Tiny fastMRI data split proposal

[mmuckley]

Although the fastMRI 2.0 release reduced data storage requirements by batching the fastMRI dataset, in some cases even the fastMRI 2.0 batches can be too small to work with (e.g., Discussion #304). This issue is to collect interest on whether even smaller, 100 GB batches might be of interest to the community.

If you think that 100 GB batches could be useful to you, please react to this post with a 👍🏼 emoji.

[stperrakis]

Great news 👍🏻 👍🏻 :)