Zhe Chen

Hi, you should compile the MSDeformAttn operator before running. ``` ln -s ../detection/ops ./ cd ops & sh make.sh # compile deformable attention ```

Hi, about how to perform instance segmentation only, you could refer to the document of [mmdetection](https://github.com/open-mmlab/mmdetection).

Thanks for your feedback. It's possible that the issue arose due to a change I made in the `drop-path-rate` while reorganizing the code. I kindly request a few days to...

I set drop-path-rate=0.1, and the results are as follows:  The best val mIoU is 51.35. Another difference is layer_scale, In the previous implementation, layer_scale is set to False, but...

Hi, what's the version of mmseg and mmcv you used? And, could you show me the command you used for training.

It's really weird, and I don't know why at the moment. Usually, I use the first command to inference a single image.

你好，权重载入的提示是不影响的， ``` RuntimeError: The size of tensor a (257) must match the size of tensor b (1025) at non-singleton dimension 3 ``` 报这个错是因为输入图像的分辨率不对，请问你推理用的图片是数据集里的吗，还是自己的数据？ 我在postdom的config里写的配置是： ``` test_pipeline = [ dict(type='LoadImageFromFile'), dict(...

因为你的系统好像是Windows, 编译Deformable Attention可能会有问题，你可以参考这个[issue](https://github.com/czczup/ViT-Adapter/issues/41)，进行一些修改。 I think the deformable attention is not compiled successfully. You can try this: replace line 11 import MultiScaleDeformableAttention as MSDA in the [ms_deform_attn_func.py](https://github.com/czczup/ViT-Adapter/blob/main/detection/ops/functions/ms_deform_attn_func.py) with from mmcv.ops.multi_scale_deform_attn import...

请问您使用的transformers版本是什么

你好，你可以参考[这个指南](https://github.com/OpenGVLab/InternVL/blob/main/internvl_chat/CONTINUED_FINETUNE.md)来微调InternVL-Chat-V1.2-Plus.