Let's make `train_epoch_length` and `eval_epoch_length` optional

[vfdev-5]

Config config.yaml files require two params: train_epoch_length and eval_epoch_length
Let's make them optional.

code-generator/src/templates/template-vision-segmentation/config.yaml

Lines 7 to 8 in a2bee40

	train_epoch_length: 1000
	eval_epoch_length: 1000

Such that epoch length is defined by the input dataloaders.

[DeepC004]

I can take up the issue

[DeepC004]

Hi, may you please elaborate how input dataloaders define epoch length. The torch.utils.data.DataLoader has no parameter defining epoch_length.

[vfdev-5]

@DeepC004 check this code for example how train_epoch_length is used:

code-generator/src/templates/template-vision-classification/main.py

Lines 107 to 111 in b25945f

	trainer.run(
	dataloader_train,
	max_epochs=config.max_epochs,
	epoch_length=config.train_epoch_length,
	)

[DeepC004]

@vfdev-5 I have a doubt regarding the optional values. Aren't the values specified for train_length_epoch and eval_length_epoch already the default ones? Incase the user does not input a custom value for a parameter then default ones will be used as defined by the code.