hi. qwen_asr is not available even though it seems to be in the runtime. you can check submission id-303160
Hi @cteceliker, can you please try again? I believe the issue should be resolved now.
Is usage of that model allowed? From the rules:
Not allowed:
Participants may not use any of the following external datasets or data providers, nor any pre-trained models that are known to have been trained on...
In its paper, model does not itemize which corpora were used for training. However, due to a raw volume of training data, there is a reasonable expectation that the datasets were in fact used.
So question is, is it ok to use open models with correct rights which obfuscate its learning dataset? There is a point in which documentation to the used model must be provided. Is corpora used for training of the external model not needed?
Yes, it’s allowed. As stated in the competition rules, participants may use external pre-trained models as long as there is no indication that those models were trained on any of the prohibited datasets identified in the guidelines, and their licenses allow for broad commercial use.
Please also note that finalist teams must disclose any external models and datasets utilized and confirm that their use complies with all competition requirements.
Let us know if you have any other concerns, and happy modeling!
1 Like