Integer limit_val_batches does not work with infinite DataLoader.

[discharged-spider]

🐛 Bug

Documentation states that you can use limit_val_batches=100 (with integer value) to limit number of batches. However, when using IterableDataset I get the following error:
pytorch_lightning.utilities.exceptions.MisconfigurationException: When using an infinite DataLoader (e.g. with an IterableDataset or when DataLoader does not implement __len__) for limit_val_batches, Trainer(limit_val_batches) must be 0.0 or 1.0.

To Reproduce

Just use:
pl.Trainer(val_check_interval=10, limit_val_batches=100)

with infinite DataLoader.

Expected behavior

Do only N (100 in this case) validation steps.

Environment

• CUDA:
  - GPU:
  - GeForce GTX 1080 Ti
  - available: True
  - version: 10.0.130
• Packages:
  - numpy: 1.17.4
  - pyTorch_debug: False
  - pyTorch_version: 1.3.1
  - pytorch-lightning: 0.8.5
  - tensorboard: 2.2.2
  - tqdm: 4.46.1
• System:
  - OS: Linux
  - architecture:
  - 64bit
  -
  - processor: x86_64
  - python: 3.7.4
  - version: Proposal for help #1 SMP Wed Mar 7 19:03:37 UTC 2018

[github-actions]

Hi! thanks for your contribution!, great first issue!

[rohitgr7]

The bug is here:
https://github.com/PyTorchLightning/pytorch-lightning/blob/6d10ac2ac84bc01f78eed381fd6152c869040b64/pytorch_lightning/trainer/data_loading.py#L303-L316