Open ghost opened 2 years ago
I don't want to use any pre-generated datasets, let's assume I have n images that contain text, and n lists of coordinates per image which have the text coordinates. I need to do something like the following:
trainer = MyTrainer(list_of_n_images, list_of_n_coordinates) trainer.train(epochs, checkpoint)
Is this something I can achieve using the code in this repo? if so, how?
I don't want to use any pre-generated datasets, let's assume I have n images that contain text, and n lists of coordinates per image which have the text coordinates. I need to do something like the following:
Is this something I can achieve using the code in this repo? if so, how?