Instructions to use docling-project/docling-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use docling-project/docling-models with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("docling-project/docling-models", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Dual licenses?
#25
by waseemr - opened
I wonder why this package is marked as dual licensed? After DocLayNet model removal from this distribution, I was expecting this package to be only apache-2.0 licensed since it has only TableFormer model now. Any explanation?
Hello Docling team,
We are evaluating TableFormer for use in a commercial AI travel assistant and have the same question regarding the dual-license labels.
Could you please clarify which license applies to the following model weights at revision fc0f2d45e2218ea24bce5045f58a389aed16dc23?
model_artifacts/tableformer/fast/tableformer_fast.safetensorsmodel_artifacts/tableformer/accurate/tableformer_accurate.safetensors
In particular:
- Are both weights covered by Apache-2.0, CDLA-Permissive-2.0, or different licenses?
- Is commercial local inference permitted?
- Can these weights be redistributed inside a commercial Docker image?
- What LICENSE, NOTICE, and attribution requirements apply?
We would appreciate an authoritative clarification for these exact model files and revision.
Thank you!
ARIADNA Development Team