zhenghuanghe

好的，感谢。

I downloaded the latest version and download dataset from google drive. When I running train.py.There are still some problems when I running train.py.  It seems that the data type...

 the type of _inputs_ is int32

 Thanks your help! There are so many place has this problem .I chang the data type from int32 to long in train.py line 114 .It can work now....

I chang np.int to np.long but result is same . 

Dear author,I change the _np.int_ to _np.int64_ and the problem was solved. My problem might be caused by numpy version. When I train the model,I find that it needs nearly...

My result is Avg recall BLEU 28.139705, avg precision BLEU 28.139705, F1 28.139705. There seems to be something wrong with my training. Can you provide a trained model ？

I am using the ViT-L/16 pretrained on SAT-493M dataset. Which layers should I use for my downstream task?

> Hi - for heads that requires pyramid feature maps we generally follow the implementation of MTP (https://arxiv.org/abs/2403.13430). MTP adds a small ViTdet style FPN to the backbone. https://github.com/ViTAE-Transformer/MTP/blob/962f7fd8781c095eb26db65ead3016e666b6d417/RS_Tasks_Finetune/Semantic_Segmentation/mmseg/models/backbones/vit_rvsa_mtp.py#L631 >...

I froze VITL-SAT493 and performed downstream segmentation fine-tuning, with the following model configuration: and use decoder: However, under this setup, the segmentation performance on open earth map is actually worse...