Benchmarks - Update result parsing in tensorrt inference

[abuccts]

• Update result parsing for newer tensorrt versions
• Update arguments when load torchvision models

[codecov[bot]]

Codecov Report

Merging #541 (62b3251) into main (f259913) will increase coverage by 0.01%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #541      +/-   ##
==========================================
+ Coverage   86.97%   86.99%   +0.01%     
==========================================
  Files          89       89              
  Lines        6006     6013       +7     
==========================================
+ Hits         5224     5231       +7     
  Misses        782      782              

Flag	Coverage Δ
cpu-python3.6-unit-test	73.32% <75.00%> (+0.01%)	:arrow_up:
cpu-python3.7-unit-test	73.32% <75.00%> (+0.01%)	:arrow_up:
cpu-python3.8-unit-test	73.80% <75.00%> (+0.01%)	:arrow_up:
cuda-unit-test	86.87% <100.00%> (+0.01%)	:arrow_up:

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

Impacted Files	Coverage Δ
...nchmarks/micro_benchmarks/_export_torch_to_onnx.py	96.00% <100.00%> (+0.16%)	:arrow_up:
...micro_benchmarks/tensorrt_inference_performance.py	89.23% <100.00%> (+0.89%)	:arrow_up: