Bug: Benchmarks - remove fp16 samples type converting time for training cnn and lstm inference

[yukirora]

Description
remove fp16 samples type converting time for training cnn and lstm inference.

[codecov]

Codecov Report

Merging #332 (a9d9902) into main (a9634ef) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main     #332   +/-   ##
=======================================
  Coverage   88.61%   88.61%           
=======================================
  Files          78       78           
  Lines        4665     4665           
=======================================
  Hits         4134     4134           
  Misses        531      531           

Flag	Coverage Δ
cpu-unit-test	73.30% <100.00%> (ø)
cuda-unit-test	88.55% <100.00%> (ø)

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

Impacted Files	Coverage Δ
...erbench/benchmarks/model_benchmarks/pytorch_cnn.py	93.40% <100.00%> (ø)
...rbench/benchmarks/model_benchmarks/pytorch_lstm.py	94.62% <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 a9634ef...a9d9902. Read the comment docs.