Shilong Liu

You can try our stronger checkpoint with SwinB trained on over 10M images. See the download and config path here: https://github.com/IDEA-Research/GroundingDINO#luggage-checkpoints

> Actually, I haved download the pretrained weight from huggingface, unfortunately, it raise some error: Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertModel: ['cls.predictions.transform.dense.weight',...

谢谢，我们会尽快测试并给出反馈

the diffusion model requires the image size(width and height) to be dividable by 64. We suggest resize images to (512, 512) for inpainting. and resize to the original size as...

Thanks for the issue. The model output boxes with corresponding `noun` in the sentence. In the example, the described words may be ignored.

plz check if the CUDA_HOME variable is set

wow it is an interesting idea. our model doesn't support yet. but it may be an interesting work to explore

of course! like this: 'cat . dog .'

Try using pycocotools-window instead of pycocotools. refer to https://github.com/cocodataset/cocoapi/issues/415

可能我没注意。 重命名一下文件应该就行了