feat: enable reward model training without dropping the last validation batch

[jveronvialard]

What does this PR do ?

This PR enables reward model training with validation_drop_last =False.

Changelog

• Please update the CHANGELOG.md under next version with high level changes in this PR.

Usage

python ${NEMO_ALIGNER_PATH}/examples/nlp/gpt/train_reward_model.py \
    ...
    ++model.data.validation_drop_last=False \
    ...

Before your PR is "Ready for review"

Pre checks:

• Make sure you read and followed Contributor guidelines
• Did you write any new necessary tests?
• Did you add or update any necessary documentation? Make sure to also update the NeMo Framework User Guide which contains the tutorials

Checklist when contributing a new algorithm

• Does the trainer resume and restore model state all states?
• Does the trainer support all parallelism techniques(PP, TP, DP)?
• Does the trainer support max_steps=-1 and validation?
• Does the trainer only call APIs defined in alignable_interface.py?
• Does the trainer have proper logging?

Additional Information

• Related to # (issue)

[odelalleau]

Great, thanks! Just a few minor suggestions then should be good to go

[odelalleau]

Good to go, thanks!