Hi all,
Just wanted to check whether these pretrained models are allowed for prize elegibility since licensing is a bit confusing:
Distil-Whisper: I noticed it was trained on TED-LIUM (CC-BY-NC-ND 3.0), which has a Non-Commercial clause. Does this disqualify it for prize eligibility under the external data rules?
Whisper large-v3: The model weights are MIT-licensed, but the training data is web-scraped with unclear provenance. Is this sufficient for prize eligibility, or does the training data need to be traceable?
Thank you!