Fix output shape of KerasLayer

[trbromley]

Context:

A minor bug was introduced in #1070. Currently, the following gives a shape of <unknown>:

layer = KerasLayer(...)
inputs = tf.keras.Input(shape=(2,))
inputs_shape = inputs.shape

output_shape = layer.compute_output_shape(inputs_shape)
print(output_shape)

(Note that inputs_shape = (None, 2)).

The output shape should be (None, 1).

Description of the Change:

This PR ensures that the output shape is as expected, i.e., (None, 1). A supporting test is added.

[github-actions]

Hello. You may have forgotten to update the changelog!
Please edit .github/CHANGELOG.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.

[trbromley]

I haven't updated the changelog because it's a bug-fixed for an unreleased change.

[codecov]

Codecov Report

Merging #1165 (f0007ed) into master (46d94db) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #1165   +/-   ##
=======================================
  Coverage   98.11%   98.11%           
=======================================
  Files         144      144           
  Lines       10775    10775           
=======================================
  Hits        10572    10572           
  Misses        203      203           

Impacted Files	Coverage Δ
pennylane/qnn/keras.py	96.92% <100.00%> (ø)

---

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 46d94db...f0007ed. Read the comment docs.

[josh146]

Nice catch!