pyhf v0.63 released

[pyhf announcements]

Dear pyhf users,

We're happy to announce that pyhf v0.6.3 is out now on PyPI and Conda-forge! This is a patch release with a few additional important changes that we didn't want to delay. As always please read the full release notes carefully while updating your code to use the new release. Below we list the most important points from the release notes.

* With the addition of writing ROOT files in uproot v4.1.0 the xmlio extra no longer requires uproot3 and all dependencies on uproot3 and uproot3-methods have been dropped.
* All backends are now fully compatible and tested with Python 3.9.
* The TensorFlow backend now supports compatibility with TensorFlow v2.2.1 and later and TensorFlow Probability v0.10.1 and later.
* The pyhf.workspace.Workspace.data() `with_aux` keyword arg has been renamed to `include_auxdata` to improve API consistency.

* The pyhf.pdf._ModelConfig `channel_nbins` dict is now sorted by keys to match the order of the `channels dict`.

* We have overhauled pyhf's Issue templates into Issue forms with the goal of making Issue reporting and question asking more streamlined for users. If the Issue forms are confusing or not helpful please let us know by opening up a Docs Issue!

There is one final patch release, v0.6.4, planned before minor release v0.7.0 with a primary focus of addressing some long standing Issues on pseudoexperiment generation. Development work on v0.6.4 has already started, but we will be taking some time to focus on catching up on the backlog of user submitted contributions, questions, and documentation that has accrued. Thank you for your patience and please report back to us any questions or problems you run into with the v0.6.X API. Please don't hesitate to contact us through the recommended channels!

Best,
The pyhf dev team (Lukas, Matthew, and Giordon)