A question on the internal principle of DataGenerator

[claudehang]

Your code and blog are wonderful! Now I still have one question. In def __getitem__(self, index), the index argument comes from where? Could your explain it further? Many thanks!

[claudehang]

And how should I understand this in your blog: A common practice is to set this value to ⌊

samples batch size ⌋ so that the model sees the training samples at most once per epoch.

[YanBC]

Same question here. I think index should be between zero and np.floor(len(self.list_IDs) / self.batch_size)

[xingyi-li]

+1, I didn’t find this until I went through errors.