fix(examples): handle missing SFT chat template

[fallintoplace]

Summary

• Guard the DeepSeek/R1 chat template rewrite when a tokenizer has no chat_template.
• Use the matching dataset prompt template string as the fallback instead of assigning the whole template map.
• Keep the existing hard failure when neither the tokenizer nor the dataset provides a valid template.

Validation

• python3 -m py_compile examples/post_training/modelopt/finetune.py
• ruff check --no-fix examples/post_training/modelopt/finetune.py
• git diff --check
• BASE_REF=main CHECK_ONLY=true SKIP_DOCS=false bash tools/autoformat.sh
• Stubbed constructor regression for missing-template fallback, unknown-dataset failure, and DeepSeek template rewrite.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[github-actions]

This PR has been automatically converted to draft because all PRs must start as drafts.

When you are ready for review, click Ready for Review to begin the review process. This will:

1. Add the oncall reviewer (optional reviewer)
2. Add required review teams based on your changes

See the contribution guide for more details.