nithin-gk / t2v-ddpm Goto Github PK

can you give a latest.pt?

Hi，thanks for your excellent work！
Could you provide the model pretrained with the ImageNet mentioned in your work？
“For the diffusion model, we use the same parameters as for the ImageNet superresolution model as used in improved diffusion [17]. We initialize the model with the imageNet pretrained weights.”
I would be appreciate

I want to specify gpu to train a model using CUDA_VISIBLE_DEVICES="1", but the model is still trained on gpu0, what should I do, thank you!

It seems that the repository of T2V-DDPM is empty, can the author open this up again? Thanks very much!

I would like to inquire about the specific format of the dataset under train

train
vis 1.jpg 2.jpg ...
nir 1.jpg 2.jpg ....

Is the infrared and visible light data one-to-one correspondence？

I have studied the code, but I sent emails to the authors of the two datasets used in the paper and asked them to use it, but I waited for a long time but didn't receive any reply. Therefore, I would like to apply to you directly for the VIS-TH dataset and ARL-VTF dataset used in the paper. Thank you very much!