Before submitting

• Was this discussed/approved via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure to update the docs?
• Did you write any new necessary tests?
• If you made a notable change (that affects users), did you update the CHANGELOG?

What does this PR do?

Fixes training with iterable dataset with ddp.
Currently if using iterable dataset, without setting sampler in ddp, self.train_dataloader.sampler will be _InfiniteConstantSampler by default in Pytorch. This sampler doesn't have attribute 'set_epoch'. This will create the error as following:

-- Process 0 terminated with the following error:
Traceback (most recent call last):
  File "/root/anaconda3/lib/python3.7/site-packages/torch/multiprocessing/spawn.py", line 20, in _wrap
    fn(i, *args)
  File "/root/anaconda3/lib/python3.7/site-packages/pytorch_lightning/trainer/distrib_data_parallel.py", line 372, in ddp_train
    self.run_pretrain_routine(model)
  File "/root/anaconda3/lib/python3.7/site-packages/pytorch_lightning/trainer/trainer.py", line 914, in run_pretrain_routine
    self.train()
  File "/root/anaconda3/lib/python3.7/site-packages/pytorch_lightning/trainer/training_loop.py", line 332, in train
    self.train_dataloader.sampler.set_epoch(epoch)
AttributeError: '_InfiniteConstantSampler' object has no attribute 'set_epoch'

The bug is caused by the logic sequence between or and and in the following code snippet:

if self.use_ddp or self.use_horovod \
    and hasattr(self.train_dataloader.sampler, 'set_epoch'):

True or False and False = True
(True or False) and False = False

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃