Update docs and examples to use Triton 23.06

[efajardo-nv]

Description

• Update docs and examples to use Triton 23.06
• Most examples are using 22.08. Significant speedup is seen in ONNX->TRT conversion that happens in first inference run of the pipeline (e.g. SID).

Checklist

[x] I am familiar with the Contributing Guidelines.
[x] New or existing tests cover these changes.
[x] The documentation is up to date with these changes.

[mdemoret-nv]

Approved. Going to merge this after running val scripts locally

[mdemoret-nv]

/merge

[mdemoret-nv]

For reference, using 23.06 has significantly sped up the validation scripts by about ~50%:

# 23.06
real    3m24.425s
user    2m46.757s
sys     0m30.639s

# 22.08
real    7m6.385s
user    6m18.460s
sys     0m57.667s

Most of this is due to the improved ONNX->TRT conversion in Triton.