davtyan</a> •
11
  <a href="https://www.cvg.unibe.ch/people/sahin">Yusuf Sahin</a> •
@@ -16,21 +18,23 @@ license: apache-2.0
16
  <a href="https://www.cvg.unibe.ch/people/favaro">Paolo Favaro</a>
17
  </p>
18
 
19
- Official pre-trained models for the paper: https://arxiv.org/abs/2602.20731
20
 
21
- Project's website: https://araachie.github.io/comit/
22
 
23
  ## Installation
24
 
25
- Follow the instructions at https://github.com/Araachie/comit
26
 
27
  ## Usage
28
 
29
  Example usage, downloading `COMiT-B` from the Hugging Face Hub:
30
 
31
  ```python
 
32
  from comit import COMiT
33
 
 
34
  model = COMiT.from_pretrained('cvg-unibe/comit-b')
35
  model.eval().to(device)
36
  ```
@@ -84,7 +88,7 @@ with torch.no_grad():
84
  ## Licensing
85
 
86
  Unless otherwise noted, the model weights are licensed under Apache license 2.0.
87
- For the code licensing, see https://github.com/Araachie/comit?tab=readme-ov-file#licensing
88
 
89
  ## Citation
90
 
 
1
  ---
2
+ license: apache-2.0
3
+ pipeline_tag: image-feature-extraction
4
  tags:
5
  - model_hub_mixin
6
  - pytorch_model_hub_mixin
 
7
  ---
8
+
9
  # Communication-Inspired Tokenization for Structured Image Representations
10
+
11
  <p align="left">
12
  <a href="https://araachie.github.io">Aram Davtyan</a> •
13
  <a href="https://www.cvg.unibe.ch/people/sahin">Yusuf Sahin</a> •
 
18
  <a href="https://www.cvg.unibe.ch/people/favaro">Paolo Favaro</a>
19
  </p>
20
 
21
+ Official pre-trained models for the paper: [Communication-Inspired Tokenization for Structured Image Representations](https://huggingface.co/papers/2602.20731).
22
 
23
+ [Project Website](https://araachie.github.io/comit/) | [GitHub Repository](https://github.com/Araachie/comit)
24
 
25
  ## Installation
26
 
27
+ Follow the instructions at [https://github.com/Araachie/comit](https://github.com/Araachie/comit)
28
 
29
  ## Usage
30
 
31
  Example usage, downloading `COMiT-B` from the Hugging Face Hub:
32
 
33
  ```python
34
+ import torch
35
  from comit import COMiT
36
 
37
+ device = "cuda" if torch.cuda.is_available() else "cpu"
38
  model = COMiT.from_pretrained('cvg-unibe/comit-b')
39
  model.eval().to(device)
40
  ```
 
88
  ## Licensing
89
 
90
  Unless otherwise noted, the model weights are licensed under Apache license 2.0.
91
+ For the code licensing, see [https://github.com/Araachie/comit?tab=readme-ov-file#licensing](https://github.com/Araachie/comit?tab=readme-ov-file#licensing)
92
 
93
  ## Citation
94