There is an AudioSet package in the previous working repo that needs to be cloned into this one for end-to-end prediction.
Should come with both the package and corresponding tests. One trick will be wgeting the VGGish model params in Travis, but this problem is already solved.
There is an
AudioSet
package in the previous working repo that needs to be cloned into this one for end-to-end prediction.Should come with both the package and corresponding tests. One trick will be
wget
ing the VGGish model params in Travis, but this problem is already solved.