Datasets:
Question on the dataset license
Hi, I have some questions about the dataset license
In the Dataset card tab this is mention "speech recognition models trained on the GigaSpeech 2 dataset may be eligible for commercial license"
But in the dataset Terms of Access this is mention "Researcher shall use the Database only for non-commercial research and educational purposes."
So if I want to train a text to speech model or fine-tune using this dataset is the model available for commercial use? or does the model train from this non-commercial only?
I am confuse on this.
Thank you
Hi,
The GigaSpeech 2 dataset itself is restricted to non-commercial research and educational purposes only. However, machine learning models trained on this dataset may potentially be used commercially, provided their usage complies with the "Fair Use" principle and does not violate any explicit copyright restrictions. SpeechColab does not own copyright to the audio files and thus cannot grant explicit commercial licenses for models derived from the dataset. Evaluating compliance with "Fair Use" for your specific use-case remains your responsibility.
For more information, please refer to: Fair Use