docling.rs's PDF/image pipeline uses ONNX graphs that are format
conversions of docling-project's own PyTorch models, not weights docling.rs
trains or modifies. They're licensed separately from docling.rs's own MIT
code (see LICENSE) under their upstream terms:
| Model | Source | License |
|---|---|---|
RT-DETR layout model (layout_heron.onnx) |
docling-project/docling-layout-heron |
Apache-2.0 |
TableFormer (tableformer/{encoder,decoder,bbox}.onnx) |
docling-project/docling-models (model_artifacts/tableformer/accurate) |
CDLA-Permissive-2.0 / Apache-2.0 |
DocumentFigureClassifier (picture_classifier.onnx) |
docling-project/DocumentFigureClassifier-v2.5 (upstream's own ONNX, re-hosted unmodified) |
Apache-2.0 |
CodeFormulaV2 (cf_{vision,embed,decoder_kv}.onnx + cf_tokenizer.json; cf_decoder_kv_int8.onnx is a post-training quantization of the same export) |
docling-project/CodeFormulaV2 |
Apache-2.0 |
scripts/install/export_layout.py, scripts/install/export_tableformer.py and
scripts/install/export_code_formula.py do the conversion (PyTorch → ONNX via
torch.onnx.export); no weights are retrained, fine-tuned, or otherwise
altered. .github/workflows/publish-models.yml runs
that conversion (and re-hosts pdfium + the OCR model alongside it) and
publishes everything as GitHub Release assets on this repo (tag models-v1),
fetched by scripts/install/download_dependencies.sh — see that script and
crates/docling-node/deps.js — purely to spare downstream users the
PyTorch/transformers/docling_ibm_models toolchain needed to export them
locally.
Both upstream licenses permit redistribution with attribution; this file is that attribution. See the linked model cards for the full license text and any additional upstream terms.