search

SeungjunNah / DeepDeblur-PyTorch

Deep Multi-scale CNN for Dynamic Scene Deblurring
MIT License
262 stars 42 forks source link

Steps to run this project on personal device is so complecated #17

Open yazeedhasan97 opened 4 years ago

yazeedhasan97 commented 4 years ago

Hi DeepDeblur Team, Thanks for your great work

I am really sorry to say that, but the steps to run this project on personal device is so complected. I have spent around 4 hours looping around, modifying and testing, still not able to run the demo. inputs are out of numbers and totally unorganized

I want to ask if there is any modification or restructuring is going to be implemented in this regards (as adding some configuration files for the input parameters)

Also, I noted that it doesn't work without GPU, is there is any intention to implement it for both cases where GPU is available and not?

SeungjunNah commented 4 years ago

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 without modifying the source code.

Can you tell me how you tried to run demo? If my explanation confuses you, I will update it.

I updated the code so that CPU execution is enabled by setting --device_type cpu as an input argument. But please beware that CPU execution is very slow.