docs: fix Reward Model links in RLHF guides

[tutao0123]

PR type

• Bug Fix
• New Feature
• Document Updates
• More Models or Datasets Support

PR information

The Reward Model examples were reorganized from the removed
examples/train/rlhf/rm.sh script into the
examples/train/rlhf/rm/ directory. PR #9831 updated the corresponding
README links, but the Chinese and English RLHF guides still referenced the
old path.

This updates the four remaining links to the current example directory. No
runtime behavior changes.

Validation

• Confirmed examples/train/rlhf/rm/ contains train.sh and infer.py.
• Confirmed no examples/train/rlhf/rm.sh references remain under
  docs/source or docs/source_en.
• pre-commit run --files docs/source/Instruction/Pre-training-and-Fine-tuning.md docs/source/Instruction/RLHF.md docs/source_en/Instruction/Pre-training-and-Fine-tuning.md docs/source_en/Instruction/RLHF.md
• git diff --check

Experiment results

Not applicable; this is a documentation-only correction.