[FAQ] Why should i train 'romp' by myself?

[vivi90]

I am missing an notice about the purpose of training romp by myself. For example: Are there any benefits of usage, if i train it again by myself instead of using the attached dataset?

Would be great to get some guidelines about this question.

[ZhengdiYu]

I am missing an notice about the purpose of training romp by myself. For example: Are there any benefits of usage, if i train it again by myself instead of using the attached dataset?

Would be great to get some guidelines about this question.

Hi, what do you mean by 'train it again by myself instead of using the attached dataset' ? I think training romp is served for 3 purposes:

1. To reproduce the results in the paper and to better understand the paper.
2. Serve as an instruction/hint to re-implement some useful plugins or some of your own work based on romp.
3. If you mean to train romp on your own dataset instead of the dataset mentioned in the paper. Modifying the existing ones would be a great start.

[vivi90]

@ZhengdiYu Yes, this was exactly the answer i missed in the README file. Thank you.

[vivi90]

@ZhengdiYu Please create an PR, if possible :)

[ZhengdiYu]

@ZhengdiYu Please create an PR, if possible :)

Do you mean that you need a pull request to add he information into README?

[vivi90]

@ZhengdiYu Why not? :)

[ZhengdiYu]

@ZhengdiYu Why not? :)

OK I will do that recently, but I have to work for now. Your profile picture is so cute, I like it.

[vivi90]

@ZhengdiYu

OK I will do that recently, [..]

ping

Your profile picture is so cute, I like it.

Thank you^^ It was one of my first exercises: https://www.deviantart.com/suk0m8u/art/Renshuu2-818674802