Improve DETR models (#19644)

* Improve DETR models

* Fix Deformable DETR loss and matcher

* Fixup

* Fix integration tests

* Improve variable names

* Apply suggestion

* Fix copies

* Fix DeformableDetrLoss

* Make Conditional DETR copy from Deformable DETR

* Copy from deformable detr's hungarian matcher

* Fix bug
This commit is contained in:
NielsRogge
2022-10-18 10:29:14 +02:00
committed by GitHub
parent 072dfdaee4
commit 90071fe42b
12 changed files with 253 additions and 257 deletions

View File

@@ -23,7 +23,7 @@ The abstract from the paper is the following:
Tips:
- One can use the [`AutoFeatureExtractor`] API to prepare images (and optional targets) for the model. This will instantiate a [`DetrFeatureExtractor`] behind the scenes.
- One can use [`DeformableDetrFeatureExtractor`] to prepare images (and optional targets) for the model.
- Training Deformable DETR is equivalent to training the original [DETR](detr) model. Demo notebooks can be found [here](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/DETR).
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/deformable_detr_architecture.png"