Remove TRT optimization from all-MiniLM-L6-v2

[efajardo-nv]

Description

• Remove automatic TensorRT optimization from all-MiniLM-L6-v2 config.pbtxt. This was causing segfault in vdb_upload example.
• Triton issue: Error when using ONNX with TensorRT (ORT-TRT) Optimization on Multi-GPU triton-inference-server/server#7885
• This PR removed optimization from sid-minibert-onnx and phishing-bert-onnx

Closes #1649

By Submitting this PR I confirm:

• I am familiar with the Contributing Guidelines.
• When the PR is ready for review, new or existing tests cover these changes.
• When the PR is ready for review, the documentation is up to date with these changes.

[dagardner-nv]

Does this PR fix #1649? If so can you update the PR to indicate that it does, and remove the reference to the bug in docs/source/extra_info/known_issues.md

[efajardo-nv]

Does this PR fix #1649? If so can you update the PR to indicate that it does, and remove the reference to the bug in docs/source/extra_info/known_issues.md
Yes appears so. Not seeing that error and no longer using tensorrt which is referenced by the error.

[dagardner-nv]

/merge