feat: update transformers to 5.x

[andrea-fasoli]

Description of the change

This PR implements support to upgrade transformer to >5.0 (tested with 5.8).

List of changes:

• upgraded requirements to transformers < 5.9
• added optional dependency group in pyproject.toml called fp8-infer to avoid loading llm_compressor (not used at inference by FMS-MO)
• use_auth_token no longer supported in the new transformers; replaced with auth (note: model_args use_auth_token remains unchanged for backward compatibility)
• special_token_map.json no longer provided by tokenizer; gated assertion to transformer version
• torchscript argument no longer supported in BertModel.from_pretrained(); gated passing this argument to transformer version
• resnet18 and vit now require torchvision; skipping related tests if not installed

Related issues or PRs

Closes #203

Was the PR tested

• I have ensured all unit tests pass (but triton tests fail locally)

Checklist for passing CI/CD:

• All commits are signed showing "Signed-off-by: Name <email@domain.com>" with git commit -signoff or equivalent
• PR title and commit messages adhere to Conventional Commits
• Contribution is formatted with tox -e fix
• Contribution passes linting with tox -e lint
• Contribution passes spellcheck with tox -e spellcheck
• Contribution passes all unit tests with tox -e unit

Note: CI/CD performs unit tests on multiple versions of Python from a fresh install. There may be differences with your local environment and the test environment.