build: Set only lower bounds on backend dependencies

[matthewfeickert]

Description

Set only lower bounds for all the pyhf backend extras. pyhf is testing daily the lower bounds, latest releases, and pre-releases, so these are empirically arrived at and tested. Note this is simply just enforcing the lower bounds already tested

$ grep 'tensor\|torch' lower-bound-requirements.txt 
# tensorflow
tensorflow==2.3.1  # tensorflow-probability v0.11.0 requires tensorflow>=2.3
tensorflow-probability==0.11.0  # c.f. PR #1657
# torch
torch==1.10.0

c.f. Henry's upcoming blog post for the deep dive on why this is optimal for libraries

Also add a note for the PR that caused these lower bounds to exist for easy reference for the core devs or the curious contributor.

Checklist Before Requesting Reviewer

• Tests are passing
• "WIP" removed from the title of the pull request
• Selected an Assignee for the PR to be responsible for the log summary

Before Merging

For the PR Assignees:

• Summarize commit messages into a comprehensive review of the PR

* Set only lower bounds on the requirements for all backend extras.
   - This is done to avoid creating installation blocking scenarios
     and in situations where users find that the latest release of
     a dependency breaks pyhf code the user can be recommended to
     install the version of that dependency that was available when
     their version of pyhf was released.
* Add notes in setup.py pointing to the PR that caused the lower bound
to be needed for easy future reference

[codecov]

Codecov Report

Merging #1698 (c36087f) into master (d5b73c0) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1698   +/-   ##
=======================================
  Coverage   98.10%   98.10%           
=======================================
  Files          64       64           
  Lines        4226     4226           
  Branches      587      587           
=======================================
  Hits         4146     4146           
  Misses         46       46           
  Partials       34       34           

Flag	Coverage Δ
contrib	25.41% <ø> (ø)
doctest	61.19% <ø> (ø)
unittests	96.42% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d5b73c0...c36087f. Read the comment docs.