talentboy2333 / remote-sensing-image-caption Goto Github PK

I would like to add a research contribution in the remote sensing image captioning field. I have tried to download the RSICD dataset but could not succeed. I would appreciate it if you provide me the RSICD dataset.

First of all thanks to the author for this excellent work,this code is very heplful for me.I tried to training this model on 2080Ti, but this seems to take a long time. I taked the batch-size to 128 and loaded the mobilennet model through the GPU, but the training still seems to be slow! 60000 iterations takes me 2 months, My current training speed is about 1000 iterations in 20 hours,which is far beyong my expectation. I would like to know if there is any way to speed up the training, or if the author can provide a pre-trained model for the image-captioning! Thanks a lot!

About the "remote-sensing-image-caption/attention-model/model.py"
mobilenet = MobileNetV2(n_class=1000)
state_dict = torch.load('./models/mobilenet_v2.pth.tar', map_location='cpu')

Where can I download the "mobilenet_v2.pth.tar"?
I've been looking for it for a long time, but I can't find it. Can you help me? Thanks！