Remaining items for Train/Fine-tune sdk

[deepanker13]

• Add working example for finetuning using s3 as the dataset location. #2005
• PVC creation as part of PyTorch job spec #1971
• Train/Fine-tune API Proposal for LLMs #1945 (comment)
• Train/Fine-tune API Proposal for LLMs #1945 (comment)

@johnugeorge @andreyvelich I am thinking of picking up the remaining tasks in the following order.

[andreyvelich]

Thanks for creating this @deepanker13!
Maybe we could convert this issue to the check-list and create individual issue for each feature ?
So we can track what we can implement before Training Operator 1.8 release.

[StefanoFioravanzo]

Should we track this here? #1994

[andreyvelich]

I am not sure if we are going to cover all action items before Kubeflow 1.9 release, but let's discus it during today's WG call: https://bit.ly/2PWVCkV