{"full_name":"facebookresearch/detr","owner":"facebookresearch","name":"detr","shard":"1","url":"https://github.com/facebookresearch/detr","description":"End-to-End Object Detection with Transformers","homepage":null,"topics":[],"tag_pages":[],"language":"Python","license":"Apache-2.0","avatar":"https://avatars.githubusercontent.com/u/16943930?v=4","created_at":"2020-05-26T23:54:52Z","pushed_at":"2024-03-12T15:58:25Z","metrics":{"stars":15344,"forks":2678,"watchers":3,"issues":526,"prs":89,"releases":1,"issues_open":240,"issues_closed":286,"prs_open":15,"prs_closed":29,"prs_merged":45,"discussions":0},"growth":{"day":{"stars":0,"forks":0,"base_day":"2026-11-15"}},"history":{"dates":["2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[15344,15345,15345,15345,15344,15344,15344],"forks":[2677,2678,2678,2678,2678,2678,2678],"watchers":[null,3,3,3,3,3,3],"issues_open":[null,240,240,240,240,240,240],"prs_open":[null,15,15,15,15,15,15]},"boards":{},"same_owner":[{"full_name":"facebookresearch/AugLy","name":"AugLy","shard":"1","boards":{}},{"full_name":"facebookresearch/ImageBind","name":"ImageBind","shard":"1","boards":{}},{"full_name":"facebookresearch/audiocraft","name":"audiocraft","shard":"1","boards":{}},{"full_name":"facebookresearch/codellama","name":"codellama","shard":"1","boards":{}},{"full_name":"facebookresearch/detectron2","name":"detectron2","shard":"1","boards":{}},{"full_name":"facebookresearch/dinov2","name":"dinov2","shard":"1","boards":{}},{"full_name":"facebookresearch/dinov3","name":"dinov3","shard":"1","boards":{}},{"full_name":"facebookresearch/faiss","name":"faiss","shard":"1","boards":{}},{"full_name":"facebookresearch/fastText","name":"fastText","shard":"1","boards":{}},{"full_name":"facebookresearch/hydra","name":"hydra","shard":"1","boards":{}},{"full_name":"facebookresearch/maskrcnn-benchmark","name":"maskrcnn-benchmark","shard":"1","boards":{}},{"full_name":"facebookresearch/mmf","name":"mmf","shard":"1","boards":{}},{"full_name":"facebookresearch/pifuhd","name":"pifuhd","shard":"1","boards":{}},{"full_name":"facebookresearch/pytorch3d","name":"pytorch3d","shard":"1","boards":{}},{"full_name":"facebookresearch/sam2","name":"sam2","shard":"1","boards":{}},{"full_name":"facebookresearch/sam3","name":"sam3","shard":"1","boards":{}},{"full_name":"facebookresearch/seamless_communication","name":"seamless_communication","shard":"1","boards":{}},{"full_name":"facebookresearch/segment-anything","name":"segment-anything","shard":"1","boards":{}},{"full_name":"facebookresearch/vggt","name":"vggt","shard":"1","boards":{}}],"issues":[{"number":9,"title":"Recommendations for training Detr on custom dataset?","comments":205,"created_at":"2020-05-28T05:01:51Z"},{"number":101,"title":"'assert (boxes1[:, 2:] >= boxes1[:, :2]).all()'  happened when training","comments":26,"created_at":"2020-06-24T06:27:50Z"},{"number":428,"title":"DETR tutorials to use it on custom data :) ","comments":21,"created_at":"2021-08-25T16:24:19Z"},{"number":211,"title":"onnx  export error and can the onnx model support tensorRT?","comments":19,"created_at":"2020-08-24T02:44:17Z"},{"number":498,"title":"model.train() and model.eval() gives different results","comments":17,"created_at":"2022-04-10T05:39:14Z"},{"number":156,"title":"How to run detr with only single class?","comments":17,"created_at":"2020-07-21T16:23:51Z"},{"number":125,"title":"Failing to converge on small datasets (Getting zeros on small custom data)","comments":16,"created_at":"2020-06-30T18:53:59Z"},{"number":29,"title":"Suggestion - change to ResNeST50 backbone  (new split attention arch)","comments":14,"created_at":"2020-06-01T05:02:56Z"}],"discussions":[],"latest_release":{"name":"Detectron2 and torchscript support, attention and panoptic notebooks, code improvements","tag":"v0.2","published_at":"2020-06-29T16:41:01Z","body":"This release contains several improvements compared to the initial release of DETR.\r\n\r\n# Highlights\r\n## Detectron2 support\r\nDETR now can be used with Detectron2 in a seamless way. It reuses the original abstractions of DETR, while making a compatibility layer so that it fits Detectron2's API. More information can be found in the [README under the `d2/` folder](https://github.com/facebookresearch/detr/blob/master/d2/README.md)\r\n\r\n## TorchScript support for all models\r\nDETR models can now be natively exported to torchscript. This enables DETR to be run in C++ via libtorch. Converting a model to torchscript is easy:\r\n```python\r\nmodel = torch.hub.load('facebookresearch/detr', 'detr_resnet50', pretrained=True)\r\nmodel = torch.jit.script(model)\r\n# now the model run entirely in torchscript!\r\n```\r\n\r\n## Notebooks for panoptic segmentation and attention visualization\r\nWe added two new notebooks in the repo, which should hopefully facilitate using and understanding how DETR works:\r\n\r\n- [DETR's hands on Colab Notebook](https://colab.research.google.com/github/facebookresearch/detr/blob/colab/notebooks/detr_attention.ipynb): Shows how to load a model from hub, generate predictions, then visualize the attention of the model (similar to the figures of the paper)\r\n- [Panoptic Colab Notebook](https://colab.research.google.com/github/facebookresearch/detr/blob/colab/notebooks/DETR_panoptic.ipynb): Demonstrates how to use DETR for panoptic segmentation and plot the predictions.\r\n\r\n## Training logs for all models\r\nFor easier reproducibility, we provide full training logs for the detection models available in torchhub. They are available together with the model weights in https://github.com/facebookresearch/detr#model-zoo\r\n\r\n# Changelog\r\n\r\n### New features\r\n- Add Detectron2 wrapper (#103, #119)\r\n- Add torchscript support for hub detection models (#51)\r\n- Add torchscript support to panoptic models (#87)\r\n- Hub config for panoptic models (#54)\r\n- Add Dockerfile (#26)\r\n- Add model logs to README (#86)\r\n\r\n### Improvements\r\n- avoid downloading the same file multiple times (#60)\r\n- improve robustness of plot_logs function, handle single directory argument for 'logs' param (#62)\r\n- Move colab to github from drive (#5)\r\n- Unifying the format of argparse (#24)\r\n\r\n### Bugfixes\r\n- Hot-fix for panoptic hub-conf (#56)\r\n- Update panoptic model hashes (#18)\r\n- update url (#8)\r\n- fix IndexError (#6)\r\n\r\n### Documentation\r\n- Add new colab link in README (#114)\r\n- Update README with details on segmentation (#100)\r\n- Changed incorrect bbox format in loss_bbox docstring (#90)\r\n- fix repo name (#2)\r\n- Add ISSUE_TEMPLATE (#31)","is_prerelease":false},"readme":true,"readme_zh":false,"news":[],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}