Seqchromloader

New `SeqChromDatasetByDataFrame` function to load tensors given a region dataframe

What's Changed
* Sphinx document integration by yztxwd in https://github.com/seqcode/seqchromloader/pull/1

New Contributors
* yztxwd made their first contribution in https://github.com/seqcode/seqchromloader/pull/1

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.2.4...0.3.0

1. Now SeqchromDatasetByWds can use device rank info to split files
2. Fix the bug occurred when there are more workers than wds files
3. Allow empty Bigwig file list as input

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.2.3...0.2.4

Fix the performance issue due to each worker of webdataset loader uncompressed all files instead of uncompressing only the files assigned

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.2.2...0.2.3

Fixes:

1. Add unique id to each sample to avoid duplicate sample name in webdataset
2. Change default dataloader kwargs to `{}`
3. Initialize Exception class before calling `__init__()`

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.2.1...0.2.2

- Consistent API for pass transform functions
- Clean the code

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.2.0...0.2.1

- Add dataloder_kws to grant loaders the ability to accept all dataloader arguments
- writer function name adjustment
- Several bug fixes

**Full Changelog**: https://github.com/seqcode/seqchromloader/compare/0.1.1...0.2.0