Instructions to use RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("zero-shot-image-classification", model="RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea") pipe( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png", candidate_labels=["animals", "humans", "landscape"], )# Load model directly from transformers import AutoProcessor, AutoModelForZeroShotImageClassification processor = AutoProcessor.from_pretrained("RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea") model = AutoModelForZeroShotImageClassification.from_pretrained("RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea") - Notebooks
- Google Colab
- Kaggle
docs: add playground link
Browse files
README.md
CHANGED
|
@@ -1,35 +1,36 @@
|
|
| 1 |
-
---
|
| 2 |
-
library_name: transformers
|
| 3 |
-
license: cc-by-nc-sa-4.0
|
| 4 |
-
tags: []
|
| 5 |
-
---
|
| 6 |
-
|
| 7 |
-
## Model Details
|
| 8 |
-
|
| 9 |
-
- Model: ReasonSigLIP2-go16-384-S0-Rea
|
| 10 |
-
- Base model: [google/siglip2-giant-opt-patch16-384](https://huggingface.co/google/siglip2-giant-opt-patch16-384)
|
| 11 |
-
- Architecture: SigLIP2 ViT-Giant-Opt/16
|
| 12 |
-
- Image resolution: 384
|
| 13 |
-
- Training stage: Stage 0 - Reasoning
|
| 14 |
-
- Training data: Only reasoning caption-image pairs from [ReasonLite-42M](https://huggingface.co/datasets/RISys-Lab/ReasonCLIPLite-42M) and [ReasonPro-16M](https://huggingface.co/datasets/RISys-Lab/ReasonCLIPPro-16M)
|
| 15 |
-
|
| 16 |
-
## Method
|
| 17 |
-
|
| 18 |
-

|
| 19 |
-
|
| 20 |
-
## Resources
|
| 21 |
-
|
| 22 |
-
- GitHub: [RISys-Lab/ReasonCLIP](https://github.com/RISys-Lab/ReasonCLIP)
|
| 23 |
-
- Paper: [arXiv:2606.26794](https://arxiv.org/abs/2606.26794)
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: cc-by-nc-sa-4.0
|
| 4 |
+
tags: []
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
## Model Details
|
| 8 |
+
|
| 9 |
+
- Model: ReasonSigLIP2-go16-384-S0-Rea
|
| 10 |
+
- Base model: [google/siglip2-giant-opt-patch16-384](https://huggingface.co/google/siglip2-giant-opt-patch16-384)
|
| 11 |
+
- Architecture: SigLIP2 ViT-Giant-Opt/16
|
| 12 |
+
- Image resolution: 384
|
| 13 |
+
- Training stage: Stage 0 - Reasoning
|
| 14 |
+
- Training data: Only reasoning caption-image pairs from [ReasonLite-42M](https://huggingface.co/datasets/RISys-Lab/ReasonCLIPLite-42M) and [ReasonPro-16M](https://huggingface.co/datasets/RISys-Lab/ReasonCLIPPro-16M)
|
| 15 |
+
|
| 16 |
+
## Method
|
| 17 |
+
|
| 18 |
+

|
| 19 |
+
|
| 20 |
+
## Resources
|
| 21 |
+
|
| 22 |
+
- GitHub: [RISys-Lab/ReasonCLIP](https://github.com/RISys-Lab/ReasonCLIP)
|
| 23 |
+
- Paper: [arXiv:2606.26794](https://arxiv.org/abs/2606.26794)
|
| 24 |
+
- Playground: [ReasonCLIP Playground](https://www.sczhang.com/reasonclip/)
|
| 25 |
+
|
| 26 |
+
## Usage
|
| 27 |
+
|
| 28 |
+
```python
|
| 29 |
+
from transformers import AutoModel, AutoProcessor
|
| 30 |
+
|
| 31 |
+
model_id = "RISys-Lab/ReasonSigLIP2-go16-384-S0-Rea"
|
| 32 |
+
model = AutoModel.from_pretrained(model_id)
|
| 33 |
+
processor = AutoProcessor.from_pretrained(model_id)
|
| 34 |
+
```
|
| 35 |
+
|
| 36 |
+
For the full checkpoint list, see the [ReasonCLIP model card](https://github.com/RISys-Lab/ReasonCLIP/blob/main/doc/model_card.md).
|