perf: optimize string tensor deserialization with high performance c++ implementation#416

wweic wants to merge 2 commits intotriton-inference-server:maintriton-inference-server/python_backend:mainfrom

wweic:wweic/optimize-string-tensor-prwweic/python_backend:wweic/optimize-string-tensor-prCopy head branch name to clipboard