Add additional metadata to the columns, dataset and benchmark

[cwognum]

Changelogs

Metadata for benchmarks

• Fixes Additional attributes for Benchmark  #63
  • Adds the task type (i.e. multi task or single task), computed automatically.
  • Adds the target type (i.e. regression or classification), can be manually specified, but if not lib tries to infer automatically using sklearn's type_of_target().
  • Adds the train set size, computed automatically.
  • Adds the test set(s) size(s), computed automatically.
  • Adds the number of classes for classification tasks, computed automatically.
  • Adds the number of test sets

Metadata for datasets

• Fixes Feat: Add precomputed fields #18
  • Adds the dtype to the column annotations, computed automatically.
  • Adds the no. datapoints and no. columns, computed automatically.
  • Adds a n optional field with a reference to the curation.

Misc

• Add additional syntax to the dataset to make interacting with it easier, similar to pandas.DataFrame. E.g. dataset[row, col] or dataset[:, "smiles"].

---

Checklist:

• Was this PR discussed in an issue? It is recommended to first discuss a new feature into a GitHub issue before opening a PR.
• Add tests to cover the fixed bug(s) or the newly introduced feature(s) (if appropriate).
• Update the API documentation if a new function is added, or an existing one is deleted.
• Write concise and explanatory changelogs above.
• If possible, assign one of the following labels to the PR: feature, fix or test (or ask a maintainer to do it for you).

[cwognum]

Worth noting that we will also have the size of the dataset (in bytes), but I will save this on the Hub side since the Hub already receives this information anyways.

[cwognum]

Quick! Change from meta-data to metadata before @jstlaurent and @hadim spot it! 👀

[zhu0619]

Thanks Cas.
It looks good to me.
I have only one minor comment.