[Feature,Quality] Raise ValueError if split_size is zero

[zhanwenchen]

Add validation to prevent split_size from being zero.

Description

Add validation to prevent split_size from being zero, which then causes CPU OOM.

Motivation and Context

Why is this change required? What problem does it solve?
verl gets CPU OOM killed when ppo_micro_batch_size < world_size in data.batch.split(ppo_micro_batch_size//world_size). That is, when TensorDict.split gets an 0 int input.

Fixes #1479.

• I have raised an issue to propose this change (required for new features and bug fixes)

Types of changes

What types of changes does your code introduce? Remove all that do not apply:

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds core functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Documentation (update in the documentation)
• Example (update in the folder of examples)

Checklist

Go over all the following points, and put an x in all the boxes that apply.
If you are unsure about any of these, don't hesitate to ask. We are here to help!

• I have read the CONTRIBUTION guide (required)
• My change requires a change to the documentation.
• I have updated the tests accordingly (required for a bug fix or a new feature).
• I have updated the documentation accordingly.

[vmoens]

Hello
Thanks for opening this! Are we confident that it will close #1479 ?
I guess that we could run a better check than this, e.g. the split size must be non-negative and within a certain range