陳羿豐 Yi-Feng Chen
陳羿豐 Yi-Feng Chen
AudioSet is video data from YouTube, which is blocked in China. If you are in China, please check for VPN settings. Also please install youtube-dl via pip to get latest...
Current translation on HackMD: https://hackmd.io/NELaVeAORWORAvz68NxAow
I use NT-Xent loss function, the same as the original paper. The validation score during training should be 1-second query accuracy, however it turned out to be way too high...
Because I used training music for dummy database during testing. I consider the case that we add songs after we train our model. In this case, the training data is...