Open-source implementation naming


#1

Hi,

I have a question about naming in source code contributions. Suppose that I want to contribute to an open-source LSTM RNN implementation, is it acceptable if I name it as EcoRNN where Eco is the name of the organization I am in, or should I name it as OpenRNN or other names?

Thank you.


#2

I can only guess, but if your RNN is something that is backed up with a publication, then the name doesn’t really matter.

If this RNN architecture is only relevant for your particular organization, then it might be not accepted by the community (because there is no documented evidences that this architecture can work on any other data, except of the data of your organization and how better it is comparing with other networks solving similar problem).

If you decide to proceed and your model relates to NLP, then it would be relevant to create a pull request to this repository: https://github.com/dmlc/gluon-nlp - this is where the models related to NLP using Gluon live.


#3

I have an implementation that is the same with cuDNN LSTM RNN (at least algorithmically), and I am hoping to open-source it. Can I name it after my own organization?