Support Inputs and Labels in the dict format

[NihalHarish]

Description of changes:

• flatten dict structure before saving them as tensors.
• Using the same logic used by tf to handle dict structures.

Style and formatting:

I have run pre-commit install to ensure that auto-formatting happens with every commit.

Issue number, if available

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[codecov-commenter]

Codecov Report

Merging #345 into master will decrease coverage by 1.87%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #345      +/-   ##
==========================================
- Coverage   87.24%   85.36%   -1.88%     
==========================================
  Files          86       86              
  Lines        6474     6477       +3     
==========================================
- Hits         5648     5529     -119     
- Misses        826      948     +122     

Impacted Files	Coverage Δ
smdebug/tensorflow/keras.py	92.81% <100.00%> (+0.03%)	:arrow_up:
smdebug/mxnet/graph.py	25.80% <0.00%> (-45.17%)	:arrow_down:
smdebug/mxnet/utils.py	59.09% <0.00%> (-31.82%)	:arrow_down:
smdebug/core/logger.py	66.12% <0.00%> (-22.59%)	:arrow_down:
smdebug/analysis/utils.py	66.66% <0.00%> (-22.23%)	:arrow_down:
smdebug/core/access_layer/base.py	64.28% <0.00%> (-21.43%)	:arrow_down:
smdebug/pytorch/utils.py	28.20% <0.00%> (-20.52%)	:arrow_down:
smdebug/core/tfevent/summary.py	81.35% <0.00%> (-15.26%)	:arrow_down:
smdebug/core/singleton_utils.py	85.29% <0.00%> (-11.77%)	:arrow_down:
smdebug/exceptions.py	67.53% <0.00%> (-10.39%)	:arrow_down:
... and 11 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 3cfa8d5...b25e5d7. Read the comment docs.