need larger dataset - receive fetch error after 40 iterations with alpaca...

[lessw2020]

The alpaca dataset is limited in size - while training the 7B to verify depthwise init was functional, received:

[rank0]: Traceback (most recent call last):
[rank0]: File "/data/users/less/local/torchtrain/train.py", line 288, in
[rank0]: main(args)
[rank0]: File "/data/users/less/local/torchtrain/train.py", line 153, in main
[rank0]: batch = next(iter(data_loader))
[rank0]: File "/home/less/local/miniconda3/envs/triton/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 631, in next
[rank0]: data = self._next_data()
[rank0]: File "/home/less/local/miniconda3/envs/triton/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 675, in _next_data
[rank0]: data = self._dataset_fetcher.fetch(index) # may raise StopIteration
[rank0]: File "/home/less/local/miniconda3/envs/triton/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 39, in fetch
[rank0]: raise StopIteration

[lessw2020]

added miniPile which resolves this (1M entries).
#88