How should the value of --num_candidates and --personality_permutations be determined?

I don't understand what you mean by command line parameter `--num_candidates` and `--personality_permutations`.(even though read help)

Should these be changed depending on the content of the training data?

For example, if training data is like this:
```
{
"personality": ["persona description"],
"utterances": [
{"candidates": [
10 data
],
"history": [
"history",
]
}
]
}
```
Should `--num_candidates` be set to 10??

And I don't fully understand about `--personality_permutations`.

I would appreciate it if you could let me know!!

Start by locating the CLI entry points that define --num_candidates and --personality_permutations, then compare their help text with the example training format in the issue. Done means documenting what each parameter controls and whether --num_candidates should match the candidate count.

Written by the indexing model from the issue text.