Open Source Image and Video Restoration Toolbox for Super-resolution, Denoise, Deblurring, etc. Currently, it includes EDSR, RCAN, SRResNet, SRGAN, ESRGAN, EDVR, BasicVSR, SwinIR, ECBSR, etc. Also support StyleGAN2, DFDNet.
Hi @xinntao,
My goal is to train an ESRGAN model to enhance the resolution of satellite images. Due to the fact that these kind of datasets are heavy for my computer, I decided to 1st train it on the DIV2K dataset to see how that goes.
I put inside the 'datasets' folder, 2 new folders: DIV2K (for testing) & Set14 (for validation).
DIV2K has:
DIV2K_train_HR
DIV2K_train_HR_sub (created by following the instructions on 'DatasetPreperation.md')
I also made the virtual environment, and run train.py through this command:
python .\basicsr\train.py -opt .\options\train\ESRGAN\train_ESRGAN_x4.yml
The error it shows me is this:
_FileNotFoundError: [Errno 2] No such file or directory: 'experiments/051_RRDBNet_PSNR_x4_f64b23_DIV2K_1000k_B16G1_wandb/models/net_g_1000000.pth'_
From what I have understood, 'experiments' is the folder inside of which the current progress of the training is saved, and validation does its work by taking info from here.
The file in the error already existed inside the provided code, so I don't understand what should be done.
Am I missing something in the process?
Congrats on your paper on ESRGAN by the way @xinntao, I am using it for my thesis!
Hi @xinntao, My goal is to train an ESRGAN model to enhance the resolution of satellite images. Due to the fact that these kind of datasets are heavy for my computer, I decided to 1st train it on the DIV2K dataset to see how that goes. I put inside the 'datasets' folder, 2 new folders: DIV2K (for testing) & Set14 (for validation). DIV2K has:
I also made the virtual environment, and run train.py through this command: python .\basicsr\train.py -opt .\options\train\ESRGAN\train_ESRGAN_x4.yml
The error it shows me is this: _FileNotFoundError: [Errno 2] No such file or directory: 'experiments/051_RRDBNet_PSNR_x4_f64b23_DIV2K_1000k_B16G1_wandb/models/net_g_1000000.pth'_
From what I have understood, 'experiments' is the folder inside of which the current progress of the training is saved, and validation does its work by taking info from here. The file in the error already existed inside the provided code, so I don't understand what should be done. Am I missing something in the process?
Congrats on your paper on ESRGAN by the way @xinntao, I am using it for my thesis!