use int64 for samples instead of bool8

[timmysilv]

Context:
The new qubit device was going to return booleans for 0-1 samples because that is what a boolean represents. However, this change is different from what we had before., and booleans still take a full byte of memory to stay aligned (let's not get too crazy here in Pythonland anyway) which is the same as np.int8.

Description of the Change:
sample_state, and consequentially qml.sample() now returns samples of dtype=np.int64 instead of np.bool8.

Benefits:

• The user will receive the more-familiar int return type.
• There are no problems with memory-alignment
• The new device API is consistent with the old one
• existing interface code still works without change

Possible Drawbacks:
np.bool8 was smaller

[github-actions]

Hello. You may have forgotten to update the changelog!
Please edit doc/releases/changelog-dev.md with:

• A one-to-two sentence description of the change. You may include a small working example for new features.
• A link back to this PR.
• Your name (or GitHub username) in the contributors section.

[codecov]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (15bcf04) 99.66% compared to head (73af749) 99.66%.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #4539   +/-   ##
=======================================
  Coverage   99.66%   99.66%           
=======================================
  Files         376      376           
  Lines       33059    33060    +1     
=======================================
+ Hits        32949    32950    +1     
  Misses        110      110           

Files Changed	Coverage Δ
pennylane/devices/qubit/sampling.py	100.00% <100.00%> (ø)

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[albi3ro]

Sad that we have to do this, but it allows us to move forward.