Object Detection
libreyolo
dino-detr
detr

LibreDINODETRr50s5

The official five-scale ResNet-50 DINO checkpoint (49.4 AP on COCO val2017), packaged for LibreYOLO.

from libreyolo import LibreYOLO

model = LibreYOLO("LibreDINODETRr50s5.pt")
results = model.predict("image.jpg")

LibreYOLO ships this museum family for inference only. Its portable deformable attention uses the pure-PyTorch grid_sample reference path; the custom CUDA extension is not packaged. Native and ONNX inference use a fixed 800 x 800 PIL-bilinear stretch with ImageNet normalization. That deployment transform intentionally differs from upstream's short-side-800/max-1333 evaluation transform.

Source

Architecture and checkpoint released by IDEA-Research/DINO at commit d84a491d41898b3befd8294d1cf2614661fc0953. Copyright (c) 2022 IDEA. All Rights Reserved. The repository declares DINO under Apache License 2.0.

The official checkpoint0011_5scale.pth comes from the authors' Google Drive release. Source SHA-256: 1ccc1b6b7139813e4d3bfbeecfcf88347ebc226829769a0bf16c4a114c275cc0.

The checkpoint binary has no separate license file or license metadata. This mirror uses Apache-2.0 implied by the releasing repository's declaration, not a publisher-confirmed checkpoint-specific grant.

Modifications

The complete upstream state dict is copied without renaming or transforming learned tensors. Conversion strictly loads it against the native architecture and adds LibreYOLO checkpoint metadata. The released 91-column COCO category-id head is retained and mapped to contiguous COCO-80 classes during postprocess.

The converted checkpoint SHA-256 is 8dd59b36fff9750835fac7eb14c07a00f244bc0ec3f205dceac74907f0ef723a. LibreYOLO's output tensors are bit-exact against the pinned upstream pure-PyTorch path (max_abs_diff == 0.0). ONNX Runtime maximum absolute differences were 9.78e-5 for logits and 4.17e-6 for boxes.

See weights/convert_dinodetr_weights.py and docs/provenance/dinodetr.md in the LibreYOLO source repository.

License

Apache License 2.0 on the releasing-project implied basis described above. See the LICENSE and NOTICE files in this repository.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Dataset used to train LibreYOLO/LibreDINODETRr50s5

Collection including LibreYOLO/LibreDINODETRr50s5