I designed and implemented a neural network model for the recognition function myself, but because of the cc by 4.0 copyright on the training data set, I received feedback when I submitted it to the Marketplace that all the open source protocols requiring attribution could not be approved.
I understood this and made positive improvements: my model was the result of a thesis on which I was the first author, This means that all information about the training can be found in this paper, including but not limited to the attribution of the training data. And then I open sourced the model weight file to the wtfpl protocol, and the submitted code plug-in contained only the interface file.
However, the auditor seemed to call back without reading it, which is still the previous statement, and gave me a warning that repeated submission will be punished. This is very disappointing to me. I do not know how the market defines the copyright of artificial intelligence models, even if CC officials only discuss the copyright of artificial intelligence models without a clear statement.
I expect a mutual discussion, not a look and see attitude of refusing to communicate 
, so will the Marketplace continue to shut out of all AI models in the future?