strong train path
Closed this issue · 1 comments
dotrungkien3210 commented
in train_synth i see pretrained_path and loss_plot_training.npy, can you help me how to download it or how to generate it
mayank-git-hub commented
If you have set pretrained=False, you don't require the pretrained_path and the loss_plot_training.npy.
You can generate it by training the model with strong supervision.
The details on how to train using strong supervision on the synthtext dataset is detailed in the readme file.
A link to the pre-trained model on synthtext dataset is also provided in the readme file.