Seungjun Nah

@mcarilli Thanks for a great job in mixed-precision! I'm trying both apex.amp and torch.cuda.amp and both of them turn out to be effective in terms of memory reduction and speed...

@mcarilli I use ```max_loss_scale``` to avoid gradient explosion when training my models in this [line](https://github.com/SeungjunNah/DeepDeblur-PyTorch/blob/master/src/optim/__init__.py#L179) of my [repository](https://github.com/SeungjunNah/DeepDeblur-PyTorch). (loss scaling [here](https://github.com/SeungjunNah/DeepDeblur-PyTorch/blob/master/src/train.py#L98)) From my experiments, I know that gradients usually explode...

Ok, skipping with 1/2000 ratio doesn't hurt practically. I wanted to see if there were ways to control the number of iterations completely, though. Thanks for the explanation!

Hi @TouqeerAhmad, Thank you for your interests in our work. We are testing and cleaning up the code. Currently, I and Sanghyun are both involved in multiple other projects, so...

Hi @shekshaa I am aware I had this issue before and I corrected the frames in GOPRO_Large_all a while ago. When did you download the datasets? Could you download the...

@shekshaa I checked the frames and they look good to me. It seems like the problem you mentioned is not the issue I thought. 1. `GOPRO_Large/train/GOPR0372_07_00/sharp/000047.png` and `GOPRO_Large_all/train/GOPR0372_07_00/000326.png` are exactly...

Did you try the data on Google Drive?

Hi @yazeedhasan97, I agree that this code is quite complicated. I tried to make it as general as possible so that using different input arguments would lead to different behavior...

The data loader assumes that the input and the target have different key strings in their full paths. By default, it is set as '**blur_gamma**' and '**sharp**' for GOPRO_Large dataset,...

The message means that the code did not locate the dataset. Did you change `--data_root` and `--data_train` accordingly? You may want to check [L40-47 in option.py](https://github.com/SeungjunNah/DeepDeblur-PyTorch/blob/master/src/option.py#L40). For example, if `args.data_root`...