Kelvin C.K. Chan
Kelvin C.K. Chan
We add padding to the end of the input. I may need to modify it. Thanks
Hello, demo does not support 2GPUs. But you can use multiple GPUs by using `dist_test.py` with the corresponding config. In particular, you can modify the data paths [here](https://github.com/ckkelvinchan/RealBasicVSR/blob/master/configs/realbasicvsr_c64b20_1x30x8_lr5e-5_150k_reds.py#L292) and use...
Sure it is in progress. I will release the models for deblur and denoise today or tomorrow. After that I will work on the demo. Hopefully will release by this...
Hi @GeorvityLabs, the Colab demo is [here](https://colab.research.google.com/drive/1I0kZMM0DQyb4ueHZw5si8fMnRCJ_eUX3?usp=sharing). Please feel free to give me some comments if there is anything to improve :)
@GeorvityLabs Thanks for your comment! I will implement that later. BasicVSR++ accepts images with various sizes. But high-resolution inputs consume more memory.
Could you provide more information?
> @lhcwork Sorry, I do not find the x2 pretrained model. do you get?? I have just uploaded the checkpoints for BasicVSR_2x. You can find the checkpoints in the same...
May I ask what exactly is the problem you encountered? Could you provide any error log?
Did you use `filename_tmpl` as shown [here](https://github.com/ckkelvinchan/BasicVSR_PlusPlus/blob/master/configs/basicvsr_plusplus_denoise.py#L28)?
(1) The error log said CUDA out of memory. You may need to reduce the batch size or patch size. (2) and (4) I will need to check whether I...