| --- |
| license: other |
| license_name: tinyformer-dinov3 |
| license_link: LICENSE |
| library_name: libreyolo |
| tags: |
| - object-detection |
| - tinyformer |
| --- |
| |
| # LibreTinyFormerx-obj2coco |
|
|
| TinyFormer-X-PBM detection weights, repackaged for LibreYOLO. |
| TinyFormer ("TinyFormer: Preserving Tiny Objects in YOLO-DETR Hybrid |
| Real-time Detectors", arXiv:2605.25046) is a DEIMv2-derived YOLO-DETR hybrid |
| specialised for tiny objects via a Spatial Semantic Adapter and a 4-scale |
| Parallel Bi-fusion neck. This size runs on a finetuned DINOv3 ViT-S/16+ tower. |
|
|
| Reported COCO val2017 mAP50-95 (Objects365 pretrain, 24-epoch COCO finetune): **60.2**. |
|
|
| This is the official Objects365-pretrained checkpoint finetuned on COCO for 24 epochs (the best-generalisation release). |
|
|
| ## Usage |
|
|
| ```python |
| from libreyolo import LibreYOLO |
| |
| model = LibreYOLO("LibreTinyFormerx-obj2coco.pt") # auto-downloads from this repo |
| results = model.predict("image.jpg") |
| ``` |
|
|
| ## Source |
|
|
| Derived from [mmpmmpmmpjosh/TinyFormer](https://github.com/mmpmmpmmpjosh/TinyFormer) |
| at commit [`9075d9f`](https://github.com/mmpmmpmmpjosh/TinyFormer/commit/9075d9fdf5ce989ca3c7e830a780dfc7fbbe4ec1), |
| sourced from the official [Google Drive checkpoint mirror](https://drive.google.com/drive/folders/1yIsUO14g5EyTFRbDeISoNJSDqNeSgT5o). |
| Copyright (C) 2026 AICVlab, National Yang Ming Chiao Tung University (NYCU). |
| The TinyFormer codebase is licensed under the Apache License, Version 2.0 |
| (DEIM/DEIMv2 lineage). |
|
|
| The backbone parameters derive from |
| [facebookresearch/dinov3](https://github.com/facebookresearch/dinov3) |
| (a finetuned DINOv3 ViT-S/16+ tower) and are licensed under the |
| [DINOv3 License](https://ai.meta.com/resources/models-and-libraries/dinov3-license/). |
| Because this checkpoint bundles DINOv3-derived parameters, the combined |
| distribution is governed by the terms of both licenses. |
|
|
| ## Modifications |
|
|
| Checkpoint metadata wrapping only — learned parameters are unchanged (the EMA |
| weights of the released solver checkpoint). The wrapper adds ``model_family``, |
| ``size``, ``nc``, and ``names`` fields so the unified ``LibreYOLO()`` factory |
| routes correctly without filename heuristics. |
| See `weights/convert_tinyformer_weights.py` in the |
| [LibreYOLO source repository](https://github.com/LibreYOLO/libreyolo). |
|
|
| ## License |
|
|
| Dual-licensed: Apache License 2.0 (TinyFormer/DEIMv2 components) and the |
| DINOv3 License (backbone parameters). The combined LICENSE file in this |
| repository contains both license texts. See also [`NOTICE`](./NOTICE) for |
| attribution detail. |
|
|
