Object Detection
ONNX
libreyolo
tinyformer
File size: 2,360 Bytes
3dd09a3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
---
license: other
license_name: tinyformer-dinov3
license_link: LICENSE
library_name: libreyolo
tags:
  - object-detection
  - tinyformer
---

# LibreTinyFormers

TinyFormer-S-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 the DEIMv2-distilled ViT-Tiny tower (DINOv3-distilled).

Reported COCO val2017 mAP50-95: **51.5**.

## Usage

```python
from libreyolo import LibreYOLO

model = LibreYOLO("LibreTinyFormers.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)
(the DEIMv2-distilled ViT-Tiny tower (DINOv3-distilled)) 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.