SESR icon indicating copy to clipboard operation
SESR copied to clipboard

Training with new data set

Open msiraj83 opened this issue 6 years ago • 1 comments

hello..

I want to train your model on my own data set. First there is no train.py file in your code. However I study the main.py file and found code train_set = DatasetFromHdf5("../lapsrn/data/data.h5") on line no 46. From where it get the data.h5 file. Can you guide me how i will train the model on my own data set.

Thanks

msiraj83 avatar May 16 '18 09:05 msiraj83

Hi, you could generate the .h5 file with these matlab files or you could rewrite your dataloader.py to fit your own data. https://github.com/twtygqyy/pytorch-LapSRN/tree/master/data You could also use the --pretrained method in main.py to finetuning from my model if you want to train faster.

opteroncx avatar May 16 '18 12:05 opteroncx