nielsr HF Staff commited on
Commit
7c92d6d
·
verified ·
1 Parent(s): 81dd8f1

Add pipeline tag and improve metadata

Browse files

Hi, I'm Niels from the community science team at Hugging Face. This PR adds the `text-generation` pipeline tag to the model metadata to help users discover it more easily. It also improves the model card layout by providing clear links to the paper, code, and project page, and adds a brief description of the GDDS framework.

Files changed (1) hide show
  1. README.md +18 -16
README.md CHANGED
@@ -1,26 +1,29 @@
1
  ---
2
- license: mit
3
  library_name: pytorch
 
 
4
  tags:
5
- - gdds
6
- - discrete-diffusion
7
- - language-modeling
8
- - research
9
- - pytorch
10
  ---
11
 
12
  # GDDS Checkpoints
13
 
14
  Official checkpoint bundle for the paper **Generalized Discrete Diffusion from Snapshots**.
15
 
 
 
16
  ## Model Sources
17
 
18
- - Paper: https://huggingface.co/papers/2603.21342
19
- - arXiv: https://arxiv.org/abs/2603.21342
20
- - Code: https://github.com/ozekri/gdds
21
- - Project page: https://oussamazekri.fr/gdds
22
 
23
- ## Included checkpoints
24
 
25
  | File | Method | Notes |
26
  | --- | --- | --- |
@@ -35,8 +38,7 @@ Official checkpoint bundle for the paper **Generalized Discrete Diffusion from S
35
 
36
  ## Usage
37
 
38
- These files are PyTorch Lightning checkpoints intended to be used with the
39
- `gdds` codebase.
40
 
41
  ```bash
42
  git clone https://github.com/ozekri/gdds.git
@@ -44,6 +46,7 @@ cd gdds
44
  pip install -r requirements.txt
45
  pip install -e .
46
 
 
47
  PYTHONPATH=src python -m discrete_diffusion.evaluations.ppl_eval \
48
  data=openwebtext \
49
  model=small \
@@ -51,8 +54,7 @@ PYTHONPATH=src python -m discrete_diffusion.evaluations.ppl_eval \
51
  eval.checkpoint_path=/path/to/checkpoints/mdlm_500k.ckpt
52
  ```
53
 
54
- For sampling and other evaluations, use the same repository and pass the
55
- relevant checkpoint path through the evaluation config.
56
 
57
  ## Citation
58
 
@@ -66,4 +68,4 @@ relevant checkpoint path through the evaluation config.
66
  primaryClass={stat.ML},
67
  url={https://arxiv.org/abs/2603.21342},
68
  }
69
- ```
 
1
  ---
 
2
  library_name: pytorch
3
+ license: mit
4
+ pipeline_tag: text-generation
5
  tags:
6
+ - gdds
7
+ - discrete-diffusion
8
+ - language-modeling
9
+ - research
10
+ - pytorch
11
  ---
12
 
13
  # GDDS Checkpoints
14
 
15
  Official checkpoint bundle for the paper **Generalized Discrete Diffusion from Snapshots**.
16
 
17
+ Generalized Discrete Diffusion from Snapshots (GDDS) is a unified framework for discrete diffusion modeling that supports arbitrary noising processes over large discrete state spaces. It introduces a training objective based on snapshot latents rather than the entire noising path, allowing for efficient training and high-quality generation.
18
+
19
  ## Model Sources
20
 
21
+ - **Paper:** [Generalized Discrete Diffusion from Snapshots](https://huggingface.co/papers/2603.21342)
22
+ - **arXiv:** [2603.21342](https://arxiv.org/abs/2603.21342)
23
+ - **Code:** [GitHub - ozekri/gdds](https://github.com/ozekri/gdds)
24
+ - **Project Page:** [https://oussamazekri.fr/gdds](https://oussamazekri.fr/gdds)
25
 
26
+ ## Included Checkpoints
27
 
28
  | File | Method | Notes |
29
  | --- | --- | --- |
 
38
 
39
  ## Usage
40
 
41
+ These files are PyTorch Lightning checkpoints intended to be used with the [`gdds`](https://github.com/ozekri/gdds) codebase.
 
42
 
43
  ```bash
44
  git clone https://github.com/ozekri/gdds.git
 
46
  pip install -r requirements.txt
47
  pip install -e .
48
 
49
+ # Example evaluation using a checkpoint
50
  PYTHONPATH=src python -m discrete_diffusion.evaluations.ppl_eval \
51
  data=openwebtext \
52
  model=small \
 
54
  eval.checkpoint_path=/path/to/checkpoints/mdlm_500k.ckpt
55
  ```
56
 
57
+ For sampling and other evaluations, use the same repository and pass the relevant checkpoint path through the Hydra evaluation config.
 
58
 
59
  ## Citation
60
 
 
68
  primaryClass={stat.ML},
69
  url={https://arxiv.org/abs/2603.21342},
70
  }
71
+ ```